diff --git "a/qwen25_0.5b_fmft_official_5e-05/checkpoint-117744/trainer_state.json" "b/qwen25_0.5b_fmft_official_5e-05/checkpoint-117744/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/qwen25_0.5b_fmft_official_5e-05/checkpoint-117744/trainer_state.json"
@@ -0,0 +1,164933 @@
+{
+  "best_metric": 0.003188470809259278,
+  "best_model_checkpoint": "./results-cc/qwen25-0.5b-instruct/qwen25_0.5b_fmft_official_5e-05/checkpoint-58872",
+  "epoch": 8.0,
+  "eval_steps": 500,
+  "global_step": 117744,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003397200706617747,
+      "grad_norm": 23.875,
+      "learning_rate": 4.9997876749558366e-05,
+      "loss": 2.399,
+      "step": 5
+    },
+    {
+      "epoch": 0.0006794401413235494,
+      "grad_norm": 11.4375,
+      "learning_rate": 4.999575349911673e-05,
+      "loss": 1.6737,
+      "step": 10
+    },
+    {
+      "epoch": 0.0010191602119853241,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.9993630248675094e-05,
+      "loss": 1.6248,
+      "step": 15
+    },
+    {
+      "epoch": 0.001358880282647099,
+      "grad_norm": 9.25,
+      "learning_rate": 4.999150699823346e-05,
+      "loss": 1.5568,
+      "step": 20
+    },
+    {
+      "epoch": 0.0016986003533088735,
+      "grad_norm": 7.625,
+      "learning_rate": 4.998938374779182e-05,
+      "loss": 1.6539,
+      "step": 25
+    },
+    {
+      "epoch": 0.0020383204239706482,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.9987260497350186e-05,
+      "loss": 1.6086,
+      "step": 30
+    },
+    {
+      "epoch": 0.002378040494632423,
+      "grad_norm": 8.25,
+      "learning_rate": 4.998513724690855e-05,
+      "loss": 1.5407,
+      "step": 35
+    },
+    {
+      "epoch": 0.002717760565294198,
+      "grad_norm": 7.28125,
+      "learning_rate": 4.9983013996466914e-05,
+      "loss": 1.5034,
+      "step": 40
+    },
+    {
+      "epoch": 0.0030574806359559724,
+      "grad_norm": 7.625,
+      "learning_rate": 4.998089074602528e-05,
+      "loss": 1.5899,
+      "step": 45
+    },
+    {
+      "epoch": 0.003397200706617747,
+      "grad_norm": 7.125,
+      "learning_rate": 4.997876749558364e-05,
+      "loss": 1.5309,
+      "step": 50
+    },
+    {
+      "epoch": 0.0037369207772795215,
+      "grad_norm": 7.75,
+      "learning_rate": 4.9976644245142006e-05,
+      "loss": 1.5443,
+      "step": 55
+    },
+    {
+      "epoch": 0.0040766408479412965,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.997452099470037e-05,
+      "loss": 1.5409,
+      "step": 60
+    },
+    {
+      "epoch": 0.0044163609186030715,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.997239774425873e-05,
+      "loss": 1.635,
+      "step": 65
+    },
+    {
+      "epoch": 0.004756080989264846,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.99702744938171e-05,
+      "loss": 1.4198,
+      "step": 70
+    },
+    {
+      "epoch": 0.005095801059926621,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.996815124337546e-05,
+      "loss": 1.5308,
+      "step": 75
+    },
+    {
+      "epoch": 0.005435521130588396,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.996602799293382e-05,
+      "loss": 1.5627,
+      "step": 80
+    },
+    {
+      "epoch": 0.00577524120125017,
+      "grad_norm": 7.25,
+      "learning_rate": 4.996390474249219e-05,
+      "loss": 1.5605,
+      "step": 85
+    },
+    {
+      "epoch": 0.006114961271911945,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.9961781492050554e-05,
+      "loss": 1.5697,
+      "step": 90
+    },
+    {
+      "epoch": 0.006454681342573719,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.995965824160891e-05,
+      "loss": 1.5472,
+      "step": 95
+    },
+    {
+      "epoch": 0.006794401413235494,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.995753499116728e-05,
+      "loss": 1.5484,
+      "step": 100
+    },
+    {
+      "epoch": 0.007134121483897269,
+      "grad_norm": 6.875,
+      "learning_rate": 4.9955411740725646e-05,
+      "loss": 1.5175,
+      "step": 105
+    },
+    {
+      "epoch": 0.007473841554559043,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.9953288490284004e-05,
+      "loss": 1.5356,
+      "step": 110
+    },
+    {
+      "epoch": 0.007813561625220818,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.9951165239842374e-05,
+      "loss": 1.6371,
+      "step": 115
+    },
+    {
+      "epoch": 0.008153281695882593,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.994904198940074e-05,
+      "loss": 1.5554,
+      "step": 120
+    },
+    {
+      "epoch": 0.008493001766544368,
+      "grad_norm": 7.0,
+      "learning_rate": 4.9946918738959096e-05,
+      "loss": 1.5367,
+      "step": 125
+    },
+    {
+      "epoch": 0.008832721837206143,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.9944795488517466e-05,
+      "loss": 1.4832,
+      "step": 130
+    },
+    {
+      "epoch": 0.009172441907867916,
+      "grad_norm": 7.375,
+      "learning_rate": 4.994267223807583e-05,
+      "loss": 1.5095,
+      "step": 135
+    },
+    {
+      "epoch": 0.009512161978529691,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.994054898763419e-05,
+      "loss": 1.5179,
+      "step": 140
+    },
+    {
+      "epoch": 0.009851882049191466,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.993842573719256e-05,
+      "loss": 1.5677,
+      "step": 145
+    },
+    {
+      "epoch": 0.010191602119853241,
+      "grad_norm": 8.375,
+      "learning_rate": 4.9936302486750916e-05,
+      "loss": 1.6052,
+      "step": 150
+    },
+    {
+      "epoch": 0.010531322190515016,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.993417923630928e-05,
+      "loss": 1.5882,
+      "step": 155
+    },
+    {
+      "epoch": 0.010871042261176791,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.993205598586765e-05,
+      "loss": 1.5258,
+      "step": 160
+    },
+    {
+      "epoch": 0.011210762331838564,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.992993273542601e-05,
+      "loss": 1.5688,
+      "step": 165
+    },
+    {
+      "epoch": 0.01155048240250034,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.992780948498437e-05,
+      "loss": 1.5488,
+      "step": 170
+    },
+    {
+      "epoch": 0.011890202473162114,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.992568623454274e-05,
+      "loss": 1.4732,
+      "step": 175
+    },
+    {
+      "epoch": 0.01222992254382389,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.99235629841011e-05,
+      "loss": 1.5702,
+      "step": 180
+    },
+    {
+      "epoch": 0.012569642614485664,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.992143973365947e-05,
+      "loss": 1.5415,
+      "step": 185
+    },
+    {
+      "epoch": 0.012909362685147438,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.9919316483217835e-05,
+      "loss": 1.5678,
+      "step": 190
+    },
+    {
+      "epoch": 0.013249082755809213,
+      "grad_norm": 6.625,
+      "learning_rate": 4.991719323277619e-05,
+      "loss": 1.6123,
+      "step": 195
+    },
+    {
+      "epoch": 0.013588802826470988,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.991506998233456e-05,
+      "loss": 1.4946,
+      "step": 200
+    },
+    {
+      "epoch": 0.013928522897132763,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.9912946731892927e-05,
+      "loss": 1.4599,
+      "step": 205
+    },
+    {
+      "epoch": 0.014268242967794538,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.9910823481451284e-05,
+      "loss": 1.5442,
+      "step": 210
+    },
+    {
+      "epoch": 0.014607963038456313,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.9908700231009655e-05,
+      "loss": 1.569,
+      "step": 215
+    },
+    {
+      "epoch": 0.014947683109118086,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.990657698056801e-05,
+      "loss": 1.6068,
+      "step": 220
+    },
+    {
+      "epoch": 0.015287403179779861,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.9904453730126376e-05,
+      "loss": 1.5767,
+      "step": 225
+    },
+    {
+      "epoch": 0.015627123250441636,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9902330479684747e-05,
+      "loss": 1.549,
+      "step": 230
+    },
+    {
+      "epoch": 0.01596684332110341,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.9900207229243104e-05,
+      "loss": 1.5223,
+      "step": 235
+    },
+    {
+      "epoch": 0.016306563391765186,
+      "grad_norm": 6.75,
+      "learning_rate": 4.989808397880147e-05,
+      "loss": 1.4527,
+      "step": 240
+    },
+    {
+      "epoch": 0.01664628346242696,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.989596072835984e-05,
+      "loss": 1.5136,
+      "step": 245
+    },
+    {
+      "epoch": 0.016986003533088736,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9893837477918196e-05,
+      "loss": 1.5361,
+      "step": 250
+    },
+    {
+      "epoch": 0.01732572360375051,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.989171422747656e-05,
+      "loss": 1.4537,
+      "step": 255
+    },
+    {
+      "epoch": 0.017665443674412286,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.988959097703493e-05,
+      "loss": 1.4849,
+      "step": 260
+    },
+    {
+      "epoch": 0.01800516374507406,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.988746772659329e-05,
+      "loss": 1.6085,
+      "step": 265
+    },
+    {
+      "epoch": 0.018344883815735832,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.988534447615165e-05,
+      "loss": 1.4448,
+      "step": 270
+    },
+    {
+      "epoch": 0.01868460388639761,
+      "grad_norm": 6.375,
+      "learning_rate": 4.988322122571002e-05,
+      "loss": 1.5382,
+      "step": 275
+    },
+    {
+      "epoch": 0.019024323957059382,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.988109797526838e-05,
+      "loss": 1.5222,
+      "step": 280
+    },
+    {
+      "epoch": 0.01936404402772116,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.9878974724826744e-05,
+      "loss": 1.48,
+      "step": 285
+    },
+    {
+      "epoch": 0.019703764098382932,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.987685147438511e-05,
+      "loss": 1.5379,
+      "step": 290
+    },
+    {
+      "epoch": 0.020043484169044706,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.987472822394347e-05,
+      "loss": 1.5766,
+      "step": 295
+    },
+    {
+      "epoch": 0.020383204239706482,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.9872604973501836e-05,
+      "loss": 1.534,
+      "step": 300
+    },
+    {
+      "epoch": 0.020722924310368256,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.98704817230602e-05,
+      "loss": 1.4431,
+      "step": 305
+    },
+    {
+      "epoch": 0.021062644381030032,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.9868358472618564e-05,
+      "loss": 1.5234,
+      "step": 310
+    },
+    {
+      "epoch": 0.021402364451691806,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.986623522217693e-05,
+      "loss": 1.5433,
+      "step": 315
+    },
+    {
+      "epoch": 0.021742084522353582,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.986411197173529e-05,
+      "loss": 1.5083,
+      "step": 320
+    },
+    {
+      "epoch": 0.022081804593015356,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.9861988721293656e-05,
+      "loss": 1.566,
+      "step": 325
+    },
+    {
+      "epoch": 0.02242152466367713,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.985986547085202e-05,
+      "loss": 1.4854,
+      "step": 330
+    },
+    {
+      "epoch": 0.022761244734338906,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.9857742220410384e-05,
+      "loss": 1.58,
+      "step": 335
+    },
+    {
+      "epoch": 0.02310096480500068,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.985561896996875e-05,
+      "loss": 1.4847,
+      "step": 340
+    },
+    {
+      "epoch": 0.023440684875662456,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.985349571952711e-05,
+      "loss": 1.5447,
+      "step": 345
+    },
+    {
+      "epoch": 0.02378040494632423,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.9851372469085476e-05,
+      "loss": 1.415,
+      "step": 350
+    },
+    {
+      "epoch": 0.024120125016986002,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.984924921864384e-05,
+      "loss": 1.4542,
+      "step": 355
+    },
+    {
+      "epoch": 0.02445984508764778,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.9847125968202204e-05,
+      "loss": 1.5284,
+      "step": 360
+    },
+    {
+      "epoch": 0.024799565158309552,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.984500271776057e-05,
+      "loss": 1.5195,
+      "step": 365
+    },
+    {
+      "epoch": 0.02513928522897133,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.984287946731893e-05,
+      "loss": 1.5618,
+      "step": 370
+    },
+    {
+      "epoch": 0.025479005299633102,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.9840756216877296e-05,
+      "loss": 1.5481,
+      "step": 375
+    },
+    {
+      "epoch": 0.025818725370294875,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.983863296643566e-05,
+      "loss": 1.5947,
+      "step": 380
+    },
+    {
+      "epoch": 0.026158445440956652,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.9836509715994024e-05,
+      "loss": 1.5959,
+      "step": 385
+    },
+    {
+      "epoch": 0.026498165511618425,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.983438646555239e-05,
+      "loss": 1.5336,
+      "step": 390
+    },
+    {
+      "epoch": 0.026837885582280202,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.983226321511075e-05,
+      "loss": 1.491,
+      "step": 395
+    },
+    {
+      "epoch": 0.027177605652941975,
+      "grad_norm": 6.875,
+      "learning_rate": 4.9830139964669116e-05,
+      "loss": 1.5064,
+      "step": 400
+    },
+    {
+      "epoch": 0.027517325723603752,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.982801671422748e-05,
+      "loss": 1.4857,
+      "step": 405
+    },
+    {
+      "epoch": 0.027857045794265525,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.9825893463785844e-05,
+      "loss": 1.5492,
+      "step": 410
+    },
+    {
+      "epoch": 0.0281967658649273,
+      "grad_norm": 7.5,
+      "learning_rate": 4.982377021334421e-05,
+      "loss": 1.5467,
+      "step": 415
+    },
+    {
+      "epoch": 0.028536485935589075,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.9821646962902565e-05,
+      "loss": 1.4709,
+      "step": 420
+    },
+    {
+      "epoch": 0.02887620600625085,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.9819523712460936e-05,
+      "loss": 1.5154,
+      "step": 425
+    },
+    {
+      "epoch": 0.029215926076912625,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.98174004620193e-05,
+      "loss": 1.4932,
+      "step": 430
+    },
+    {
+      "epoch": 0.0295556461475744,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.981527721157766e-05,
+      "loss": 1.4704,
+      "step": 435
+    },
+    {
+      "epoch": 0.029895366218236172,
+      "grad_norm": 6.125,
+      "learning_rate": 4.981315396113603e-05,
+      "loss": 1.4731,
+      "step": 440
+    },
+    {
+      "epoch": 0.03023508628889795,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.981103071069439e-05,
+      "loss": 1.5004,
+      "step": 445
+    },
+    {
+      "epoch": 0.030574806359559722,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.980890746025275e-05,
+      "loss": 1.4275,
+      "step": 450
+    },
+    {
+      "epoch": 0.0309145264302215,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.980678420981112e-05,
+      "loss": 1.5783,
+      "step": 455
+    },
+    {
+      "epoch": 0.03125424650088327,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.9804660959369484e-05,
+      "loss": 1.4955,
+      "step": 460
+    },
+    {
+      "epoch": 0.03159396657154505,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.980253770892784e-05,
+      "loss": 1.5208,
+      "step": 465
+    },
+    {
+      "epoch": 0.03193368664220682,
+      "grad_norm": 6.0,
+      "learning_rate": 4.980041445848621e-05,
+      "loss": 1.5058,
+      "step": 470
+    },
+    {
+      "epoch": 0.032273406712868595,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9798291208044576e-05,
+      "loss": 1.4908,
+      "step": 475
+    },
+    {
+      "epoch": 0.03261312678353037,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.979616795760293e-05,
+      "loss": 1.4957,
+      "step": 480
+    },
+    {
+      "epoch": 0.03295284685419215,
+      "grad_norm": 6.125,
+      "learning_rate": 4.9794044707161304e-05,
+      "loss": 1.5149,
+      "step": 485
+    },
+    {
+      "epoch": 0.03329256692485392,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.979192145671966e-05,
+      "loss": 1.4924,
+      "step": 490
+    },
+    {
+      "epoch": 0.033632286995515695,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9789798206278025e-05,
+      "loss": 1.4713,
+      "step": 495
+    },
+    {
+      "epoch": 0.03397200706617747,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.9787674955836396e-05,
+      "loss": 1.5022,
+      "step": 500
+    },
+    {
+      "epoch": 0.03431172713683924,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.978555170539475e-05,
+      "loss": 1.494,
+      "step": 505
+    },
+    {
+      "epoch": 0.03465144720750102,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.978342845495312e-05,
+      "loss": 1.4304,
+      "step": 510
+    },
+    {
+      "epoch": 0.034991167278162795,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.978130520451149e-05,
+      "loss": 1.4458,
+      "step": 515
+    },
+    {
+      "epoch": 0.03533088734882457,
+      "grad_norm": 6.625,
+      "learning_rate": 4.9779181954069845e-05,
+      "loss": 1.5779,
+      "step": 520
+    },
+    {
+      "epoch": 0.03567060741948634,
+      "grad_norm": 7.875,
+      "learning_rate": 4.9777058703628216e-05,
+      "loss": 1.594,
+      "step": 525
+    },
+    {
+      "epoch": 0.03601032749014812,
+      "grad_norm": 7.375,
+      "learning_rate": 4.977493545318658e-05,
+      "loss": 1.4908,
+      "step": 530
+    },
+    {
+      "epoch": 0.036350047560809895,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.977281220274494e-05,
+      "loss": 1.5638,
+      "step": 535
+    },
+    {
+      "epoch": 0.036689767631471665,
+      "grad_norm": 6.5,
+      "learning_rate": 4.977068895230331e-05,
+      "loss": 1.6649,
+      "step": 540
+    },
+    {
+      "epoch": 0.03702948770213344,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.976856570186167e-05,
+      "loss": 1.4847,
+      "step": 545
+    },
+    {
+      "epoch": 0.03736920777279522,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.976644245142003e-05,
+      "loss": 1.5145,
+      "step": 550
+    },
+    {
+      "epoch": 0.03770892784345699,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.97643192009784e-05,
+      "loss": 1.5118,
+      "step": 555
+    },
+    {
+      "epoch": 0.038048647914118765,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.9762195950536764e-05,
+      "loss": 1.5993,
+      "step": 560
+    },
+    {
+      "epoch": 0.03838836798478054,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.976007270009512e-05,
+      "loss": 1.5077,
+      "step": 565
+    },
+    {
+      "epoch": 0.03872808805544232,
+      "grad_norm": 6.875,
+      "learning_rate": 4.975794944965349e-05,
+      "loss": 1.5224,
+      "step": 570
+    },
+    {
+      "epoch": 0.03906780812610409,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.975582619921185e-05,
+      "loss": 1.4884,
+      "step": 575
+    },
+    {
+      "epoch": 0.039407528196765865,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.975370294877021e-05,
+      "loss": 1.5013,
+      "step": 580
+    },
+    {
+      "epoch": 0.03974724826742764,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.9751579698328584e-05,
+      "loss": 1.4974,
+      "step": 585
+    },
+    {
+      "epoch": 0.04008696833808941,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.974945644788694e-05,
+      "loss": 1.4737,
+      "step": 590
+    },
+    {
+      "epoch": 0.04042668840875119,
+      "grad_norm": 6.25,
+      "learning_rate": 4.9747333197445305e-05,
+      "loss": 1.4579,
+      "step": 595
+    },
+    {
+      "epoch": 0.040766408479412965,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.9745209947003676e-05,
+      "loss": 1.3987,
+      "step": 600
+    },
+    {
+      "epoch": 0.04110612855007474,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.974308669656203e-05,
+      "loss": 1.4832,
+      "step": 605
+    },
+    {
+      "epoch": 0.04144584862073651,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.97409634461204e-05,
+      "loss": 1.5074,
+      "step": 610
+    },
+    {
+      "epoch": 0.04178556869139829,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.973884019567877e-05,
+      "loss": 1.5505,
+      "step": 615
+    },
+    {
+      "epoch": 0.042125288762060065,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.9736716945237125e-05,
+      "loss": 1.5158,
+      "step": 620
+    },
+    {
+      "epoch": 0.042465008832721834,
+      "grad_norm": 5.625,
+      "learning_rate": 4.973459369479549e-05,
+      "loss": 1.4338,
+      "step": 625
+    },
+    {
+      "epoch": 0.04280472890338361,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.973247044435386e-05,
+      "loss": 1.3955,
+      "step": 630
+    },
+    {
+      "epoch": 0.04314444897404539,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.973034719391222e-05,
+      "loss": 1.5275,
+      "step": 635
+    },
+    {
+      "epoch": 0.043484169044707165,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.972822394347058e-05,
+      "loss": 1.4382,
+      "step": 640
+    },
+    {
+      "epoch": 0.043823889115368934,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.9726100693028945e-05,
+      "loss": 1.5352,
+      "step": 645
+    },
+    {
+      "epoch": 0.04416360918603071,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.972397744258731e-05,
+      "loss": 1.4863,
+      "step": 650
+    },
+    {
+      "epoch": 0.04450332925669249,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.972185419214567e-05,
+      "loss": 1.5064,
+      "step": 655
+    },
+    {
+      "epoch": 0.04484304932735426,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.971973094170404e-05,
+      "loss": 1.4521,
+      "step": 660
+    },
+    {
+      "epoch": 0.045182769398016034,
+      "grad_norm": 6.0,
+      "learning_rate": 4.97176076912624e-05,
+      "loss": 1.4796,
+      "step": 665
+    },
+    {
+      "epoch": 0.04552248946867781,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9715484440820765e-05,
+      "loss": 1.5539,
+      "step": 670
+    },
+    {
+      "epoch": 0.04586220953933958,
+      "grad_norm": 6.875,
+      "learning_rate": 4.971336119037913e-05,
+      "loss": 1.4506,
+      "step": 675
+    },
+    {
+      "epoch": 0.04620192961000136,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.971123793993749e-05,
+      "loss": 1.4188,
+      "step": 680
+    },
+    {
+      "epoch": 0.046541649680663134,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.970911468949586e-05,
+      "loss": 1.4797,
+      "step": 685
+    },
+    {
+      "epoch": 0.04688136975132491,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.970699143905422e-05,
+      "loss": 1.5831,
+      "step": 690
+    },
+    {
+      "epoch": 0.04722108982198668,
+      "grad_norm": 6.625,
+      "learning_rate": 4.9704868188612585e-05,
+      "loss": 1.5128,
+      "step": 695
+    },
+    {
+      "epoch": 0.04756080989264846,
+      "grad_norm": 6.875,
+      "learning_rate": 4.970274493817095e-05,
+      "loss": 1.5558,
+      "step": 700
+    },
+    {
+      "epoch": 0.047900529963310234,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.970062168772931e-05,
+      "loss": 1.5671,
+      "step": 705
+    },
+    {
+      "epoch": 0.048240250033972004,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.969849843728768e-05,
+      "loss": 1.5466,
+      "step": 710
+    },
+    {
+      "epoch": 0.04857997010463378,
+      "grad_norm": 6.625,
+      "learning_rate": 4.969637518684604e-05,
+      "loss": 1.5864,
+      "step": 715
+    },
+    {
+      "epoch": 0.04891969017529556,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.9694251936404405e-05,
+      "loss": 1.5189,
+      "step": 720
+    },
+    {
+      "epoch": 0.049259410245957334,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.969212868596277e-05,
+      "loss": 1.3615,
+      "step": 725
+    },
+    {
+      "epoch": 0.049599130316619104,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.9690005435521133e-05,
+      "loss": 1.5156,
+      "step": 730
+    },
+    {
+      "epoch": 0.04993885038728088,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.96878821850795e-05,
+      "loss": 1.4286,
+      "step": 735
+    },
+    {
+      "epoch": 0.05027857045794266,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.968575893463786e-05,
+      "loss": 1.4971,
+      "step": 740
+    },
+    {
+      "epoch": 0.05061829052860443,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9683635684196225e-05,
+      "loss": 1.4613,
+      "step": 745
+    },
+    {
+      "epoch": 0.050958010599266204,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.968151243375459e-05,
+      "loss": 1.5054,
+      "step": 750
+    },
+    {
+      "epoch": 0.05129773066992798,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.9679389183312953e-05,
+      "loss": 1.5876,
+      "step": 755
+    },
+    {
+      "epoch": 0.05163745074058975,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.967726593287132e-05,
+      "loss": 1.51,
+      "step": 760
+    },
+    {
+      "epoch": 0.05197717081125153,
+      "grad_norm": 6.25,
+      "learning_rate": 4.967514268242968e-05,
+      "loss": 1.5307,
+      "step": 765
+    },
+    {
+      "epoch": 0.052316890881913304,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.9673019431988045e-05,
+      "loss": 1.4992,
+      "step": 770
+    },
+    {
+      "epoch": 0.05265661095257508,
+      "grad_norm": 6.0,
+      "learning_rate": 4.96708961815464e-05,
+      "loss": 1.6272,
+      "step": 775
+    },
+    {
+      "epoch": 0.05299633102323685,
+      "grad_norm": 5.75,
+      "learning_rate": 4.9668772931104773e-05,
+      "loss": 1.5342,
+      "step": 780
+    },
+    {
+      "epoch": 0.05333605109389863,
+      "grad_norm": 6.0,
+      "learning_rate": 4.966664968066314e-05,
+      "loss": 1.3954,
+      "step": 785
+    },
+    {
+      "epoch": 0.053675771164560404,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.9664526430221495e-05,
+      "loss": 1.5686,
+      "step": 790
+    },
+    {
+      "epoch": 0.054015491235222174,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9662403179779865e-05,
+      "loss": 1.4803,
+      "step": 795
+    },
+    {
+      "epoch": 0.05435521130588395,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.966027992933823e-05,
+      "loss": 1.5869,
+      "step": 800
+    },
+    {
+      "epoch": 0.05469493137654573,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.965815667889659e-05,
+      "loss": 1.4846,
+      "step": 805
+    },
+    {
+      "epoch": 0.055034651447207504,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.965603342845496e-05,
+      "loss": 1.5697,
+      "step": 810
+    },
+    {
+      "epoch": 0.055374371517869274,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.965391017801332e-05,
+      "loss": 1.3527,
+      "step": 815
+    },
+    {
+      "epoch": 0.05571409158853105,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.965178692757168e-05,
+      "loss": 1.4965,
+      "step": 820
+    },
+    {
+      "epoch": 0.05605381165919283,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.964966367713005e-05,
+      "loss": 1.498,
+      "step": 825
+    },
+    {
+      "epoch": 0.0563935317298546,
+      "grad_norm": 6.25,
+      "learning_rate": 4.9647540426688413e-05,
+      "loss": 1.5917,
+      "step": 830
+    },
+    {
+      "epoch": 0.056733251800516374,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.964541717624677e-05,
+      "loss": 1.4338,
+      "step": 835
+    },
+    {
+      "epoch": 0.05707297187117815,
+      "grad_norm": 6.375,
+      "learning_rate": 4.964329392580514e-05,
+      "loss": 1.4895,
+      "step": 840
+    },
+    {
+      "epoch": 0.05741269194183993,
+      "grad_norm": 6.375,
+      "learning_rate": 4.96411706753635e-05,
+      "loss": 1.5225,
+      "step": 845
+    },
+    {
+      "epoch": 0.0577524120125017,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.963904742492186e-05,
+      "loss": 1.5133,
+      "step": 850
+    },
+    {
+      "epoch": 0.058092132083163474,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9636924174480234e-05,
+      "loss": 1.42,
+      "step": 855
+    },
+    {
+      "epoch": 0.05843185215382525,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.963480092403859e-05,
+      "loss": 1.5467,
+      "step": 860
+    },
+    {
+      "epoch": 0.05877157222448702,
+      "grad_norm": 6.25,
+      "learning_rate": 4.963267767359696e-05,
+      "loss": 1.5391,
+      "step": 865
+    },
+    {
+      "epoch": 0.0591112922951488,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.9630554423155326e-05,
+      "loss": 1.4394,
+      "step": 870
+    },
+    {
+      "epoch": 0.059451012365810574,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.962843117271368e-05,
+      "loss": 1.5973,
+      "step": 875
+    },
+    {
+      "epoch": 0.059790732436472344,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9626307922272054e-05,
+      "loss": 1.4694,
+      "step": 880
+    },
+    {
+      "epoch": 0.06013045250713412,
+      "grad_norm": 6.625,
+      "learning_rate": 4.962418467183042e-05,
+      "loss": 1.5052,
+      "step": 885
+    },
+    {
+      "epoch": 0.0604701725777959,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9622061421388775e-05,
+      "loss": 1.5257,
+      "step": 890
+    },
+    {
+      "epoch": 0.060809892648457674,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.9619938170947146e-05,
+      "loss": 1.449,
+      "step": 895
+    },
+    {
+      "epoch": 0.061149612719119444,
+      "grad_norm": 6.0,
+      "learning_rate": 4.961781492050551e-05,
+      "loss": 1.5326,
+      "step": 900
+    },
+    {
+      "epoch": 0.06148933278978122,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.961569167006387e-05,
+      "loss": 1.5239,
+      "step": 905
+    },
+    {
+      "epoch": 0.061829052860443,
+      "grad_norm": 6.25,
+      "learning_rate": 4.961356841962224e-05,
+      "loss": 1.5428,
+      "step": 910
+    },
+    {
+      "epoch": 0.06216877293110477,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9611445169180595e-05,
+      "loss": 1.4652,
+      "step": 915
+    },
+    {
+      "epoch": 0.06250849300176654,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.960932191873896e-05,
+      "loss": 1.475,
+      "step": 920
+    },
+    {
+      "epoch": 0.06284821307242831,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.960719866829733e-05,
+      "loss": 1.5127,
+      "step": 925
+    },
+    {
+      "epoch": 0.0631879331430901,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.960507541785569e-05,
+      "loss": 1.4485,
+      "step": 930
+    },
+    {
+      "epoch": 0.06352765321375187,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.960295216741405e-05,
+      "loss": 1.4812,
+      "step": 935
+    },
+    {
+      "epoch": 0.06386737328441364,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.960082891697242e-05,
+      "loss": 1.4899,
+      "step": 940
+    },
+    {
+      "epoch": 0.06420709335507542,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.959870566653078e-05,
+      "loss": 1.4876,
+      "step": 945
+    },
+    {
+      "epoch": 0.06454681342573719,
+      "grad_norm": 5.75,
+      "learning_rate": 4.959658241608914e-05,
+      "loss": 1.4932,
+      "step": 950
+    },
+    {
+      "epoch": 0.06488653349639897,
+      "grad_norm": 6.75,
+      "learning_rate": 4.9594459165647514e-05,
+      "loss": 1.5012,
+      "step": 955
+    },
+    {
+      "epoch": 0.06522625356706074,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.959233591520587e-05,
+      "loss": 1.4216,
+      "step": 960
+    },
+    {
+      "epoch": 0.06556597363772251,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9590212664764235e-05,
+      "loss": 1.5415,
+      "step": 965
+    },
+    {
+      "epoch": 0.0659056937083843,
+      "grad_norm": 6.25,
+      "learning_rate": 4.9588089414322606e-05,
+      "loss": 1.5126,
+      "step": 970
+    },
+    {
+      "epoch": 0.06624541377904607,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.958596616388096e-05,
+      "loss": 1.5099,
+      "step": 975
+    },
+    {
+      "epoch": 0.06658513384970784,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.958384291343933e-05,
+      "loss": 1.5168,
+      "step": 980
+    },
+    {
+      "epoch": 0.06692485392036962,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.95817196629977e-05,
+      "loss": 1.3469,
+      "step": 985
+    },
+    {
+      "epoch": 0.06726457399103139,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.9579596412556055e-05,
+      "loss": 1.5757,
+      "step": 990
+    },
+    {
+      "epoch": 0.06760429406169316,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.957747316211442e-05,
+      "loss": 1.4989,
+      "step": 995
+    },
+    {
+      "epoch": 0.06794401413235494,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.957534991167278e-05,
+      "loss": 1.3401,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06828373420301671,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.957322666123115e-05,
+      "loss": 1.457,
+      "step": 1005
+    },
+    {
+      "epoch": 0.06862345427367848,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.957110341078951e-05,
+      "loss": 1.4846,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06896317434434027,
+      "grad_norm": 6.375,
+      "learning_rate": 4.9568980160347875e-05,
+      "loss": 1.5839,
+      "step": 1015
+    },
+    {
+      "epoch": 0.06930289441500204,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.956685690990624e-05,
+      "loss": 1.4352,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0696426144856638,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.95647336594646e-05,
+      "loss": 1.4391,
+      "step": 1025
+    },
+    {
+      "epoch": 0.06998233455632559,
+      "grad_norm": 6.125,
+      "learning_rate": 4.956261040902297e-05,
+      "loss": 1.4521,
+      "step": 1030
+    },
+    {
+      "epoch": 0.07032205462698736,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.956048715858133e-05,
+      "loss": 1.4946,
+      "step": 1035
+    },
+    {
+      "epoch": 0.07066177469764914,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.9558363908139695e-05,
+      "loss": 1.5151,
+      "step": 1040
+    },
+    {
+      "epoch": 0.07100149476831091,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.955624065769806e-05,
+      "loss": 1.5,
+      "step": 1045
+    },
+    {
+      "epoch": 0.07134121483897268,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.955411740725642e-05,
+      "loss": 1.4604,
+      "step": 1050
+    },
+    {
+      "epoch": 0.07168093490963447,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.955199415681479e-05,
+      "loss": 1.4397,
+      "step": 1055
+    },
+    {
+      "epoch": 0.07202065498029624,
+      "grad_norm": 6.25,
+      "learning_rate": 4.954987090637315e-05,
+      "loss": 1.3952,
+      "step": 1060
+    },
+    {
+      "epoch": 0.072360375050958,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.9547747655931515e-05,
+      "loss": 1.3983,
+      "step": 1065
+    },
+    {
+      "epoch": 0.07270009512161979,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.954562440548988e-05,
+      "loss": 1.4716,
+      "step": 1070
+    },
+    {
+      "epoch": 0.07303981519228156,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.954350115504824e-05,
+      "loss": 1.3926,
+      "step": 1075
+    },
+    {
+      "epoch": 0.07337953526294333,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.954137790460661e-05,
+      "loss": 1.452,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07371925533360511,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.953925465416497e-05,
+      "loss": 1.4609,
+      "step": 1085
+    },
+    {
+      "epoch": 0.07405897540426688,
+      "grad_norm": 5.625,
+      "learning_rate": 4.9537131403723335e-05,
+      "loss": 1.4669,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07439869547492865,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.95350081532817e-05,
+      "loss": 1.4886,
+      "step": 1095
+    },
+    {
+      "epoch": 0.07473841554559044,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.953288490284006e-05,
+      "loss": 1.4238,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0750781356162522,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.953076165239843e-05,
+      "loss": 1.478,
+      "step": 1105
+    },
+    {
+      "epoch": 0.07541785568691398,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.952863840195679e-05,
+      "loss": 1.4913,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07575757575757576,
+      "grad_norm": 5.875,
+      "learning_rate": 4.952651515151515e-05,
+      "loss": 1.5082,
+      "step": 1115
+    },
+    {
+      "epoch": 0.07609729582823753,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.952439190107352e-05,
+      "loss": 1.4377,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07643701589889931,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.952226865063188e-05,
+      "loss": 1.4659,
+      "step": 1125
+    },
+    {
+      "epoch": 0.07677673596956108,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.952014540019024e-05,
+      "loss": 1.3741,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07711645604022285,
+      "grad_norm": 9.4375,
+      "learning_rate": 4.951802214974861e-05,
+      "loss": 1.5028,
+      "step": 1135
+    },
+    {
+      "epoch": 0.07745617611088464,
+      "grad_norm": 6.25,
+      "learning_rate": 4.9515898899306975e-05,
+      "loss": 1.3455,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0777958961815464,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.951377564886533e-05,
+      "loss": 1.5219,
+      "step": 1145
+    },
+    {
+      "epoch": 0.07813561625220818,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.95116523984237e-05,
+      "loss": 1.4049,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07847533632286996,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.950952914798207e-05,
+      "loss": 1.4605,
+      "step": 1155
+    },
+    {
+      "epoch": 0.07881505639353173,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.9507405897540424e-05,
+      "loss": 1.4104,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0791547764641935,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.9505282647098795e-05,
+      "loss": 1.4648,
+      "step": 1165
+    },
+    {
+      "epoch": 0.07949449653485528,
+      "grad_norm": 6.125,
+      "learning_rate": 4.950315939665716e-05,
+      "loss": 1.4747,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07983421660551705,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9501036146215516e-05,
+      "loss": 1.5105,
+      "step": 1175
+    },
+    {
+      "epoch": 0.08017393667617882,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.949891289577389e-05,
+      "loss": 1.5128,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0805136567468406,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.949678964533225e-05,
+      "loss": 1.5111,
+      "step": 1185
+    },
+    {
+      "epoch": 0.08085337681750238,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.949466639489061e-05,
+      "loss": 1.4738,
+      "step": 1190
+    },
+    {
+      "epoch": 0.08119309688816416,
+      "grad_norm": 6.0,
+      "learning_rate": 4.949254314444898e-05,
+      "loss": 1.3451,
+      "step": 1195
+    },
+    {
+      "epoch": 0.08153281695882593,
+      "grad_norm": 6.0,
+      "learning_rate": 4.9490419894007336e-05,
+      "loss": 1.4248,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0818725370294877,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.948829664356571e-05,
+      "loss": 1.5538,
+      "step": 1205
+    },
+    {
+      "epoch": 0.08221225710014948,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.948617339312407e-05,
+      "loss": 1.4164,
+      "step": 1210
+    },
+    {
+      "epoch": 0.08255197717081125,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.948405014268243e-05,
+      "loss": 1.4695,
+      "step": 1215
+    },
+    {
+      "epoch": 0.08289169724147302,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.94819268922408e-05,
+      "loss": 1.3407,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0832314173121348,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.947980364179916e-05,
+      "loss": 1.4778,
+      "step": 1225
+    },
+    {
+      "epoch": 0.08357113738279658,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.947768039135752e-05,
+      "loss": 1.4411,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08391085745345835,
+      "grad_norm": 5.875,
+      "learning_rate": 4.947555714091589e-05,
+      "loss": 1.5428,
+      "step": 1235
+    },
+    {
+      "epoch": 0.08425057752412013,
+      "grad_norm": 6.125,
+      "learning_rate": 4.9473433890474255e-05,
+      "loss": 1.3777,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0845902975947819,
+      "grad_norm": 5.75,
+      "learning_rate": 4.947131064003261e-05,
+      "loss": 1.3424,
+      "step": 1245
+    },
+    {
+      "epoch": 0.08493001766544367,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.946918738959098e-05,
+      "loss": 1.3985,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08526973773610545,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.946706413914935e-05,
+      "loss": 1.4626,
+      "step": 1255
+    },
+    {
+      "epoch": 0.08560945780676722,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9464940888707704e-05,
+      "loss": 1.4337,
+      "step": 1260
+    },
+    {
+      "epoch": 0.08594917787742899,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9462817638266075e-05,
+      "loss": 1.5086,
+      "step": 1265
+    },
+    {
+      "epoch": 0.08628889794809078,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.946069438782443e-05,
+      "loss": 1.4588,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08662861801875255,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9458571137382796e-05,
+      "loss": 1.4231,
+      "step": 1275
+    },
+    {
+      "epoch": 0.08696833808941433,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.945644788694117e-05,
+      "loss": 1.5086,
+      "step": 1280
+    },
+    {
+      "epoch": 0.0873080581600761,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9454324636499524e-05,
+      "loss": 1.5472,
+      "step": 1285
+    },
+    {
+      "epoch": 0.08764777823073787,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.945220138605789e-05,
+      "loss": 1.4754,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08798749830139965,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.945007813561626e-05,
+      "loss": 1.4756,
+      "step": 1295
+    },
+    {
+      "epoch": 0.08832721837206142,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.9447954885174616e-05,
+      "loss": 1.4227,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08866693844272319,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.944583163473298e-05,
+      "loss": 1.4386,
+      "step": 1305
+    },
+    {
+      "epoch": 0.08900665851338498,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.944370838429135e-05,
+      "loss": 1.5088,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08934637858404675,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.944158513384971e-05,
+      "loss": 1.3731,
+      "step": 1315
+    },
+    {
+      "epoch": 0.08968609865470852,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.943946188340807e-05,
+      "loss": 1.3379,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0900258187253703,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.943733863296644e-05,
+      "loss": 1.4324,
+      "step": 1325
+    },
+    {
+      "epoch": 0.09036553879603207,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.94352153825248e-05,
+      "loss": 1.4296,
+      "step": 1330
+    },
+    {
+      "epoch": 0.09070525886669384,
+      "grad_norm": 6.375,
+      "learning_rate": 4.9433092132083164e-05,
+      "loss": 1.522,
+      "step": 1335
+    },
+    {
+      "epoch": 0.09104497893735562,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9430968881641535e-05,
+      "loss": 1.5473,
+      "step": 1340
+    },
+    {
+      "epoch": 0.09138469900801739,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.942884563119989e-05,
+      "loss": 1.4028,
+      "step": 1345
+    },
+    {
+      "epoch": 0.09172441907867916,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.9426722380758256e-05,
+      "loss": 1.4582,
+      "step": 1350
+    },
+    {
+      "epoch": 0.09206413914934095,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.942459913031662e-05,
+      "loss": 1.4234,
+      "step": 1355
+    },
+    {
+      "epoch": 0.09240385922000272,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.9422475879874984e-05,
+      "loss": 1.4109,
+      "step": 1360
+    },
+    {
+      "epoch": 0.0927435792906645,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.942035262943335e-05,
+      "loss": 1.406,
+      "step": 1365
+    },
+    {
+      "epoch": 0.09308329936132627,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.941822937899171e-05,
+      "loss": 1.4408,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09342301943198804,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.9416106128550076e-05,
+      "loss": 1.3366,
+      "step": 1375
+    },
+    {
+      "epoch": 0.09376273950264982,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.941398287810844e-05,
+      "loss": 1.5042,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09410245957331159,
+      "grad_norm": 5.375,
+      "learning_rate": 4.9411859627666804e-05,
+      "loss": 1.3929,
+      "step": 1385
+    },
+    {
+      "epoch": 0.09444217964397336,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.940973637722517e-05,
+      "loss": 1.5103,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09478189971463515,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.940761312678353e-05,
+      "loss": 1.4239,
+      "step": 1395
+    },
+    {
+      "epoch": 0.09512161978529692,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.9405489876341896e-05,
+      "loss": 1.5538,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09546133985595869,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.940336662590026e-05,
+      "loss": 1.5307,
+      "step": 1405
+    },
+    {
+      "epoch": 0.09580105992662047,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9401243375458624e-05,
+      "loss": 1.4439,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09614077999728224,
+      "grad_norm": 6.0,
+      "learning_rate": 4.939912012501699e-05,
+      "loss": 1.499,
+      "step": 1415
+    },
+    {
+      "epoch": 0.09648050006794401,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.939699687457535e-05,
+      "loss": 1.5387,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09682022013860579,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9394873624133716e-05,
+      "loss": 1.533,
+      "step": 1425
+    },
+    {
+      "epoch": 0.09715994020926756,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.939275037369208e-05,
+      "loss": 1.4586,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09749966027992933,
+      "grad_norm": 6.375,
+      "learning_rate": 4.9390627123250444e-05,
+      "loss": 1.4538,
+      "step": 1435
+    },
+    {
+      "epoch": 0.09783938035059112,
+      "grad_norm": 5.5,
+      "learning_rate": 4.938850387280881e-05,
+      "loss": 1.4045,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09817910042125289,
+      "grad_norm": 5.625,
+      "learning_rate": 4.938638062236717e-05,
+      "loss": 1.4065,
+      "step": 1445
+    },
+    {
+      "epoch": 0.09851882049191467,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9384257371925536e-05,
+      "loss": 1.521,
+      "step": 1450
+    },
+    {
+      "epoch": 0.09885854056257644,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.93821341214839e-05,
+      "loss": 1.4516,
+      "step": 1455
+    },
+    {
+      "epoch": 0.09919826063323821,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.9380010871042264e-05,
+      "loss": 1.5305,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09953798070389999,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.937788762060063e-05,
+      "loss": 1.4258,
+      "step": 1465
+    },
+    {
+      "epoch": 0.09987770077456176,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9375764370158986e-05,
+      "loss": 1.4221,
+      "step": 1470
+    },
+    {
+      "epoch": 0.10021742084522353,
+      "grad_norm": 6.0,
+      "learning_rate": 4.9373641119717356e-05,
+      "loss": 1.4577,
+      "step": 1475
+    },
+    {
+      "epoch": 0.10055714091588532,
+      "grad_norm": 7.0,
+      "learning_rate": 4.937151786927572e-05,
+      "loss": 1.4198,
+      "step": 1480
+    },
+    {
+      "epoch": 0.10089686098654709,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.936939461883408e-05,
+      "loss": 1.4547,
+      "step": 1485
+    },
+    {
+      "epoch": 0.10123658105720885,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.936727136839245e-05,
+      "loss": 1.4111,
+      "step": 1490
+    },
+    {
+      "epoch": 0.10157630112787064,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.936514811795081e-05,
+      "loss": 1.4435,
+      "step": 1495
+    },
+    {
+      "epoch": 0.10191602119853241,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.936302486750917e-05,
+      "loss": 1.3358,
+      "step": 1500
+    },
+    {
+      "epoch": 0.10225574126919418,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.936090161706754e-05,
+      "loss": 1.398,
+      "step": 1505
+    },
+    {
+      "epoch": 0.10259546133985596,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.9358778366625904e-05,
+      "loss": 1.3668,
+      "step": 1510
+    },
+    {
+      "epoch": 0.10293518141051773,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.935665511618426e-05,
+      "loss": 1.3964,
+      "step": 1515
+    },
+    {
+      "epoch": 0.1032749014811795,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.935453186574263e-05,
+      "loss": 1.4515,
+      "step": 1520
+    },
+    {
+      "epoch": 0.10361462155184128,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.9352408615300996e-05,
+      "loss": 1.4612,
+      "step": 1525
+    },
+    {
+      "epoch": 0.10395434162250305,
+      "grad_norm": 5.375,
+      "learning_rate": 4.9350285364859354e-05,
+      "loss": 1.4231,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10429406169316484,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.9348162114417724e-05,
+      "loss": 1.3935,
+      "step": 1535
+    },
+    {
+      "epoch": 0.10463378176382661,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.934603886397609e-05,
+      "loss": 1.4261,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10497350183448838,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.934391561353445e-05,
+      "loss": 1.3994,
+      "step": 1545
+    },
+    {
+      "epoch": 0.10531322190515016,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.9341792363092816e-05,
+      "loss": 1.4366,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10565294197581193,
+      "grad_norm": 6.375,
+      "learning_rate": 4.9339669112651174e-05,
+      "loss": 1.3738,
+      "step": 1555
+    },
+    {
+      "epoch": 0.1059926620464737,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.9337545862209544e-05,
+      "loss": 1.325,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10633238211713548,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.933542261176791e-05,
+      "loss": 1.3739,
+      "step": 1565
+    },
+    {
+      "epoch": 0.10667210218779725,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.9333299361326266e-05,
+      "loss": 1.4269,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10701182225845902,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.9331176110884637e-05,
+      "loss": 1.4907,
+      "step": 1575
+    },
+    {
+      "epoch": 0.10735154232912081,
+      "grad_norm": 5.5,
+      "learning_rate": 4.9329052860443e-05,
+      "loss": 1.4403,
+      "step": 1580
+    },
+    {
+      "epoch": 0.10769126239978258,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.932692961000136e-05,
+      "loss": 1.3537,
+      "step": 1585
+    },
+    {
+      "epoch": 0.10803098247044435,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.932480635955973e-05,
+      "loss": 1.4384,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10837070254110613,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.932268310911809e-05,
+      "loss": 1.4647,
+      "step": 1595
+    },
+    {
+      "epoch": 0.1087104226117679,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.932055985867645e-05,
+      "loss": 1.4761,
+      "step": 1600
+    },
+    {
+      "epoch": 0.10905014268242967,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.931843660823482e-05,
+      "loss": 1.4881,
+      "step": 1605
+    },
+    {
+      "epoch": 0.10938986275309145,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.9316313357793185e-05,
+      "loss": 1.4669,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10972958282375322,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.931419010735154e-05,
+      "loss": 1.3818,
+      "step": 1615
+    },
+    {
+      "epoch": 0.11006930289441501,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.931206685690991e-05,
+      "loss": 1.4509,
+      "step": 1620
+    },
+    {
+      "epoch": 0.11040902296507678,
+      "grad_norm": 5.75,
+      "learning_rate": 4.930994360646827e-05,
+      "loss": 1.3628,
+      "step": 1625
+    },
+    {
+      "epoch": 0.11074874303573855,
+      "grad_norm": 5.875,
+      "learning_rate": 4.9307820356026634e-05,
+      "loss": 1.4654,
+      "step": 1630
+    },
+    {
+      "epoch": 0.11108846310640033,
+      "grad_norm": 5.875,
+      "learning_rate": 4.9305697105585005e-05,
+      "loss": 1.4621,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1114281831770621,
+      "grad_norm": 5.625,
+      "learning_rate": 4.930357385514336e-05,
+      "loss": 1.4481,
+      "step": 1640
+    },
+    {
+      "epoch": 0.11176790324772387,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.9301450604701726e-05,
+      "loss": 1.4581,
+      "step": 1645
+    },
+    {
+      "epoch": 0.11210762331838565,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.9299327354260097e-05,
+      "loss": 1.433,
+      "step": 1650
+    },
+    {
+      "epoch": 0.11244734338904742,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.9297204103818454e-05,
+      "loss": 1.5547,
+      "step": 1655
+    },
+    {
+      "epoch": 0.1127870634597092,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.929508085337682e-05,
+      "loss": 1.4062,
+      "step": 1660
+    },
+    {
+      "epoch": 0.11312678353037098,
+      "grad_norm": 6.0,
+      "learning_rate": 4.929295760293519e-05,
+      "loss": 1.3747,
+      "step": 1665
+    },
+    {
+      "epoch": 0.11346650360103275,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.9290834352493546e-05,
+      "loss": 1.4854,
+      "step": 1670
+    },
+    {
+      "epoch": 0.11380622367169452,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.928871110205191e-05,
+      "loss": 1.3355,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1141459437423563,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.928658785161028e-05,
+      "loss": 1.4826,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11448566381301807,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.928446460116864e-05,
+      "loss": 1.3356,
+      "step": 1685
+    },
+    {
+      "epoch": 0.11482538388367985,
+      "grad_norm": 5.875,
+      "learning_rate": 4.9282341350727e-05,
+      "loss": 1.3329,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11516510395434162,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.9280218100285366e-05,
+      "loss": 1.4435,
+      "step": 1695
+    },
+    {
+      "epoch": 0.1155048240250034,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.927809484984373e-05,
+      "loss": 1.4012,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11584454409566518,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.9275971599402094e-05,
+      "loss": 1.4594,
+      "step": 1705
+    },
+    {
+      "epoch": 0.11618426416632695,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.927384834896046e-05,
+      "loss": 1.4137,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11652398423698872,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.927172509851882e-05,
+      "loss": 1.3094,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1168637043076505,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9269601848077186e-05,
+      "loss": 1.4247,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11720342437831227,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.926747859763555e-05,
+      "loss": 1.3586,
+      "step": 1725
+    },
+    {
+      "epoch": 0.11754314444897404,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9265355347193914e-05,
+      "loss": 1.4542,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11788286451963582,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.926323209675228e-05,
+      "loss": 1.4004,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1182225845902976,
+      "grad_norm": 6.25,
+      "learning_rate": 4.926110884631064e-05,
+      "loss": 1.4749,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11856230466095936,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.9258985595869006e-05,
+      "loss": 1.3441,
+      "step": 1745
+    },
+    {
+      "epoch": 0.11890202473162115,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.925686234542737e-05,
+      "loss": 1.3893,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11924174480228292,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.9254739094985734e-05,
+      "loss": 1.4374,
+      "step": 1755
+    },
+    {
+      "epoch": 0.11958146487294469,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.92526158445441e-05,
+      "loss": 1.5779,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11992118494360647,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.925049259410246e-05,
+      "loss": 1.409,
+      "step": 1765
+    },
+    {
+      "epoch": 0.12026090501426824,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.9248369343660826e-05,
+      "loss": 1.3534,
+      "step": 1770
+    },
+    {
+      "epoch": 0.12060062508493002,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.924624609321919e-05,
+      "loss": 1.422,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1209403451555918,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.9244122842777554e-05,
+      "loss": 1.4274,
+      "step": 1780
+    },
+    {
+      "epoch": 0.12128006522625356,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.924199959233592e-05,
+      "loss": 1.432,
+      "step": 1785
+    },
+    {
+      "epoch": 0.12161978529691535,
+      "grad_norm": 5.5,
+      "learning_rate": 4.923987634189428e-05,
+      "loss": 1.3771,
+      "step": 1790
+    },
+    {
+      "epoch": 0.12195950536757712,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.9237753091452646e-05,
+      "loss": 1.3636,
+      "step": 1795
+    },
+    {
+      "epoch": 0.12229922543823889,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.923562984101101e-05,
+      "loss": 1.4086,
+      "step": 1800
+    },
+    {
+      "epoch": 0.12263894550890067,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9233506590569374e-05,
+      "loss": 1.3856,
+      "step": 1805
+    },
+    {
+      "epoch": 0.12297866557956244,
+      "grad_norm": 6.0,
+      "learning_rate": 4.923138334012774e-05,
+      "loss": 1.4567,
+      "step": 1810
+    },
+    {
+      "epoch": 0.12331838565022421,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.92292600896861e-05,
+      "loss": 1.4377,
+      "step": 1815
+    },
+    {
+      "epoch": 0.123658105720886,
+      "grad_norm": 6.5,
+      "learning_rate": 4.9227136839244466e-05,
+      "loss": 1.5207,
+      "step": 1820
+    },
+    {
+      "epoch": 0.12399782579154776,
+      "grad_norm": 5.75,
+      "learning_rate": 4.922501358880282e-05,
+      "loss": 1.4023,
+      "step": 1825
+    },
+    {
+      "epoch": 0.12433754586220953,
+      "grad_norm": 5.25,
+      "learning_rate": 4.9222890338361194e-05,
+      "loss": 1.3761,
+      "step": 1830
+    },
+    {
+      "epoch": 0.12467726593287132,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.922076708791956e-05,
+      "loss": 1.4324,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1250169860035331,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.9218643837477915e-05,
+      "loss": 1.4536,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12535670607419486,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.9216520587036286e-05,
+      "loss": 1.4282,
+      "step": 1845
+    },
+    {
+      "epoch": 0.12569642614485663,
+      "grad_norm": 5.75,
+      "learning_rate": 4.921439733659465e-05,
+      "loss": 1.4352,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12603614621551842,
+      "grad_norm": 5.875,
+      "learning_rate": 4.921227408615301e-05,
+      "loss": 1.4017,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1263758662861802,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.921015083571138e-05,
+      "loss": 1.4429,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12671558635684196,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.920802758526974e-05,
+      "loss": 1.2769,
+      "step": 1865
+    },
+    {
+      "epoch": 0.12705530642750373,
+      "grad_norm": 6.25,
+      "learning_rate": 4.92059043348281e-05,
+      "loss": 1.509,
+      "step": 1870
+    },
+    {
+      "epoch": 0.1273950264981655,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.920378108438647e-05,
+      "loss": 1.4352,
+      "step": 1875
+    },
+    {
+      "epoch": 0.12773474656882727,
+      "grad_norm": 6.125,
+      "learning_rate": 4.9201657833944834e-05,
+      "loss": 1.4919,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12807446663948907,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.91995345835032e-05,
+      "loss": 1.4828,
+      "step": 1885
+    },
+    {
+      "epoch": 0.12841418671015084,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.919741133306156e-05,
+      "loss": 1.4392,
+      "step": 1890
+    },
+    {
+      "epoch": 0.1287539067808126,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.919528808261992e-05,
+      "loss": 1.3723,
+      "step": 1895
+    },
+    {
+      "epoch": 0.12909362685147438,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.919316483217829e-05,
+      "loss": 1.4304,
+      "step": 1900
+    },
+    {
+      "epoch": 0.12943334692213615,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.9191041581736654e-05,
+      "loss": 1.4374,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12977306699279795,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.918891833129501e-05,
+      "loss": 1.3218,
+      "step": 1910
+    },
+    {
+      "epoch": 0.13011278706345972,
+      "grad_norm": 5.375,
+      "learning_rate": 4.918679508085338e-05,
+      "loss": 1.417,
+      "step": 1915
+    },
+    {
+      "epoch": 0.1304525071341215,
+      "grad_norm": 5.75,
+      "learning_rate": 4.9184671830411746e-05,
+      "loss": 1.3625,
+      "step": 1920
+    },
+    {
+      "epoch": 0.13079222720478326,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.91825485799701e-05,
+      "loss": 1.3866,
+      "step": 1925
+    },
+    {
+      "epoch": 0.13113194727544503,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.9180425329528474e-05,
+      "loss": 1.3726,
+      "step": 1930
+    },
+    {
+      "epoch": 0.1314716673461068,
+      "grad_norm": 5.125,
+      "learning_rate": 4.917830207908684e-05,
+      "loss": 1.4216,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1318113874167686,
+      "grad_norm": 5.375,
+      "learning_rate": 4.9176178828645195e-05,
+      "loss": 1.3618,
+      "step": 1940
+    },
+    {
+      "epoch": 0.13215110748743036,
+      "grad_norm": 5.375,
+      "learning_rate": 4.9174055578203566e-05,
+      "loss": 1.3482,
+      "step": 1945
+    },
+    {
+      "epoch": 0.13249082755809213,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.917193232776193e-05,
+      "loss": 1.4721,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1328305476287539,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.916980907732029e-05,
+      "loss": 1.4929,
+      "step": 1955
+    },
+    {
+      "epoch": 0.13317026769941567,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.916768582687866e-05,
+      "loss": 1.4713,
+      "step": 1960
+    },
+    {
+      "epoch": 0.13350998777007744,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.916556257643702e-05,
+      "loss": 1.4098,
+      "step": 1965
+    },
+    {
+      "epoch": 0.13384970784073924,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.916343932599538e-05,
+      "loss": 1.4804,
+      "step": 1970
+    },
+    {
+      "epoch": 0.134189427911401,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.916131607555375e-05,
+      "loss": 1.3562,
+      "step": 1975
+    },
+    {
+      "epoch": 0.13452914798206278,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.915919282511211e-05,
+      "loss": 1.3681,
+      "step": 1980
+    },
+    {
+      "epoch": 0.13486886805272455,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.915706957467047e-05,
+      "loss": 1.4642,
+      "step": 1985
+    },
+    {
+      "epoch": 0.13520858812338632,
+      "grad_norm": 5.375,
+      "learning_rate": 4.915494632422884e-05,
+      "loss": 1.298,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13554830819404812,
+      "grad_norm": 5.75,
+      "learning_rate": 4.91528230737872e-05,
+      "loss": 1.4662,
+      "step": 1995
+    },
+    {
+      "epoch": 0.1358880282647099,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.915069982334556e-05,
+      "loss": 1.3921,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13622774833537166,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.9148576572903934e-05,
+      "loss": 1.3357,
+      "step": 2005
+    },
+    {
+      "epoch": 0.13656746840603343,
+      "grad_norm": 5.75,
+      "learning_rate": 4.914645332246229e-05,
+      "loss": 1.51,
+      "step": 2010
+    },
+    {
+      "epoch": 0.1369071884766952,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.9144330072020655e-05,
+      "loss": 1.5117,
+      "step": 2015
+    },
+    {
+      "epoch": 0.13724690854735697,
+      "grad_norm": 5.75,
+      "learning_rate": 4.9142206821579026e-05,
+      "loss": 1.519,
+      "step": 2020
+    },
+    {
+      "epoch": 0.13758662861801876,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.914008357113738e-05,
+      "loss": 1.4227,
+      "step": 2025
+    },
+    {
+      "epoch": 0.13792634868868053,
+      "grad_norm": 5.75,
+      "learning_rate": 4.913796032069575e-05,
+      "loss": 1.3288,
+      "step": 2030
+    },
+    {
+      "epoch": 0.1382660687593423,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.913583707025412e-05,
+      "loss": 1.404,
+      "step": 2035
+    },
+    {
+      "epoch": 0.13860578883000407,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.9133713819812475e-05,
+      "loss": 1.4619,
+      "step": 2040
+    },
+    {
+      "epoch": 0.13894550890066584,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.913159056937084e-05,
+      "loss": 1.5248,
+      "step": 2045
+    },
+    {
+      "epoch": 0.1392852289713276,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.91294673189292e-05,
+      "loss": 1.4558,
+      "step": 2050
+    },
+    {
+      "epoch": 0.1396249490419894,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.912734406848757e-05,
+      "loss": 1.3518,
+      "step": 2055
+    },
+    {
+      "epoch": 0.13996466911265118,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.912522081804593e-05,
+      "loss": 1.3561,
+      "step": 2060
+    },
+    {
+      "epoch": 0.14030438918331295,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.9123097567604295e-05,
+      "loss": 1.458,
+      "step": 2065
+    },
+    {
+      "epoch": 0.14064410925397472,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.912097431716266e-05,
+      "loss": 1.4471,
+      "step": 2070
+    },
+    {
+      "epoch": 0.1409838293246365,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.911885106672102e-05,
+      "loss": 1.4005,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1413235493952983,
+      "grad_norm": 5.375,
+      "learning_rate": 4.911672781627939e-05,
+      "loss": 1.449,
+      "step": 2080
+    },
+    {
+      "epoch": 0.14166326946596006,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.911460456583775e-05,
+      "loss": 1.4674,
+      "step": 2085
+    },
+    {
+      "epoch": 0.14200298953662183,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.9112481315396115e-05,
+      "loss": 1.3984,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1423427096072836,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.911035806495448e-05,
+      "loss": 1.4049,
+      "step": 2095
+    },
+    {
+      "epoch": 0.14268242967794537,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.910823481451284e-05,
+      "loss": 1.3679,
+      "step": 2100
+    },
+    {
+      "epoch": 0.14302214974860714,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.910611156407121e-05,
+      "loss": 1.4456,
+      "step": 2105
+    },
+    {
+      "epoch": 0.14336186981926893,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.910398831362957e-05,
+      "loss": 1.4652,
+      "step": 2110
+    },
+    {
+      "epoch": 0.1437015898899307,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.9101865063187935e-05,
+      "loss": 1.4086,
+      "step": 2115
+    },
+    {
+      "epoch": 0.14404130996059247,
+      "grad_norm": 5.875,
+      "learning_rate": 4.90997418127463e-05,
+      "loss": 1.4586,
+      "step": 2120
+    },
+    {
+      "epoch": 0.14438103003125424,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.909761856230466e-05,
+      "loss": 1.3595,
+      "step": 2125
+    },
+    {
+      "epoch": 0.144720750101916,
+      "grad_norm": 6.375,
+      "learning_rate": 4.909549531186303e-05,
+      "loss": 1.3764,
+      "step": 2130
+    },
+    {
+      "epoch": 0.14506047017257778,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.909337206142139e-05,
+      "loss": 1.3726,
+      "step": 2135
+    },
+    {
+      "epoch": 0.14540019024323958,
+      "grad_norm": 5.875,
+      "learning_rate": 4.9091248810979755e-05,
+      "loss": 1.4388,
+      "step": 2140
+    },
+    {
+      "epoch": 0.14573991031390135,
+      "grad_norm": 5.5,
+      "learning_rate": 4.908912556053812e-05,
+      "loss": 1.4348,
+      "step": 2145
+    },
+    {
+      "epoch": 0.14607963038456312,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.9087002310096483e-05,
+      "loss": 1.2065,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1464193504552249,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.908487905965485e-05,
+      "loss": 1.553,
+      "step": 2155
+    },
+    {
+      "epoch": 0.14675907052588666,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.908275580921321e-05,
+      "loss": 1.3571,
+      "step": 2160
+    },
+    {
+      "epoch": 0.14709879059654846,
+      "grad_norm": 5.75,
+      "learning_rate": 4.9080632558771575e-05,
+      "loss": 1.3664,
+      "step": 2165
+    },
+    {
+      "epoch": 0.14743851066721023,
+      "grad_norm": 5.25,
+      "learning_rate": 4.907850930832994e-05,
+      "loss": 1.3474,
+      "step": 2170
+    },
+    {
+      "epoch": 0.147778230737872,
+      "grad_norm": 5.75,
+      "learning_rate": 4.9076386057888303e-05,
+      "loss": 1.5126,
+      "step": 2175
+    },
+    {
+      "epoch": 0.14811795080853377,
+      "grad_norm": 5.625,
+      "learning_rate": 4.907426280744666e-05,
+      "loss": 1.3294,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14845767087919554,
+      "grad_norm": 5.125,
+      "learning_rate": 4.907213955700503e-05,
+      "loss": 1.5188,
+      "step": 2185
+    },
+    {
+      "epoch": 0.1487973909498573,
+      "grad_norm": 5.5,
+      "learning_rate": 4.9070016306563395e-05,
+      "loss": 1.4242,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1491371110205191,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.906789305612175e-05,
+      "loss": 1.323,
+      "step": 2195
+    },
+    {
+      "epoch": 0.14947683109118087,
+      "grad_norm": 5.5,
+      "learning_rate": 4.9065769805680123e-05,
+      "loss": 1.4291,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14981655116184264,
+      "grad_norm": 5.375,
+      "learning_rate": 4.906364655523849e-05,
+      "loss": 1.3417,
+      "step": 2205
+    },
+    {
+      "epoch": 0.1501562712325044,
+      "grad_norm": 6.125,
+      "learning_rate": 4.9061523304796845e-05,
+      "loss": 1.4625,
+      "step": 2210
+    },
+    {
+      "epoch": 0.15049599130316618,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.9059400054355215e-05,
+      "loss": 1.4297,
+      "step": 2215
+    },
+    {
+      "epoch": 0.15083571137382795,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.905727680391358e-05,
+      "loss": 1.4224,
+      "step": 2220
+    },
+    {
+      "epoch": 0.15117543144448975,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.9055153553471943e-05,
+      "loss": 1.4499,
+      "step": 2225
+    },
+    {
+      "epoch": 0.15151515151515152,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.905303030303031e-05,
+      "loss": 1.3842,
+      "step": 2230
+    },
+    {
+      "epoch": 0.1518548715858133,
+      "grad_norm": 5.625,
+      "learning_rate": 4.905090705258867e-05,
+      "loss": 1.4205,
+      "step": 2235
+    },
+    {
+      "epoch": 0.15219459165647506,
+      "grad_norm": 5.5,
+      "learning_rate": 4.9048783802147035e-05,
+      "loss": 1.4996,
+      "step": 2240
+    },
+    {
+      "epoch": 0.15253431172713683,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.90466605517054e-05,
+      "loss": 1.4591,
+      "step": 2245
+    },
+    {
+      "epoch": 0.15287403179779863,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.904453730126376e-05,
+      "loss": 1.3338,
+      "step": 2250
+    },
+    {
+      "epoch": 0.1532137518684604,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.904241405082213e-05,
+      "loss": 1.2996,
+      "step": 2255
+    },
+    {
+      "epoch": 0.15355347193912217,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.904029080038049e-05,
+      "loss": 1.4458,
+      "step": 2260
+    },
+    {
+      "epoch": 0.15389319200978394,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.903816754993885e-05,
+      "loss": 1.4621,
+      "step": 2265
+    },
+    {
+      "epoch": 0.1542329120804457,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.903604429949722e-05,
+      "loss": 1.3946,
+      "step": 2270
+    },
+    {
+      "epoch": 0.15457263215110748,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.9033921049055583e-05,
+      "loss": 1.4055,
+      "step": 2275
+    },
+    {
+      "epoch": 0.15491235222176927,
+      "grad_norm": 6.0,
+      "learning_rate": 4.903179779861394e-05,
+      "loss": 1.3789,
+      "step": 2280
+    },
+    {
+      "epoch": 0.15525207229243104,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.902967454817231e-05,
+      "loss": 1.3769,
+      "step": 2285
+    },
+    {
+      "epoch": 0.1555917923630928,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.9027551297730676e-05,
+      "loss": 1.4272,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15593151243375458,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.902542804728903e-05,
+      "loss": 1.3834,
+      "step": 2295
+    },
+    {
+      "epoch": 0.15627123250441635,
+      "grad_norm": 5.875,
+      "learning_rate": 4.9023304796847404e-05,
+      "loss": 1.4676,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15661095257507815,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.902118154640577e-05,
+      "loss": 1.4398,
+      "step": 2305
+    },
+    {
+      "epoch": 0.15695067264573992,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.9019058295964125e-05,
+      "loss": 1.4111,
+      "step": 2310
+    },
+    {
+      "epoch": 0.1572903927164017,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.9016935045522496e-05,
+      "loss": 1.3662,
+      "step": 2315
+    },
+    {
+      "epoch": 0.15763011278706346,
+      "grad_norm": 5.875,
+      "learning_rate": 4.901481179508085e-05,
+      "loss": 1.4072,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15796983285772523,
+      "grad_norm": 5.25,
+      "learning_rate": 4.901268854463922e-05,
+      "loss": 1.4029,
+      "step": 2325
+    },
+    {
+      "epoch": 0.158309552928387,
+      "grad_norm": 6.25,
+      "learning_rate": 4.901056529419759e-05,
+      "loss": 1.4458,
+      "step": 2330
+    },
+    {
+      "epoch": 0.1586492729990488,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.9008442043755945e-05,
+      "loss": 1.4288,
+      "step": 2335
+    },
+    {
+      "epoch": 0.15898899306971057,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.900631879331431e-05,
+      "loss": 1.4598,
+      "step": 2340
+    },
+    {
+      "epoch": 0.15932871314037234,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.900419554287268e-05,
+      "loss": 1.4297,
+      "step": 2345
+    },
+    {
+      "epoch": 0.1596684332110341,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.900207229243104e-05,
+      "loss": 1.3467,
+      "step": 2350
+    },
+    {
+      "epoch": 0.16000815328169588,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.89999490419894e-05,
+      "loss": 1.42,
+      "step": 2355
+    },
+    {
+      "epoch": 0.16034787335235764,
+      "grad_norm": 4.875,
+      "learning_rate": 4.899782579154777e-05,
+      "loss": 1.4065,
+      "step": 2360
+    },
+    {
+      "epoch": 0.16068759342301944,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.899570254110613e-05,
+      "loss": 1.4457,
+      "step": 2365
+    },
+    {
+      "epoch": 0.1610273134936812,
+      "grad_norm": 5.5,
+      "learning_rate": 4.899357929066449e-05,
+      "loss": 1.3662,
+      "step": 2370
+    },
+    {
+      "epoch": 0.16136703356434298,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.8991456040222864e-05,
+      "loss": 1.3622,
+      "step": 2375
+    },
+    {
+      "epoch": 0.16170675363500475,
+      "grad_norm": 4.875,
+      "learning_rate": 4.898933278978122e-05,
+      "loss": 1.3229,
+      "step": 2380
+    },
+    {
+      "epoch": 0.16204647370566652,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8987209539339585e-05,
+      "loss": 1.3789,
+      "step": 2385
+    },
+    {
+      "epoch": 0.16238619377632832,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.8985086288897956e-05,
+      "loss": 1.4885,
+      "step": 2390
+    },
+    {
+      "epoch": 0.1627259138469901,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.898296303845631e-05,
+      "loss": 1.3974,
+      "step": 2395
+    },
+    {
+      "epoch": 0.16306563391765186,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.898083978801468e-05,
+      "loss": 1.3658,
+      "step": 2400
+    },
+    {
+      "epoch": 0.16340535398831363,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.897871653757304e-05,
+      "loss": 1.2744,
+      "step": 2405
+    },
+    {
+      "epoch": 0.1637450740589754,
+      "grad_norm": 5.5,
+      "learning_rate": 4.8976593287131405e-05,
+      "loss": 1.4581,
+      "step": 2410
+    },
+    {
+      "epoch": 0.16408479412963717,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.897447003668977e-05,
+      "loss": 1.4984,
+      "step": 2415
+    },
+    {
+      "epoch": 0.16442451420029897,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.897234678624813e-05,
+      "loss": 1.3995,
+      "step": 2420
+    },
+    {
+      "epoch": 0.16476423427096074,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.89702235358065e-05,
+      "loss": 1.4105,
+      "step": 2425
+    },
+    {
+      "epoch": 0.1651039543416225,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.896810028536486e-05,
+      "loss": 1.3637,
+      "step": 2430
+    },
+    {
+      "epoch": 0.16544367441228428,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.8965977034923225e-05,
+      "loss": 1.6643,
+      "step": 2435
+    },
+    {
+      "epoch": 0.16578339448294604,
+      "grad_norm": 5.75,
+      "learning_rate": 4.896385378448159e-05,
+      "loss": 1.4433,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16612311455360781,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.896173053403995e-05,
+      "loss": 1.5031,
+      "step": 2445
+    },
+    {
+      "epoch": 0.1664628346242696,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.895960728359832e-05,
+      "loss": 1.4013,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16680255469493138,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.895748403315668e-05,
+      "loss": 1.3251,
+      "step": 2455
+    },
+    {
+      "epoch": 0.16714227476559315,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.8955360782715045e-05,
+      "loss": 1.4519,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16748199483625492,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.895323753227341e-05,
+      "loss": 1.3501,
+      "step": 2465
+    },
+    {
+      "epoch": 0.1678217149069167,
+      "grad_norm": 5.0,
+      "learning_rate": 4.895111428183177e-05,
+      "loss": 1.3826,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1681614349775785,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.894899103139014e-05,
+      "loss": 1.3633,
+      "step": 2475
+    },
+    {
+      "epoch": 0.16850115504824026,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.89468677809485e-05,
+      "loss": 1.3845,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16884087511890203,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.8944744530506865e-05,
+      "loss": 1.4917,
+      "step": 2485
+    },
+    {
+      "epoch": 0.1691805951895638,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.894262128006523e-05,
+      "loss": 1.4568,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16952031526022557,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.894049802962359e-05,
+      "loss": 1.4477,
+      "step": 2495
+    },
+    {
+      "epoch": 0.16986003533088734,
+      "grad_norm": 5.625,
+      "learning_rate": 4.893837477918196e-05,
+      "loss": 1.4597,
+      "step": 2500
+    },
+    {
+      "epoch": 0.17019975540154914,
+      "grad_norm": 6.125,
+      "learning_rate": 4.893625152874032e-05,
+      "loss": 1.405,
+      "step": 2505
+    },
+    {
+      "epoch": 0.1705394754722109,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8934128278298685e-05,
+      "loss": 1.3316,
+      "step": 2510
+    },
+    {
+      "epoch": 0.17087919554287267,
+      "grad_norm": 5.5,
+      "learning_rate": 4.893200502785705e-05,
+      "loss": 1.3869,
+      "step": 2515
+    },
+    {
+      "epoch": 0.17121891561353444,
+      "grad_norm": 5.625,
+      "learning_rate": 4.8929881777415406e-05,
+      "loss": 1.4218,
+      "step": 2520
+    },
+    {
+      "epoch": 0.17155863568419621,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.892775852697378e-05,
+      "loss": 1.4238,
+      "step": 2525
+    },
+    {
+      "epoch": 0.17189835575485798,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.892563527653214e-05,
+      "loss": 1.5291,
+      "step": 2530
+    },
+    {
+      "epoch": 0.17223807582551978,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.89235120260905e-05,
+      "loss": 1.4135,
+      "step": 2535
+    },
+    {
+      "epoch": 0.17257779589618155,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.892138877564887e-05,
+      "loss": 1.4728,
+      "step": 2540
+    },
+    {
+      "epoch": 0.17291751596684332,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.891926552520723e-05,
+      "loss": 1.2515,
+      "step": 2545
+    },
+    {
+      "epoch": 0.1732572360375051,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.891714227476559e-05,
+      "loss": 1.3533,
+      "step": 2550
+    },
+    {
+      "epoch": 0.17359695610816686,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.891501902432396e-05,
+      "loss": 1.3775,
+      "step": 2555
+    },
+    {
+      "epoch": 0.17393667617882866,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8912895773882325e-05,
+      "loss": 1.3471,
+      "step": 2560
+    },
+    {
+      "epoch": 0.17427639624949043,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.891077252344069e-05,
+      "loss": 1.2939,
+      "step": 2565
+    },
+    {
+      "epoch": 0.1746161163201522,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.890864927299905e-05,
+      "loss": 1.3962,
+      "step": 2570
+    },
+    {
+      "epoch": 0.17495583639081397,
+      "grad_norm": 5.625,
+      "learning_rate": 4.890652602255742e-05,
+      "loss": 1.359,
+      "step": 2575
+    },
+    {
+      "epoch": 0.17529555646147574,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.890440277211578e-05,
+      "loss": 1.3453,
+      "step": 2580
+    },
+    {
+      "epoch": 0.1756352765321375,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8902279521674145e-05,
+      "loss": 1.2882,
+      "step": 2585
+    },
+    {
+      "epoch": 0.1759749966027993,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.890015627123251e-05,
+      "loss": 1.3783,
+      "step": 2590
+    },
+    {
+      "epoch": 0.17631471667346107,
+      "grad_norm": 5.625,
+      "learning_rate": 4.889803302079087e-05,
+      "loss": 1.4008,
+      "step": 2595
+    },
+    {
+      "epoch": 0.17665443674412284,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.889590977034924e-05,
+      "loss": 1.3989,
+      "step": 2600
+    },
+    {
+      "epoch": 0.17699415681478461,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.8893786519907594e-05,
+      "loss": 1.443,
+      "step": 2605
+    },
+    {
+      "epoch": 0.17733387688544638,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.8891663269465965e-05,
+      "loss": 1.3488,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17767359695610815,
+      "grad_norm": 5.625,
+      "learning_rate": 4.888954001902433e-05,
+      "loss": 1.4066,
+      "step": 2615
+    },
+    {
+      "epoch": 0.17801331702676995,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.8887416768582686e-05,
+      "loss": 1.4261,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17835303709743172,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.888529351814106e-05,
+      "loss": 1.3387,
+      "step": 2625
+    },
+    {
+      "epoch": 0.1786927571680935,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.888317026769942e-05,
+      "loss": 1.3489,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17903247723875526,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.888104701725778e-05,
+      "loss": 1.4955,
+      "step": 2635
+    },
+    {
+      "epoch": 0.17937219730941703,
+      "grad_norm": 5.875,
+      "learning_rate": 4.887892376681615e-05,
+      "loss": 1.4182,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17971191738007883,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.887680051637451e-05,
+      "loss": 1.4002,
+      "step": 2645
+    },
+    {
+      "epoch": 0.1800516374507406,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.887467726593287e-05,
+      "loss": 1.5055,
+      "step": 2650
+    },
+    {
+      "epoch": 0.18039135752140237,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.887255401549124e-05,
+      "loss": 1.4424,
+      "step": 2655
+    },
+    {
+      "epoch": 0.18073107759206414,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8870430765049605e-05,
+      "loss": 1.3598,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1810707976627259,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.886830751460796e-05,
+      "loss": 1.3473,
+      "step": 2665
+    },
+    {
+      "epoch": 0.18141051773338768,
+      "grad_norm": 6.375,
+      "learning_rate": 4.886618426416633e-05,
+      "loss": 1.3547,
+      "step": 2670
+    },
+    {
+      "epoch": 0.18175023780404947,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.886406101372469e-05,
+      "loss": 1.3971,
+      "step": 2675
+    },
+    {
+      "epoch": 0.18208995787471124,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8861937763283054e-05,
+      "loss": 1.4738,
+      "step": 2680
+    },
+    {
+      "epoch": 0.18242967794537301,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8859814512841425e-05,
+      "loss": 1.4292,
+      "step": 2685
+    },
+    {
+      "epoch": 0.18276939801603478,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.885769126239978e-05,
+      "loss": 1.3899,
+      "step": 2690
+    },
+    {
+      "epoch": 0.18310911808669655,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.8855568011958146e-05,
+      "loss": 1.4088,
+      "step": 2695
+    },
+    {
+      "epoch": 0.18344883815735832,
+      "grad_norm": 5.25,
+      "learning_rate": 4.885344476151652e-05,
+      "loss": 1.3877,
+      "step": 2700
+    },
+    {
+      "epoch": 0.18378855822802012,
+      "grad_norm": 5.125,
+      "learning_rate": 4.8851321511074874e-05,
+      "loss": 1.3276,
+      "step": 2705
+    },
+    {
+      "epoch": 0.1841282782986819,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.884919826063324e-05,
+      "loss": 1.34,
+      "step": 2710
+    },
+    {
+      "epoch": 0.18446799836934366,
+      "grad_norm": 6.625,
+      "learning_rate": 4.884707501019161e-05,
+      "loss": 1.4025,
+      "step": 2715
+    },
+    {
+      "epoch": 0.18480771844000543,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.8844951759749966e-05,
+      "loss": 1.3493,
+      "step": 2720
+    },
+    {
+      "epoch": 0.1851474385106672,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.884282850930833e-05,
+      "loss": 1.4122,
+      "step": 2725
+    },
+    {
+      "epoch": 0.185487158581329,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.88407052588667e-05,
+      "loss": 1.443,
+      "step": 2730
+    },
+    {
+      "epoch": 0.18582687865199077,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.883858200842506e-05,
+      "loss": 1.4167,
+      "step": 2735
+    },
+    {
+      "epoch": 0.18616659872265254,
+      "grad_norm": 5.625,
+      "learning_rate": 4.883645875798342e-05,
+      "loss": 1.483,
+      "step": 2740
+    },
+    {
+      "epoch": 0.1865063187933143,
+      "grad_norm": 5.5,
+      "learning_rate": 4.8834335507541786e-05,
+      "loss": 1.3239,
+      "step": 2745
+    },
+    {
+      "epoch": 0.18684603886397608,
+      "grad_norm": 5.875,
+      "learning_rate": 4.883221225710015e-05,
+      "loss": 1.3136,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18718575893463785,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.8830089006658514e-05,
+      "loss": 1.4607,
+      "step": 2755
+    },
+    {
+      "epoch": 0.18752547900529964,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.882796575621688e-05,
+      "loss": 1.5126,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18786519907596141,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.882584250577524e-05,
+      "loss": 1.4018,
+      "step": 2765
+    },
+    {
+      "epoch": 0.18820491914662318,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8823719255333606e-05,
+      "loss": 1.4569,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18854463921728495,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.882159600489197e-05,
+      "loss": 1.5811,
+      "step": 2775
+    },
+    {
+      "epoch": 0.18888435928794672,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8819472754450334e-05,
+      "loss": 1.4193,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1892240793586085,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.88173495040087e-05,
+      "loss": 1.4671,
+      "step": 2785
+    },
+    {
+      "epoch": 0.1895637994292703,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.881522625356706e-05,
+      "loss": 1.4372,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18990351949993206,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8813103003125426e-05,
+      "loss": 1.3231,
+      "step": 2795
+    },
+    {
+      "epoch": 0.19024323957059383,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.881097975268379e-05,
+      "loss": 1.3135,
+      "step": 2800
+    },
+    {
+      "epoch": 0.1905829596412556,
+      "grad_norm": 5.0,
+      "learning_rate": 4.8808856502242154e-05,
+      "loss": 1.2982,
+      "step": 2805
+    },
+    {
+      "epoch": 0.19092267971191737,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.880673325180052e-05,
+      "loss": 1.3464,
+      "step": 2810
+    },
+    {
+      "epoch": 0.19126239978257917,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.880461000135888e-05,
+      "loss": 1.3853,
+      "step": 2815
+    },
+    {
+      "epoch": 0.19160211985324094,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.8802486750917246e-05,
+      "loss": 1.4531,
+      "step": 2820
+    },
+    {
+      "epoch": 0.1919418399239027,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.880036350047561e-05,
+      "loss": 1.4193,
+      "step": 2825
+    },
+    {
+      "epoch": 0.19228155999456448,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.8798240250033974e-05,
+      "loss": 1.427,
+      "step": 2830
+    },
+    {
+      "epoch": 0.19262128006522625,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.879611699959234e-05,
+      "loss": 1.3426,
+      "step": 2835
+    },
+    {
+      "epoch": 0.19296100013588802,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.87939937491507e-05,
+      "loss": 1.3372,
+      "step": 2840
+    },
+    {
+      "epoch": 0.19330072020654981,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8791870498709066e-05,
+      "loss": 1.4008,
+      "step": 2845
+    },
+    {
+      "epoch": 0.19364044027721158,
+      "grad_norm": 5.5,
+      "learning_rate": 4.878974724826743e-05,
+      "loss": 1.4745,
+      "step": 2850
+    },
+    {
+      "epoch": 0.19398016034787335,
+      "grad_norm": 5.125,
+      "learning_rate": 4.8787623997825794e-05,
+      "loss": 1.4476,
+      "step": 2855
+    },
+    {
+      "epoch": 0.19431988041853512,
+      "grad_norm": 5.75,
+      "learning_rate": 4.878550074738416e-05,
+      "loss": 1.4936,
+      "step": 2860
+    },
+    {
+      "epoch": 0.1946596004891969,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.878337749694252e-05,
+      "loss": 1.3583,
+      "step": 2865
+    },
+    {
+      "epoch": 0.19499932055985866,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.8781254246500886e-05,
+      "loss": 1.3941,
+      "step": 2870
+    },
+    {
+      "epoch": 0.19533904063052046,
+      "grad_norm": 6.0,
+      "learning_rate": 4.8779130996059244e-05,
+      "loss": 1.3808,
+      "step": 2875
+    },
+    {
+      "epoch": 0.19567876070118223,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8777007745617614e-05,
+      "loss": 1.3909,
+      "step": 2880
+    },
+    {
+      "epoch": 0.196018480771844,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.877488449517598e-05,
+      "loss": 1.4054,
+      "step": 2885
+    },
+    {
+      "epoch": 0.19635820084250577,
+      "grad_norm": 4.875,
+      "learning_rate": 4.8772761244734336e-05,
+      "loss": 1.3184,
+      "step": 2890
+    },
+    {
+      "epoch": 0.19669792091316754,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.8770637994292706e-05,
+      "loss": 1.2951,
+      "step": 2895
+    },
+    {
+      "epoch": 0.19703764098382934,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.876851474385107e-05,
+      "loss": 1.3221,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1973773610544911,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.8766391493409434e-05,
+      "loss": 1.3748,
+      "step": 2905
+    },
+    {
+      "epoch": 0.19771708112515288,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.87642682429678e-05,
+      "loss": 1.4658,
+      "step": 2910
+    },
+    {
+      "epoch": 0.19805680119581465,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.876214499252616e-05,
+      "loss": 1.3682,
+      "step": 2915
+    },
+    {
+      "epoch": 0.19839652126647642,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.8760021742084526e-05,
+      "loss": 1.3523,
+      "step": 2920
+    },
+    {
+      "epoch": 0.1987362413371382,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.875789849164289e-05,
+      "loss": 1.3319,
+      "step": 2925
+    },
+    {
+      "epoch": 0.19907596140779998,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.8755775241201254e-05,
+      "loss": 1.2725,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19941568147846175,
+      "grad_norm": 5.5,
+      "learning_rate": 4.875365199075962e-05,
+      "loss": 1.3979,
+      "step": 2935
+    },
+    {
+      "epoch": 0.19975540154912352,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.875152874031798e-05,
+      "loss": 1.2965,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2000951216197853,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.874940548987634e-05,
+      "loss": 1.3774,
+      "step": 2945
+    },
+    {
+      "epoch": 0.20043484169044706,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.874728223943471e-05,
+      "loss": 1.4207,
+      "step": 2950
+    },
+    {
+      "epoch": 0.20077456176110883,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.8745158988993074e-05,
+      "loss": 1.4413,
+      "step": 2955
+    },
+    {
+      "epoch": 0.20111428183177063,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.874303573855143e-05,
+      "loss": 1.4259,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2014540019024324,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.87409124881098e-05,
+      "loss": 1.3845,
+      "step": 2965
+    },
+    {
+      "epoch": 0.20179372197309417,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.8738789237668166e-05,
+      "loss": 1.374,
+      "step": 2970
+    },
+    {
+      "epoch": 0.20213344204375594,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.8736665987226524e-05,
+      "loss": 1.4781,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2024731621144177,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8734542736784894e-05,
+      "loss": 1.4312,
+      "step": 2980
+    },
+    {
+      "epoch": 0.2028128821850795,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.873241948634326e-05,
+      "loss": 1.4507,
+      "step": 2985
+    },
+    {
+      "epoch": 0.20315260225574128,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.8730296235901616e-05,
+      "loss": 1.4297,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20349232232640305,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.8728172985459986e-05,
+      "loss": 1.3676,
+      "step": 2995
+    },
+    {
+      "epoch": 0.20383204239706482,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.872604973501835e-05,
+      "loss": 1.3296,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2041717624677266,
+      "grad_norm": 6.25,
+      "learning_rate": 4.872392648457671e-05,
+      "loss": 1.3762,
+      "step": 3005
+    },
+    {
+      "epoch": 0.20451148253838836,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.872180323413508e-05,
+      "loss": 1.3101,
+      "step": 3010
+    },
+    {
+      "epoch": 0.20485120260905015,
+      "grad_norm": 5.75,
+      "learning_rate": 4.871967998369344e-05,
+      "loss": 1.4317,
+      "step": 3015
+    },
+    {
+      "epoch": 0.20519092267971192,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.87175567332518e-05,
+      "loss": 1.3736,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2055306427503737,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.871543348281017e-05,
+      "loss": 1.3797,
+      "step": 3025
+    },
+    {
+      "epoch": 0.20587036282103546,
+      "grad_norm": 6.125,
+      "learning_rate": 4.871331023236853e-05,
+      "loss": 1.4357,
+      "step": 3030
+    },
+    {
+      "epoch": 0.20621008289169723,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.871118698192689e-05,
+      "loss": 1.532,
+      "step": 3035
+    },
+    {
+      "epoch": 0.206549802962359,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.870906373148526e-05,
+      "loss": 1.3176,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2068895230330208,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.870694048104362e-05,
+      "loss": 1.4286,
+      "step": 3045
+    },
+    {
+      "epoch": 0.20722924310368257,
+      "grad_norm": 5.0,
+      "learning_rate": 4.8704817230601984e-05,
+      "loss": 1.3084,
+      "step": 3050
+    },
+    {
+      "epoch": 0.20756896317434434,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.8702693980160355e-05,
+      "loss": 1.3142,
+      "step": 3055
+    },
+    {
+      "epoch": 0.2079086832450061,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.870057072971871e-05,
+      "loss": 1.3869,
+      "step": 3060
+    },
+    {
+      "epoch": 0.20824840331566788,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.8698447479277076e-05,
+      "loss": 1.3131,
+      "step": 3065
+    },
+    {
+      "epoch": 0.20858812338632968,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.8696324228835447e-05,
+      "loss": 1.4044,
+      "step": 3070
+    },
+    {
+      "epoch": 0.20892784345699145,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.8694200978393804e-05,
+      "loss": 1.3625,
+      "step": 3075
+    },
+    {
+      "epoch": 0.20926756352765322,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.869207772795217e-05,
+      "loss": 1.3381,
+      "step": 3080
+    },
+    {
+      "epoch": 0.209607283598315,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.868995447751054e-05,
+      "loss": 1.4845,
+      "step": 3085
+    },
+    {
+      "epoch": 0.20994700366897676,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8687831227068896e-05,
+      "loss": 1.3352,
+      "step": 3090
+    },
+    {
+      "epoch": 0.21028672373963853,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.868570797662726e-05,
+      "loss": 1.4521,
+      "step": 3095
+    },
+    {
+      "epoch": 0.21062644381030032,
+      "grad_norm": 6.125,
+      "learning_rate": 4.8683584726185624e-05,
+      "loss": 1.4325,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2109661638809621,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.868146147574399e-05,
+      "loss": 1.2212,
+      "step": 3105
+    },
+    {
+      "epoch": 0.21130588395162386,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.867933822530235e-05,
+      "loss": 1.4217,
+      "step": 3110
+    },
+    {
+      "epoch": 0.21164560402228563,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.8677214974860716e-05,
+      "loss": 1.4087,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2119853240929474,
+      "grad_norm": 6.5,
+      "learning_rate": 4.867509172441908e-05,
+      "loss": 1.359,
+      "step": 3120
+    },
+    {
+      "epoch": 0.21232504416360917,
+      "grad_norm": 5.25,
+      "learning_rate": 4.8672968473977444e-05,
+      "loss": 1.3966,
+      "step": 3125
+    },
+    {
+      "epoch": 0.21266476423427097,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.867084522353581e-05,
+      "loss": 1.3308,
+      "step": 3130
+    },
+    {
+      "epoch": 0.21300448430493274,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.866872197309417e-05,
+      "loss": 1.3716,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2133442043755945,
+      "grad_norm": 5.625,
+      "learning_rate": 4.8666598722652536e-05,
+      "loss": 1.2703,
+      "step": 3140
+    },
+    {
+      "epoch": 0.21368392444625628,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.86644754722109e-05,
+      "loss": 1.3483,
+      "step": 3145
+    },
+    {
+      "epoch": 0.21402364451691805,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.8662352221769264e-05,
+      "loss": 1.4144,
+      "step": 3150
+    },
+    {
+      "epoch": 0.21436336458757985,
+      "grad_norm": 6.25,
+      "learning_rate": 4.866022897132763e-05,
+      "loss": 1.3439,
+      "step": 3155
+    },
+    {
+      "epoch": 0.21470308465824162,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.865810572088599e-05,
+      "loss": 1.406,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2150428047289034,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8655982470444356e-05,
+      "loss": 1.4858,
+      "step": 3165
+    },
+    {
+      "epoch": 0.21538252479956516,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.865385922000272e-05,
+      "loss": 1.2449,
+      "step": 3170
+    },
+    {
+      "epoch": 0.21572224487022693,
+      "grad_norm": 4.875,
+      "learning_rate": 4.8651735969561084e-05,
+      "loss": 1.4289,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2160619649408887,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.864961271911945e-05,
+      "loss": 1.3328,
+      "step": 3180
+    },
+    {
+      "epoch": 0.2164016850115505,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.864748946867781e-05,
+      "loss": 1.3146,
+      "step": 3185
+    },
+    {
+      "epoch": 0.21674140508221226,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8645366218236176e-05,
+      "loss": 1.3286,
+      "step": 3190
+    },
+    {
+      "epoch": 0.21708112515287403,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.864324296779454e-05,
+      "loss": 1.413,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2174208452235358,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.8641119717352904e-05,
+      "loss": 1.4719,
+      "step": 3200
+    },
+    {
+      "epoch": 0.21776056529419757,
+      "grad_norm": 7.0,
+      "learning_rate": 4.863899646691127e-05,
+      "loss": 1.3069,
+      "step": 3205
+    },
+    {
+      "epoch": 0.21810028536485934,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.863687321646963e-05,
+      "loss": 1.3724,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21844000543552114,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.8634749966027996e-05,
+      "loss": 1.3567,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2187797255061829,
+      "grad_norm": 5.625,
+      "learning_rate": 4.863262671558636e-05,
+      "loss": 1.3963,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21911944557684468,
+      "grad_norm": 5.5,
+      "learning_rate": 4.8630503465144724e-05,
+      "loss": 1.323,
+      "step": 3225
+    },
+    {
+      "epoch": 0.21945916564750645,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.862838021470308e-05,
+      "loss": 1.3635,
+      "step": 3230
+    },
+    {
+      "epoch": 0.21979888571816822,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.862625696426145e-05,
+      "loss": 1.3725,
+      "step": 3235
+    },
+    {
+      "epoch": 0.22013860578883002,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.8624133713819816e-05,
+      "loss": 1.3978,
+      "step": 3240
+    },
+    {
+      "epoch": 0.22047832585949179,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.862201046337818e-05,
+      "loss": 1.432,
+      "step": 3245
+    },
+    {
+      "epoch": 0.22081804593015356,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.8619887212936544e-05,
+      "loss": 1.4217,
+      "step": 3250
+    },
+    {
+      "epoch": 0.22115776600081533,
+      "grad_norm": 5.5,
+      "learning_rate": 4.861776396249491e-05,
+      "loss": 1.3897,
+      "step": 3255
+    },
+    {
+      "epoch": 0.2214974860714771,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.861564071205327e-05,
+      "loss": 1.4355,
+      "step": 3260
+    },
+    {
+      "epoch": 0.22183720614213887,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.8613517461611636e-05,
+      "loss": 1.4283,
+      "step": 3265
+    },
+    {
+      "epoch": 0.22217692621280066,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.861139421117e-05,
+      "loss": 1.2823,
+      "step": 3270
+    },
+    {
+      "epoch": 0.22251664628346243,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.8609270960728364e-05,
+      "loss": 1.3796,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2228563663541242,
+      "grad_norm": 5.5,
+      "learning_rate": 4.860714771028673e-05,
+      "loss": 1.4239,
+      "step": 3280
+    },
+    {
+      "epoch": 0.22319608642478597,
+      "grad_norm": 5.5,
+      "learning_rate": 4.860502445984509e-05,
+      "loss": 1.3966,
+      "step": 3285
+    },
+    {
+      "epoch": 0.22353580649544774,
+      "grad_norm": 6.125,
+      "learning_rate": 4.8602901209403456e-05,
+      "loss": 1.3576,
+      "step": 3290
+    },
+    {
+      "epoch": 0.22387552656610954,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.860077795896182e-05,
+      "loss": 1.2688,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2242152466367713,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.859865470852018e-05,
+      "loss": 1.4104,
+      "step": 3300
+    },
+    {
+      "epoch": 0.22455496670743308,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.859653145807855e-05,
+      "loss": 1.5236,
+      "step": 3305
+    },
+    {
+      "epoch": 0.22489468677809485,
+      "grad_norm": 4.75,
+      "learning_rate": 4.859440820763691e-05,
+      "loss": 1.2704,
+      "step": 3310
+    },
+    {
+      "epoch": 0.22523440684875662,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.859228495719527e-05,
+      "loss": 1.332,
+      "step": 3315
+    },
+    {
+      "epoch": 0.2255741269194184,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.859016170675364e-05,
+      "loss": 1.3078,
+      "step": 3320
+    },
+    {
+      "epoch": 0.22591384699008019,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.8588038456312004e-05,
+      "loss": 1.3394,
+      "step": 3325
+    },
+    {
+      "epoch": 0.22625356706074196,
+      "grad_norm": 5.25,
+      "learning_rate": 4.858591520587036e-05,
+      "loss": 1.4843,
+      "step": 3330
+    },
+    {
+      "epoch": 0.22659328713140373,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.858379195542873e-05,
+      "loss": 1.3613,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2269330072020655,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.8581668704987096e-05,
+      "loss": 1.4019,
+      "step": 3340
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.857954545454545e-05,
+      "loss": 1.3346,
+      "step": 3345
+    },
+    {
+      "epoch": 0.22761244734338903,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.8577422204103824e-05,
+      "loss": 1.4178,
+      "step": 3350
+    },
+    {
+      "epoch": 0.22795216741405083,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.857529895366219e-05,
+      "loss": 1.3297,
+      "step": 3355
+    },
+    {
+      "epoch": 0.2282918874847126,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.8573175703220545e-05,
+      "loss": 1.4144,
+      "step": 3360
+    },
+    {
+      "epoch": 0.22863160755537437,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.8571052452778916e-05,
+      "loss": 1.4315,
+      "step": 3365
+    },
+    {
+      "epoch": 0.22897132762603614,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.856892920233727e-05,
+      "loss": 1.3227,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2293110476966979,
+      "grad_norm": 5.625,
+      "learning_rate": 4.856680595189564e-05,
+      "loss": 1.2725,
+      "step": 3375
+    },
+    {
+      "epoch": 0.2296507677673597,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.856468270145401e-05,
+      "loss": 1.3907,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22999048783802148,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8562559451012365e-05,
+      "loss": 1.4225,
+      "step": 3385
+    },
+    {
+      "epoch": 0.23033020790868325,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.856043620057073e-05,
+      "loss": 1.4277,
+      "step": 3390
+    },
+    {
+      "epoch": 0.23066992797934502,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.85583129501291e-05,
+      "loss": 1.32,
+      "step": 3395
+    },
+    {
+      "epoch": 0.2310096480500068,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.855618969968746e-05,
+      "loss": 1.3818,
+      "step": 3400
+    },
+    {
+      "epoch": 0.23134936812066856,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.855406644924582e-05,
+      "loss": 1.4453,
+      "step": 3405
+    },
+    {
+      "epoch": 0.23168908819133036,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.855194319880419e-05,
+      "loss": 1.2651,
+      "step": 3410
+    },
+    {
+      "epoch": 0.23202880826199213,
+      "grad_norm": 5.5,
+      "learning_rate": 4.854981994836255e-05,
+      "loss": 1.3796,
+      "step": 3415
+    },
+    {
+      "epoch": 0.2323685283326539,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.854769669792091e-05,
+      "loss": 1.3456,
+      "step": 3420
+    },
+    {
+      "epoch": 0.23270824840331567,
+      "grad_norm": 6.0,
+      "learning_rate": 4.8545573447479284e-05,
+      "loss": 1.378,
+      "step": 3425
+    },
+    {
+      "epoch": 0.23304796847397743,
+      "grad_norm": 5.375,
+      "learning_rate": 4.854345019703764e-05,
+      "loss": 1.4281,
+      "step": 3430
+    },
+    {
+      "epoch": 0.2333876885446392,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.8541326946596005e-05,
+      "loss": 1.3989,
+      "step": 3435
+    },
+    {
+      "epoch": 0.233727408615301,
+      "grad_norm": 6.25,
+      "learning_rate": 4.8539203696154376e-05,
+      "loss": 1.2927,
+      "step": 3440
+    },
+    {
+      "epoch": 0.23406712868596277,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.853708044571273e-05,
+      "loss": 1.2958,
+      "step": 3445
+    },
+    {
+      "epoch": 0.23440684875662454,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.85349571952711e-05,
+      "loss": 1.4032,
+      "step": 3450
+    },
+    {
+      "epoch": 0.2347465688272863,
+      "grad_norm": 6.0,
+      "learning_rate": 4.853283394482946e-05,
+      "loss": 1.4232,
+      "step": 3455
+    },
+    {
+      "epoch": 0.23508628889794808,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.8530710694387825e-05,
+      "loss": 1.2945,
+      "step": 3460
+    },
+    {
+      "epoch": 0.23542600896860988,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.852858744394619e-05,
+      "loss": 1.4133,
+      "step": 3465
+    },
+    {
+      "epoch": 0.23576572903927165,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.852646419350455e-05,
+      "loss": 1.3067,
+      "step": 3470
+    },
+    {
+      "epoch": 0.23610544910993342,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.852434094306292e-05,
+      "loss": 1.3765,
+      "step": 3475
+    },
+    {
+      "epoch": 0.2364451691805952,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.852221769262128e-05,
+      "loss": 1.3512,
+      "step": 3480
+    },
+    {
+      "epoch": 0.23678488925125696,
+      "grad_norm": 4.375,
+      "learning_rate": 4.8520094442179645e-05,
+      "loss": 1.2725,
+      "step": 3485
+    },
+    {
+      "epoch": 0.23712460932191873,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.851797119173801e-05,
+      "loss": 1.2809,
+      "step": 3490
+    },
+    {
+      "epoch": 0.23746432939258053,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.851584794129637e-05,
+      "loss": 1.3228,
+      "step": 3495
+    },
+    {
+      "epoch": 0.2378040494632423,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.851372469085474e-05,
+      "loss": 1.35,
+      "step": 3500
+    },
+    {
+      "epoch": 0.23814376953390406,
+      "grad_norm": 6.25,
+      "learning_rate": 4.85116014404131e-05,
+      "loss": 1.3944,
+      "step": 3505
+    },
+    {
+      "epoch": 0.23848348960456583,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8509478189971465e-05,
+      "loss": 1.3254,
+      "step": 3510
+    },
+    {
+      "epoch": 0.2388232096752276,
+      "grad_norm": 5.25,
+      "learning_rate": 4.850735493952983e-05,
+      "loss": 1.3673,
+      "step": 3515
+    },
+    {
+      "epoch": 0.23916292974588937,
+      "grad_norm": 5.5,
+      "learning_rate": 4.850523168908819e-05,
+      "loss": 1.3679,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23950264981655117,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.850310843864656e-05,
+      "loss": 1.3871,
+      "step": 3525
+    },
+    {
+      "epoch": 0.23984236988721294,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.850098518820492e-05,
+      "loss": 1.3236,
+      "step": 3530
+    },
+    {
+      "epoch": 0.2401820899578747,
+      "grad_norm": 6.375,
+      "learning_rate": 4.8498861937763285e-05,
+      "loss": 1.3749,
+      "step": 3535
+    },
+    {
+      "epoch": 0.24052181002853648,
+      "grad_norm": 5.375,
+      "learning_rate": 4.849673868732165e-05,
+      "loss": 1.38,
+      "step": 3540
+    },
+    {
+      "epoch": 0.24086153009919825,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.849461543688001e-05,
+      "loss": 1.3086,
+      "step": 3545
+    },
+    {
+      "epoch": 0.24120125016986005,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.849249218643838e-05,
+      "loss": 1.4226,
+      "step": 3550
+    },
+    {
+      "epoch": 0.24154097024052182,
+      "grad_norm": 5.5,
+      "learning_rate": 4.849036893599674e-05,
+      "loss": 1.4477,
+      "step": 3555
+    },
+    {
+      "epoch": 0.2418806903111836,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8488245685555105e-05,
+      "loss": 1.3254,
+      "step": 3560
+    },
+    {
+      "epoch": 0.24222041038184536,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.848612243511347e-05,
+      "loss": 1.3132,
+      "step": 3565
+    },
+    {
+      "epoch": 0.24256013045250713,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.848399918467183e-05,
+      "loss": 1.3314,
+      "step": 3570
+    },
+    {
+      "epoch": 0.2428998505231689,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.84818759342302e-05,
+      "loss": 1.4421,
+      "step": 3575
+    },
+    {
+      "epoch": 0.2432395705938307,
+      "grad_norm": 5.25,
+      "learning_rate": 4.847975268378856e-05,
+      "loss": 1.6417,
+      "step": 3580
+    },
+    {
+      "epoch": 0.24357929066449246,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.8477629433346925e-05,
+      "loss": 1.4163,
+      "step": 3585
+    },
+    {
+      "epoch": 0.24391901073515423,
+      "grad_norm": 5.375,
+      "learning_rate": 4.847550618290529e-05,
+      "loss": 1.2936,
+      "step": 3590
+    },
+    {
+      "epoch": 0.244258730805816,
+      "grad_norm": 5.5,
+      "learning_rate": 4.8473382932463653e-05,
+      "loss": 1.3723,
+      "step": 3595
+    },
+    {
+      "epoch": 0.24459845087647777,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.847125968202202e-05,
+      "loss": 1.3334,
+      "step": 3600
+    },
+    {
+      "epoch": 0.24493817094713954,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.846913643158038e-05,
+      "loss": 1.3704,
+      "step": 3605
+    },
+    {
+      "epoch": 0.24527789101780134,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8467013181138745e-05,
+      "loss": 1.3522,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2456176110884631,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.846488993069711e-05,
+      "loss": 1.2763,
+      "step": 3615
+    },
+    {
+      "epoch": 0.24595733115912488,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.8462766680255473e-05,
+      "loss": 1.3042,
+      "step": 3620
+    },
+    {
+      "epoch": 0.24629705122978665,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.846064342981384e-05,
+      "loss": 1.2931,
+      "step": 3625
+    },
+    {
+      "epoch": 0.24663677130044842,
+      "grad_norm": 5.625,
+      "learning_rate": 4.84585201793722e-05,
+      "loss": 1.3337,
+      "step": 3630
+    },
+    {
+      "epoch": 0.24697649137111022,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8456396928930565e-05,
+      "loss": 1.4034,
+      "step": 3635
+    },
+    {
+      "epoch": 0.247316211441772,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.845427367848893e-05,
+      "loss": 1.4244,
+      "step": 3640
+    },
+    {
+      "epoch": 0.24765593151243376,
+      "grad_norm": 5.0,
+      "learning_rate": 4.8452150428047293e-05,
+      "loss": 1.2727,
+      "step": 3645
+    },
+    {
+      "epoch": 0.24799565158309553,
+      "grad_norm": 4.875,
+      "learning_rate": 4.845002717760566e-05,
+      "loss": 1.3127,
+      "step": 3650
+    },
+    {
+      "epoch": 0.2483353716537573,
+      "grad_norm": 5.25,
+      "learning_rate": 4.8447903927164015e-05,
+      "loss": 1.2276,
+      "step": 3655
+    },
+    {
+      "epoch": 0.24867509172441907,
+      "grad_norm": 5.75,
+      "learning_rate": 4.8445780676722385e-05,
+      "loss": 1.3198,
+      "step": 3660
+    },
+    {
+      "epoch": 0.24901481179508086,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.844365742628075e-05,
+      "loss": 1.3543,
+      "step": 3665
+    },
+    {
+      "epoch": 0.24935453186574263,
+      "grad_norm": 6.0,
+      "learning_rate": 4.844153417583911e-05,
+      "loss": 1.3102,
+      "step": 3670
+    },
+    {
+      "epoch": 0.2496942519364044,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.843941092539748e-05,
+      "loss": 1.4132,
+      "step": 3675
+    },
+    {
+      "epoch": 0.2500339720070662,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.843728767495584e-05,
+      "loss": 1.3746,
+      "step": 3680
+    },
+    {
+      "epoch": 0.25037369207772797,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.84351644245142e-05,
+      "loss": 1.2817,
+      "step": 3685
+    },
+    {
+      "epoch": 0.2507134121483897,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.843304117407257e-05,
+      "loss": 1.3564,
+      "step": 3690
+    },
+    {
+      "epoch": 0.2510531322190515,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.8430917923630933e-05,
+      "loss": 1.4147,
+      "step": 3695
+    },
+    {
+      "epoch": 0.25139285228971325,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.842879467318929e-05,
+      "loss": 1.3388,
+      "step": 3700
+    },
+    {
+      "epoch": 0.25173257236037505,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.842667142274766e-05,
+      "loss": 1.3674,
+      "step": 3705
+    },
+    {
+      "epoch": 0.25207229243103685,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.8424548172306025e-05,
+      "loss": 1.4028,
+      "step": 3710
+    },
+    {
+      "epoch": 0.2524120125016986,
+      "grad_norm": 6.125,
+      "learning_rate": 4.842242492186438e-05,
+      "loss": 1.3856,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2527517325723604,
+      "grad_norm": 5.25,
+      "learning_rate": 4.8420301671422754e-05,
+      "loss": 1.287,
+      "step": 3720
+    },
+    {
+      "epoch": 0.25309145264302213,
+      "grad_norm": 5.0,
+      "learning_rate": 4.841817842098111e-05,
+      "loss": 1.3544,
+      "step": 3725
+    },
+    {
+      "epoch": 0.2534311727136839,
+      "grad_norm": 5.625,
+      "learning_rate": 4.8416055170539475e-05,
+      "loss": 1.1769,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2537708927843457,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.8413931920097846e-05,
+      "loss": 1.4207,
+      "step": 3735
+    },
+    {
+      "epoch": 0.25411061285500747,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.84118086696562e-05,
+      "loss": 1.3439,
+      "step": 3740
+    },
+    {
+      "epoch": 0.25445033292566926,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.840968541921457e-05,
+      "loss": 1.348,
+      "step": 3745
+    },
+    {
+      "epoch": 0.254790052996331,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.840756216877294e-05,
+      "loss": 1.4218,
+      "step": 3750
+    },
+    {
+      "epoch": 0.2551297730669928,
+      "grad_norm": 5.625,
+      "learning_rate": 4.8405438918331295e-05,
+      "loss": 1.4408,
+      "step": 3755
+    },
+    {
+      "epoch": 0.25546949313765455,
+      "grad_norm": 5.375,
+      "learning_rate": 4.840331566788966e-05,
+      "loss": 1.3364,
+      "step": 3760
+    },
+    {
+      "epoch": 0.25580921320831634,
+      "grad_norm": 4.625,
+      "learning_rate": 4.840119241744803e-05,
+      "loss": 1.3278,
+      "step": 3765
+    },
+    {
+      "epoch": 0.25614893327897814,
+      "grad_norm": 5.375,
+      "learning_rate": 4.839906916700639e-05,
+      "loss": 1.336,
+      "step": 3770
+    },
+    {
+      "epoch": 0.2564886533496399,
+      "grad_norm": 5.125,
+      "learning_rate": 4.839694591656475e-05,
+      "loss": 1.4268,
+      "step": 3775
+    },
+    {
+      "epoch": 0.2568283734203017,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.839482266612312e-05,
+      "loss": 1.3637,
+      "step": 3780
+    },
+    {
+      "epoch": 0.2571680934909634,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.839269941568148e-05,
+      "loss": 1.4176,
+      "step": 3785
+    },
+    {
+      "epoch": 0.2575078135616252,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.839057616523984e-05,
+      "loss": 1.402,
+      "step": 3790
+    },
+    {
+      "epoch": 0.257847533632287,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.8388452914798214e-05,
+      "loss": 1.2878,
+      "step": 3795
+    },
+    {
+      "epoch": 0.25818725370294876,
+      "grad_norm": 5.25,
+      "learning_rate": 4.838632966435657e-05,
+      "loss": 1.3567,
+      "step": 3800
+    },
+    {
+      "epoch": 0.25852697377361056,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8384206413914935e-05,
+      "loss": 1.3219,
+      "step": 3805
+    },
+    {
+      "epoch": 0.2588666938442723,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.83820831634733e-05,
+      "loss": 1.3536,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2592064139149341,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.837995991303166e-05,
+      "loss": 1.3576,
+      "step": 3815
+    },
+    {
+      "epoch": 0.2595461339855959,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.837783666259003e-05,
+      "loss": 1.3512,
+      "step": 3820
+    },
+    {
+      "epoch": 0.25988585405625764,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.837571341214839e-05,
+      "loss": 1.4162,
+      "step": 3825
+    },
+    {
+      "epoch": 0.26022557412691943,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.8373590161706755e-05,
+      "loss": 1.2637,
+      "step": 3830
+    },
+    {
+      "epoch": 0.2605652941975812,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.837146691126512e-05,
+      "loss": 1.2523,
+      "step": 3835
+    },
+    {
+      "epoch": 0.260905014268243,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.836934366082348e-05,
+      "loss": 1.4031,
+      "step": 3840
+    },
+    {
+      "epoch": 0.2612447343389047,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.836722041038185e-05,
+      "loss": 1.4064,
+      "step": 3845
+    },
+    {
+      "epoch": 0.2615844544095665,
+      "grad_norm": 5.5,
+      "learning_rate": 4.836509715994021e-05,
+      "loss": 1.4108,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2619241744802283,
+      "grad_norm": 5.875,
+      "learning_rate": 4.8362973909498575e-05,
+      "loss": 1.3559,
+      "step": 3855
+    },
+    {
+      "epoch": 0.26226389455089005,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.836085065905694e-05,
+      "loss": 1.403,
+      "step": 3860
+    },
+    {
+      "epoch": 0.26260361462155185,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.83587274086153e-05,
+      "loss": 1.3531,
+      "step": 3865
+    },
+    {
+      "epoch": 0.2629433346922136,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.835660415817367e-05,
+      "loss": 1.2937,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2632830547628754,
+      "grad_norm": 6.125,
+      "learning_rate": 4.835448090773203e-05,
+      "loss": 1.359,
+      "step": 3875
+    },
+    {
+      "epoch": 0.2636227748335372,
+      "grad_norm": 5.625,
+      "learning_rate": 4.8352357657290395e-05,
+      "loss": 1.2956,
+      "step": 3880
+    },
+    {
+      "epoch": 0.26396249490419893,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.835023440684876e-05,
+      "loss": 1.3559,
+      "step": 3885
+    },
+    {
+      "epoch": 0.2643022149748607,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.834811115640712e-05,
+      "loss": 1.3947,
+      "step": 3890
+    },
+    {
+      "epoch": 0.26464193504552247,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.834598790596549e-05,
+      "loss": 1.3483,
+      "step": 3895
+    },
+    {
+      "epoch": 0.26498165511618427,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.834386465552385e-05,
+      "loss": 1.3093,
+      "step": 3900
+    },
+    {
+      "epoch": 0.26532137518684606,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.8341741405082215e-05,
+      "loss": 1.3617,
+      "step": 3905
+    },
+    {
+      "epoch": 0.2656610952575078,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.833961815464058e-05,
+      "loss": 1.4009,
+      "step": 3910
+    },
+    {
+      "epoch": 0.2660008153281696,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.833749490419894e-05,
+      "loss": 1.3137,
+      "step": 3915
+    },
+    {
+      "epoch": 0.26634053539883135,
+      "grad_norm": 6.625,
+      "learning_rate": 4.833537165375731e-05,
+      "loss": 1.4031,
+      "step": 3920
+    },
+    {
+      "epoch": 0.26668025546949314,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.833324840331567e-05,
+      "loss": 1.3548,
+      "step": 3925
+    },
+    {
+      "epoch": 0.2670199755401549,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8331125152874035e-05,
+      "loss": 1.3613,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2673596956108167,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.83290019024324e-05,
+      "loss": 1.3354,
+      "step": 3935
+    },
+    {
+      "epoch": 0.2676994156814785,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.832687865199076e-05,
+      "loss": 1.3427,
+      "step": 3940
+    },
+    {
+      "epoch": 0.2680391357521402,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.832475540154913e-05,
+      "loss": 1.4245,
+      "step": 3945
+    },
+    {
+      "epoch": 0.268378855822802,
+      "grad_norm": 5.375,
+      "learning_rate": 4.832263215110749e-05,
+      "loss": 1.4335,
+      "step": 3950
+    },
+    {
+      "epoch": 0.26871857589346376,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8320508900665855e-05,
+      "loss": 1.2394,
+      "step": 3955
+    },
+    {
+      "epoch": 0.26905829596412556,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.831838565022422e-05,
+      "loss": 1.3884,
+      "step": 3960
+    },
+    {
+      "epoch": 0.26939801603478736,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.831626239978258e-05,
+      "loss": 1.3912,
+      "step": 3965
+    },
+    {
+      "epoch": 0.2697377361054491,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.831413914934095e-05,
+      "loss": 1.3446,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2700774561761109,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.831201589889931e-05,
+      "loss": 1.4568,
+      "step": 3975
+    },
+    {
+      "epoch": 0.27041717624677264,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.8309892648457675e-05,
+      "loss": 1.3715,
+      "step": 3980
+    },
+    {
+      "epoch": 0.27075689631743444,
+      "grad_norm": 6.0,
+      "learning_rate": 4.830776939801604e-05,
+      "loss": 1.42,
+      "step": 3985
+    },
+    {
+      "epoch": 0.27109661638809623,
+      "grad_norm": 5.375,
+      "learning_rate": 4.83056461475744e-05,
+      "loss": 1.3536,
+      "step": 3990
+    },
+    {
+      "epoch": 0.271436336458758,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.830352289713277e-05,
+      "loss": 1.3848,
+      "step": 3995
+    },
+    {
+      "epoch": 0.2717760565294198,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.830139964669113e-05,
+      "loss": 1.4557,
+      "step": 4000
+    },
+    {
+      "epoch": 0.2721157766000815,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.8299276396249495e-05,
+      "loss": 1.4271,
+      "step": 4005
+    },
+    {
+      "epoch": 0.2724554966707433,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.829715314580785e-05,
+      "loss": 1.3589,
+      "step": 4010
+    },
+    {
+      "epoch": 0.27279521674140506,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.829502989536622e-05,
+      "loss": 1.404,
+      "step": 4015
+    },
+    {
+      "epoch": 0.27313493681206685,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.829290664492459e-05,
+      "loss": 1.336,
+      "step": 4020
+    },
+    {
+      "epoch": 0.27347465688272865,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.8290783394482944e-05,
+      "loss": 1.4174,
+      "step": 4025
+    },
+    {
+      "epoch": 0.2738143769533904,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.8288660144041315e-05,
+      "loss": 1.3324,
+      "step": 4030
+    },
+    {
+      "epoch": 0.2741540970240522,
+      "grad_norm": 7.375,
+      "learning_rate": 4.828653689359968e-05,
+      "loss": 1.4031,
+      "step": 4035
+    },
+    {
+      "epoch": 0.27449381709471393,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.8284413643158036e-05,
+      "loss": 1.3659,
+      "step": 4040
+    },
+    {
+      "epoch": 0.27483353716537573,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.828229039271641e-05,
+      "loss": 1.4568,
+      "step": 4045
+    },
+    {
+      "epoch": 0.2751732572360375,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.828016714227477e-05,
+      "loss": 1.3825,
+      "step": 4050
+    },
+    {
+      "epoch": 0.27551297730669927,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.827804389183313e-05,
+      "loss": 1.38,
+      "step": 4055
+    },
+    {
+      "epoch": 0.27585269737736107,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.82759206413915e-05,
+      "loss": 1.2439,
+      "step": 4060
+    },
+    {
+      "epoch": 0.2761924174480228,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.827379739094986e-05,
+      "loss": 1.3082,
+      "step": 4065
+    },
+    {
+      "epoch": 0.2765321375186846,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.827167414050822e-05,
+      "loss": 1.3194,
+      "step": 4070
+    },
+    {
+      "epoch": 0.2768718575893464,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.826955089006659e-05,
+      "loss": 1.3495,
+      "step": 4075
+    },
+    {
+      "epoch": 0.27721157766000815,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.826742763962495e-05,
+      "loss": 1.3372,
+      "step": 4080
+    },
+    {
+      "epoch": 0.27755129773066994,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.826530438918331e-05,
+      "loss": 1.3639,
+      "step": 4085
+    },
+    {
+      "epoch": 0.2778910178013317,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.826318113874168e-05,
+      "loss": 1.3866,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2782307378719935,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.826105788830004e-05,
+      "loss": 1.4029,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2785704579426552,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.8258934637858404e-05,
+      "loss": 1.3996,
+      "step": 4100
+    },
+    {
+      "epoch": 0.278910178013317,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.8256811387416775e-05,
+      "loss": 1.3921,
+      "step": 4105
+    },
+    {
+      "epoch": 0.2792498980839788,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.825468813697513e-05,
+      "loss": 1.3157,
+      "step": 4110
+    },
+    {
+      "epoch": 0.27958961815464056,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8252564886533496e-05,
+      "loss": 1.3111,
+      "step": 4115
+    },
+    {
+      "epoch": 0.27992933822530236,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.825044163609187e-05,
+      "loss": 1.4403,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2802690582959641,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.8248318385650224e-05,
+      "loss": 1.3318,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2806087783666259,
+      "grad_norm": 5.5,
+      "learning_rate": 4.824619513520859e-05,
+      "loss": 1.317,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2809484984372877,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.824407188476696e-05,
+      "loss": 1.3872,
+      "step": 4135
+    },
+    {
+      "epoch": 0.28128821850794944,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.8241948634325316e-05,
+      "loss": 1.3353,
+      "step": 4140
+    },
+    {
+      "epoch": 0.28162793857861124,
+      "grad_norm": 5.375,
+      "learning_rate": 4.823982538388368e-05,
+      "loss": 1.2713,
+      "step": 4145
+    },
+    {
+      "epoch": 0.281967658649273,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.8237702133442044e-05,
+      "loss": 1.318,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2823073787199348,
+      "grad_norm": 5.75,
+      "learning_rate": 4.823557888300041e-05,
+      "loss": 1.4281,
+      "step": 4155
+    },
+    {
+      "epoch": 0.2826470987905966,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.823345563255877e-05,
+      "loss": 1.4508,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2829868188612583,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.8231332382117136e-05,
+      "loss": 1.2525,
+      "step": 4165
+    },
+    {
+      "epoch": 0.2833265389319201,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.82292091316755e-05,
+      "loss": 1.3115,
+      "step": 4170
+    },
+    {
+      "epoch": 0.28366625900258186,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.8227085881233864e-05,
+      "loss": 1.3978,
+      "step": 4175
+    },
+    {
+      "epoch": 0.28400597907324365,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.822496263079223e-05,
+      "loss": 1.254,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2843456991439054,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.822283938035059e-05,
+      "loss": 1.3349,
+      "step": 4185
+    },
+    {
+      "epoch": 0.2846854192145672,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.8220716129908956e-05,
+      "loss": 1.3063,
+      "step": 4190
+    },
+    {
+      "epoch": 0.285025139285229,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.821859287946732e-05,
+      "loss": 1.3221,
+      "step": 4195
+    },
+    {
+      "epoch": 0.28536485935589073,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.8216469629025684e-05,
+      "loss": 1.3192,
+      "step": 4200
+    },
+    {
+      "epoch": 0.28570457942655253,
+      "grad_norm": 5.25,
+      "learning_rate": 4.821434637858405e-05,
+      "loss": 1.3963,
+      "step": 4205
+    },
+    {
+      "epoch": 0.28604429949721427,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.821222312814241e-05,
+      "loss": 1.4414,
+      "step": 4210
+    },
+    {
+      "epoch": 0.28638401956787607,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.8210099877700776e-05,
+      "loss": 1.3353,
+      "step": 4215
+    },
+    {
+      "epoch": 0.28672373963853787,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.820797662725914e-05,
+      "loss": 1.3122,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2870634597091996,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.8205853376817504e-05,
+      "loss": 1.4122,
+      "step": 4225
+    },
+    {
+      "epoch": 0.2874031797798614,
+      "grad_norm": 5.375,
+      "learning_rate": 4.820373012637587e-05,
+      "loss": 1.2998,
+      "step": 4230
+    },
+    {
+      "epoch": 0.28774289985052315,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.820160687593423e-05,
+      "loss": 1.3261,
+      "step": 4235
+    },
+    {
+      "epoch": 0.28808261992118495,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.8199483625492596e-05,
+      "loss": 1.3242,
+      "step": 4240
+    },
+    {
+      "epoch": 0.28842233999184674,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.819736037505096e-05,
+      "loss": 1.4308,
+      "step": 4245
+    },
+    {
+      "epoch": 0.2887620600625085,
+      "grad_norm": 5.25,
+      "learning_rate": 4.8195237124609324e-05,
+      "loss": 1.3349,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2891017801331703,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.819311387416769e-05,
+      "loss": 1.3042,
+      "step": 4255
+    },
+    {
+      "epoch": 0.289441500203832,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.819099062372605e-05,
+      "loss": 1.4178,
+      "step": 4260
+    },
+    {
+      "epoch": 0.2897812202744938,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.8188867373284416e-05,
+      "loss": 1.3886,
+      "step": 4265
+    },
+    {
+      "epoch": 0.29012094034515556,
+      "grad_norm": 5.0,
+      "learning_rate": 4.818674412284278e-05,
+      "loss": 1.2715,
+      "step": 4270
+    },
+    {
+      "epoch": 0.29046066041581736,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.8184620872401144e-05,
+      "loss": 1.3728,
+      "step": 4275
+    },
+    {
+      "epoch": 0.29080038048647916,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.818249762195951e-05,
+      "loss": 1.2917,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2911401005571409,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.818037437151787e-05,
+      "loss": 1.349,
+      "step": 4285
+    },
+    {
+      "epoch": 0.2914798206278027,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.8178251121076236e-05,
+      "loss": 1.3046,
+      "step": 4290
+    },
+    {
+      "epoch": 0.29181954069846444,
+      "grad_norm": 5.875,
+      "learning_rate": 4.81761278706346e-05,
+      "loss": 1.4068,
+      "step": 4295
+    },
+    {
+      "epoch": 0.29215926076912624,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.8174004620192964e-05,
+      "loss": 1.4445,
+      "step": 4300
+    },
+    {
+      "epoch": 0.29249898083978804,
+      "grad_norm": 5.25,
+      "learning_rate": 4.817188136975133e-05,
+      "loss": 1.3517,
+      "step": 4305
+    },
+    {
+      "epoch": 0.2928387009104498,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.816975811930969e-05,
+      "loss": 1.3864,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2931784209811116,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.8167634868868056e-05,
+      "loss": 1.3162,
+      "step": 4315
+    },
+    {
+      "epoch": 0.2935181410517733,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.816551161842642e-05,
+      "loss": 1.2557,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2938578611224351,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.8163388367984784e-05,
+      "loss": 1.317,
+      "step": 4325
+    },
+    {
+      "epoch": 0.2941975811930969,
+      "grad_norm": 5.875,
+      "learning_rate": 4.816126511754315e-05,
+      "loss": 1.3421,
+      "step": 4330
+    },
+    {
+      "epoch": 0.29453730126375866,
+      "grad_norm": 5.25,
+      "learning_rate": 4.815914186710151e-05,
+      "loss": 1.4658,
+      "step": 4335
+    },
+    {
+      "epoch": 0.29487702133442045,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8157018616659876e-05,
+      "loss": 1.3181,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2952167414050822,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.815489536621824e-05,
+      "loss": 1.2738,
+      "step": 4345
+    },
+    {
+      "epoch": 0.295556461475744,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.81527721157766e-05,
+      "loss": 1.3923,
+      "step": 4350
+    },
+    {
+      "epoch": 0.29589618154640573,
+      "grad_norm": 5.5,
+      "learning_rate": 4.815064886533497e-05,
+      "loss": 1.3211,
+      "step": 4355
+    },
+    {
+      "epoch": 0.29623590161706753,
+      "grad_norm": 5.375,
+      "learning_rate": 4.814852561489333e-05,
+      "loss": 1.269,
+      "step": 4360
+    },
+    {
+      "epoch": 0.29657562168772933,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.814640236445169e-05,
+      "loss": 1.3657,
+      "step": 4365
+    },
+    {
+      "epoch": 0.29691534175839107,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.814427911401006e-05,
+      "loss": 1.428,
+      "step": 4370
+    },
+    {
+      "epoch": 0.29725506182905287,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.8142155863568424e-05,
+      "loss": 1.2222,
+      "step": 4375
+    },
+    {
+      "epoch": 0.2975947818997146,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.814003261312678e-05,
+      "loss": 1.3278,
+      "step": 4380
+    },
+    {
+      "epoch": 0.2979345019703764,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.813790936268515e-05,
+      "loss": 1.3048,
+      "step": 4385
+    },
+    {
+      "epoch": 0.2982742220410382,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8135786112243516e-05,
+      "loss": 1.3789,
+      "step": 4390
+    },
+    {
+      "epoch": 0.29861394211169995,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.8133662861801874e-05,
+      "loss": 1.3892,
+      "step": 4395
+    },
+    {
+      "epoch": 0.29895366218236175,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8131539611360244e-05,
+      "loss": 1.444,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2992933822530235,
+      "grad_norm": 4.75,
+      "learning_rate": 4.812941636091861e-05,
+      "loss": 1.3144,
+      "step": 4405
+    },
+    {
+      "epoch": 0.2996331023236853,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.8127293110476966e-05,
+      "loss": 1.2034,
+      "step": 4410
+    },
+    {
+      "epoch": 0.2999728223943471,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.8125169860035336e-05,
+      "loss": 1.2861,
+      "step": 4415
+    },
+    {
+      "epoch": 0.3003125424650088,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.81230466095937e-05,
+      "loss": 1.3811,
+      "step": 4420
+    },
+    {
+      "epoch": 0.3006522625356706,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.812092335915206e-05,
+      "loss": 1.4439,
+      "step": 4425
+    },
+    {
+      "epoch": 0.30099198260633236,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.811880010871043e-05,
+      "loss": 1.4964,
+      "step": 4430
+    },
+    {
+      "epoch": 0.30133170267699416,
+      "grad_norm": 5.0,
+      "learning_rate": 4.8116676858268786e-05,
+      "loss": 1.3786,
+      "step": 4435
+    },
+    {
+      "epoch": 0.3016714227476559,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.811455360782715e-05,
+      "loss": 1.3227,
+      "step": 4440
+    },
+    {
+      "epoch": 0.3020111428183177,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.811243035738552e-05,
+      "loss": 1.3853,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3023508628889795,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.811030710694388e-05,
+      "loss": 1.2724,
+      "step": 4450
+    },
+    {
+      "epoch": 0.30269058295964124,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.810818385650224e-05,
+      "loss": 1.2652,
+      "step": 4455
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 5.75,
+      "learning_rate": 4.810606060606061e-05,
+      "loss": 1.2974,
+      "step": 4460
+    },
+    {
+      "epoch": 0.3033700231009648,
+      "grad_norm": 6.375,
+      "learning_rate": 4.810393735561897e-05,
+      "loss": 1.3347,
+      "step": 4465
+    },
+    {
+      "epoch": 0.3037097431716266,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.8101814105177334e-05,
+      "loss": 1.3663,
+      "step": 4470
+    },
+    {
+      "epoch": 0.3040494632422884,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.8099690854735705e-05,
+      "loss": 1.4181,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3043891833129501,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.809756760429406e-05,
+      "loss": 1.343,
+      "step": 4480
+    },
+    {
+      "epoch": 0.3047289033836119,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.8095444353852426e-05,
+      "loss": 1.3758,
+      "step": 4485
+    },
+    {
+      "epoch": 0.30506862345427366,
+      "grad_norm": 5.75,
+      "learning_rate": 4.8093321103410797e-05,
+      "loss": 1.4673,
+      "step": 4490
+    },
+    {
+      "epoch": 0.30540834352493546,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.8091197852969154e-05,
+      "loss": 1.2665,
+      "step": 4495
+    },
+    {
+      "epoch": 0.30574806359559725,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.808907460252752e-05,
+      "loss": 1.3895,
+      "step": 4500
+    },
+    {
+      "epoch": 0.306087783666259,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.808695135208588e-05,
+      "loss": 1.2689,
+      "step": 4505
+    },
+    {
+      "epoch": 0.3064275037369208,
+      "grad_norm": 5.375,
+      "learning_rate": 4.8084828101644246e-05,
+      "loss": 1.3794,
+      "step": 4510
+    },
+    {
+      "epoch": 0.30676722380758253,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.808270485120261e-05,
+      "loss": 1.3689,
+      "step": 4515
+    },
+    {
+      "epoch": 0.30710694387824433,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.8080581600760974e-05,
+      "loss": 1.3005,
+      "step": 4520
+    },
+    {
+      "epoch": 0.3074466639489061,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.807845835031934e-05,
+      "loss": 1.3087,
+      "step": 4525
+    },
+    {
+      "epoch": 0.30778638401956787,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.80763350998777e-05,
+      "loss": 1.3812,
+      "step": 4530
+    },
+    {
+      "epoch": 0.30812610409022967,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.8074211849436066e-05,
+      "loss": 1.3208,
+      "step": 4535
+    },
+    {
+      "epoch": 0.3084658241608914,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.807208859899443e-05,
+      "loss": 1.3439,
+      "step": 4540
+    },
+    {
+      "epoch": 0.3088055442315532,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8069965348552794e-05,
+      "loss": 1.3258,
+      "step": 4545
+    },
+    {
+      "epoch": 0.30914526430221495,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.806784209811116e-05,
+      "loss": 1.3654,
+      "step": 4550
+    },
+    {
+      "epoch": 0.30948498437287675,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.806571884766952e-05,
+      "loss": 1.3196,
+      "step": 4555
+    },
+    {
+      "epoch": 0.30982470444353855,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8063595597227886e-05,
+      "loss": 1.3445,
+      "step": 4560
+    },
+    {
+      "epoch": 0.3101644245142003,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.806147234678625e-05,
+      "loss": 1.3775,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3105041445848621,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.8059349096344614e-05,
+      "loss": 1.3479,
+      "step": 4570
+    },
+    {
+      "epoch": 0.3108438646555238,
+      "grad_norm": 5.375,
+      "learning_rate": 4.805722584590298e-05,
+      "loss": 1.2369,
+      "step": 4575
+    },
+    {
+      "epoch": 0.3111835847261856,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.805510259546134e-05,
+      "loss": 1.4058,
+      "step": 4580
+    },
+    {
+      "epoch": 0.3115233047968474,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.8052979345019706e-05,
+      "loss": 1.4298,
+      "step": 4585
+    },
+    {
+      "epoch": 0.31186302486750916,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.805085609457807e-05,
+      "loss": 1.2817,
+      "step": 4590
+    },
+    {
+      "epoch": 0.31220274493817096,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.8048732844136434e-05,
+      "loss": 1.3549,
+      "step": 4595
+    },
+    {
+      "epoch": 0.3125424650088327,
+      "grad_norm": 5.75,
+      "learning_rate": 4.80466095936948e-05,
+      "loss": 1.363,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3128821850794945,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.804448634325316e-05,
+      "loss": 1.3491,
+      "step": 4605
+    },
+    {
+      "epoch": 0.3132219051501563,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.8042363092811526e-05,
+      "loss": 1.3296,
+      "step": 4610
+    },
+    {
+      "epoch": 0.31356162522081804,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.804023984236989e-05,
+      "loss": 1.3464,
+      "step": 4615
+    },
+    {
+      "epoch": 0.31390134529147984,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.8038116591928254e-05,
+      "loss": 1.2998,
+      "step": 4620
+    },
+    {
+      "epoch": 0.3142410653621416,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.803599334148662e-05,
+      "loss": 1.2691,
+      "step": 4625
+    },
+    {
+      "epoch": 0.3145807854328034,
+      "grad_norm": 5.625,
+      "learning_rate": 4.803387009104498e-05,
+      "loss": 1.3044,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3149205055034651,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.8031746840603346e-05,
+      "loss": 1.3089,
+      "step": 4635
+    },
+    {
+      "epoch": 0.3152602255741269,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.802962359016171e-05,
+      "loss": 1.304,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3155999456447887,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.8027500339720074e-05,
+      "loss": 1.3863,
+      "step": 4645
+    },
+    {
+      "epoch": 0.31593966571545046,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.802537708927844e-05,
+      "loss": 1.2616,
+      "step": 4650
+    },
+    {
+      "epoch": 0.31627938578611225,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.80232538388368e-05,
+      "loss": 1.3552,
+      "step": 4655
+    },
+    {
+      "epoch": 0.316619105856774,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.8021130588395166e-05,
+      "loss": 1.337,
+      "step": 4660
+    },
+    {
+      "epoch": 0.3169588259274358,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.801900733795353e-05,
+      "loss": 1.3205,
+      "step": 4665
+    },
+    {
+      "epoch": 0.3172985459980976,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.8016884087511894e-05,
+      "loss": 1.3292,
+      "step": 4670
+    },
+    {
+      "epoch": 0.31763826606875933,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.801476083707026e-05,
+      "loss": 1.521,
+      "step": 4675
+    },
+    {
+      "epoch": 0.31797798613942113,
+      "grad_norm": 5.375,
+      "learning_rate": 4.801263758662862e-05,
+      "loss": 1.243,
+      "step": 4680
+    },
+    {
+      "epoch": 0.3183177062100829,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.8010514336186986e-05,
+      "loss": 1.4572,
+      "step": 4685
+    },
+    {
+      "epoch": 0.31865742628074467,
+      "grad_norm": 5.0,
+      "learning_rate": 4.800839108574535e-05,
+      "loss": 1.3746,
+      "step": 4690
+    },
+    {
+      "epoch": 0.31899714635140647,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.8006267835303714e-05,
+      "loss": 1.34,
+      "step": 4695
+    },
+    {
+      "epoch": 0.3193368664220682,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.800414458486208e-05,
+      "loss": 1.2935,
+      "step": 4700
+    },
+    {
+      "epoch": 0.31967658649273,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.8002021334420435e-05,
+      "loss": 1.3257,
+      "step": 4705
+    },
+    {
+      "epoch": 0.32001630656339175,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.7999898083978806e-05,
+      "loss": 1.2629,
+      "step": 4710
+    },
+    {
+      "epoch": 0.32035602663405355,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.799777483353717e-05,
+      "loss": 1.3403,
+      "step": 4715
+    },
+    {
+      "epoch": 0.3206957467047153,
+      "grad_norm": 5.625,
+      "learning_rate": 4.799565158309553e-05,
+      "loss": 1.3079,
+      "step": 4720
+    },
+    {
+      "epoch": 0.3210354667753771,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.79935283326539e-05,
+      "loss": 1.3562,
+      "step": 4725
+    },
+    {
+      "epoch": 0.3213751868460389,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.799140508221226e-05,
+      "loss": 1.3501,
+      "step": 4730
+    },
+    {
+      "epoch": 0.3217149069167006,
+      "grad_norm": 5.5,
+      "learning_rate": 4.798928183177062e-05,
+      "loss": 1.2974,
+      "step": 4735
+    },
+    {
+      "epoch": 0.3220546269873624,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.798715858132899e-05,
+      "loss": 1.3271,
+      "step": 4740
+    },
+    {
+      "epoch": 0.32239434705802417,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.7985035330887354e-05,
+      "loss": 1.3367,
+      "step": 4745
+    },
+    {
+      "epoch": 0.32273406712868596,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.798291208044571e-05,
+      "loss": 1.3409,
+      "step": 4750
+    },
+    {
+      "epoch": 0.32307378719934776,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.798078883000408e-05,
+      "loss": 1.3708,
+      "step": 4755
+    },
+    {
+      "epoch": 0.3234135072700095,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.7978665579562446e-05,
+      "loss": 1.3216,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3237532273406713,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.79765423291208e-05,
+      "loss": 1.3877,
+      "step": 4765
+    },
+    {
+      "epoch": 0.32409294741133304,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7974419078679174e-05,
+      "loss": 1.3473,
+      "step": 4770
+    },
+    {
+      "epoch": 0.32443266748199484,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.797229582823753e-05,
+      "loss": 1.2974,
+      "step": 4775
+    },
+    {
+      "epoch": 0.32477238755265664,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7970172577795895e-05,
+      "loss": 1.4199,
+      "step": 4780
+    },
+    {
+      "epoch": 0.3251121076233184,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.7968049327354266e-05,
+      "loss": 1.2832,
+      "step": 4785
+    },
+    {
+      "epoch": 0.3254518276939802,
+      "grad_norm": 5.375,
+      "learning_rate": 4.796592607691262e-05,
+      "loss": 1.3179,
+      "step": 4790
+    },
+    {
+      "epoch": 0.3257915477646419,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.796380282647099e-05,
+      "loss": 1.3088,
+      "step": 4795
+    },
+    {
+      "epoch": 0.3261312678353037,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.796167957602936e-05,
+      "loss": 1.3761,
+      "step": 4800
+    },
+    {
+      "epoch": 0.32647098790596546,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.7959556325587715e-05,
+      "loss": 1.2301,
+      "step": 4805
+    },
+    {
+      "epoch": 0.32681070797662726,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.795743307514608e-05,
+      "loss": 1.371,
+      "step": 4810
+    },
+    {
+      "epoch": 0.32715042804728905,
+      "grad_norm": 5.75,
+      "learning_rate": 4.795530982470445e-05,
+      "loss": 1.3202,
+      "step": 4815
+    },
+    {
+      "epoch": 0.3274901481179508,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.795318657426281e-05,
+      "loss": 1.354,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3278298681886126,
+      "grad_norm": 4.75,
+      "learning_rate": 4.795106332382117e-05,
+      "loss": 1.4265,
+      "step": 4825
+    },
+    {
+      "epoch": 0.32816958825927434,
+      "grad_norm": 5.625,
+      "learning_rate": 4.794894007337954e-05,
+      "loss": 1.2297,
+      "step": 4830
+    },
+    {
+      "epoch": 0.32850930832993613,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.79468168229379e-05,
+      "loss": 1.2867,
+      "step": 4835
+    },
+    {
+      "epoch": 0.32884902840059793,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.794469357249626e-05,
+      "loss": 1.3344,
+      "step": 4840
+    },
+    {
+      "epoch": 0.3291887484712597,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.7942570322054634e-05,
+      "loss": 1.3525,
+      "step": 4845
+    },
+    {
+      "epoch": 0.32952846854192147,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.794044707161299e-05,
+      "loss": 1.305,
+      "step": 4850
+    },
+    {
+      "epoch": 0.3298681886125832,
+      "grad_norm": 6.125,
+      "learning_rate": 4.7938323821171355e-05,
+      "loss": 1.2196,
+      "step": 4855
+    },
+    {
+      "epoch": 0.330207908683245,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.793620057072972e-05,
+      "loss": 1.302,
+      "step": 4860
+    },
+    {
+      "epoch": 0.3305476287539068,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.793407732028808e-05,
+      "loss": 1.3325,
+      "step": 4865
+    },
+    {
+      "epoch": 0.33088734882456855,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.793195406984645e-05,
+      "loss": 1.4289,
+      "step": 4870
+    },
+    {
+      "epoch": 0.33122706889523035,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.792983081940481e-05,
+      "loss": 1.2599,
+      "step": 4875
+    },
+    {
+      "epoch": 0.3315667889658921,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.7927707568963175e-05,
+      "loss": 1.2664,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3319065090365539,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.792558431852154e-05,
+      "loss": 1.4152,
+      "step": 4885
+    },
+    {
+      "epoch": 0.33224622910721563,
+      "grad_norm": 5.625,
+      "learning_rate": 4.79234610680799e-05,
+      "loss": 1.3211,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3325859491778774,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.792133781763827e-05,
+      "loss": 1.2759,
+      "step": 4895
+    },
+    {
+      "epoch": 0.3329256692485392,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.791921456719663e-05,
+      "loss": 1.4358,
+      "step": 4900
+    },
+    {
+      "epoch": 0.33326538931920097,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.7917091316754995e-05,
+      "loss": 1.2857,
+      "step": 4905
+    },
+    {
+      "epoch": 0.33360510938986276,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.791496806631336e-05,
+      "loss": 1.3304,
+      "step": 4910
+    },
+    {
+      "epoch": 0.3339448294605245,
+      "grad_norm": 5.75,
+      "learning_rate": 4.791284481587172e-05,
+      "loss": 1.4421,
+      "step": 4915
+    },
+    {
+      "epoch": 0.3342845495311863,
+      "grad_norm": 5.125,
+      "learning_rate": 4.791072156543009e-05,
+      "loss": 1.2826,
+      "step": 4920
+    },
+    {
+      "epoch": 0.3346242696018481,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.790859831498845e-05,
+      "loss": 1.3931,
+      "step": 4925
+    },
+    {
+      "epoch": 0.33496398967250984,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7906475064546815e-05,
+      "loss": 1.5172,
+      "step": 4930
+    },
+    {
+      "epoch": 0.33530370974317164,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.790435181410518e-05,
+      "loss": 1.3625,
+      "step": 4935
+    },
+    {
+      "epoch": 0.3356434298138334,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.790222856366354e-05,
+      "loss": 1.4408,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3359831498844952,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.790010531322191e-05,
+      "loss": 1.4181,
+      "step": 4945
+    },
+    {
+      "epoch": 0.336322869955157,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.789798206278027e-05,
+      "loss": 1.3327,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3366625900258187,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7895858812338635e-05,
+      "loss": 1.3977,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3370023100964805,
+      "grad_norm": 5.0,
+      "learning_rate": 4.7893735561897e-05,
+      "loss": 1.2197,
+      "step": 4960
+    },
+    {
+      "epoch": 0.33734203016714226,
+      "grad_norm": 5.125,
+      "learning_rate": 4.789161231145536e-05,
+      "loss": 1.3645,
+      "step": 4965
+    },
+    {
+      "epoch": 0.33768175023780406,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.788948906101373e-05,
+      "loss": 1.3305,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3380214703084658,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.788736581057209e-05,
+      "loss": 1.4352,
+      "step": 4975
+    },
+    {
+      "epoch": 0.3383611903791276,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.7885242560130455e-05,
+      "loss": 1.3183,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3387009104497894,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.788311930968882e-05,
+      "loss": 1.3657,
+      "step": 4985
+    },
+    {
+      "epoch": 0.33904063052045114,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.788099605924718e-05,
+      "loss": 1.2962,
+      "step": 4990
+    },
+    {
+      "epoch": 0.33938035059111293,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.787887280880555e-05,
+      "loss": 1.4286,
+      "step": 4995
+    },
+    {
+      "epoch": 0.3397200706617747,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.787674955836391e-05,
+      "loss": 1.3331,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3400597907324365,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.7874626307922275e-05,
+      "loss": 1.3514,
+      "step": 5005
+    },
+    {
+      "epoch": 0.34039951080309827,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.787250305748064e-05,
+      "loss": 1.2857,
+      "step": 5010
+    },
+    {
+      "epoch": 0.34073923087376,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7870379807039003e-05,
+      "loss": 1.3099,
+      "step": 5015
+    },
+    {
+      "epoch": 0.3410789509444218,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.786825655659737e-05,
+      "loss": 1.2544,
+      "step": 5020
+    },
+    {
+      "epoch": 0.34141867101508355,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.786613330615573e-05,
+      "loss": 1.2858,
+      "step": 5025
+    },
+    {
+      "epoch": 0.34175839108574535,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.7864010055714095e-05,
+      "loss": 1.3162,
+      "step": 5030
+    },
+    {
+      "epoch": 0.34209811115640715,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.786188680527246e-05,
+      "loss": 1.3,
+      "step": 5035
+    },
+    {
+      "epoch": 0.3424378312270689,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7859763554830823e-05,
+      "loss": 1.3643,
+      "step": 5040
+    },
+    {
+      "epoch": 0.3427775512977307,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.785764030438919e-05,
+      "loss": 1.3629,
+      "step": 5045
+    },
+    {
+      "epoch": 0.34311727136839243,
+      "grad_norm": 5.375,
+      "learning_rate": 4.785551705394755e-05,
+      "loss": 1.2546,
+      "step": 5050
+    },
+    {
+      "epoch": 0.3434569914390542,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.7853393803505915e-05,
+      "loss": 1.4042,
+      "step": 5055
+    },
+    {
+      "epoch": 0.34379671150971597,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.785127055306427e-05,
+      "loss": 1.197,
+      "step": 5060
+    },
+    {
+      "epoch": 0.34413643158037777,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.7849147302622643e-05,
+      "loss": 1.1813,
+      "step": 5065
+    },
+    {
+      "epoch": 0.34447615165103956,
+      "grad_norm": 5.5,
+      "learning_rate": 4.784702405218101e-05,
+      "loss": 1.4065,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3448158717217013,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.7844900801739365e-05,
+      "loss": 1.2553,
+      "step": 5075
+    },
+    {
+      "epoch": 0.3451555917923631,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.7842777551297735e-05,
+      "loss": 1.3585,
+      "step": 5080
+    },
+    {
+      "epoch": 0.34549531186302485,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.78406543008561e-05,
+      "loss": 1.2633,
+      "step": 5085
+    },
+    {
+      "epoch": 0.34583503193368664,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.783853105041446e-05,
+      "loss": 1.3164,
+      "step": 5090
+    },
+    {
+      "epoch": 0.34617475200434844,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.783640779997283e-05,
+      "loss": 1.3777,
+      "step": 5095
+    },
+    {
+      "epoch": 0.3465144720750102,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.783428454953119e-05,
+      "loss": 1.3047,
+      "step": 5100
+    },
+    {
+      "epoch": 0.346854192145672,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.783216129908955e-05,
+      "loss": 1.3226,
+      "step": 5105
+    },
+    {
+      "epoch": 0.3471939122163337,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.783003804864792e-05,
+      "loss": 1.2537,
+      "step": 5110
+    },
+    {
+      "epoch": 0.3475336322869955,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.7827914798206283e-05,
+      "loss": 1.297,
+      "step": 5115
+    },
+    {
+      "epoch": 0.3478733523576573,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.782579154776464e-05,
+      "loss": 1.4026,
+      "step": 5120
+    },
+    {
+      "epoch": 0.34821307242831906,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.782366829732301e-05,
+      "loss": 1.4684,
+      "step": 5125
+    },
+    {
+      "epoch": 0.34855279249898086,
+      "grad_norm": 5.125,
+      "learning_rate": 4.782154504688137e-05,
+      "loss": 1.3442,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3488925125696426,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.781942179643973e-05,
+      "loss": 1.3205,
+      "step": 5135
+    },
+    {
+      "epoch": 0.3492322326403044,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.7817298545998103e-05,
+      "loss": 1.355,
+      "step": 5140
+    },
+    {
+      "epoch": 0.34957195271096614,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.781517529555646e-05,
+      "loss": 1.2925,
+      "step": 5145
+    },
+    {
+      "epoch": 0.34991167278162794,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.7813052045114825e-05,
+      "loss": 1.2692,
+      "step": 5150
+    },
+    {
+      "epoch": 0.35025139285228973,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7810928794673196e-05,
+      "loss": 1.351,
+      "step": 5155
+    },
+    {
+      "epoch": 0.3505911129229515,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.780880554423155e-05,
+      "loss": 1.3462,
+      "step": 5160
+    },
+    {
+      "epoch": 0.3509308329936133,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.780668229378992e-05,
+      "loss": 1.3128,
+      "step": 5165
+    },
+    {
+      "epoch": 0.351270553064275,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.780455904334829e-05,
+      "loss": 1.3615,
+      "step": 5170
+    },
+    {
+      "epoch": 0.3516102731349368,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.7802435792906645e-05,
+      "loss": 1.1858,
+      "step": 5175
+    },
+    {
+      "epoch": 0.3519499932055986,
+      "grad_norm": 5.125,
+      "learning_rate": 4.780031254246501e-05,
+      "loss": 1.3019,
+      "step": 5180
+    },
+    {
+      "epoch": 0.35228971327626035,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.779818929202338e-05,
+      "loss": 1.3025,
+      "step": 5185
+    },
+    {
+      "epoch": 0.35262943334692215,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.779606604158174e-05,
+      "loss": 1.4087,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3529691534175839,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.77939427911401e-05,
+      "loss": 1.2449,
+      "step": 5195
+    },
+    {
+      "epoch": 0.3533088734882457,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.7791819540698465e-05,
+      "loss": 1.3617,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3536485935589075,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.778969629025683e-05,
+      "loss": 1.3328,
+      "step": 5205
+    },
+    {
+      "epoch": 0.35398831362956923,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.778757303981519e-05,
+      "loss": 1.2707,
+      "step": 5210
+    },
+    {
+      "epoch": 0.354328033700231,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.778544978937356e-05,
+      "loss": 1.3296,
+      "step": 5215
+    },
+    {
+      "epoch": 0.35466775377089277,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.778332653893192e-05,
+      "loss": 1.4546,
+      "step": 5220
+    },
+    {
+      "epoch": 0.35500747384155457,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.7781203288490285e-05,
+      "loss": 1.3147,
+      "step": 5225
+    },
+    {
+      "epoch": 0.3553471939122163,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.777908003804865e-05,
+      "loss": 1.328,
+      "step": 5230
+    },
+    {
+      "epoch": 0.3556869139828781,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.777695678760701e-05,
+      "loss": 1.3234,
+      "step": 5235
+    },
+    {
+      "epoch": 0.3560266340535399,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.777483353716538e-05,
+      "loss": 1.309,
+      "step": 5240
+    },
+    {
+      "epoch": 0.35636635412420165,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.777271028672374e-05,
+      "loss": 1.3218,
+      "step": 5245
+    },
+    {
+      "epoch": 0.35670607419486344,
+      "grad_norm": 6.25,
+      "learning_rate": 4.7770587036282105e-05,
+      "loss": 1.4306,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3570457942655252,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.776846378584047e-05,
+      "loss": 1.2789,
+      "step": 5255
+    },
+    {
+      "epoch": 0.357385514336187,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.776634053539883e-05,
+      "loss": 1.3324,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3577252344068488,
+      "grad_norm": 5.5,
+      "learning_rate": 4.77642172849572e-05,
+      "loss": 1.1933,
+      "step": 5265
+    },
+    {
+      "epoch": 0.3580649544775105,
+      "grad_norm": 5.875,
+      "learning_rate": 4.776209403451556e-05,
+      "loss": 1.3767,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3584046745481723,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7759970784073925e-05,
+      "loss": 1.4372,
+      "step": 5275
+    },
+    {
+      "epoch": 0.35874439461883406,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.775784753363229e-05,
+      "loss": 1.3571,
+      "step": 5280
+    },
+    {
+      "epoch": 0.35908411468949586,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.775572428319065e-05,
+      "loss": 1.232,
+      "step": 5285
+    },
+    {
+      "epoch": 0.35942383476015766,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.775360103274902e-05,
+      "loss": 1.3018,
+      "step": 5290
+    },
+    {
+      "epoch": 0.3597635548308194,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.775147778230738e-05,
+      "loss": 1.3713,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3601032749014812,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.7749354531865745e-05,
+      "loss": 1.3461,
+      "step": 5300
+    },
+    {
+      "epoch": 0.36044299497214294,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.774723128142411e-05,
+      "loss": 1.3825,
+      "step": 5305
+    },
+    {
+      "epoch": 0.36078271504280474,
+      "grad_norm": 6.125,
+      "learning_rate": 4.774510803098247e-05,
+      "loss": 1.2815,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3611224351134665,
+      "grad_norm": 5.25,
+      "learning_rate": 4.774298478054084e-05,
+      "loss": 1.3699,
+      "step": 5315
+    },
+    {
+      "epoch": 0.3614621551841283,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.77408615300992e-05,
+      "loss": 1.1994,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3618018752547901,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.7738738279657565e-05,
+      "loss": 1.396,
+      "step": 5325
+    },
+    {
+      "epoch": 0.3621415953254518,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.773661502921593e-05,
+      "loss": 1.2344,
+      "step": 5330
+    },
+    {
+      "epoch": 0.3624813153961136,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.773449177877429e-05,
+      "loss": 1.3707,
+      "step": 5335
+    },
+    {
+      "epoch": 0.36282103546677535,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.773236852833266e-05,
+      "loss": 1.4147,
+      "step": 5340
+    },
+    {
+      "epoch": 0.36316075553743715,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.773024527789102e-05,
+      "loss": 1.3274,
+      "step": 5345
+    },
+    {
+      "epoch": 0.36350047560809895,
+      "grad_norm": 5.625,
+      "learning_rate": 4.7728122027449385e-05,
+      "loss": 1.3148,
+      "step": 5350
+    },
+    {
+      "epoch": 0.3638401956787607,
+      "grad_norm": 5.125,
+      "learning_rate": 4.772599877700775e-05,
+      "loss": 1.3469,
+      "step": 5355
+    },
+    {
+      "epoch": 0.3641799157494225,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.772387552656611e-05,
+      "loss": 1.199,
+      "step": 5360
+    },
+    {
+      "epoch": 0.36451963582008423,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.772175227612448e-05,
+      "loss": 1.3049,
+      "step": 5365
+    },
+    {
+      "epoch": 0.36485935589074603,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.771962902568284e-05,
+      "loss": 1.2786,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3651990759614078,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.7717505775241205e-05,
+      "loss": 1.3795,
+      "step": 5375
+    },
+    {
+      "epoch": 0.36553879603206957,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.771538252479957e-05,
+      "loss": 1.3225,
+      "step": 5380
+    },
+    {
+      "epoch": 0.36587851610273137,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.771325927435793e-05,
+      "loss": 1.2573,
+      "step": 5385
+    },
+    {
+      "epoch": 0.3662182361733931,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.77111360239163e-05,
+      "loss": 1.2764,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3665579562440549,
+      "grad_norm": 4.75,
+      "learning_rate": 4.770901277347466e-05,
+      "loss": 1.3044,
+      "step": 5395
+    },
+    {
+      "epoch": 0.36689767631471665,
+      "grad_norm": 5.25,
+      "learning_rate": 4.770688952303302e-05,
+      "loss": 1.3625,
+      "step": 5400
+    },
+    {
+      "epoch": 0.36723739638537845,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.770476627259139e-05,
+      "loss": 1.2714,
+      "step": 5405
+    },
+    {
+      "epoch": 0.36757711645604024,
+      "grad_norm": 5.875,
+      "learning_rate": 4.770264302214975e-05,
+      "loss": 1.3115,
+      "step": 5410
+    },
+    {
+      "epoch": 0.367916836526702,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.770051977170811e-05,
+      "loss": 1.3612,
+      "step": 5415
+    },
+    {
+      "epoch": 0.3682565565973638,
+      "grad_norm": 5.5,
+      "learning_rate": 4.769839652126648e-05,
+      "loss": 1.2656,
+      "step": 5420
+    },
+    {
+      "epoch": 0.3685962766680255,
+      "grad_norm": 5.75,
+      "learning_rate": 4.7696273270824845e-05,
+      "loss": 1.3505,
+      "step": 5425
+    },
+    {
+      "epoch": 0.3689359967386873,
+      "grad_norm": 5.375,
+      "learning_rate": 4.76941500203832e-05,
+      "loss": 1.3165,
+      "step": 5430
+    },
+    {
+      "epoch": 0.3692757168093491,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.769202676994157e-05,
+      "loss": 1.2563,
+      "step": 5435
+    },
+    {
+      "epoch": 0.36961543688001086,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.768990351949994e-05,
+      "loss": 1.33,
+      "step": 5440
+    },
+    {
+      "epoch": 0.36995515695067266,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7687780269058294e-05,
+      "loss": 1.3941,
+      "step": 5445
+    },
+    {
+      "epoch": 0.3702948770213344,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.7685657018616665e-05,
+      "loss": 1.3202,
+      "step": 5450
+    },
+    {
+      "epoch": 0.3706345970919962,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.768353376817503e-05,
+      "loss": 1.3907,
+      "step": 5455
+    },
+    {
+      "epoch": 0.370974317162658,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.7681410517733386e-05,
+      "loss": 1.3693,
+      "step": 5460
+    },
+    {
+      "epoch": 0.37131403723331974,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.767928726729176e-05,
+      "loss": 1.3263,
+      "step": 5465
+    },
+    {
+      "epoch": 0.37165375730398154,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.767716401685012e-05,
+      "loss": 1.3522,
+      "step": 5470
+    },
+    {
+      "epoch": 0.3719934773746433,
+      "grad_norm": 4.875,
+      "learning_rate": 4.767504076640848e-05,
+      "loss": 1.2569,
+      "step": 5475
+    },
+    {
+      "epoch": 0.3723331974453051,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.767291751596685e-05,
+      "loss": 1.3007,
+      "step": 5480
+    },
+    {
+      "epoch": 0.3726729175159668,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.7670794265525206e-05,
+      "loss": 1.3692,
+      "step": 5485
+    },
+    {
+      "epoch": 0.3730126375866286,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.766867101508357e-05,
+      "loss": 1.2842,
+      "step": 5490
+    },
+    {
+      "epoch": 0.3733523576572904,
+      "grad_norm": 5.375,
+      "learning_rate": 4.766654776464194e-05,
+      "loss": 1.3347,
+      "step": 5495
+    },
+    {
+      "epoch": 0.37369207772795215,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.76644245142003e-05,
+      "loss": 1.2912,
+      "step": 5500
+    },
+    {
+      "epoch": 0.37403179779861395,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.766230126375866e-05,
+      "loss": 1.3923,
+      "step": 5505
+    },
+    {
+      "epoch": 0.3743715178692757,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.766017801331703e-05,
+      "loss": 1.259,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3747112379399375,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.765805476287539e-05,
+      "loss": 1.4049,
+      "step": 5515
+    },
+    {
+      "epoch": 0.3750509580105993,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.7655931512433754e-05,
+      "loss": 1.2978,
+      "step": 5520
+    },
+    {
+      "epoch": 0.37539067808126103,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.7653808261992125e-05,
+      "loss": 1.3069,
+      "step": 5525
+    },
+    {
+      "epoch": 0.37573039815192283,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.765168501155048e-05,
+      "loss": 1.4116,
+      "step": 5530
+    },
+    {
+      "epoch": 0.37607011822258457,
+      "grad_norm": 5.5,
+      "learning_rate": 4.7649561761108846e-05,
+      "loss": 1.4074,
+      "step": 5535
+    },
+    {
+      "epoch": 0.37640983829324637,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.764743851066722e-05,
+      "loss": 1.257,
+      "step": 5540
+    },
+    {
+      "epoch": 0.37674955836390817,
+      "grad_norm": 5.0,
+      "learning_rate": 4.7645315260225574e-05,
+      "loss": 1.3712,
+      "step": 5545
+    },
+    {
+      "epoch": 0.3770892784345699,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.764319200978394e-05,
+      "loss": 1.3443,
+      "step": 5550
+    },
+    {
+      "epoch": 0.3774289985052317,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.76410687593423e-05,
+      "loss": 1.3378,
+      "step": 5555
+    },
+    {
+      "epoch": 0.37776871857589345,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.7638945508900666e-05,
+      "loss": 1.2967,
+      "step": 5560
+    },
+    {
+      "epoch": 0.37810843864655524,
+      "grad_norm": 5.125,
+      "learning_rate": 4.763682225845903e-05,
+      "loss": 1.2376,
+      "step": 5565
+    },
+    {
+      "epoch": 0.378448158717217,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7634699008017394e-05,
+      "loss": 1.4123,
+      "step": 5570
+    },
+    {
+      "epoch": 0.3787878787878788,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.763257575757576e-05,
+      "loss": 1.4059,
+      "step": 5575
+    },
+    {
+      "epoch": 0.3791275988585406,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.763045250713412e-05,
+      "loss": 1.371,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3794673189292023,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.7628329256692486e-05,
+      "loss": 1.3756,
+      "step": 5585
+    },
+    {
+      "epoch": 0.3798070389998641,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.762620600625085e-05,
+      "loss": 1.2893,
+      "step": 5590
+    },
+    {
+      "epoch": 0.38014675907052586,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.7624082755809214e-05,
+      "loss": 1.2228,
+      "step": 5595
+    },
+    {
+      "epoch": 0.38048647914118766,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.762195950536758e-05,
+      "loss": 1.4572,
+      "step": 5600
+    },
+    {
+      "epoch": 0.38082619921184946,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.761983625492594e-05,
+      "loss": 1.3545,
+      "step": 5605
+    },
+    {
+      "epoch": 0.3811659192825112,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.7617713004484306e-05,
+      "loss": 1.3604,
+      "step": 5610
+    },
+    {
+      "epoch": 0.381505639353173,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.761558975404267e-05,
+      "loss": 1.2964,
+      "step": 5615
+    },
+    {
+      "epoch": 0.38184535942383474,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.7613466503601034e-05,
+      "loss": 1.3764,
+      "step": 5620
+    },
+    {
+      "epoch": 0.38218507949449654,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.76113432531594e-05,
+      "loss": 1.227,
+      "step": 5625
+    },
+    {
+      "epoch": 0.38252479956515834,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.760922000271776e-05,
+      "loss": 1.3618,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3828645196358201,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.7607096752276126e-05,
+      "loss": 1.3072,
+      "step": 5635
+    },
+    {
+      "epoch": 0.3832042397064819,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.760497350183449e-05,
+      "loss": 1.3941,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3835439597771436,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.7602850251392854e-05,
+      "loss": 1.3616,
+      "step": 5645
+    },
+    {
+      "epoch": 0.3838836798478054,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.760072700095122e-05,
+      "loss": 1.2938,
+      "step": 5650
+    },
+    {
+      "epoch": 0.38422339991846716,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.759860375050958e-05,
+      "loss": 1.2371,
+      "step": 5655
+    },
+    {
+      "epoch": 0.38456311998912895,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.7596480500067946e-05,
+      "loss": 1.3022,
+      "step": 5660
+    },
+    {
+      "epoch": 0.38490284005979075,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.759435724962631e-05,
+      "loss": 1.2511,
+      "step": 5665
+    },
+    {
+      "epoch": 0.3852425601304525,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.7592233999184674e-05,
+      "loss": 1.3186,
+      "step": 5670
+    },
+    {
+      "epoch": 0.3855822802011143,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.759011074874304e-05,
+      "loss": 1.3305,
+      "step": 5675
+    },
+    {
+      "epoch": 0.38592200027177603,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.75879874983014e-05,
+      "loss": 1.2506,
+      "step": 5680
+    },
+    {
+      "epoch": 0.38626172034243783,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.7585864247859766e-05,
+      "loss": 1.3068,
+      "step": 5685
+    },
+    {
+      "epoch": 0.38660144041309963,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.758374099741813e-05,
+      "loss": 1.4278,
+      "step": 5690
+    },
+    {
+      "epoch": 0.38694116048376137,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.7581617746976494e-05,
+      "loss": 1.2985,
+      "step": 5695
+    },
+    {
+      "epoch": 0.38728088055442317,
+      "grad_norm": 6.25,
+      "learning_rate": 4.757949449653486e-05,
+      "loss": 1.2665,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3876206006250849,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.757737124609322e-05,
+      "loss": 1.3261,
+      "step": 5705
+    },
+    {
+      "epoch": 0.3879603206957467,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.7575247995651586e-05,
+      "loss": 1.2954,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3883000407664085,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.757312474520995e-05,
+      "loss": 1.3123,
+      "step": 5715
+    },
+    {
+      "epoch": 0.38863976083707025,
+      "grad_norm": 5.875,
+      "learning_rate": 4.7571001494768314e-05,
+      "loss": 1.3556,
+      "step": 5720
+    },
+    {
+      "epoch": 0.38897948090773204,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.756887824432668e-05,
+      "loss": 1.3254,
+      "step": 5725
+    },
+    {
+      "epoch": 0.3893192009783938,
+      "grad_norm": 5.375,
+      "learning_rate": 4.756675499388504e-05,
+      "loss": 1.2906,
+      "step": 5730
+    },
+    {
+      "epoch": 0.3896589210490556,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.7564631743443406e-05,
+      "loss": 1.327,
+      "step": 5735
+    },
+    {
+      "epoch": 0.3899986411197173,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.756250849300177e-05,
+      "loss": 1.4019,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3903383611903791,
+      "grad_norm": 5.75,
+      "learning_rate": 4.7560385242560134e-05,
+      "loss": 1.3809,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3906780812610409,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.75582619921185e-05,
+      "loss": 1.3244,
+      "step": 5750
+    },
+    {
+      "epoch": 0.39101780133170266,
+      "grad_norm": 6.625,
+      "learning_rate": 4.7556138741676856e-05,
+      "loss": 1.3437,
+      "step": 5755
+    },
+    {
+      "epoch": 0.39135752140236446,
+      "grad_norm": 5.375,
+      "learning_rate": 4.7554015491235226e-05,
+      "loss": 1.2966,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3916972414730262,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.755189224079359e-05,
+      "loss": 1.2953,
+      "step": 5765
+    },
+    {
+      "epoch": 0.392036961543688,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.754976899035195e-05,
+      "loss": 1.2484,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3923766816143498,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.754764573991032e-05,
+      "loss": 1.3957,
+      "step": 5775
+    },
+    {
+      "epoch": 0.39271640168501154,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.754552248946868e-05,
+      "loss": 1.1887,
+      "step": 5780
+    },
+    {
+      "epoch": 0.39305612175567334,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.754339923902704e-05,
+      "loss": 1.352,
+      "step": 5785
+    },
+    {
+      "epoch": 0.3933958418263351,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.754127598858541e-05,
+      "loss": 1.2955,
+      "step": 5790
+    },
+    {
+      "epoch": 0.3937355618969969,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7539152738143774e-05,
+      "loss": 1.3416,
+      "step": 5795
+    },
+    {
+      "epoch": 0.3940752819676587,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.753702948770213e-05,
+      "loss": 1.2293,
+      "step": 5800
+    },
+    {
+      "epoch": 0.3944150020383204,
+      "grad_norm": 5.875,
+      "learning_rate": 4.75349062372605e-05,
+      "loss": 1.3553,
+      "step": 5805
+    },
+    {
+      "epoch": 0.3947547221089822,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.7532782986818866e-05,
+      "loss": 1.3859,
+      "step": 5810
+    },
+    {
+      "epoch": 0.39509444217964396,
+      "grad_norm": 5.0,
+      "learning_rate": 4.7530659736377224e-05,
+      "loss": 1.2067,
+      "step": 5815
+    },
+    {
+      "epoch": 0.39543416225030575,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.7528536485935594e-05,
+      "loss": 1.3463,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3957738823209675,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.752641323549395e-05,
+      "loss": 1.383,
+      "step": 5825
+    },
+    {
+      "epoch": 0.3961136023916293,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7524289985052316e-05,
+      "loss": 1.3269,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3964533224622911,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.7522166734610686e-05,
+      "loss": 1.3356,
+      "step": 5835
+    },
+    {
+      "epoch": 0.39679304253295283,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.7520043484169044e-05,
+      "loss": 1.3399,
+      "step": 5840
+    },
+    {
+      "epoch": 0.39713276260361463,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.751792023372741e-05,
+      "loss": 1.3909,
+      "step": 5845
+    },
+    {
+      "epoch": 0.3974724826742764,
+      "grad_norm": 5.0,
+      "learning_rate": 4.751579698328578e-05,
+      "loss": 1.2426,
+      "step": 5850
+    },
+    {
+      "epoch": 0.39781220274493817,
+      "grad_norm": 5.125,
+      "learning_rate": 4.7513673732844136e-05,
+      "loss": 1.306,
+      "step": 5855
+    },
+    {
+      "epoch": 0.39815192281559997,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.75115504824025e-05,
+      "loss": 1.3206,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3984916428862617,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.750942723196087e-05,
+      "loss": 1.3727,
+      "step": 5865
+    },
+    {
+      "epoch": 0.3988313629569235,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.750730398151923e-05,
+      "loss": 1.2958,
+      "step": 5870
+    },
+    {
+      "epoch": 0.39917108302758525,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.750518073107759e-05,
+      "loss": 1.308,
+      "step": 5875
+    },
+    {
+      "epoch": 0.39951080309824705,
+      "grad_norm": 5.375,
+      "learning_rate": 4.750305748063596e-05,
+      "loss": 1.2731,
+      "step": 5880
+    },
+    {
+      "epoch": 0.39985052316890884,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.750093423019432e-05,
+      "loss": 1.3518,
+      "step": 5885
+    },
+    {
+      "epoch": 0.4001902432395706,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.7498810979752684e-05,
+      "loss": 1.3382,
+      "step": 5890
+    },
+    {
+      "epoch": 0.4005299633102324,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.7496687729311055e-05,
+      "loss": 1.2218,
+      "step": 5895
+    },
+    {
+      "epoch": 0.4008696833808941,
+      "grad_norm": 4.875,
+      "learning_rate": 4.749456447886941e-05,
+      "loss": 1.258,
+      "step": 5900
+    },
+    {
+      "epoch": 0.4012094034515559,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.7492441228427776e-05,
+      "loss": 1.2436,
+      "step": 5905
+    },
+    {
+      "epoch": 0.40154912352221767,
+      "grad_norm": 4.875,
+      "learning_rate": 4.749031797798614e-05,
+      "loss": 1.3063,
+      "step": 5910
+    },
+    {
+      "epoch": 0.40188884359287946,
+      "grad_norm": 5.75,
+      "learning_rate": 4.7488194727544504e-05,
+      "loss": 1.3971,
+      "step": 5915
+    },
+    {
+      "epoch": 0.40222856366354126,
+      "grad_norm": 5.875,
+      "learning_rate": 4.748607147710287e-05,
+      "loss": 1.3156,
+      "step": 5920
+    },
+    {
+      "epoch": 0.402568283734203,
+      "grad_norm": 5.25,
+      "learning_rate": 4.748394822666123e-05,
+      "loss": 1.2985,
+      "step": 5925
+    },
+    {
+      "epoch": 0.4029080038048648,
+      "grad_norm": 5.375,
+      "learning_rate": 4.7481824976219596e-05,
+      "loss": 1.3203,
+      "step": 5930
+    },
+    {
+      "epoch": 0.40324772387552654,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.747970172577796e-05,
+      "loss": 1.3931,
+      "step": 5935
+    },
+    {
+      "epoch": 0.40358744394618834,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.7477578475336324e-05,
+      "loss": 1.4216,
+      "step": 5940
+    },
+    {
+      "epoch": 0.40392716401685014,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.747545522489469e-05,
+      "loss": 1.2324,
+      "step": 5945
+    },
+    {
+      "epoch": 0.4042668840875119,
+      "grad_norm": 5.625,
+      "learning_rate": 4.747333197445305e-05,
+      "loss": 1.365,
+      "step": 5950
+    },
+    {
+      "epoch": 0.4046066041581737,
+      "grad_norm": 5.25,
+      "learning_rate": 4.7471208724011416e-05,
+      "loss": 1.275,
+      "step": 5955
+    },
+    {
+      "epoch": 0.4049463242288354,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.746908547356978e-05,
+      "loss": 1.4302,
+      "step": 5960
+    },
+    {
+      "epoch": 0.4052860442994972,
+      "grad_norm": 4.875,
+      "learning_rate": 4.746696222312815e-05,
+      "loss": 1.2703,
+      "step": 5965
+    },
+    {
+      "epoch": 0.405625764370159,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.746483897268651e-05,
+      "loss": 1.3843,
+      "step": 5970
+    },
+    {
+      "epoch": 0.40596548444082076,
+      "grad_norm": 5.375,
+      "learning_rate": 4.746271572224487e-05,
+      "loss": 1.2557,
+      "step": 5975
+    },
+    {
+      "epoch": 0.40630520451148255,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.7460592471803236e-05,
+      "loss": 1.3725,
+      "step": 5980
+    },
+    {
+      "epoch": 0.4066449245821443,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.74584692213616e-05,
+      "loss": 1.2374,
+      "step": 5985
+    },
+    {
+      "epoch": 0.4069846446528061,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7456345970919964e-05,
+      "loss": 1.2199,
+      "step": 5990
+    },
+    {
+      "epoch": 0.40732436472346784,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.745422272047833e-05,
+      "loss": 1.2891,
+      "step": 5995
+    },
+    {
+      "epoch": 0.40766408479412963,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.745209947003669e-05,
+      "loss": 1.2009,
+      "step": 6000
+    },
+    {
+      "epoch": 0.40800380486479143,
+      "grad_norm": 5.125,
+      "learning_rate": 4.7449976219595056e-05,
+      "loss": 1.2556,
+      "step": 6005
+    },
+    {
+      "epoch": 0.4083435249354532,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.744785296915342e-05,
+      "loss": 1.2772,
+      "step": 6010
+    },
+    {
+      "epoch": 0.40868324500611497,
+      "grad_norm": 5.875,
+      "learning_rate": 4.7445729718711784e-05,
+      "loss": 1.3103,
+      "step": 6015
+    },
+    {
+      "epoch": 0.4090229650767767,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.744360646827015e-05,
+      "loss": 1.3751,
+      "step": 6020
+    },
+    {
+      "epoch": 0.4093626851474385,
+      "grad_norm": 5.625,
+      "learning_rate": 4.744148321782851e-05,
+      "loss": 1.2884,
+      "step": 6025
+    },
+    {
+      "epoch": 0.4097024052181003,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.7439359967386876e-05,
+      "loss": 1.2956,
+      "step": 6030
+    },
+    {
+      "epoch": 0.41004212528876205,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.743723671694524e-05,
+      "loss": 1.1667,
+      "step": 6035
+    },
+    {
+      "epoch": 0.41038184535942385,
+      "grad_norm": 5.5,
+      "learning_rate": 4.7435113466503604e-05,
+      "loss": 1.2476,
+      "step": 6040
+    },
+    {
+      "epoch": 0.4107215654300856,
+      "grad_norm": 5.5,
+      "learning_rate": 4.743299021606197e-05,
+      "loss": 1.3299,
+      "step": 6045
+    },
+    {
+      "epoch": 0.4110612855007474,
+      "grad_norm": 5.75,
+      "learning_rate": 4.743086696562033e-05,
+      "loss": 1.2662,
+      "step": 6050
+    },
+    {
+      "epoch": 0.4114010055714092,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7428743715178696e-05,
+      "loss": 1.275,
+      "step": 6055
+    },
+    {
+      "epoch": 0.4117407256420709,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.742662046473706e-05,
+      "loss": 1.2727,
+      "step": 6060
+    },
+    {
+      "epoch": 0.4120804457127327,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.7424497214295424e-05,
+      "loss": 1.4267,
+      "step": 6065
+    },
+    {
+      "epoch": 0.41242016578339447,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.742237396385379e-05,
+      "loss": 1.3747,
+      "step": 6070
+    },
+    {
+      "epoch": 0.41275988585405626,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.742025071341215e-05,
+      "loss": 1.216,
+      "step": 6075
+    },
+    {
+      "epoch": 0.413099605924718,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.7418127462970516e-05,
+      "loss": 1.294,
+      "step": 6080
+    },
+    {
+      "epoch": 0.4134393259953798,
+      "grad_norm": 6.0,
+      "learning_rate": 4.741600421252888e-05,
+      "loss": 1.3391,
+      "step": 6085
+    },
+    {
+      "epoch": 0.4137790460660416,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.7413880962087244e-05,
+      "loss": 1.3428,
+      "step": 6090
+    },
+    {
+      "epoch": 0.41411876613670334,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.741175771164561e-05,
+      "loss": 1.3131,
+      "step": 6095
+    },
+    {
+      "epoch": 0.41445848620736514,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.740963446120397e-05,
+      "loss": 1.2684,
+      "step": 6100
+    },
+    {
+      "epoch": 0.4147982062780269,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.7407511210762336e-05,
+      "loss": 1.3578,
+      "step": 6105
+    },
+    {
+      "epoch": 0.4151379263486887,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.740538796032069e-05,
+      "loss": 1.4157,
+      "step": 6110
+    },
+    {
+      "epoch": 0.4154776464193505,
+      "grad_norm": 6.0,
+      "learning_rate": 4.7403264709879064e-05,
+      "loss": 1.3943,
+      "step": 6115
+    },
+    {
+      "epoch": 0.4158173664900122,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.740114145943743e-05,
+      "loss": 1.3434,
+      "step": 6120
+    },
+    {
+      "epoch": 0.416157086560674,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7399018208995785e-05,
+      "loss": 1.3296,
+      "step": 6125
+    },
+    {
+      "epoch": 0.41649680663133576,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.7396894958554156e-05,
+      "loss": 1.4148,
+      "step": 6130
+    },
+    {
+      "epoch": 0.41683652670199756,
+      "grad_norm": 5.25,
+      "learning_rate": 4.739477170811252e-05,
+      "loss": 1.3176,
+      "step": 6135
+    },
+    {
+      "epoch": 0.41717624677265935,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.739264845767088e-05,
+      "loss": 1.272,
+      "step": 6140
+    },
+    {
+      "epoch": 0.4175159668433211,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.739052520722925e-05,
+      "loss": 1.3075,
+      "step": 6145
+    },
+    {
+      "epoch": 0.4178556869139829,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.738840195678761e-05,
+      "loss": 1.3396,
+      "step": 6150
+    },
+    {
+      "epoch": 0.41819540698464464,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.738627870634597e-05,
+      "loss": 1.2105,
+      "step": 6155
+    },
+    {
+      "epoch": 0.41853512705530643,
+      "grad_norm": 5.375,
+      "learning_rate": 4.738415545590434e-05,
+      "loss": 1.3038,
+      "step": 6160
+    },
+    {
+      "epoch": 0.4188748471259682,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7382032205462704e-05,
+      "loss": 1.3691,
+      "step": 6165
+    },
+    {
+      "epoch": 0.41921456719663,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.737990895502106e-05,
+      "loss": 1.2614,
+      "step": 6170
+    },
+    {
+      "epoch": 0.41955428726729177,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.737778570457943e-05,
+      "loss": 1.2571,
+      "step": 6175
+    },
+    {
+      "epoch": 0.4198940073379535,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.737566245413779e-05,
+      "loss": 1.3345,
+      "step": 6180
+    },
+    {
+      "epoch": 0.4202337274086153,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.737353920369615e-05,
+      "loss": 1.2878,
+      "step": 6185
+    },
+    {
+      "epoch": 0.42057344747927705,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.7371415953254524e-05,
+      "loss": 1.3641,
+      "step": 6190
+    },
+    {
+      "epoch": 0.42091316754993885,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.736929270281288e-05,
+      "loss": 1.396,
+      "step": 6195
+    },
+    {
+      "epoch": 0.42125288762060065,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7367169452371245e-05,
+      "loss": 1.2986,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4215926076912624,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.7365046201929616e-05,
+      "loss": 1.2892,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4219323277619242,
+      "grad_norm": 5.25,
+      "learning_rate": 4.736292295148797e-05,
+      "loss": 1.3369,
+      "step": 6210
+    },
+    {
+      "epoch": 0.42227204783258593,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.736079970104634e-05,
+      "loss": 1.2454,
+      "step": 6215
+    },
+    {
+      "epoch": 0.4226117679032477,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.735867645060471e-05,
+      "loss": 1.2707,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4229514879739095,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.7356553200163065e-05,
+      "loss": 1.3465,
+      "step": 6225
+    },
+    {
+      "epoch": 0.42329120804457127,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.735442994972143e-05,
+      "loss": 1.396,
+      "step": 6230
+    },
+    {
+      "epoch": 0.42363092811523306,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.73523066992798e-05,
+      "loss": 1.308,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4239706481858948,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.735018344883816e-05,
+      "loss": 1.3892,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4243103682565566,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.734806019839652e-05,
+      "loss": 1.2969,
+      "step": 6245
+    },
+    {
+      "epoch": 0.42465008832721834,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.734593694795489e-05,
+      "loss": 1.2572,
+      "step": 6250
+    },
+    {
+      "epoch": 0.42498980839788014,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.734381369751325e-05,
+      "loss": 1.2398,
+      "step": 6255
+    },
+    {
+      "epoch": 0.42532952846854194,
+      "grad_norm": 5.25,
+      "learning_rate": 4.734169044707161e-05,
+      "loss": 1.2455,
+      "step": 6260
+    },
+    {
+      "epoch": 0.4256692485392037,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.733956719662998e-05,
+      "loss": 1.2196,
+      "step": 6265
+    },
+    {
+      "epoch": 0.4260089686098655,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.733744394618834e-05,
+      "loss": 1.3571,
+      "step": 6270
+    },
+    {
+      "epoch": 0.4263486886805272,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.7335320695746705e-05,
+      "loss": 1.2766,
+      "step": 6275
+    },
+    {
+      "epoch": 0.426688408751189,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.733319744530507e-05,
+      "loss": 1.3486,
+      "step": 6280
+    },
+    {
+      "epoch": 0.4270281288218508,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.733107419486343e-05,
+      "loss": 1.3336,
+      "step": 6285
+    },
+    {
+      "epoch": 0.42736784889251256,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.73289509444218e-05,
+      "loss": 1.2875,
+      "step": 6290
+    },
+    {
+      "epoch": 0.42770756896317436,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.732682769398016e-05,
+      "loss": 1.3185,
+      "step": 6295
+    },
+    {
+      "epoch": 0.4280472890338361,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.7324704443538525e-05,
+      "loss": 1.4517,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4283870091044979,
+      "grad_norm": 5.625,
+      "learning_rate": 4.7322581193096896e-05,
+      "loss": 1.2639,
+      "step": 6305
+    },
+    {
+      "epoch": 0.4287267291751597,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.732045794265525e-05,
+      "loss": 1.3088,
+      "step": 6310
+    },
+    {
+      "epoch": 0.42906644924582144,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.731833469221362e-05,
+      "loss": 1.3546,
+      "step": 6315
+    },
+    {
+      "epoch": 0.42940616931648323,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.731621144177199e-05,
+      "loss": 1.241,
+      "step": 6320
+    },
+    {
+      "epoch": 0.429745889387145,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.7314088191330345e-05,
+      "loss": 1.25,
+      "step": 6325
+    },
+    {
+      "epoch": 0.4300856094578068,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.731196494088871e-05,
+      "loss": 1.2611,
+      "step": 6330
+    },
+    {
+      "epoch": 0.4304253295284685,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.730984169044707e-05,
+      "loss": 1.3946,
+      "step": 6335
+    },
+    {
+      "epoch": 0.4307650495991303,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.730771844000544e-05,
+      "loss": 1.2698,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4311047696697921,
+      "grad_norm": 5.25,
+      "learning_rate": 4.73055951895638e-05,
+      "loss": 1.3262,
+      "step": 6345
+    },
+    {
+      "epoch": 0.43144448974045385,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.7303471939122165e-05,
+      "loss": 1.3052,
+      "step": 6350
+    },
+    {
+      "epoch": 0.43178420981111565,
+      "grad_norm": 5.875,
+      "learning_rate": 4.730134868868053e-05,
+      "loss": 1.3058,
+      "step": 6355
+    },
+    {
+      "epoch": 0.4321239298817774,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.729922543823889e-05,
+      "loss": 1.2679,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4324636499524392,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.729710218779726e-05,
+      "loss": 1.3299,
+      "step": 6365
+    },
+    {
+      "epoch": 0.432803370023101,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.729497893735562e-05,
+      "loss": 1.2336,
+      "step": 6370
+    },
+    {
+      "epoch": 0.43314309009376273,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7292855686913985e-05,
+      "loss": 1.386,
+      "step": 6375
+    },
+    {
+      "epoch": 0.4334828101644245,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.729073243647235e-05,
+      "loss": 1.3756,
+      "step": 6380
+    },
+    {
+      "epoch": 0.43382253023508627,
+      "grad_norm": 5.375,
+      "learning_rate": 4.728860918603071e-05,
+      "loss": 1.3117,
+      "step": 6385
+    },
+    {
+      "epoch": 0.43416225030574807,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.728648593558908e-05,
+      "loss": 1.3333,
+      "step": 6390
+    },
+    {
+      "epoch": 0.43450197037640986,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.728436268514744e-05,
+      "loss": 1.3109,
+      "step": 6395
+    },
+    {
+      "epoch": 0.4348416904470716,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.7282239434705805e-05,
+      "loss": 1.2212,
+      "step": 6400
+    },
+    {
+      "epoch": 0.4351814105177334,
+      "grad_norm": 5.0,
+      "learning_rate": 4.728011618426417e-05,
+      "loss": 1.4259,
+      "step": 6405
+    },
+    {
+      "epoch": 0.43552113058839514,
+      "grad_norm": 5.125,
+      "learning_rate": 4.727799293382253e-05,
+      "loss": 1.3337,
+      "step": 6410
+    },
+    {
+      "epoch": 0.43586085065905694,
+      "grad_norm": 5.5,
+      "learning_rate": 4.72758696833809e-05,
+      "loss": 1.2744,
+      "step": 6415
+    },
+    {
+      "epoch": 0.4362005707297187,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.727374643293926e-05,
+      "loss": 1.2286,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4365402908003805,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.7271623182497625e-05,
+      "loss": 1.3164,
+      "step": 6425
+    },
+    {
+      "epoch": 0.4368800108710423,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.726949993205599e-05,
+      "loss": 1.3641,
+      "step": 6430
+    },
+    {
+      "epoch": 0.437219730941704,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.7267376681614353e-05,
+      "loss": 1.2333,
+      "step": 6435
+    },
+    {
+      "epoch": 0.4375594510123658,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.726525343117272e-05,
+      "loss": 1.3042,
+      "step": 6440
+    },
+    {
+      "epoch": 0.43789917108302756,
+      "grad_norm": 5.75,
+      "learning_rate": 4.726313018073108e-05,
+      "loss": 1.2605,
+      "step": 6445
+    },
+    {
+      "epoch": 0.43823889115368936,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.7261006930289445e-05,
+      "loss": 1.4174,
+      "step": 6450
+    },
+    {
+      "epoch": 0.43857861122435116,
+      "grad_norm": 4.75,
+      "learning_rate": 4.725888367984781e-05,
+      "loss": 1.338,
+      "step": 6455
+    },
+    {
+      "epoch": 0.4389183312950129,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.7256760429406173e-05,
+      "loss": 1.3203,
+      "step": 6460
+    },
+    {
+      "epoch": 0.4392580513656747,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.725463717896453e-05,
+      "loss": 1.338,
+      "step": 6465
+    },
+    {
+      "epoch": 0.43959777143633644,
+      "grad_norm": 5.25,
+      "learning_rate": 4.72525139285229e-05,
+      "loss": 1.504,
+      "step": 6470
+    },
+    {
+      "epoch": 0.43993749150699824,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7250390678081265e-05,
+      "loss": 1.3181,
+      "step": 6475
+    },
+    {
+      "epoch": 0.44027721157766003,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.724826742763962e-05,
+      "loss": 1.3299,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4406169316483218,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.7246144177197993e-05,
+      "loss": 1.3697,
+      "step": 6485
+    },
+    {
+      "epoch": 0.44095665171898357,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.724402092675636e-05,
+      "loss": 1.3806,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4412963717896453,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7241897676314715e-05,
+      "loss": 1.3054,
+      "step": 6495
+    },
+    {
+      "epoch": 0.4416360918603071,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.7239774425873085e-05,
+      "loss": 1.281,
+      "step": 6500
+    },
+    {
+      "epoch": 0.4419758119309689,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.723765117543145e-05,
+      "loss": 1.2409,
+      "step": 6505
+    },
+    {
+      "epoch": 0.44231553200163065,
+      "grad_norm": 4.75,
+      "learning_rate": 4.723552792498981e-05,
+      "loss": 1.3892,
+      "step": 6510
+    },
+    {
+      "epoch": 0.44265525207229245,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.723340467454818e-05,
+      "loss": 1.2411,
+      "step": 6515
+    },
+    {
+      "epoch": 0.4429949721429542,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.723128142410654e-05,
+      "loss": 1.2632,
+      "step": 6520
+    },
+    {
+      "epoch": 0.443334692213616,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.72291581736649e-05,
+      "loss": 1.241,
+      "step": 6525
+    },
+    {
+      "epoch": 0.44367441228427773,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.722703492322327e-05,
+      "loss": 1.2148,
+      "step": 6530
+    },
+    {
+      "epoch": 0.44401413235493953,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.722491167278163e-05,
+      "loss": 1.259,
+      "step": 6535
+    },
+    {
+      "epoch": 0.4443538524256013,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.722278842233999e-05,
+      "loss": 1.297,
+      "step": 6540
+    },
+    {
+      "epoch": 0.44469357249626307,
+      "grad_norm": 5.375,
+      "learning_rate": 4.722066517189836e-05,
+      "loss": 1.3552,
+      "step": 6545
+    },
+    {
+      "epoch": 0.44503329256692487,
+      "grad_norm": 5.375,
+      "learning_rate": 4.721854192145672e-05,
+      "loss": 1.2243,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4453730126375866,
+      "grad_norm": 5.375,
+      "learning_rate": 4.721641867101508e-05,
+      "loss": 1.308,
+      "step": 6555
+    },
+    {
+      "epoch": 0.4457127327082484,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7214295420573453e-05,
+      "loss": 1.3065,
+      "step": 6560
+    },
+    {
+      "epoch": 0.4460524527789102,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.721217217013181e-05,
+      "loss": 1.2988,
+      "step": 6565
+    },
+    {
+      "epoch": 0.44639217284957194,
+      "grad_norm": 5.25,
+      "learning_rate": 4.7210048919690175e-05,
+      "loss": 1.2596,
+      "step": 6570
+    },
+    {
+      "epoch": 0.44673189292023374,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7207925669248545e-05,
+      "loss": 1.3296,
+      "step": 6575
+    },
+    {
+      "epoch": 0.4470716129908955,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.72058024188069e-05,
+      "loss": 1.3417,
+      "step": 6580
+    },
+    {
+      "epoch": 0.4474113330615573,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.720367916836527e-05,
+      "loss": 1.2729,
+      "step": 6585
+    },
+    {
+      "epoch": 0.4477510531322191,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.720155591792364e-05,
+      "loss": 1.3447,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4480907732028808,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.7199432667481995e-05,
+      "loss": 1.359,
+      "step": 6595
+    },
+    {
+      "epoch": 0.4484304932735426,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.719730941704036e-05,
+      "loss": 1.2902,
+      "step": 6600
+    },
+    {
+      "epoch": 0.44877021334420436,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.719518616659872e-05,
+      "loss": 1.3281,
+      "step": 6605
+    },
+    {
+      "epoch": 0.44910993341486616,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.719306291615709e-05,
+      "loss": 1.4038,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4494496534855279,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.719093966571545e-05,
+      "loss": 1.3164,
+      "step": 6615
+    },
+    {
+      "epoch": 0.4497893735561897,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.7188816415273815e-05,
+      "loss": 1.3789,
+      "step": 6620
+    },
+    {
+      "epoch": 0.4501290936268515,
+      "grad_norm": 5.875,
+      "learning_rate": 4.718669316483218e-05,
+      "loss": 1.2988,
+      "step": 6625
+    },
+    {
+      "epoch": 0.45046881369751324,
+      "grad_norm": 5.5,
+      "learning_rate": 4.718456991439054e-05,
+      "loss": 1.258,
+      "step": 6630
+    },
+    {
+      "epoch": 0.45080853376817503,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.718244666394891e-05,
+      "loss": 1.3553,
+      "step": 6635
+    },
+    {
+      "epoch": 0.4511482538388368,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.718032341350727e-05,
+      "loss": 1.3359,
+      "step": 6640
+    },
+    {
+      "epoch": 0.4514879739094986,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.717820016306564e-05,
+      "loss": 1.2785,
+      "step": 6645
+    },
+    {
+      "epoch": 0.45182769398016037,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.7176076912624e-05,
+      "loss": 1.2377,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4521674140508221,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.717395366218236e-05,
+      "loss": 1.2628,
+      "step": 6655
+    },
+    {
+      "epoch": 0.4525071341214839,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7171830411740734e-05,
+      "loss": 1.3298,
+      "step": 6660
+    },
+    {
+      "epoch": 0.45284685419214565,
+      "grad_norm": 5.25,
+      "learning_rate": 4.716970716129909e-05,
+      "loss": 1.2086,
+      "step": 6665
+    },
+    {
+      "epoch": 0.45318657426280745,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.7167583910857455e-05,
+      "loss": 1.2607,
+      "step": 6670
+    },
+    {
+      "epoch": 0.45352629433346925,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.7165460660415826e-05,
+      "loss": 1.3351,
+      "step": 6675
+    },
+    {
+      "epoch": 0.453866014404131,
+      "grad_norm": 5.75,
+      "learning_rate": 4.716333740997418e-05,
+      "loss": 1.3499,
+      "step": 6680
+    },
+    {
+      "epoch": 0.4542057344747928,
+      "grad_norm": 4.75,
+      "learning_rate": 4.716121415953255e-05,
+      "loss": 1.2843,
+      "step": 6685
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.715909090909091e-05,
+      "loss": 1.2159,
+      "step": 6690
+    },
+    {
+      "epoch": 0.45488517461611633,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7156967658649275e-05,
+      "loss": 1.4426,
+      "step": 6695
+    },
+    {
+      "epoch": 0.45522489468677807,
+      "grad_norm": 5.625,
+      "learning_rate": 4.715484440820764e-05,
+      "loss": 1.1756,
+      "step": 6700
+    },
+    {
+      "epoch": 0.45556461475743987,
+      "grad_norm": 5.75,
+      "learning_rate": 4.7152721157766e-05,
+      "loss": 1.2793,
+      "step": 6705
+    },
+    {
+      "epoch": 0.45590433482810166,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.715059790732437e-05,
+      "loss": 1.3247,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4562440548987634,
+      "grad_norm": 5.375,
+      "learning_rate": 4.714847465688273e-05,
+      "loss": 1.2811,
+      "step": 6715
+    },
+    {
+      "epoch": 0.4565837749694252,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.7146351406441095e-05,
+      "loss": 1.3514,
+      "step": 6720
+    },
+    {
+      "epoch": 0.45692349504008695,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.714422815599946e-05,
+      "loss": 1.2167,
+      "step": 6725
+    },
+    {
+      "epoch": 0.45726321511074874,
+      "grad_norm": 6.0,
+      "learning_rate": 4.714210490555782e-05,
+      "loss": 1.3799,
+      "step": 6730
+    },
+    {
+      "epoch": 0.45760293518141054,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.713998165511619e-05,
+      "loss": 1.238,
+      "step": 6735
+    },
+    {
+      "epoch": 0.4579426552520723,
+      "grad_norm": 5.375,
+      "learning_rate": 4.713785840467455e-05,
+      "loss": 1.3195,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4582823753227341,
+      "grad_norm": 5.5,
+      "learning_rate": 4.7135735154232915e-05,
+      "loss": 1.3515,
+      "step": 6745
+    },
+    {
+      "epoch": 0.4586220953933958,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.713361190379128e-05,
+      "loss": 1.2924,
+      "step": 6750
+    },
+    {
+      "epoch": 0.4589618154640576,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.713148865334964e-05,
+      "loss": 1.4272,
+      "step": 6755
+    },
+    {
+      "epoch": 0.4593015355347194,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.712936540290801e-05,
+      "loss": 1.25,
+      "step": 6760
+    },
+    {
+      "epoch": 0.45964125560538116,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.712724215246637e-05,
+      "loss": 1.3047,
+      "step": 6765
+    },
+    {
+      "epoch": 0.45998097567604296,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.7125118902024735e-05,
+      "loss": 1.3505,
+      "step": 6770
+    },
+    {
+      "epoch": 0.4603206957467047,
+      "grad_norm": 5.5,
+      "learning_rate": 4.71229956515831e-05,
+      "loss": 1.3343,
+      "step": 6775
+    },
+    {
+      "epoch": 0.4606604158173665,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.712087240114146e-05,
+      "loss": 1.3529,
+      "step": 6780
+    },
+    {
+      "epoch": 0.46100013588802824,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.711874915069983e-05,
+      "loss": 1.2822,
+      "step": 6785
+    },
+    {
+      "epoch": 0.46133985595869004,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.711662590025819e-05,
+      "loss": 1.3786,
+      "step": 6790
+    },
+    {
+      "epoch": 0.46167957602935183,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.7114502649816555e-05,
+      "loss": 1.2473,
+      "step": 6795
+    },
+    {
+      "epoch": 0.4620192961000136,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.711237939937492e-05,
+      "loss": 1.3538,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4623590161706754,
+      "grad_norm": 6.0,
+      "learning_rate": 4.7110256148933276e-05,
+      "loss": 1.2829,
+      "step": 6805
+    },
+    {
+      "epoch": 0.4626987362413371,
+      "grad_norm": 5.125,
+      "learning_rate": 4.710813289849165e-05,
+      "loss": 1.3017,
+      "step": 6810
+    },
+    {
+      "epoch": 0.4630384563119989,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.710600964805001e-05,
+      "loss": 1.2314,
+      "step": 6815
+    },
+    {
+      "epoch": 0.4633781763826607,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.710388639760837e-05,
+      "loss": 1.2864,
+      "step": 6820
+    },
+    {
+      "epoch": 0.46371789645332245,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.710176314716674e-05,
+      "loss": 1.286,
+      "step": 6825
+    },
+    {
+      "epoch": 0.46405761652398425,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.70996398967251e-05,
+      "loss": 1.288,
+      "step": 6830
+    },
+    {
+      "epoch": 0.464397336594646,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.709751664628346e-05,
+      "loss": 1.3043,
+      "step": 6835
+    },
+    {
+      "epoch": 0.4647370566653078,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.709539339584183e-05,
+      "loss": 1.4001,
+      "step": 6840
+    },
+    {
+      "epoch": 0.4650767767359696,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.7093270145400195e-05,
+      "loss": 1.3007,
+      "step": 6845
+    },
+    {
+      "epoch": 0.46541649680663133,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.709114689495855e-05,
+      "loss": 1.4387,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4657562168772931,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.708902364451692e-05,
+      "loss": 1.3207,
+      "step": 6855
+    },
+    {
+      "epoch": 0.46609593694795487,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.708690039407529e-05,
+      "loss": 1.281,
+      "step": 6860
+    },
+    {
+      "epoch": 0.46643565701861667,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.7084777143633644e-05,
+      "loss": 1.3556,
+      "step": 6865
+    },
+    {
+      "epoch": 0.4667753770892784,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.7082653893192015e-05,
+      "loss": 1.2445,
+      "step": 6870
+    },
+    {
+      "epoch": 0.4671150971599402,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.708053064275038e-05,
+      "loss": 1.1666,
+      "step": 6875
+    },
+    {
+      "epoch": 0.467454817230602,
+      "grad_norm": 5.75,
+      "learning_rate": 4.7078407392308736e-05,
+      "loss": 1.3496,
+      "step": 6880
+    },
+    {
+      "epoch": 0.46779453730126375,
+      "grad_norm": 5.5,
+      "learning_rate": 4.707628414186711e-05,
+      "loss": 1.32,
+      "step": 6885
+    },
+    {
+      "epoch": 0.46813425737192554,
+      "grad_norm": 5.625,
+      "learning_rate": 4.7074160891425464e-05,
+      "loss": 1.3667,
+      "step": 6890
+    },
+    {
+      "epoch": 0.4684739774425873,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.707203764098383e-05,
+      "loss": 1.3171,
+      "step": 6895
+    },
+    {
+      "epoch": 0.4688136975132491,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.70699143905422e-05,
+      "loss": 1.3375,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4691534175839109,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.7067791140100556e-05,
+      "loss": 1.2619,
+      "step": 6905
+    },
+    {
+      "epoch": 0.4694931376545726,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.706566788965892e-05,
+      "loss": 1.3229,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4698328577252344,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.706354463921729e-05,
+      "loss": 1.3236,
+      "step": 6915
+    },
+    {
+      "epoch": 0.47017257779589616,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.706142138877565e-05,
+      "loss": 1.2894,
+      "step": 6920
+    },
+    {
+      "epoch": 0.47051229786655796,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.705929813833401e-05,
+      "loss": 1.3496,
+      "step": 6925
+    },
+    {
+      "epoch": 0.47085201793721976,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.705717488789238e-05,
+      "loss": 1.2698,
+      "step": 6930
+    },
+    {
+      "epoch": 0.4711917380078815,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.705505163745074e-05,
+      "loss": 1.3206,
+      "step": 6935
+    },
+    {
+      "epoch": 0.4715314580785433,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.7052928387009104e-05,
+      "loss": 1.2838,
+      "step": 6940
+    },
+    {
+      "epoch": 0.47187117814920504,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.7050805136567475e-05,
+      "loss": 1.3741,
+      "step": 6945
+    },
+    {
+      "epoch": 0.47221089821986684,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.704868188612583e-05,
+      "loss": 1.2023,
+      "step": 6950
+    },
+    {
+      "epoch": 0.4725506182905286,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.7046558635684196e-05,
+      "loss": 1.4043,
+      "step": 6955
+    },
+    {
+      "epoch": 0.4728903383611904,
+      "grad_norm": 7.125,
+      "learning_rate": 4.704443538524256e-05,
+      "loss": 1.3741,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4732300584318522,
+      "grad_norm": 5.75,
+      "learning_rate": 4.7042312134800924e-05,
+      "loss": 1.324,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4735697785025139,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.704018888435929e-05,
+      "loss": 1.2993,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4739094985731757,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.703806563391765e-05,
+      "loss": 1.3558,
+      "step": 6975
+    },
+    {
+      "epoch": 0.47424921864383746,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.7035942383476016e-05,
+      "loss": 1.2404,
+      "step": 6980
+    },
+    {
+      "epoch": 0.47458893871449925,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.703381913303439e-05,
+      "loss": 1.1858,
+      "step": 6985
+    },
+    {
+      "epoch": 0.47492865878516105,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.7031695882592744e-05,
+      "loss": 1.3316,
+      "step": 6990
+    },
+    {
+      "epoch": 0.4752683788558228,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.702957263215111e-05,
+      "loss": 1.2112,
+      "step": 6995
+    },
+    {
+      "epoch": 0.4756080989264846,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.702744938170948e-05,
+      "loss": 1.2575,
+      "step": 7000
+    },
+    {
+      "epoch": 0.47594781899714633,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.7025326131267836e-05,
+      "loss": 1.3531,
+      "step": 7005
+    },
+    {
+      "epoch": 0.47628753906780813,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.70232028808262e-05,
+      "loss": 1.3421,
+      "step": 7010
+    },
+    {
+      "epoch": 0.4766272591384699,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.702107963038457e-05,
+      "loss": 1.3886,
+      "step": 7015
+    },
+    {
+      "epoch": 0.47696697920913167,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.701895637994293e-05,
+      "loss": 1.2832,
+      "step": 7020
+    },
+    {
+      "epoch": 0.47730669927979347,
+      "grad_norm": 4.75,
+      "learning_rate": 4.701683312950129e-05,
+      "loss": 1.2235,
+      "step": 7025
+    },
+    {
+      "epoch": 0.4776464193504552,
+      "grad_norm": 6.75,
+      "learning_rate": 4.7014709879059656e-05,
+      "loss": 1.34,
+      "step": 7030
+    },
+    {
+      "epoch": 0.477986139421117,
+      "grad_norm": 5.25,
+      "learning_rate": 4.701258662861802e-05,
+      "loss": 1.2756,
+      "step": 7035
+    },
+    {
+      "epoch": 0.47832585949177875,
+      "grad_norm": 5.5,
+      "learning_rate": 4.7010463378176384e-05,
+      "loss": 1.2436,
+      "step": 7040
+    },
+    {
+      "epoch": 0.47866557956244055,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.700834012773475e-05,
+      "loss": 1.2985,
+      "step": 7045
+    },
+    {
+      "epoch": 0.47900529963310234,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.700621687729311e-05,
+      "loss": 1.222,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4793450197037641,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.7004093626851476e-05,
+      "loss": 1.2719,
+      "step": 7055
+    },
+    {
+      "epoch": 0.4796847397744259,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.700197037640984e-05,
+      "loss": 1.3371,
+      "step": 7060
+    },
+    {
+      "epoch": 0.4800244598450876,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.6999847125968204e-05,
+      "loss": 1.266,
+      "step": 7065
+    },
+    {
+      "epoch": 0.4803641799157494,
+      "grad_norm": 4.625,
+      "learning_rate": 4.699772387552657e-05,
+      "loss": 1.2712,
+      "step": 7070
+    },
+    {
+      "epoch": 0.4807038999864112,
+      "grad_norm": 5.375,
+      "learning_rate": 4.699560062508493e-05,
+      "loss": 1.3265,
+      "step": 7075
+    },
+    {
+      "epoch": 0.48104362005707296,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.6993477374643296e-05,
+      "loss": 1.3347,
+      "step": 7080
+    },
+    {
+      "epoch": 0.48138334012773476,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.699135412420166e-05,
+      "loss": 1.3057,
+      "step": 7085
+    },
+    {
+      "epoch": 0.4817230601983965,
+      "grad_norm": 5.25,
+      "learning_rate": 4.6989230873760024e-05,
+      "loss": 1.3254,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4820627802690583,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.698710762331839e-05,
+      "loss": 1.2284,
+      "step": 7095
+    },
+    {
+      "epoch": 0.4824025003397201,
+      "grad_norm": 5.5,
+      "learning_rate": 4.698498437287675e-05,
+      "loss": 1.3287,
+      "step": 7100
+    },
+    {
+      "epoch": 0.48274222041038184,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.6982861122435116e-05,
+      "loss": 1.2777,
+      "step": 7105
+    },
+    {
+      "epoch": 0.48308194048104364,
+      "grad_norm": 5.5,
+      "learning_rate": 4.698073787199348e-05,
+      "loss": 1.3045,
+      "step": 7110
+    },
+    {
+      "epoch": 0.4834216605517054,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.6978614621551844e-05,
+      "loss": 1.1692,
+      "step": 7115
+    },
+    {
+      "epoch": 0.4837613806223672,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.697649137111021e-05,
+      "loss": 1.3605,
+      "step": 7120
+    },
+    {
+      "epoch": 0.4841011006930289,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.697436812066857e-05,
+      "loss": 1.3168,
+      "step": 7125
+    },
+    {
+      "epoch": 0.4844408207636907,
+      "grad_norm": 5.875,
+      "learning_rate": 4.6972244870226936e-05,
+      "loss": 1.332,
+      "step": 7130
+    },
+    {
+      "epoch": 0.4847805408343525,
+      "grad_norm": 4.75,
+      "learning_rate": 4.69701216197853e-05,
+      "loss": 1.251,
+      "step": 7135
+    },
+    {
+      "epoch": 0.48512026090501426,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.6967998369343664e-05,
+      "loss": 1.2728,
+      "step": 7140
+    },
+    {
+      "epoch": 0.48545998097567605,
+      "grad_norm": 5.75,
+      "learning_rate": 4.696587511890203e-05,
+      "loss": 1.2869,
+      "step": 7145
+    },
+    {
+      "epoch": 0.4857997010463378,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.696375186846039e-05,
+      "loss": 1.2976,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4861394211169996,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.6961628618018756e-05,
+      "loss": 1.3018,
+      "step": 7155
+    },
+    {
+      "epoch": 0.4864791411876614,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.6959505367577114e-05,
+      "loss": 1.3394,
+      "step": 7160
+    },
+    {
+      "epoch": 0.48681886125832313,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6957382117135484e-05,
+      "loss": 1.2324,
+      "step": 7165
+    },
+    {
+      "epoch": 0.48715858132898493,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.695525886669385e-05,
+      "loss": 1.2774,
+      "step": 7170
+    },
+    {
+      "epoch": 0.48749830139964667,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6953135616252206e-05,
+      "loss": 1.3315,
+      "step": 7175
+    },
+    {
+      "epoch": 0.48783802147030847,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.6951012365810576e-05,
+      "loss": 1.2412,
+      "step": 7180
+    },
+    {
+      "epoch": 0.48817774154097027,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.694888911536894e-05,
+      "loss": 1.2396,
+      "step": 7185
+    },
+    {
+      "epoch": 0.488517461611632,
+      "grad_norm": 5.625,
+      "learning_rate": 4.69467658649273e-05,
+      "loss": 1.2769,
+      "step": 7190
+    },
+    {
+      "epoch": 0.4888571816822938,
+      "grad_norm": 5.625,
+      "learning_rate": 4.694464261448567e-05,
+      "loss": 1.4312,
+      "step": 7195
+    },
+    {
+      "epoch": 0.48919690175295555,
+      "grad_norm": 5.0,
+      "learning_rate": 4.694251936404403e-05,
+      "loss": 1.3402,
+      "step": 7200
+    },
+    {
+      "epoch": 0.48953662182361735,
+      "grad_norm": 5.5,
+      "learning_rate": 4.694039611360239e-05,
+      "loss": 1.2857,
+      "step": 7205
+    },
+    {
+      "epoch": 0.4898763418942791,
+      "grad_norm": 4.875,
+      "learning_rate": 4.693827286316076e-05,
+      "loss": 1.2846,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4902160619649409,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6936149612719124e-05,
+      "loss": 1.298,
+      "step": 7215
+    },
+    {
+      "epoch": 0.4905557820356027,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.693402636227748e-05,
+      "loss": 1.2757,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4908955021062644,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.693190311183585e-05,
+      "loss": 1.2893,
+      "step": 7225
+    },
+    {
+      "epoch": 0.4912352221769262,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.692977986139421e-05,
+      "loss": 1.2903,
+      "step": 7230
+    },
+    {
+      "epoch": 0.49157494224758796,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6927656610952574e-05,
+      "loss": 1.2712,
+      "step": 7235
+    },
+    {
+      "epoch": 0.49191466231824976,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.6925533360510944e-05,
+      "loss": 1.1913,
+      "step": 7240
+    },
+    {
+      "epoch": 0.49225438238891156,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.69234101100693e-05,
+      "loss": 1.2635,
+      "step": 7245
+    },
+    {
+      "epoch": 0.4925941024595733,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.6921286859627666e-05,
+      "loss": 1.3965,
+      "step": 7250
+    },
+    {
+      "epoch": 0.4929338225302351,
+      "grad_norm": 4.875,
+      "learning_rate": 4.6919163609186036e-05,
+      "loss": 1.2395,
+      "step": 7255
+    },
+    {
+      "epoch": 0.49327354260089684,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.6917040358744394e-05,
+      "loss": 1.4052,
+      "step": 7260
+    },
+    {
+      "epoch": 0.49361326267155864,
+      "grad_norm": 5.75,
+      "learning_rate": 4.691491710830276e-05,
+      "loss": 1.297,
+      "step": 7265
+    },
+    {
+      "epoch": 0.49395298274222044,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.691279385786113e-05,
+      "loss": 1.2718,
+      "step": 7270
+    },
+    {
+      "epoch": 0.4942927028128822,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.6910670607419486e-05,
+      "loss": 1.3518,
+      "step": 7275
+    },
+    {
+      "epoch": 0.494632422883544,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.690854735697785e-05,
+      "loss": 1.2943,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4949721429542057,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.690642410653622e-05,
+      "loss": 1.3686,
+      "step": 7285
+    },
+    {
+      "epoch": 0.4953118630248675,
+      "grad_norm": 5.625,
+      "learning_rate": 4.690430085609458e-05,
+      "loss": 1.2685,
+      "step": 7290
+    },
+    {
+      "epoch": 0.49565158309552926,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.690217760565294e-05,
+      "loss": 1.2237,
+      "step": 7295
+    },
+    {
+      "epoch": 0.49599130316619106,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.690005435521131e-05,
+      "loss": 1.3154,
+      "step": 7300
+    },
+    {
+      "epoch": 0.49633102323685285,
+      "grad_norm": 4.875,
+      "learning_rate": 4.689793110476967e-05,
+      "loss": 1.2653,
+      "step": 7305
+    },
+    {
+      "epoch": 0.4966707433075146,
+      "grad_norm": 5.625,
+      "learning_rate": 4.6895807854328034e-05,
+      "loss": 1.3204,
+      "step": 7310
+    },
+    {
+      "epoch": 0.4970104633781764,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.68936846038864e-05,
+      "loss": 1.2931,
+      "step": 7315
+    },
+    {
+      "epoch": 0.49735018344883813,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.689156135344476e-05,
+      "loss": 1.3637,
+      "step": 7320
+    },
+    {
+      "epoch": 0.49768990351949993,
+      "grad_norm": 5.875,
+      "learning_rate": 4.688943810300313e-05,
+      "loss": 1.3219,
+      "step": 7325
+    },
+    {
+      "epoch": 0.49802962359016173,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.688731485256149e-05,
+      "loss": 1.2893,
+      "step": 7330
+    },
+    {
+      "epoch": 0.49836934366082347,
+      "grad_norm": 5.0,
+      "learning_rate": 4.6885191602119854e-05,
+      "loss": 1.3411,
+      "step": 7335
+    },
+    {
+      "epoch": 0.49870906373148527,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.6883068351678225e-05,
+      "loss": 1.3093,
+      "step": 7340
+    },
+    {
+      "epoch": 0.499048783802147,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.688094510123658e-05,
+      "loss": 1.3128,
+      "step": 7345
+    },
+    {
+      "epoch": 0.4993885038728088,
+      "grad_norm": 5.625,
+      "learning_rate": 4.6878821850794946e-05,
+      "loss": 1.1918,
+      "step": 7350
+    },
+    {
+      "epoch": 0.4997282239434706,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.6876698600353317e-05,
+      "loss": 1.235,
+      "step": 7355
+    },
+    {
+      "epoch": 0.5000679440141323,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.6874575349911674e-05,
+      "loss": 1.3317,
+      "step": 7360
+    },
+    {
+      "epoch": 0.5004076640847941,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.687245209947004e-05,
+      "loss": 1.2978,
+      "step": 7365
+    },
+    {
+      "epoch": 0.5007473841554559,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.687032884902841e-05,
+      "loss": 1.2214,
+      "step": 7370
+    },
+    {
+      "epoch": 0.5010871042261177,
+      "grad_norm": 3.96875,
+      "learning_rate": 4.6868205598586766e-05,
+      "loss": 1.1583,
+      "step": 7375
+    },
+    {
+      "epoch": 0.5014268242967794,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.686608234814513e-05,
+      "loss": 1.3489,
+      "step": 7380
+    },
+    {
+      "epoch": 0.5017665443674413,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.6863959097703494e-05,
+      "loss": 1.3956,
+      "step": 7385
+    },
+    {
+      "epoch": 0.502106264438103,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.686183584726186e-05,
+      "loss": 1.2598,
+      "step": 7390
+    },
+    {
+      "epoch": 0.5024459845087648,
+      "grad_norm": 5.5,
+      "learning_rate": 4.685971259682022e-05,
+      "loss": 1.2744,
+      "step": 7395
+    },
+    {
+      "epoch": 0.5027857045794265,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.6857589346378586e-05,
+      "loss": 1.355,
+      "step": 7400
+    },
+    {
+      "epoch": 0.5031254246500884,
+      "grad_norm": 5.625,
+      "learning_rate": 4.685546609593695e-05,
+      "loss": 1.3461,
+      "step": 7405
+    },
+    {
+      "epoch": 0.5034651447207501,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.6853342845495314e-05,
+      "loss": 1.2279,
+      "step": 7410
+    },
+    {
+      "epoch": 0.5038048647914118,
+      "grad_norm": 5.5,
+      "learning_rate": 4.685121959505368e-05,
+      "loss": 1.3635,
+      "step": 7415
+    },
+    {
+      "epoch": 0.5041445848620737,
+      "grad_norm": 5.875,
+      "learning_rate": 4.684909634461204e-05,
+      "loss": 1.3418,
+      "step": 7420
+    },
+    {
+      "epoch": 0.5044843049327354,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.6846973094170406e-05,
+      "loss": 1.2256,
+      "step": 7425
+    },
+    {
+      "epoch": 0.5048240250033972,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.684484984372877e-05,
+      "loss": 1.3205,
+      "step": 7430
+    },
+    {
+      "epoch": 0.5051637450740589,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.6842726593287134e-05,
+      "loss": 1.3831,
+      "step": 7435
+    },
+    {
+      "epoch": 0.5055034651447208,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.68406033428455e-05,
+      "loss": 1.3024,
+      "step": 7440
+    },
+    {
+      "epoch": 0.5058431852153825,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.683848009240386e-05,
+      "loss": 1.2466,
+      "step": 7445
+    },
+    {
+      "epoch": 0.5061829052860443,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.6836356841962226e-05,
+      "loss": 1.2762,
+      "step": 7450
+    },
+    {
+      "epoch": 0.5065226253567061,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.683423359152059e-05,
+      "loss": 1.2843,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5068623454273679,
+      "grad_norm": 4.75,
+      "learning_rate": 4.6832110341078954e-05,
+      "loss": 1.3396,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5072020654980296,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.682998709063732e-05,
+      "loss": 1.3952,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5075417855686915,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.682786384019568e-05,
+      "loss": 1.2361,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5078815056393532,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.6825740589754046e-05,
+      "loss": 1.3747,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5082212257100149,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.682361733931241e-05,
+      "loss": 1.2149,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5085609457806767,
+      "grad_norm": 5.125,
+      "learning_rate": 4.6821494088870774e-05,
+      "loss": 1.3429,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5089006658513385,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.681937083842914e-05,
+      "loss": 1.2213,
+      "step": 7490
+    },
+    {
+      "epoch": 0.5092403859220003,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.68172475879875e-05,
+      "loss": 1.35,
+      "step": 7495
+    },
+    {
+      "epoch": 0.509580105992662,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.6815124337545866e-05,
+      "loss": 1.277,
+      "step": 7500
+    },
+    {
+      "epoch": 0.5099198260633239,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.681300108710423e-05,
+      "loss": 1.4028,
+      "step": 7505
+    },
+    {
+      "epoch": 0.5102595461339856,
+      "grad_norm": 5.375,
+      "learning_rate": 4.6810877836662594e-05,
+      "loss": 1.3159,
+      "step": 7510
+    },
+    {
+      "epoch": 0.5105992662046474,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.680875458622095e-05,
+      "loss": 1.3331,
+      "step": 7515
+    },
+    {
+      "epoch": 0.5109389862753091,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.680663133577932e-05,
+      "loss": 1.3643,
+      "step": 7520
+    },
+    {
+      "epoch": 0.511278706345971,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6804508085337686e-05,
+      "loss": 1.3489,
+      "step": 7525
+    },
+    {
+      "epoch": 0.5116184264166327,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.680238483489604e-05,
+      "loss": 1.2743,
+      "step": 7530
+    },
+    {
+      "epoch": 0.5119581464872944,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.6800261584454414e-05,
+      "loss": 1.1851,
+      "step": 7535
+    },
+    {
+      "epoch": 0.5122978665579563,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.679813833401278e-05,
+      "loss": 1.4073,
+      "step": 7540
+    },
+    {
+      "epoch": 0.512637586628618,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.6796015083571135e-05,
+      "loss": 1.2592,
+      "step": 7545
+    },
+    {
+      "epoch": 0.5129773066992798,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6793891833129506e-05,
+      "loss": 1.384,
+      "step": 7550
+    },
+    {
+      "epoch": 0.5133170267699416,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.679176858268787e-05,
+      "loss": 1.319,
+      "step": 7555
+    },
+    {
+      "epoch": 0.5136567468406034,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.678964533224623e-05,
+      "loss": 1.4105,
+      "step": 7560
+    },
+    {
+      "epoch": 0.5139964669112651,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.67875220818046e-05,
+      "loss": 1.3048,
+      "step": 7565
+    },
+    {
+      "epoch": 0.5143361869819268,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.678539883136296e-05,
+      "loss": 1.3025,
+      "step": 7570
+    },
+    {
+      "epoch": 0.5146759070525887,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.678327558092132e-05,
+      "loss": 1.1663,
+      "step": 7575
+    },
+    {
+      "epoch": 0.5150156271232504,
+      "grad_norm": 4.625,
+      "learning_rate": 4.678115233047969e-05,
+      "loss": 1.2606,
+      "step": 7580
+    },
+    {
+      "epoch": 0.5153553471939122,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.677902908003805e-05,
+      "loss": 1.3336,
+      "step": 7585
+    },
+    {
+      "epoch": 0.515695067264574,
+      "grad_norm": 5.875,
+      "learning_rate": 4.677690582959641e-05,
+      "loss": 1.1627,
+      "step": 7590
+    },
+    {
+      "epoch": 0.5160347873352358,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.677478257915478e-05,
+      "loss": 1.2364,
+      "step": 7595
+    },
+    {
+      "epoch": 0.5163745074058975,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.677265932871314e-05,
+      "loss": 1.2579,
+      "step": 7600
+    },
+    {
+      "epoch": 0.5167142274765593,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.67705360782715e-05,
+      "loss": 1.2904,
+      "step": 7605
+    },
+    {
+      "epoch": 0.5170539475472211,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6768412827829874e-05,
+      "loss": 1.2914,
+      "step": 7610
+    },
+    {
+      "epoch": 0.5173936676178829,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.676628957738823e-05,
+      "loss": 1.1284,
+      "step": 7615
+    },
+    {
+      "epoch": 0.5177333876885446,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.6764166326946595e-05,
+      "loss": 1.3836,
+      "step": 7620
+    },
+    {
+      "epoch": 0.5180731077592065,
+      "grad_norm": 5.125,
+      "learning_rate": 4.6762043076504966e-05,
+      "loss": 1.3724,
+      "step": 7625
+    },
+    {
+      "epoch": 0.5184128278298682,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.675991982606332e-05,
+      "loss": 1.3356,
+      "step": 7630
+    },
+    {
+      "epoch": 0.5187525479005299,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.675779657562169e-05,
+      "loss": 1.3022,
+      "step": 7635
+    },
+    {
+      "epoch": 0.5190922679711918,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.675567332518006e-05,
+      "loss": 1.3948,
+      "step": 7640
+    },
+    {
+      "epoch": 0.5194319880418535,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.6753550074738415e-05,
+      "loss": 1.2125,
+      "step": 7645
+    },
+    {
+      "epoch": 0.5197717081125153,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.675142682429678e-05,
+      "loss": 1.2638,
+      "step": 7650
+    },
+    {
+      "epoch": 0.520111428183177,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.674930357385514e-05,
+      "loss": 1.2351,
+      "step": 7655
+    },
+    {
+      "epoch": 0.5204511482538389,
+      "grad_norm": 5.5,
+      "learning_rate": 4.674718032341351e-05,
+      "loss": 1.2748,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5207908683245006,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.674505707297187e-05,
+      "loss": 1.2226,
+      "step": 7665
+    },
+    {
+      "epoch": 0.5211305883951624,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.6742933822530235e-05,
+      "loss": 1.3146,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5214703084658242,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.67408105720886e-05,
+      "loss": 1.2334,
+      "step": 7675
+    },
+    {
+      "epoch": 0.521810028536486,
+      "grad_norm": 5.25,
+      "learning_rate": 4.673868732164697e-05,
+      "loss": 1.2141,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5221497486071477,
+      "grad_norm": 6.0,
+      "learning_rate": 4.673656407120533e-05,
+      "loss": 1.3066,
+      "step": 7685
+    },
+    {
+      "epoch": 0.5224894686778094,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.673444082076369e-05,
+      "loss": 1.3399,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5228291887484713,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.673231757032206e-05,
+      "loss": 1.251,
+      "step": 7695
+    },
+    {
+      "epoch": 0.523168908819133,
+      "grad_norm": 5.5,
+      "learning_rate": 4.673019431988042e-05,
+      "loss": 1.293,
+      "step": 7700
+    },
+    {
+      "epoch": 0.5235086288897948,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.672807106943878e-05,
+      "loss": 1.39,
+      "step": 7705
+    },
+    {
+      "epoch": 0.5238483489604566,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.6725947818997154e-05,
+      "loss": 1.196,
+      "step": 7710
+    },
+    {
+      "epoch": 0.5241880690311184,
+      "grad_norm": 5.875,
+      "learning_rate": 4.672382456855551e-05,
+      "loss": 1.2742,
+      "step": 7715
+    },
+    {
+      "epoch": 0.5245277891017801,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.6721701318113875e-05,
+      "loss": 1.4282,
+      "step": 7720
+    },
+    {
+      "epoch": 0.524867509172442,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6719578067672246e-05,
+      "loss": 1.2611,
+      "step": 7725
+    },
+    {
+      "epoch": 0.5252072292431037,
+      "grad_norm": 5.5,
+      "learning_rate": 4.67174548172306e-05,
+      "loss": 1.3835,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5255469493137654,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.671533156678897e-05,
+      "loss": 1.2825,
+      "step": 7735
+    },
+    {
+      "epoch": 0.5258866693844272,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.671320831634733e-05,
+      "loss": 1.2561,
+      "step": 7740
+    },
+    {
+      "epoch": 0.526226389455089,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6711085065905695e-05,
+      "loss": 1.2714,
+      "step": 7745
+    },
+    {
+      "epoch": 0.5265661095257508,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.670896181546406e-05,
+      "loss": 1.4177,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5269058295964125,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.670683856502242e-05,
+      "loss": 1.4452,
+      "step": 7755
+    },
+    {
+      "epoch": 0.5272455496670744,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.670471531458079e-05,
+      "loss": 1.2595,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5275852697377361,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.670259206413915e-05,
+      "loss": 1.2169,
+      "step": 7765
+    },
+    {
+      "epoch": 0.5279249898083979,
+      "grad_norm": 6.5,
+      "learning_rate": 4.6700468813697515e-05,
+      "loss": 1.1865,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5282647098790596,
+      "grad_norm": 5.625,
+      "learning_rate": 4.669834556325588e-05,
+      "loss": 1.3333,
+      "step": 7775
+    },
+    {
+      "epoch": 0.5286044299497215,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.669622231281424e-05,
+      "loss": 1.1646,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5289441500203832,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.669409906237261e-05,
+      "loss": 1.3839,
+      "step": 7785
+    },
+    {
+      "epoch": 0.5292838700910449,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.669197581193097e-05,
+      "loss": 1.3406,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5296235901617068,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.6689852561489335e-05,
+      "loss": 1.3643,
+      "step": 7795
+    },
+    {
+      "epoch": 0.5299633102323685,
+      "grad_norm": 5.5,
+      "learning_rate": 4.66877293110477e-05,
+      "loss": 1.3102,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5303030303030303,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.668560606060606e-05,
+      "loss": 1.276,
+      "step": 7805
+    },
+    {
+      "epoch": 0.5306427503736921,
+      "grad_norm": 5.375,
+      "learning_rate": 4.668348281016443e-05,
+      "loss": 1.2105,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5309824704443539,
+      "grad_norm": 5.25,
+      "learning_rate": 4.668135955972279e-05,
+      "loss": 1.2575,
+      "step": 7815
+    },
+    {
+      "epoch": 0.5313221905150156,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6679236309281155e-05,
+      "loss": 1.3448,
+      "step": 7820
+    },
+    {
+      "epoch": 0.5316619105856774,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.667711305883952e-05,
+      "loss": 1.2896,
+      "step": 7825
+    },
+    {
+      "epoch": 0.5320016306563392,
+      "grad_norm": 5.375,
+      "learning_rate": 4.667498980839788e-05,
+      "loss": 1.2275,
+      "step": 7830
+    },
+    {
+      "epoch": 0.532341350727001,
+      "grad_norm": 5.5,
+      "learning_rate": 4.667286655795625e-05,
+      "loss": 1.3161,
+      "step": 7835
+    },
+    {
+      "epoch": 0.5326810707976627,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.667074330751461e-05,
+      "loss": 1.2453,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5330207908683245,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6668620057072975e-05,
+      "loss": 1.2869,
+      "step": 7845
+    },
+    {
+      "epoch": 0.5333605109389863,
+      "grad_norm": 5.0,
+      "learning_rate": 4.666649680663134e-05,
+      "loss": 1.3197,
+      "step": 7850
+    },
+    {
+      "epoch": 0.533700231009648,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.6664373556189697e-05,
+      "loss": 1.3099,
+      "step": 7855
+    },
+    {
+      "epoch": 0.5340399510803098,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.666225030574807e-05,
+      "loss": 1.2687,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5343796711509716,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.666012705530643e-05,
+      "loss": 1.3325,
+      "step": 7865
+    },
+    {
+      "epoch": 0.5347193912216334,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.665800380486479e-05,
+      "loss": 1.3155,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5350591112922951,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.665588055442316e-05,
+      "loss": 1.2244,
+      "step": 7875
+    },
+    {
+      "epoch": 0.535398831362957,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.6653757303981523e-05,
+      "loss": 1.2245,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5357385514336187,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.665163405353988e-05,
+      "loss": 1.2607,
+      "step": 7885
+    },
+    {
+      "epoch": 0.5360782715042804,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.664951080309825e-05,
+      "loss": 1.3979,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5364179915749423,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.6647387552656615e-05,
+      "loss": 1.3198,
+      "step": 7895
+    },
+    {
+      "epoch": 0.536757711645604,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.664526430221497e-05,
+      "loss": 1.1913,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5370974317162658,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.6643141051773343e-05,
+      "loss": 1.3168,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5374371517869275,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.664101780133171e-05,
+      "loss": 1.4437,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5377768718575894,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.6638894550890065e-05,
+      "loss": 1.4159,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5381165919282511,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.6636771300448435e-05,
+      "loss": 1.3251,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5384563119989129,
+      "grad_norm": 4.875,
+      "learning_rate": 4.66346480500068e-05,
+      "loss": 1.2805,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5387960320695747,
+      "grad_norm": 5.25,
+      "learning_rate": 4.663252479956516e-05,
+      "loss": 1.2594,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5391357521402365,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.663040154912353e-05,
+      "loss": 1.3841,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5394754722108982,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.6628278298681885e-05,
+      "loss": 1.2689,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5398151922815599,
+      "grad_norm": 5.5,
+      "learning_rate": 4.662615504824025e-05,
+      "loss": 1.3057,
+      "step": 7945
+    },
+    {
+      "epoch": 0.5401549123522218,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.662403179779862e-05,
+      "loss": 1.3278,
+      "step": 7950
+    },
+    {
+      "epoch": 0.5404946324228835,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.662190854735698e-05,
+      "loss": 1.2556,
+      "step": 7955
+    },
+    {
+      "epoch": 0.5408343524935453,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.661978529691534e-05,
+      "loss": 1.242,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5411740725642071,
+      "grad_norm": 5.125,
+      "learning_rate": 4.661766204647371e-05,
+      "loss": 1.3265,
+      "step": 7965
+    },
+    {
+      "epoch": 0.5415137926348689,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.661553879603207e-05,
+      "loss": 1.3411,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5418535127055306,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.661341554559043e-05,
+      "loss": 1.2867,
+      "step": 7975
+    },
+    {
+      "epoch": 0.5421932327761925,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.6611292295148803e-05,
+      "loss": 1.2729,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5425329528468542,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.660916904470716e-05,
+      "loss": 1.3116,
+      "step": 7985
+    },
+    {
+      "epoch": 0.542872672917516,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.6607045794265525e-05,
+      "loss": 1.2386,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5432123929881777,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6604922543823895e-05,
+      "loss": 1.253,
+      "step": 7995
+    },
+    {
+      "epoch": 0.5435521130588395,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.660279929338225e-05,
+      "loss": 1.2944,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5438918331295013,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.660067604294062e-05,
+      "loss": 1.2432,
+      "step": 8005
+    },
+    {
+      "epoch": 0.544231553200163,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.659855279249898e-05,
+      "loss": 1.3753,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5445712732708249,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.6596429542057345e-05,
+      "loss": 1.2964,
+      "step": 8015
+    },
+    {
+      "epoch": 0.5449109933414866,
+      "grad_norm": 6.0,
+      "learning_rate": 4.6594306291615716e-05,
+      "loss": 1.2705,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5452507134121484,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.659218304117407e-05,
+      "loss": 1.2985,
+      "step": 8025
+    },
+    {
+      "epoch": 0.5455904334828101,
+      "grad_norm": 5.375,
+      "learning_rate": 4.659005979073244e-05,
+      "loss": 1.288,
+      "step": 8030
+    },
+    {
+      "epoch": 0.545930153553472,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.658793654029081e-05,
+      "loss": 1.3556,
+      "step": 8035
+    },
+    {
+      "epoch": 0.5462698736241337,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.6585813289849165e-05,
+      "loss": 1.2254,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5466095936947954,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.658369003940753e-05,
+      "loss": 1.2796,
+      "step": 8045
+    },
+    {
+      "epoch": 0.5469493137654573,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.65815667889659e-05,
+      "loss": 1.3308,
+      "step": 8050
+    },
+    {
+      "epoch": 0.547289033836119,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.657944353852426e-05,
+      "loss": 1.2487,
+      "step": 8055
+    },
+    {
+      "epoch": 0.5476287539067808,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.657732028808262e-05,
+      "loss": 1.3357,
+      "step": 8060
+    },
+    {
+      "epoch": 0.5479684739774426,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.657519703764099e-05,
+      "loss": 1.161,
+      "step": 8065
+    },
+    {
+      "epoch": 0.5483081940481044,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.657307378719935e-05,
+      "loss": 1.2932,
+      "step": 8070
+    },
+    {
+      "epoch": 0.5486479141187661,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.657095053675771e-05,
+      "loss": 1.3162,
+      "step": 8075
+    },
+    {
+      "epoch": 0.5489876341894279,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.656882728631608e-05,
+      "loss": 1.2358,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5493273542600897,
+      "grad_norm": 5.375,
+      "learning_rate": 4.656670403587444e-05,
+      "loss": 1.2796,
+      "step": 8085
+    },
+    {
+      "epoch": 0.5496670743307515,
+      "grad_norm": 5.625,
+      "learning_rate": 4.6564580785432805e-05,
+      "loss": 1.2897,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5500067944014132,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.656245753499117e-05,
+      "loss": 1.2609,
+      "step": 8095
+    },
+    {
+      "epoch": 0.550346514472075,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.656033428454953e-05,
+      "loss": 1.2902,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5506862345427368,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.65582110341079e-05,
+      "loss": 1.2672,
+      "step": 8105
+    },
+    {
+      "epoch": 0.5510259546133985,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.655608778366626e-05,
+      "loss": 1.2602,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5513656746840603,
+      "grad_norm": 5.25,
+      "learning_rate": 4.6553964533224625e-05,
+      "loss": 1.2874,
+      "step": 8115
+    },
+    {
+      "epoch": 0.5517053947547221,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.655184128278299e-05,
+      "loss": 1.4087,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5520451148253839,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.654971803234135e-05,
+      "loss": 1.3549,
+      "step": 8125
+    },
+    {
+      "epoch": 0.5523848348960456,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.654759478189972e-05,
+      "loss": 1.3326,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5527245549667075,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.654547153145808e-05,
+      "loss": 1.4051,
+      "step": 8135
+    },
+    {
+      "epoch": 0.5530642750373692,
+      "grad_norm": 5.0,
+      "learning_rate": 4.6543348281016445e-05,
+      "loss": 1.3521,
+      "step": 8140
+    },
+    {
+      "epoch": 0.553403995108031,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.654122503057481e-05,
+      "loss": 1.3661,
+      "step": 8145
+    },
+    {
+      "epoch": 0.5537437151786928,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.653910178013317e-05,
+      "loss": 1.2213,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5540834352493546,
+      "grad_norm": 5.5,
+      "learning_rate": 4.653697852969154e-05,
+      "loss": 1.3094,
+      "step": 8155
+    },
+    {
+      "epoch": 0.5544231553200163,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.65348552792499e-05,
+      "loss": 1.2753,
+      "step": 8160
+    },
+    {
+      "epoch": 0.554762875390678,
+      "grad_norm": 5.625,
+      "learning_rate": 4.6532732028808265e-05,
+      "loss": 1.3721,
+      "step": 8165
+    },
+    {
+      "epoch": 0.5551025954613399,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.653060877836663e-05,
+      "loss": 1.3338,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5554423155320016,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.652848552792499e-05,
+      "loss": 1.1977,
+      "step": 8175
+    },
+    {
+      "epoch": 0.5557820356026634,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.652636227748336e-05,
+      "loss": 1.258,
+      "step": 8180
+    },
+    {
+      "epoch": 0.5561217556733252,
+      "grad_norm": 5.5,
+      "learning_rate": 4.652423902704172e-05,
+      "loss": 1.1799,
+      "step": 8185
+    },
+    {
+      "epoch": 0.556461475743987,
+      "grad_norm": 5.375,
+      "learning_rate": 4.6522115776600085e-05,
+      "loss": 1.2917,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5568011958146487,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.651999252615845e-05,
+      "loss": 1.251,
+      "step": 8195
+    },
+    {
+      "epoch": 0.5571409158853105,
+      "grad_norm": 5.375,
+      "learning_rate": 4.651786927571681e-05,
+      "loss": 1.3086,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5574806359559723,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.651574602527518e-05,
+      "loss": 1.271,
+      "step": 8205
+    },
+    {
+      "epoch": 0.557820356026634,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.6513622774833534e-05,
+      "loss": 1.2018,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5581600760972958,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.6511499524391905e-05,
+      "loss": 1.2058,
+      "step": 8215
+    },
+    {
+      "epoch": 0.5584997961679576,
+      "grad_norm": 4.3125,
+      "learning_rate": 4.650937627395027e-05,
+      "loss": 1.2101,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5588395162386194,
+      "grad_norm": 5.125,
+      "learning_rate": 4.6507253023508626e-05,
+      "loss": 1.2368,
+      "step": 8225
+    },
+    {
+      "epoch": 0.5591792363092811,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.6505129773067e-05,
+      "loss": 1.3648,
+      "step": 8230
+    },
+    {
+      "epoch": 0.559518956379943,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.650300652262536e-05,
+      "loss": 1.2243,
+      "step": 8235
+    },
+    {
+      "epoch": 0.5598586764506047,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.650088327218372e-05,
+      "loss": 1.386,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5601983965212665,
+      "grad_norm": 5.75,
+      "learning_rate": 4.649876002174209e-05,
+      "loss": 1.3844,
+      "step": 8245
+    },
+    {
+      "epoch": 0.5605381165919282,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.649663677130045e-05,
+      "loss": 1.1891,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5608778366625901,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.649451352085881e-05,
+      "loss": 1.1625,
+      "step": 8255
+    },
+    {
+      "epoch": 0.5612175567332518,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.649239027041718e-05,
+      "loss": 1.3048,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5615572768039135,
+      "grad_norm": 5.75,
+      "learning_rate": 4.6490267019975545e-05,
+      "loss": 1.2669,
+      "step": 8265
+    },
+    {
+      "epoch": 0.5618969968745754,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.64881437695339e-05,
+      "loss": 1.2515,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5622367169452371,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.648602051909227e-05,
+      "loss": 1.2501,
+      "step": 8275
+    },
+    {
+      "epoch": 0.5625764370158989,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.648389726865063e-05,
+      "loss": 1.2646,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5629161570865606,
+      "grad_norm": 5.0,
+      "learning_rate": 4.6481774018208994e-05,
+      "loss": 1.2902,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5632558771572225,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.6479650767767365e-05,
+      "loss": 1.3661,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5635955972278842,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.647752751732572e-05,
+      "loss": 1.2837,
+      "step": 8295
+    },
+    {
+      "epoch": 0.563935317298546,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.6475404266884086e-05,
+      "loss": 1.3401,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5642750373692078,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.647328101644246e-05,
+      "loss": 1.2485,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5646147574398696,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.6471157766000814e-05,
+      "loss": 1.2946,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5649544775105313,
+      "grad_norm": 5.875,
+      "learning_rate": 4.646903451555918e-05,
+      "loss": 1.2701,
+      "step": 8315
+    },
+    {
+      "epoch": 0.5652941975811931,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.646691126511755e-05,
+      "loss": 1.3007,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5656339176518549,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.6464788014675906e-05,
+      "loss": 1.3946,
+      "step": 8325
+    },
+    {
+      "epoch": 0.5659736377225166,
+      "grad_norm": 5.75,
+      "learning_rate": 4.646266476423427e-05,
+      "loss": 1.3249,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5663133577931784,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.646054151379264e-05,
+      "loss": 1.4042,
+      "step": 8335
+    },
+    {
+      "epoch": 0.5666530778638402,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.6458418263351e-05,
+      "loss": 1.2502,
+      "step": 8340
+    },
+    {
+      "epoch": 0.566992797934502,
+      "grad_norm": 5.125,
+      "learning_rate": 4.645629501290936e-05,
+      "loss": 1.3403,
+      "step": 8345
+    },
+    {
+      "epoch": 0.5673325180051637,
+      "grad_norm": 5.875,
+      "learning_rate": 4.645417176246773e-05,
+      "loss": 1.3016,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5676722380758256,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.645204851202609e-05,
+      "loss": 1.3349,
+      "step": 8355
+    },
+    {
+      "epoch": 0.5680119581464873,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.644992526158446e-05,
+      "loss": 1.3395,
+      "step": 8360
+    },
+    {
+      "epoch": 0.568351678217149,
+      "grad_norm": 5.25,
+      "learning_rate": 4.644780201114282e-05,
+      "loss": 1.2914,
+      "step": 8365
+    },
+    {
+      "epoch": 0.5686913982878108,
+      "grad_norm": 5.875,
+      "learning_rate": 4.644567876070118e-05,
+      "loss": 1.2517,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5690311183584726,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.644355551025955e-05,
+      "loss": 1.213,
+      "step": 8375
+    },
+    {
+      "epoch": 0.5693708384291344,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.644143225981791e-05,
+      "loss": 1.3537,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5697105584997961,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.6439309009376274e-05,
+      "loss": 1.2939,
+      "step": 8385
+    },
+    {
+      "epoch": 0.570050278570458,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6437185758934645e-05,
+      "loss": 1.2738,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5703899986411197,
+      "grad_norm": 5.75,
+      "learning_rate": 4.6435062508493e-05,
+      "loss": 1.3351,
+      "step": 8395
+    },
+    {
+      "epoch": 0.5707297187117815,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.6432939258051366e-05,
+      "loss": 1.2879,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5710694387824433,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.643081600760974e-05,
+      "loss": 1.3266,
+      "step": 8405
+    },
+    {
+      "epoch": 0.5714091588531051,
+      "grad_norm": 5.625,
+      "learning_rate": 4.6428692757168094e-05,
+      "loss": 1.1938,
+      "step": 8410
+    },
+    {
+      "epoch": 0.5717488789237668,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.642656950672646e-05,
+      "loss": 1.1965,
+      "step": 8415
+    },
+    {
+      "epoch": 0.5720885989944285,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.642444625628483e-05,
+      "loss": 1.3661,
+      "step": 8420
+    },
+    {
+      "epoch": 0.5724283190650904,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.6422323005843186e-05,
+      "loss": 1.354,
+      "step": 8425
+    },
+    {
+      "epoch": 0.5727680391357521,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.642019975540155e-05,
+      "loss": 1.3011,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5731077592064139,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6418076504959914e-05,
+      "loss": 1.2609,
+      "step": 8435
+    },
+    {
+      "epoch": 0.5734474792770757,
+      "grad_norm": 4.875,
+      "learning_rate": 4.641595325451828e-05,
+      "loss": 1.2508,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5737871993477375,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.641383000407664e-05,
+      "loss": 1.3164,
+      "step": 8445
+    },
+    {
+      "epoch": 0.5741269194183992,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.6411706753635006e-05,
+      "loss": 1.2454,
+      "step": 8450
+    },
+    {
+      "epoch": 0.574466639489061,
+      "grad_norm": 5.25,
+      "learning_rate": 4.640958350319337e-05,
+      "loss": 1.1877,
+      "step": 8455
+    },
+    {
+      "epoch": 0.5748063595597228,
+      "grad_norm": 5.5,
+      "learning_rate": 4.6407460252751734e-05,
+      "loss": 1.2887,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5751460796303846,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.64053370023101e-05,
+      "loss": 1.2995,
+      "step": 8465
+    },
+    {
+      "epoch": 0.5754857997010463,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.640321375186846e-05,
+      "loss": 1.3044,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5758255197717081,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.6401090501426826e-05,
+      "loss": 1.2253,
+      "step": 8475
+    },
+    {
+      "epoch": 0.5761652398423699,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.639896725098519e-05,
+      "loss": 1.2796,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5765049599130316,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.6396844000543554e-05,
+      "loss": 1.2163,
+      "step": 8485
+    },
+    {
+      "epoch": 0.5768446799836935,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.639472075010192e-05,
+      "loss": 1.2452,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5771844000543552,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.639259749966028e-05,
+      "loss": 1.2519,
+      "step": 8495
+    },
+    {
+      "epoch": 0.577524120125017,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.6390474249218646e-05,
+      "loss": 1.3208,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5778638401956787,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.638835099877701e-05,
+      "loss": 1.2545,
+      "step": 8505
+    },
+    {
+      "epoch": 0.5782035602663406,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.6386227748335374e-05,
+      "loss": 1.2588,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5785432803370023,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.638410449789374e-05,
+      "loss": 1.2683,
+      "step": 8515
+    },
+    {
+      "epoch": 0.578883000407664,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.63819812474521e-05,
+      "loss": 1.2997,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5792227204783259,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.6379857997010466e-05,
+      "loss": 1.2535,
+      "step": 8525
+    },
+    {
+      "epoch": 0.5795624405489876,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.637773474656883e-05,
+      "loss": 1.2443,
+      "step": 8530
+    },
+    {
+      "epoch": 0.5799021606196494,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6375611496127194e-05,
+      "loss": 1.2408,
+      "step": 8535
+    },
+    {
+      "epoch": 0.5802418806903111,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.637348824568556e-05,
+      "loss": 1.258,
+      "step": 8540
+    },
+    {
+      "epoch": 0.580581600760973,
+      "grad_norm": 4.75,
+      "learning_rate": 4.637136499524392e-05,
+      "loss": 1.3946,
+      "step": 8545
+    },
+    {
+      "epoch": 0.5809213208316347,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.6369241744802286e-05,
+      "loss": 1.3208,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5812610409022965,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.636711849436065e-05,
+      "loss": 1.243,
+      "step": 8555
+    },
+    {
+      "epoch": 0.5816007609729583,
+      "grad_norm": 5.375,
+      "learning_rate": 4.6364995243919014e-05,
+      "loss": 1.2655,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5819404810436201,
+      "grad_norm": 5.0,
+      "learning_rate": 4.636287199347737e-05,
+      "loss": 1.244,
+      "step": 8565
+    },
+    {
+      "epoch": 0.5822802011142818,
+      "grad_norm": 5.75,
+      "learning_rate": 4.636074874303574e-05,
+      "loss": 1.3307,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5826199211849437,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.6358625492594106e-05,
+      "loss": 1.3296,
+      "step": 8575
+    },
+    {
+      "epoch": 0.5829596412556054,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.6356502242152464e-05,
+      "loss": 1.3259,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5832993613262671,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.6354378991710834e-05,
+      "loss": 1.2397,
+      "step": 8585
+    },
+    {
+      "epoch": 0.5836390813969289,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.63522557412692e-05,
+      "loss": 1.3567,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5839788014675907,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.6350132490827556e-05,
+      "loss": 1.2662,
+      "step": 8595
+    },
+    {
+      "epoch": 0.5843185215382525,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.6348009240385926e-05,
+      "loss": 1.1275,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5846582416089142,
+      "grad_norm": 5.5,
+      "learning_rate": 4.634588598994429e-05,
+      "loss": 1.4346,
+      "step": 8605
+    },
+    {
+      "epoch": 0.5849979616795761,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.634376273950265e-05,
+      "loss": 1.3095,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5853376817502378,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.634163948906102e-05,
+      "loss": 1.311,
+      "step": 8615
+    },
+    {
+      "epoch": 0.5856774018208996,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.633951623861938e-05,
+      "loss": 1.294,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5860171218915613,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.633739298817774e-05,
+      "loss": 1.3598,
+      "step": 8625
+    },
+    {
+      "epoch": 0.5863568419622232,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.633526973773611e-05,
+      "loss": 1.3204,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5866965620328849,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.633314648729447e-05,
+      "loss": 1.3825,
+      "step": 8635
+    },
+    {
+      "epoch": 0.5870362821035466,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.633102323685283e-05,
+      "loss": 1.1467,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5873760021742085,
+      "grad_norm": 5.5,
+      "learning_rate": 4.63288999864112e-05,
+      "loss": 1.317,
+      "step": 8645
+    },
+    {
+      "epoch": 0.5877157222448702,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.632677673596956e-05,
+      "loss": 1.4617,
+      "step": 8650
+    },
+    {
+      "epoch": 0.588055442315532,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.6324653485527924e-05,
+      "loss": 1.2301,
+      "step": 8655
+    },
+    {
+      "epoch": 0.5883951623861938,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.6322530235086294e-05,
+      "loss": 1.3363,
+      "step": 8660
+    },
+    {
+      "epoch": 0.5887348824568556,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.632040698464465e-05,
+      "loss": 1.2764,
+      "step": 8665
+    },
+    {
+      "epoch": 0.5890746025275173,
+      "grad_norm": 5.0,
+      "learning_rate": 4.6318283734203016e-05,
+      "loss": 1.3054,
+      "step": 8670
+    },
+    {
+      "epoch": 0.589414322598179,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.6316160483761386e-05,
+      "loss": 1.4148,
+      "step": 8675
+    },
+    {
+      "epoch": 0.5897540426688409,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.6314037233319744e-05,
+      "loss": 1.32,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5900937627395026,
+      "grad_norm": 5.125,
+      "learning_rate": 4.631191398287811e-05,
+      "loss": 1.3022,
+      "step": 8685
+    },
+    {
+      "epoch": 0.5904334828101644,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.630979073243648e-05,
+      "loss": 1.1626,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5907732028808262,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.6307667481994836e-05,
+      "loss": 1.1883,
+      "step": 8695
+    },
+    {
+      "epoch": 0.591112922951488,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.6305544231553206e-05,
+      "loss": 1.3224,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5914526430221497,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.630342098111157e-05,
+      "loss": 1.2544,
+      "step": 8705
+    },
+    {
+      "epoch": 0.5917923630928115,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.630129773066993e-05,
+      "loss": 1.3141,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5921320831634733,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.62991744802283e-05,
+      "loss": 1.15,
+      "step": 8715
+    },
+    {
+      "epoch": 0.5924718032341351,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.6297051229786656e-05,
+      "loss": 1.2404,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5928115233047968,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.629492797934502e-05,
+      "loss": 1.3024,
+      "step": 8725
+    },
+    {
+      "epoch": 0.5931512433754587,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.629280472890339e-05,
+      "loss": 1.2826,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5934909634461204,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.629068147846175e-05,
+      "loss": 1.3584,
+      "step": 8735
+    },
+    {
+      "epoch": 0.5938306835167821,
+      "grad_norm": 5.125,
+      "learning_rate": 4.628855822802011e-05,
+      "loss": 1.2784,
+      "step": 8740
+    },
+    {
+      "epoch": 0.594170403587444,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.628643497757848e-05,
+      "loss": 1.2886,
+      "step": 8745
+    },
+    {
+      "epoch": 0.5945101236581057,
+      "grad_norm": 5.125,
+      "learning_rate": 4.628431172713684e-05,
+      "loss": 1.3594,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5948498437287675,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.6282188476695204e-05,
+      "loss": 1.2569,
+      "step": 8755
+    },
+    {
+      "epoch": 0.5951895637994292,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.6280065226253575e-05,
+      "loss": 1.3054,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5955292838700911,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.627794197581193e-05,
+      "loss": 1.2826,
+      "step": 8765
+    },
+    {
+      "epoch": 0.5958690039407528,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.6275818725370296e-05,
+      "loss": 1.2719,
+      "step": 8770
+    },
+    {
+      "epoch": 0.5962087240114146,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6273695474928667e-05,
+      "loss": 1.2745,
+      "step": 8775
+    },
+    {
+      "epoch": 0.5965484440820764,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6271572224487024e-05,
+      "loss": 1.2781,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5968881641527382,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.626944897404539e-05,
+      "loss": 1.2127,
+      "step": 8785
+    },
+    {
+      "epoch": 0.5972278842233999,
+      "grad_norm": 4.625,
+      "learning_rate": 4.626732572360375e-05,
+      "loss": 1.3155,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5975676042940616,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.6265202473162116e-05,
+      "loss": 1.3428,
+      "step": 8795
+    },
+    {
+      "epoch": 0.5979073243647235,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.626307922272048e-05,
+      "loss": 1.2393,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5982470444353852,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.6260955972278844e-05,
+      "loss": 1.288,
+      "step": 8805
+    },
+    {
+      "epoch": 0.598586764506047,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.625883272183721e-05,
+      "loss": 1.2869,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5989264845767088,
+      "grad_norm": 5.5,
+      "learning_rate": 4.625670947139557e-05,
+      "loss": 1.2777,
+      "step": 8815
+    },
+    {
+      "epoch": 0.5992662046473706,
+      "grad_norm": 5.125,
+      "learning_rate": 4.6254586220953936e-05,
+      "loss": 1.2057,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5996059247180323,
+      "grad_norm": 4.875,
+      "learning_rate": 4.62524629705123e-05,
+      "loss": 1.2184,
+      "step": 8825
+    },
+    {
+      "epoch": 0.5999456447886942,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.6250339720070664e-05,
+      "loss": 1.223,
+      "step": 8830
+    },
+    {
+      "epoch": 0.6002853648593559,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.624821646962903e-05,
+      "loss": 1.3641,
+      "step": 8835
+    },
+    {
+      "epoch": 0.6006250849300176,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.624609321918739e-05,
+      "loss": 1.3557,
+      "step": 8840
+    },
+    {
+      "epoch": 0.6009648050006794,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.6243969968745756e-05,
+      "loss": 1.3111,
+      "step": 8845
+    },
+    {
+      "epoch": 0.6013045250713412,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.624184671830412e-05,
+      "loss": 1.3215,
+      "step": 8850
+    },
+    {
+      "epoch": 0.601644245142003,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.6239723467862484e-05,
+      "loss": 1.2664,
+      "step": 8855
+    },
+    {
+      "epoch": 0.6019839652126647,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.623760021742085e-05,
+      "loss": 1.3083,
+      "step": 8860
+    },
+    {
+      "epoch": 0.6023236852833266,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.623547696697921e-05,
+      "loss": 1.2243,
+      "step": 8865
+    },
+    {
+      "epoch": 0.6026634053539883,
+      "grad_norm": 5.375,
+      "learning_rate": 4.6233353716537576e-05,
+      "loss": 1.3224,
+      "step": 8870
+    },
+    {
+      "epoch": 0.6030031254246501,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.623123046609594e-05,
+      "loss": 1.2478,
+      "step": 8875
+    },
+    {
+      "epoch": 0.6033428454953118,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.6229107215654304e-05,
+      "loss": 1.1578,
+      "step": 8880
+    },
+    {
+      "epoch": 0.6036825655659737,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.622698396521267e-05,
+      "loss": 1.2538,
+      "step": 8885
+    },
+    {
+      "epoch": 0.6040222856366354,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.622486071477103e-05,
+      "loss": 1.3647,
+      "step": 8890
+    },
+    {
+      "epoch": 0.6043620057072971,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.6222737464329396e-05,
+      "loss": 1.2043,
+      "step": 8895
+    },
+    {
+      "epoch": 0.604701725777959,
+      "grad_norm": 5.75,
+      "learning_rate": 4.622061421388776e-05,
+      "loss": 1.3247,
+      "step": 8900
+    },
+    {
+      "epoch": 0.6050414458486207,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.6218490963446124e-05,
+      "loss": 1.2236,
+      "step": 8905
+    },
+    {
+      "epoch": 0.6053811659192825,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.621636771300449e-05,
+      "loss": 1.309,
+      "step": 8910
+    },
+    {
+      "epoch": 0.6057208859899443,
+      "grad_norm": 5.25,
+      "learning_rate": 4.621424446256285e-05,
+      "loss": 1.1321,
+      "step": 8915
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.621212121212121e-05,
+      "loss": 1.3273,
+      "step": 8920
+    },
+    {
+      "epoch": 0.6064003261312678,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.620999796167958e-05,
+      "loss": 1.2599,
+      "step": 8925
+    },
+    {
+      "epoch": 0.6067400462019296,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.6207874711237944e-05,
+      "loss": 1.1954,
+      "step": 8930
+    },
+    {
+      "epoch": 0.6070797662725914,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.62057514607963e-05,
+      "loss": 1.3217,
+      "step": 8935
+    },
+    {
+      "epoch": 0.6074194863432532,
+      "grad_norm": 5.375,
+      "learning_rate": 4.620362821035467e-05,
+      "loss": 1.3086,
+      "step": 8940
+    },
+    {
+      "epoch": 0.6077592064139149,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.6201504959913036e-05,
+      "loss": 1.2125,
+      "step": 8945
+    },
+    {
+      "epoch": 0.6080989264845768,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.619938170947139e-05,
+      "loss": 1.2311,
+      "step": 8950
+    },
+    {
+      "epoch": 0.6084386465552385,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6197258459029764e-05,
+      "loss": 1.2788,
+      "step": 8955
+    },
+    {
+      "epoch": 0.6087783666259002,
+      "grad_norm": 5.375,
+      "learning_rate": 4.619513520858813e-05,
+      "loss": 1.2621,
+      "step": 8960
+    },
+    {
+      "epoch": 0.609118086696562,
+      "grad_norm": 5.75,
+      "learning_rate": 4.6193011958146485e-05,
+      "loss": 1.2195,
+      "step": 8965
+    },
+    {
+      "epoch": 0.6094578067672238,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.6190888707704856e-05,
+      "loss": 1.3592,
+      "step": 8970
+    },
+    {
+      "epoch": 0.6097975268378856,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.618876545726322e-05,
+      "loss": 1.2146,
+      "step": 8975
+    },
+    {
+      "epoch": 0.6101372469085473,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.618664220682158e-05,
+      "loss": 1.2243,
+      "step": 8980
+    },
+    {
+      "epoch": 0.6104769669792092,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.618451895637995e-05,
+      "loss": 1.2375,
+      "step": 8985
+    },
+    {
+      "epoch": 0.6108166870498709,
+      "grad_norm": 4.875,
+      "learning_rate": 4.6182395705938305e-05,
+      "loss": 1.3177,
+      "step": 8990
+    },
+    {
+      "epoch": 0.6111564071205327,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.618027245549667e-05,
+      "loss": 1.2269,
+      "step": 8995
+    },
+    {
+      "epoch": 0.6114961271911945,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.617814920505504e-05,
+      "loss": 1.2988,
+      "step": 9000
+    },
+    {
+      "epoch": 0.6118358472618562,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.61760259546134e-05,
+      "loss": 1.2661,
+      "step": 9005
+    },
+    {
+      "epoch": 0.612175567332518,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.617390270417176e-05,
+      "loss": 1.2634,
+      "step": 9010
+    },
+    {
+      "epoch": 0.6125152874031797,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.617177945373013e-05,
+      "loss": 1.2626,
+      "step": 9015
+    },
+    {
+      "epoch": 0.6128550074738416,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.616965620328849e-05,
+      "loss": 1.1738,
+      "step": 9020
+    },
+    {
+      "epoch": 0.6131947275445033,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.616753295284685e-05,
+      "loss": 1.1746,
+      "step": 9025
+    },
+    {
+      "epoch": 0.6135344476151651,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.6165409702405224e-05,
+      "loss": 1.2452,
+      "step": 9030
+    },
+    {
+      "epoch": 0.6138741676858269,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.616328645196358e-05,
+      "loss": 1.3139,
+      "step": 9035
+    },
+    {
+      "epoch": 0.6142138877564887,
+      "grad_norm": 5.125,
+      "learning_rate": 4.616116320152195e-05,
+      "loss": 1.3551,
+      "step": 9040
+    },
+    {
+      "epoch": 0.6145536078271504,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.6159039951080316e-05,
+      "loss": 1.3148,
+      "step": 9045
+    },
+    {
+      "epoch": 0.6148933278978121,
+      "grad_norm": 5.25,
+      "learning_rate": 4.615691670063867e-05,
+      "loss": 1.293,
+      "step": 9050
+    },
+    {
+      "epoch": 0.615233047968474,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.6154793450197044e-05,
+      "loss": 1.427,
+      "step": 9055
+    },
+    {
+      "epoch": 0.6155727680391357,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.61526701997554e-05,
+      "loss": 1.2767,
+      "step": 9060
+    },
+    {
+      "epoch": 0.6159124881097975,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.6150546949313765e-05,
+      "loss": 1.3096,
+      "step": 9065
+    },
+    {
+      "epoch": 0.6162522081804593,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.6148423698872136e-05,
+      "loss": 1.2027,
+      "step": 9070
+    },
+    {
+      "epoch": 0.6165919282511211,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.614630044843049e-05,
+      "loss": 1.1803,
+      "step": 9075
+    },
+    {
+      "epoch": 0.6169316483217828,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.614417719798886e-05,
+      "loss": 1.257,
+      "step": 9080
+    },
+    {
+      "epoch": 0.6172713683924447,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.614205394754723e-05,
+      "loss": 1.1965,
+      "step": 9085
+    },
+    {
+      "epoch": 0.6176110884631064,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.6139930697105585e-05,
+      "loss": 1.2094,
+      "step": 9090
+    },
+    {
+      "epoch": 0.6179508085337682,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.613780744666395e-05,
+      "loss": 1.2616,
+      "step": 9095
+    },
+    {
+      "epoch": 0.6182905286044299,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.613568419622232e-05,
+      "loss": 1.3752,
+      "step": 9100
+    },
+    {
+      "epoch": 0.6186302486750918,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.613356094578068e-05,
+      "loss": 1.2376,
+      "step": 9105
+    },
+    {
+      "epoch": 0.6189699687457535,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.613143769533904e-05,
+      "loss": 1.1798,
+      "step": 9110
+    },
+    {
+      "epoch": 0.6193096888164152,
+      "grad_norm": 6.0,
+      "learning_rate": 4.612931444489741e-05,
+      "loss": 1.2912,
+      "step": 9115
+    },
+    {
+      "epoch": 0.6196494088870771,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.612719119445577e-05,
+      "loss": 1.3155,
+      "step": 9120
+    },
+    {
+      "epoch": 0.6199891289577388,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.612506794401413e-05,
+      "loss": 1.2837,
+      "step": 9125
+    },
+    {
+      "epoch": 0.6203288490284006,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.6122944693572504e-05,
+      "loss": 1.2835,
+      "step": 9130
+    },
+    {
+      "epoch": 0.6206685690990623,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.612082144313086e-05,
+      "loss": 1.3101,
+      "step": 9135
+    },
+    {
+      "epoch": 0.6210082891697242,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.6118698192689225e-05,
+      "loss": 1.2643,
+      "step": 9140
+    },
+    {
+      "epoch": 0.6213480092403859,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.611657494224759e-05,
+      "loss": 1.2447,
+      "step": 9145
+    },
+    {
+      "epoch": 0.6216877293110477,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.611445169180595e-05,
+      "loss": 1.3615,
+      "step": 9150
+    },
+    {
+      "epoch": 0.6220274493817095,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.611232844136432e-05,
+      "loss": 1.278,
+      "step": 9155
+    },
+    {
+      "epoch": 0.6223671694523712,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.611020519092268e-05,
+      "loss": 1.3364,
+      "step": 9160
+    },
+    {
+      "epoch": 0.622706889523033,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.6108081940481045e-05,
+      "loss": 1.2677,
+      "step": 9165
+    },
+    {
+      "epoch": 0.6230466095936948,
+      "grad_norm": 5.25,
+      "learning_rate": 4.610595869003941e-05,
+      "loss": 1.2083,
+      "step": 9170
+    },
+    {
+      "epoch": 0.6233863296643566,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.610383543959777e-05,
+      "loss": 1.2874,
+      "step": 9175
+    },
+    {
+      "epoch": 0.6237260497350183,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.610171218915614e-05,
+      "loss": 1.2554,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6240657698056801,
+      "grad_norm": 6.375,
+      "learning_rate": 4.60995889387145e-05,
+      "loss": 1.3699,
+      "step": 9185
+    },
+    {
+      "epoch": 0.6244054898763419,
+      "grad_norm": 5.375,
+      "learning_rate": 4.6097465688272865e-05,
+      "loss": 1.3591,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6247452099470037,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.609534243783123e-05,
+      "loss": 1.3554,
+      "step": 9195
+    },
+    {
+      "epoch": 0.6250849300176654,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.609321918738959e-05,
+      "loss": 1.2051,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6254246500883273,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.609109593694796e-05,
+      "loss": 1.3077,
+      "step": 9205
+    },
+    {
+      "epoch": 0.625764370158989,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.608897268650632e-05,
+      "loss": 1.3452,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6261040902296507,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.6086849436064685e-05,
+      "loss": 1.2494,
+      "step": 9215
+    },
+    {
+      "epoch": 0.6264438103003126,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.608472618562305e-05,
+      "loss": 1.3769,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6267835303709743,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.608260293518141e-05,
+      "loss": 1.3091,
+      "step": 9225
+    },
+    {
+      "epoch": 0.6271232504416361,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.608047968473978e-05,
+      "loss": 1.2878,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6274629705122978,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.607835643429814e-05,
+      "loss": 1.2847,
+      "step": 9235
+    },
+    {
+      "epoch": 0.6278026905829597,
+      "grad_norm": 4.75,
+      "learning_rate": 4.6076233183856505e-05,
+      "loss": 1.2532,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6281424106536214,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.607410993341487e-05,
+      "loss": 1.3262,
+      "step": 9245
+    },
+    {
+      "epoch": 0.6284821307242832,
+      "grad_norm": 10.0625,
+      "learning_rate": 4.607198668297323e-05,
+      "loss": 1.2079,
+      "step": 9250
+    },
+    {
+      "epoch": 0.628821850794945,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.60698634325316e-05,
+      "loss": 1.2295,
+      "step": 9255
+    },
+    {
+      "epoch": 0.6291615708656068,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.6067740182089955e-05,
+      "loss": 1.3491,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6295012909362685,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.6065616931648325e-05,
+      "loss": 1.1676,
+      "step": 9265
+    },
+    {
+      "epoch": 0.6298410110069302,
+      "grad_norm": 5.75,
+      "learning_rate": 4.606349368120669e-05,
+      "loss": 1.2518,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6301807310775921,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.6061370430765047e-05,
+      "loss": 1.2595,
+      "step": 9275
+    },
+    {
+      "epoch": 0.6305204511482538,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.605924718032342e-05,
+      "loss": 1.26,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6308601712189156,
+      "grad_norm": 5.375,
+      "learning_rate": 4.605712392988178e-05,
+      "loss": 1.2998,
+      "step": 9285
+    },
+    {
+      "epoch": 0.6311998912895774,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.605500067944014e-05,
+      "loss": 1.2555,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6315396113602392,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.605287742899851e-05,
+      "loss": 1.3485,
+      "step": 9295
+    },
+    {
+      "epoch": 0.6318793314309009,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.6050754178556873e-05,
+      "loss": 1.2661,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6322190515015628,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.604863092811523e-05,
+      "loss": 1.1961,
+      "step": 9305
+    },
+    {
+      "epoch": 0.6325587715722245,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.60465076776736e-05,
+      "loss": 1.2369,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6328984916428863,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.6044384427231965e-05,
+      "loss": 1.2094,
+      "step": 9315
+    },
+    {
+      "epoch": 0.633238211713548,
+      "grad_norm": 5.125,
+      "learning_rate": 4.604226117679032e-05,
+      "loss": 1.215,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6335779317842098,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.6040137926348693e-05,
+      "loss": 1.3042,
+      "step": 9325
+    },
+    {
+      "epoch": 0.6339176518548716,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.603801467590706e-05,
+      "loss": 1.3264,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6342573719255333,
+      "grad_norm": 5.625,
+      "learning_rate": 4.6035891425465415e-05,
+      "loss": 1.2523,
+      "step": 9335
+    },
+    {
+      "epoch": 0.6345970919961952,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.6033768175023785e-05,
+      "loss": 1.2863,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6349368120668569,
+      "grad_norm": 6.125,
+      "learning_rate": 4.603164492458214e-05,
+      "loss": 1.2609,
+      "step": 9345
+    },
+    {
+      "epoch": 0.6352765321375187,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.602952167414051e-05,
+      "loss": 1.2191,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6356162522081804,
+      "grad_norm": 5.375,
+      "learning_rate": 4.602739842369888e-05,
+      "loss": 1.2493,
+      "step": 9355
+    },
+    {
+      "epoch": 0.6359559722788423,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.6025275173257235e-05,
+      "loss": 1.252,
+      "step": 9360
+    },
+    {
+      "epoch": 0.636295692349504,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.60231519228156e-05,
+      "loss": 1.19,
+      "step": 9365
+    },
+    {
+      "epoch": 0.6366354124201657,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.602102867237397e-05,
+      "loss": 1.299,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6369751324908276,
+      "grad_norm": 5.0,
+      "learning_rate": 4.601890542193233e-05,
+      "loss": 1.1535,
+      "step": 9375
+    },
+    {
+      "epoch": 0.6373148525614893,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.60167821714907e-05,
+      "loss": 1.2618,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6376545726321511,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.601465892104906e-05,
+      "loss": 1.2647,
+      "step": 9385
+    },
+    {
+      "epoch": 0.6379942927028129,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.601253567060742e-05,
+      "loss": 1.2413,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6383340127734747,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.601041242016579e-05,
+      "loss": 1.2101,
+      "step": 9395
+    },
+    {
+      "epoch": 0.6386737328441364,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.6008289169724153e-05,
+      "loss": 1.2573,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6390134529147982,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.600616591928251e-05,
+      "loss": 1.244,
+      "step": 9405
+    },
+    {
+      "epoch": 0.63935317298546,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.600404266884088e-05,
+      "loss": 1.2899,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6396928930561218,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.600191941839924e-05,
+      "loss": 1.2236,
+      "step": 9415
+    },
+    {
+      "epoch": 0.6400326131267835,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.59997961679576e-05,
+      "loss": 1.2567,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6403723331974454,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.5997672917515973e-05,
+      "loss": 1.3312,
+      "step": 9425
+    },
+    {
+      "epoch": 0.6407120532681071,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.599554966707433e-05,
+      "loss": 1.242,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6410517733387688,
+      "grad_norm": 5.75,
+      "learning_rate": 4.5993426416632695e-05,
+      "loss": 1.1519,
+      "step": 9435
+    },
+    {
+      "epoch": 0.6413914934094306,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.5991303166191066e-05,
+      "loss": 1.2113,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6417312134800924,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.598917991574942e-05,
+      "loss": 1.3994,
+      "step": 9445
+    },
+    {
+      "epoch": 0.6420709335507542,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.598705666530779e-05,
+      "loss": 1.245,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6424106536214159,
+      "grad_norm": 5.5,
+      "learning_rate": 4.598493341486616e-05,
+      "loss": 1.2482,
+      "step": 9455
+    },
+    {
+      "epoch": 0.6427503736920778,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.5982810164424515e-05,
+      "loss": 1.3309,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6430900937627395,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.598068691398288e-05,
+      "loss": 1.1821,
+      "step": 9465
+    },
+    {
+      "epoch": 0.6434298138334013,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.597856366354125e-05,
+      "loss": 1.311,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6437695339040631,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.597644041309961e-05,
+      "loss": 1.2295,
+      "step": 9475
+    },
+    {
+      "epoch": 0.6441092539747248,
+      "grad_norm": 5.375,
+      "learning_rate": 4.597431716265797e-05,
+      "loss": 1.2747,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6444489740453866,
+      "grad_norm": 5.75,
+      "learning_rate": 4.5972193912216335e-05,
+      "loss": 1.3607,
+      "step": 9485
+    },
+    {
+      "epoch": 0.6447886941160483,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.59700706617747e-05,
+      "loss": 1.2933,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6451284141867102,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.596794741133306e-05,
+      "loss": 1.2617,
+      "step": 9495
+    },
+    {
+      "epoch": 0.6454681342573719,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.596582416089143e-05,
+      "loss": 1.2561,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6458078543280337,
+      "grad_norm": 4.875,
+      "learning_rate": 4.596370091044979e-05,
+      "loss": 1.2,
+      "step": 9505
+    },
+    {
+      "epoch": 0.6461475743986955,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5961577660008155e-05,
+      "loss": 1.2551,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6464872944693573,
+      "grad_norm": 5.75,
+      "learning_rate": 4.595945440956652e-05,
+      "loss": 1.3668,
+      "step": 9515
+    },
+    {
+      "epoch": 0.646827014540019,
+      "grad_norm": 5.0,
+      "learning_rate": 4.595733115912488e-05,
+      "loss": 1.2466,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6471667346106807,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.595520790868325e-05,
+      "loss": 1.238,
+      "step": 9525
+    },
+    {
+      "epoch": 0.6475064546813426,
+      "grad_norm": 5.125,
+      "learning_rate": 4.595308465824161e-05,
+      "loss": 1.2265,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6478461747520043,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.5950961407799975e-05,
+      "loss": 1.2898,
+      "step": 9535
+    },
+    {
+      "epoch": 0.6481858948226661,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.594883815735834e-05,
+      "loss": 1.3234,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6485256148933279,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.59467149069167e-05,
+      "loss": 1.2897,
+      "step": 9545
+    },
+    {
+      "epoch": 0.6488653349639897,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.594459165647507e-05,
+      "loss": 1.3401,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6492050550346514,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.594246840603343e-05,
+      "loss": 1.2884,
+      "step": 9555
+    },
+    {
+      "epoch": 0.6495447751053133,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.5940345155591795e-05,
+      "loss": 1.1979,
+      "step": 9560
+    },
+    {
+      "epoch": 0.649884495175975,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.593822190515016e-05,
+      "loss": 1.3187,
+      "step": 9565
+    },
+    {
+      "epoch": 0.6502242152466368,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.593609865470852e-05,
+      "loss": 1.2712,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6505639353172985,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.593397540426689e-05,
+      "loss": 1.2773,
+      "step": 9575
+    },
+    {
+      "epoch": 0.6509036553879604,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.593185215382525e-05,
+      "loss": 1.2523,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6512433754586221,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.5929728903383615e-05,
+      "loss": 1.2234,
+      "step": 9585
+    },
+    {
+      "epoch": 0.6515830955292838,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.592760565294198e-05,
+      "loss": 1.2468,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6519228155999457,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.592548240250034e-05,
+      "loss": 1.3407,
+      "step": 9595
+    },
+    {
+      "epoch": 0.6522625356706074,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.592335915205871e-05,
+      "loss": 1.306,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6526022557412692,
+      "grad_norm": 5.0,
+      "learning_rate": 4.592123590161707e-05,
+      "loss": 1.2424,
+      "step": 9605
+    },
+    {
+      "epoch": 0.6529419758119309,
+      "grad_norm": 5.375,
+      "learning_rate": 4.5919112651175435e-05,
+      "loss": 1.2708,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6532816958825928,
+      "grad_norm": 5.375,
+      "learning_rate": 4.591698940073379e-05,
+      "loss": 1.2516,
+      "step": 9615
+    },
+    {
+      "epoch": 0.6536214159532545,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.591486615029216e-05,
+      "loss": 1.2337,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6539611360239163,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.591274289985053e-05,
+      "loss": 1.3061,
+      "step": 9625
+    },
+    {
+      "epoch": 0.6543008560945781,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5910619649408884e-05,
+      "loss": 1.2971,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6546405761652399,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5908496398967255e-05,
+      "loss": 1.2618,
+      "step": 9635
+    },
+    {
+      "epoch": 0.6549802962359016,
+      "grad_norm": 5.875,
+      "learning_rate": 4.590637314852562e-05,
+      "loss": 1.2658,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6553200163065634,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.5904249898083976e-05,
+      "loss": 1.2508,
+      "step": 9645
+    },
+    {
+      "epoch": 0.6556597363772252,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.590212664764235e-05,
+      "loss": 1.2738,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6559994564478869,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.590000339720071e-05,
+      "loss": 1.3039,
+      "step": 9655
+    },
+    {
+      "epoch": 0.6563391765185487,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.589788014675907e-05,
+      "loss": 1.3194,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6566788965892105,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.589575689631744e-05,
+      "loss": 1.3242,
+      "step": 9665
+    },
+    {
+      "epoch": 0.6570186166598723,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.58936336458758e-05,
+      "loss": 1.2012,
+      "step": 9670
+    },
+    {
+      "epoch": 0.657358336730534,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.589151039543416e-05,
+      "loss": 1.2053,
+      "step": 9675
+    },
+    {
+      "epoch": 0.6576980568011959,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.588938714499253e-05,
+      "loss": 1.2362,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6580377768718576,
+      "grad_norm": 5.625,
+      "learning_rate": 4.588726389455089e-05,
+      "loss": 1.1943,
+      "step": 9685
+    },
+    {
+      "epoch": 0.6583774969425193,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.588514064410925e-05,
+      "loss": 1.2278,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6587172170131811,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.588301739366762e-05,
+      "loss": 1.2271,
+      "step": 9695
+    },
+    {
+      "epoch": 0.6590569370838429,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.588089414322598e-05,
+      "loss": 1.2975,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6593966571545047,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.5878770892784344e-05,
+      "loss": 1.2387,
+      "step": 9705
+    },
+    {
+      "epoch": 0.6597363772251664,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.5876647642342715e-05,
+      "loss": 1.267,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6600760972958283,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.587452439190107e-05,
+      "loss": 1.2751,
+      "step": 9715
+    },
+    {
+      "epoch": 0.66041581736649,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.587240114145944e-05,
+      "loss": 1.3598,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6607555374371518,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.587027789101781e-05,
+      "loss": 1.3073,
+      "step": 9725
+    },
+    {
+      "epoch": 0.6610952575078136,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.5868154640576164e-05,
+      "loss": 1.1325,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6614349775784754,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.5866031390134535e-05,
+      "loss": 1.2744,
+      "step": 9735
+    },
+    {
+      "epoch": 0.6617746976491371,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.58639081396929e-05,
+      "loss": 1.3354,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6621144177197988,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.5861784889251256e-05,
+      "loss": 1.2749,
+      "step": 9745
+    },
+    {
+      "epoch": 0.6624541377904607,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.585966163880963e-05,
+      "loss": 1.2611,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6627938578611224,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.585753838836799e-05,
+      "loss": 1.2918,
+      "step": 9755
+    },
+    {
+      "epoch": 0.6631335779317842,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.585541513792635e-05,
+      "loss": 1.285,
+      "step": 9760
+    },
+    {
+      "epoch": 0.663473298002446,
+      "grad_norm": 5.125,
+      "learning_rate": 4.585329188748472e-05,
+      "loss": 1.158,
+      "step": 9765
+    },
+    {
+      "epoch": 0.6638130180731078,
+      "grad_norm": 5.875,
+      "learning_rate": 4.5851168637043076e-05,
+      "loss": 1.3793,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6641527381437695,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.584904538660144e-05,
+      "loss": 1.28,
+      "step": 9775
+    },
+    {
+      "epoch": 0.6644924582144313,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.584692213615981e-05,
+      "loss": 1.2922,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6648321782850931,
+      "grad_norm": 5.875,
+      "learning_rate": 4.584479888571817e-05,
+      "loss": 1.3759,
+      "step": 9785
+    },
+    {
+      "epoch": 0.6651718983557549,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.584267563527653e-05,
+      "loss": 1.288,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6655116184264166,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.58405523848349e-05,
+      "loss": 1.1265,
+      "step": 9795
+    },
+    {
+      "epoch": 0.6658513384970784,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.583842913439326e-05,
+      "loss": 1.1952,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6661910585677402,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.5836305883951624e-05,
+      "loss": 1.2109,
+      "step": 9805
+    },
+    {
+      "epoch": 0.6665307786384019,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.5834182633509995e-05,
+      "loss": 1.2692,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6668704987090638,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.583205938306835e-05,
+      "loss": 1.2534,
+      "step": 9815
+    },
+    {
+      "epoch": 0.6672102187797255,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5829936132626716e-05,
+      "loss": 1.2357,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6675499388503873,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.582781288218509e-05,
+      "loss": 1.2261,
+      "step": 9825
+    },
+    {
+      "epoch": 0.667889658921049,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.5825689631743444e-05,
+      "loss": 1.2738,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6682293789917109,
+      "grad_norm": 5.5,
+      "learning_rate": 4.582356638130181e-05,
+      "loss": 1.2145,
+      "step": 9835
+    },
+    {
+      "epoch": 0.6685690990623726,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.582144313086017e-05,
+      "loss": 1.3403,
+      "step": 9840
+    },
+    {
+      "epoch": 0.6689088191330343,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.5819319880418536e-05,
+      "loss": 1.2045,
+      "step": 9845
+    },
+    {
+      "epoch": 0.6692485392036962,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.58171966299769e-05,
+      "loss": 1.2238,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6695882592743579,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.5815073379535264e-05,
+      "loss": 1.3085,
+      "step": 9855
+    },
+    {
+      "epoch": 0.6699279793450197,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.581295012909363e-05,
+      "loss": 1.3845,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6702676994156814,
+      "grad_norm": 5.25,
+      "learning_rate": 4.581082687865199e-05,
+      "loss": 1.31,
+      "step": 9865
+    },
+    {
+      "epoch": 0.6706074194863433,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5808703628210356e-05,
+      "loss": 1.2779,
+      "step": 9870
+    },
+    {
+      "epoch": 0.670947139557005,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.580658037776872e-05,
+      "loss": 1.3958,
+      "step": 9875
+    },
+    {
+      "epoch": 0.6712868596276668,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.5804457127327084e-05,
+      "loss": 1.1384,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6716265796983286,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.580233387688545e-05,
+      "loss": 1.3754,
+      "step": 9885
+    },
+    {
+      "epoch": 0.6719662997689904,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.580021062644381e-05,
+      "loss": 1.286,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6723060198396521,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.5798087376002176e-05,
+      "loss": 1.1528,
+      "step": 9895
+    },
+    {
+      "epoch": 0.672645739910314,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.579596412556054e-05,
+      "loss": 1.3263,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6729854599809757,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.5793840875118904e-05,
+      "loss": 1.2865,
+      "step": 9905
+    },
+    {
+      "epoch": 0.6733251800516374,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.579171762467727e-05,
+      "loss": 1.293,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6736649001222992,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.578959437423563e-05,
+      "loss": 1.4205,
+      "step": 9915
+    },
+    {
+      "epoch": 0.674004620192961,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.5787471123793996e-05,
+      "loss": 1.255,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6743443402636228,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.578534787335236e-05,
+      "loss": 1.2826,
+      "step": 9925
+    },
+    {
+      "epoch": 0.6746840603342845,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.5783224622910724e-05,
+      "loss": 1.3954,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6750237804049464,
+      "grad_norm": 5.625,
+      "learning_rate": 4.578110137246909e-05,
+      "loss": 1.3249,
+      "step": 9935
+    },
+    {
+      "epoch": 0.6753635004756081,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.577897812202745e-05,
+      "loss": 1.2982,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6757032205462699,
+      "grad_norm": 5.375,
+      "learning_rate": 4.5776854871585816e-05,
+      "loss": 1.1891,
+      "step": 9945
+    },
+    {
+      "epoch": 0.6760429406169316,
+      "grad_norm": 5.125,
+      "learning_rate": 4.577473162114418e-05,
+      "loss": 1.3427,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6763826606875935,
+      "grad_norm": 5.125,
+      "learning_rate": 4.5772608370702544e-05,
+      "loss": 1.2026,
+      "step": 9955
+    },
+    {
+      "epoch": 0.6767223807582552,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.577048512026091e-05,
+      "loss": 1.1532,
+      "step": 9960
+    },
+    {
+      "epoch": 0.6770621008289169,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.576836186981927e-05,
+      "loss": 1.2601,
+      "step": 9965
+    },
+    {
+      "epoch": 0.6774018208995788,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.576623861937763e-05,
+      "loss": 1.2297,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6777415409702405,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5764115368936e-05,
+      "loss": 1.2468,
+      "step": 9975
+    },
+    {
+      "epoch": 0.6780812610409023,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.5761992118494364e-05,
+      "loss": 1.2502,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6784209811115641,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.575986886805272e-05,
+      "loss": 1.252,
+      "step": 9985
+    },
+    {
+      "epoch": 0.6787607011822259,
+      "grad_norm": 6.125,
+      "learning_rate": 4.575774561761109e-05,
+      "loss": 1.2785,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6791004212528876,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5755622367169456e-05,
+      "loss": 1.2802,
+      "step": 9995
+    },
+    {
+      "epoch": 0.6794401413235494,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.5753499116727814e-05,
+      "loss": 1.164,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6797798613942112,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.5751375866286184e-05,
+      "loss": 1.1912,
+      "step": 10005
+    },
+    {
+      "epoch": 0.680119581464873,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.574925261584455e-05,
+      "loss": 1.1964,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6804593015355347,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.5747129365402906e-05,
+      "loss": 1.3063,
+      "step": 10015
+    },
+    {
+      "epoch": 0.6807990216061965,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.5745006114961276e-05,
+      "loss": 1.2389,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6811387416768583,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.574288286451964e-05,
+      "loss": 1.2448,
+      "step": 10025
+    },
+    {
+      "epoch": 0.68147846174752,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.5740759614078e-05,
+      "loss": 1.1832,
+      "step": 10030
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 5.0,
+      "learning_rate": 4.573863636363637e-05,
+      "loss": 1.3057,
+      "step": 10035
+    },
+    {
+      "epoch": 0.6821579018888436,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.5736513113194726e-05,
+      "loss": 1.2938,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6824976219595054,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.573438986275309e-05,
+      "loss": 1.2715,
+      "step": 10045
+    },
+    {
+      "epoch": 0.6828373420301671,
+      "grad_norm": 5.375,
+      "learning_rate": 4.573226661231146e-05,
+      "loss": 1.1949,
+      "step": 10050
+    },
+    {
+      "epoch": 0.683177062100829,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.573014336186982e-05,
+      "loss": 1.1902,
+      "step": 10055
+    },
+    {
+      "epoch": 0.6835167821714907,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.572802011142819e-05,
+      "loss": 1.2855,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6838565022421524,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.572589686098655e-05,
+      "loss": 1.3178,
+      "step": 10065
+    },
+    {
+      "epoch": 0.6841962223128143,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.572377361054491e-05,
+      "loss": 1.299,
+      "step": 10070
+    },
+    {
+      "epoch": 0.684535942383476,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.572165036010328e-05,
+      "loss": 1.2114,
+      "step": 10075
+    },
+    {
+      "epoch": 0.6848756624541378,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5719527109661644e-05,
+      "loss": 1.2986,
+      "step": 10080
+    },
+    {
+      "epoch": 0.6852153825247995,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.571740385922e-05,
+      "loss": 1.3334,
+      "step": 10085
+    },
+    {
+      "epoch": 0.6855551025954614,
+      "grad_norm": 5.375,
+      "learning_rate": 4.571528060877837e-05,
+      "loss": 1.2277,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6858948226661231,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5713157358336736e-05,
+      "loss": 1.1869,
+      "step": 10095
+    },
+    {
+      "epoch": 0.6862345427367849,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5711034107895094e-05,
+      "loss": 1.2481,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6865742628074467,
+      "grad_norm": 5.0,
+      "learning_rate": 4.5708910857453464e-05,
+      "loss": 1.2824,
+      "step": 10105
+    },
+    {
+      "epoch": 0.6869139828781085,
+      "grad_norm": 5.0,
+      "learning_rate": 4.570678760701182e-05,
+      "loss": 1.1939,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6872537029487702,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5704664356570186e-05,
+      "loss": 1.2985,
+      "step": 10115
+    },
+    {
+      "epoch": 0.6875934230194319,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5702541106128556e-05,
+      "loss": 1.1774,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6879331430900938,
+      "grad_norm": 6.125,
+      "learning_rate": 4.5700417855686914e-05,
+      "loss": 1.3325,
+      "step": 10125
+    },
+    {
+      "epoch": 0.6882728631607555,
+      "grad_norm": 5.75,
+      "learning_rate": 4.569829460524528e-05,
+      "loss": 1.2715,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6886125832314173,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.569617135480365e-05,
+      "loss": 1.3227,
+      "step": 10135
+    },
+    {
+      "epoch": 0.6889523033020791,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5694048104362006e-05,
+      "loss": 1.3211,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6892920233727409,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.569192485392037e-05,
+      "loss": 1.3196,
+      "step": 10145
+    },
+    {
+      "epoch": 0.6896317434434026,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.568980160347874e-05,
+      "loss": 1.2015,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6899714635140645,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.56876783530371e-05,
+      "loss": 1.2614,
+      "step": 10155
+    },
+    {
+      "epoch": 0.6903111835847262,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.568555510259546e-05,
+      "loss": 1.2567,
+      "step": 10160
+    },
+    {
+      "epoch": 0.690650903655388,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.568343185215383e-05,
+      "loss": 1.2996,
+      "step": 10165
+    },
+    {
+      "epoch": 0.6909906237260497,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.568130860171219e-05,
+      "loss": 1.3715,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6913303437967115,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5679185351270554e-05,
+      "loss": 1.2282,
+      "step": 10175
+    },
+    {
+      "epoch": 0.6916700638673733,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.5677062100828925e-05,
+      "loss": 1.3319,
+      "step": 10180
+    },
+    {
+      "epoch": 0.692009783938035,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.567493885038728e-05,
+      "loss": 1.2085,
+      "step": 10185
+    },
+    {
+      "epoch": 0.6923495040086969,
+      "grad_norm": 4.0,
+      "learning_rate": 4.5672815599945646e-05,
+      "loss": 1.2394,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6926892240793586,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.567069234950401e-05,
+      "loss": 1.3128,
+      "step": 10195
+    },
+    {
+      "epoch": 0.6930289441500204,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5668569099062374e-05,
+      "loss": 1.2777,
+      "step": 10200
+    },
+    {
+      "epoch": 0.6933686642206821,
+      "grad_norm": 6.0,
+      "learning_rate": 4.566644584862074e-05,
+      "loss": 1.1837,
+      "step": 10205
+    },
+    {
+      "epoch": 0.693708384291344,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.56643225981791e-05,
+      "loss": 1.2459,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6940481043620057,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5662199347737466e-05,
+      "loss": 1.2757,
+      "step": 10215
+    },
+    {
+      "epoch": 0.6943878244326674,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.566007609729583e-05,
+      "loss": 1.3059,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6947275445033293,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.5657952846854194e-05,
+      "loss": 1.2508,
+      "step": 10225
+    },
+    {
+      "epoch": 0.695067264573991,
+      "grad_norm": 5.75,
+      "learning_rate": 4.565582959641256e-05,
+      "loss": 1.3,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6954069846446528,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.565370634597092e-05,
+      "loss": 1.3008,
+      "step": 10235
+    },
+    {
+      "epoch": 0.6957467047153146,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.5651583095529286e-05,
+      "loss": 1.2314,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6960864247859764,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.564945984508765e-05,
+      "loss": 1.3254,
+      "step": 10245
+    },
+    {
+      "epoch": 0.6964261448566381,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5647336594646014e-05,
+      "loss": 1.2573,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6967658649272999,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.564521334420438e-05,
+      "loss": 1.2348,
+      "step": 10255
+    },
+    {
+      "epoch": 0.6971055849979617,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.564309009376274e-05,
+      "loss": 1.2579,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6974453050686235,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.5640966843321106e-05,
+      "loss": 1.2431,
+      "step": 10265
+    },
+    {
+      "epoch": 0.6977850251392852,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.563884359287947e-05,
+      "loss": 1.2777,
+      "step": 10270
+    },
+    {
+      "epoch": 0.698124745209947,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.5636720342437834e-05,
+      "loss": 1.1663,
+      "step": 10275
+    },
+    {
+      "epoch": 0.6984644652806088,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.56345970919962e-05,
+      "loss": 1.1742,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6988041853512705,
+      "grad_norm": 5.5,
+      "learning_rate": 4.563247384155456e-05,
+      "loss": 1.344,
+      "step": 10285
+    },
+    {
+      "epoch": 0.6991439054219323,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.5630350591112926e-05,
+      "loss": 1.3715,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6994836254925941,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.562822734067129e-05,
+      "loss": 1.1226,
+      "step": 10295
+    },
+    {
+      "epoch": 0.6998233455632559,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.5626104090229654e-05,
+      "loss": 1.2804,
+      "step": 10300
+    },
+    {
+      "epoch": 0.7001630656339176,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.562398083978802e-05,
+      "loss": 1.2366,
+      "step": 10305
+    },
+    {
+      "epoch": 0.7005027857045795,
+      "grad_norm": 5.875,
+      "learning_rate": 4.5621857589346375e-05,
+      "loss": 1.2173,
+      "step": 10310
+    },
+    {
+      "epoch": 0.7008425057752412,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.5619734338904746e-05,
+      "loss": 1.233,
+      "step": 10315
+    },
+    {
+      "epoch": 0.701182225845903,
+      "grad_norm": 5.125,
+      "learning_rate": 4.561761108846311e-05,
+      "loss": 1.2059,
+      "step": 10320
+    },
+    {
+      "epoch": 0.7015219459165648,
+      "grad_norm": 5.625,
+      "learning_rate": 4.561548783802147e-05,
+      "loss": 1.2715,
+      "step": 10325
+    },
+    {
+      "epoch": 0.7018616659872265,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.561336458757984e-05,
+      "loss": 1.2521,
+      "step": 10330
+    },
+    {
+      "epoch": 0.7022013860578883,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.56112413371382e-05,
+      "loss": 1.2024,
+      "step": 10335
+    },
+    {
+      "epoch": 0.70254110612855,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.560911808669656e-05,
+      "loss": 1.2755,
+      "step": 10340
+    },
+    {
+      "epoch": 0.7028808261992119,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.560699483625493e-05,
+      "loss": 1.2655,
+      "step": 10345
+    },
+    {
+      "epoch": 0.7032205462698736,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5604871585813294e-05,
+      "loss": 1.3467,
+      "step": 10350
+    },
+    {
+      "epoch": 0.7035602663405354,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.560274833537165e-05,
+      "loss": 1.1767,
+      "step": 10355
+    },
+    {
+      "epoch": 0.7038999864111972,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.560062508493002e-05,
+      "loss": 1.2332,
+      "step": 10360
+    },
+    {
+      "epoch": 0.704239706481859,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.5598501834488386e-05,
+      "loss": 1.3015,
+      "step": 10365
+    },
+    {
+      "epoch": 0.7045794265525207,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.559637858404674e-05,
+      "loss": 1.2431,
+      "step": 10370
+    },
+    {
+      "epoch": 0.7049191466231824,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5594255333605114e-05,
+      "loss": 1.2682,
+      "step": 10375
+    },
+    {
+      "epoch": 0.7052588666938443,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.559213208316348e-05,
+      "loss": 1.3109,
+      "step": 10380
+    },
+    {
+      "epoch": 0.705598586764506,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.5590008832721835e-05,
+      "loss": 1.3583,
+      "step": 10385
+    },
+    {
+      "epoch": 0.7059383068351678,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.5587885582280206e-05,
+      "loss": 1.2557,
+      "step": 10390
+    },
+    {
+      "epoch": 0.7062780269058296,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.558576233183856e-05,
+      "loss": 1.2472,
+      "step": 10395
+    },
+    {
+      "epoch": 0.7066177469764914,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.5583639081396934e-05,
+      "loss": 1.2051,
+      "step": 10400
+    },
+    {
+      "epoch": 0.7069574670471531,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.55815158309553e-05,
+      "loss": 1.1987,
+      "step": 10405
+    },
+    {
+      "epoch": 0.707297187117815,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5579392580513655e-05,
+      "loss": 1.3148,
+      "step": 10410
+    },
+    {
+      "epoch": 0.7076369071884767,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.5577269330072026e-05,
+      "loss": 1.2894,
+      "step": 10415
+    },
+    {
+      "epoch": 0.7079766272591385,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.557514607963039e-05,
+      "loss": 1.314,
+      "step": 10420
+    },
+    {
+      "epoch": 0.7083163473298002,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.557302282918875e-05,
+      "loss": 1.393,
+      "step": 10425
+    },
+    {
+      "epoch": 0.708656067400462,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.557089957874712e-05,
+      "loss": 1.3065,
+      "step": 10430
+    },
+    {
+      "epoch": 0.7089957874711238,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.556877632830548e-05,
+      "loss": 1.3249,
+      "step": 10435
+    },
+    {
+      "epoch": 0.7093355075417855,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.556665307786384e-05,
+      "loss": 1.2996,
+      "step": 10440
+    },
+    {
+      "epoch": 0.7096752276124474,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.556452982742221e-05,
+      "loss": 1.3837,
+      "step": 10445
+    },
+    {
+      "epoch": 0.7100149476831091,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.5562406576980574e-05,
+      "loss": 1.2588,
+      "step": 10450
+    },
+    {
+      "epoch": 0.7103546677537709,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.556028332653893e-05,
+      "loss": 1.1709,
+      "step": 10455
+    },
+    {
+      "epoch": 0.7106943878244326,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.55581600760973e-05,
+      "loss": 1.3316,
+      "step": 10460
+    },
+    {
+      "epoch": 0.7110341078950945,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.555603682565566e-05,
+      "loss": 1.278,
+      "step": 10465
+    },
+    {
+      "epoch": 0.7113738279657562,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.555391357521402e-05,
+      "loss": 1.3689,
+      "step": 10470
+    },
+    {
+      "epoch": 0.711713548036418,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.5551790324772394e-05,
+      "loss": 1.2759,
+      "step": 10475
+    },
+    {
+      "epoch": 0.7120532681070798,
+      "grad_norm": 5.875,
+      "learning_rate": 4.554966707433075e-05,
+      "loss": 1.2785,
+      "step": 10480
+    },
+    {
+      "epoch": 0.7123929881777415,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5547543823889115e-05,
+      "loss": 1.1038,
+      "step": 10485
+    },
+    {
+      "epoch": 0.7127327082484033,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.5545420573447486e-05,
+      "loss": 1.3011,
+      "step": 10490
+    },
+    {
+      "epoch": 0.7130724283190651,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.554329732300584e-05,
+      "loss": 1.1444,
+      "step": 10495
+    },
+    {
+      "epoch": 0.7134121483897269,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.554117407256421e-05,
+      "loss": 1.2007,
+      "step": 10500
+    },
+    {
+      "epoch": 0.7137518684603886,
+      "grad_norm": 6.375,
+      "learning_rate": 4.553905082212258e-05,
+      "loss": 1.3057,
+      "step": 10505
+    },
+    {
+      "epoch": 0.7140915885310504,
+      "grad_norm": 5.125,
+      "learning_rate": 4.5536927571680935e-05,
+      "loss": 1.3409,
+      "step": 10510
+    },
+    {
+      "epoch": 0.7144313086017122,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.55348043212393e-05,
+      "loss": 1.3828,
+      "step": 10515
+    },
+    {
+      "epoch": 0.714771028672374,
+      "grad_norm": 5.375,
+      "learning_rate": 4.553268107079767e-05,
+      "loss": 1.2547,
+      "step": 10520
+    },
+    {
+      "epoch": 0.7151107487430357,
+      "grad_norm": 5.5,
+      "learning_rate": 4.553055782035603e-05,
+      "loss": 1.2209,
+      "step": 10525
+    },
+    {
+      "epoch": 0.7154504688136976,
+      "grad_norm": 5.375,
+      "learning_rate": 4.552843456991439e-05,
+      "loss": 1.2611,
+      "step": 10530
+    },
+    {
+      "epoch": 0.7157901888843593,
+      "grad_norm": 4.09375,
+      "learning_rate": 4.5526311319472755e-05,
+      "loss": 1.1357,
+      "step": 10535
+    },
+    {
+      "epoch": 0.716129908955021,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.552418806903112e-05,
+      "loss": 1.2194,
+      "step": 10540
+    },
+    {
+      "epoch": 0.7164696290256828,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.552206481858948e-05,
+      "loss": 1.2109,
+      "step": 10545
+    },
+    {
+      "epoch": 0.7168093490963446,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.551994156814785e-05,
+      "loss": 1.2435,
+      "step": 10550
+    },
+    {
+      "epoch": 0.7171490691670064,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.551781831770621e-05,
+      "loss": 1.2057,
+      "step": 10555
+    },
+    {
+      "epoch": 0.7174887892376681,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5515695067264575e-05,
+      "loss": 1.3731,
+      "step": 10560
+    },
+    {
+      "epoch": 0.71782850930833,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.551357181682294e-05,
+      "loss": 1.353,
+      "step": 10565
+    },
+    {
+      "epoch": 0.7181682293789917,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.55114485663813e-05,
+      "loss": 1.3868,
+      "step": 10570
+    },
+    {
+      "epoch": 0.7185079494496535,
+      "grad_norm": 6.0,
+      "learning_rate": 4.550932531593967e-05,
+      "loss": 1.3627,
+      "step": 10575
+    },
+    {
+      "epoch": 0.7188476695203153,
+      "grad_norm": 5.875,
+      "learning_rate": 4.550720206549803e-05,
+      "loss": 1.3212,
+      "step": 10580
+    },
+    {
+      "epoch": 0.719187389590977,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.5505078815056395e-05,
+      "loss": 1.316,
+      "step": 10585
+    },
+    {
+      "epoch": 0.7195271096616388,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.550295556461476e-05,
+      "loss": 1.3269,
+      "step": 10590
+    },
+    {
+      "epoch": 0.7198668297323005,
+      "grad_norm": 5.375,
+      "learning_rate": 4.550083231417312e-05,
+      "loss": 1.2001,
+      "step": 10595
+    },
+    {
+      "epoch": 0.7202065498029624,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.549870906373149e-05,
+      "loss": 1.253,
+      "step": 10600
+    },
+    {
+      "epoch": 0.7205462698736241,
+      "grad_norm": 5.125,
+      "learning_rate": 4.549658581328985e-05,
+      "loss": 1.257,
+      "step": 10605
+    },
+    {
+      "epoch": 0.7208859899442859,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5494462562848215e-05,
+      "loss": 1.2806,
+      "step": 10610
+    },
+    {
+      "epoch": 0.7212257100149477,
+      "grad_norm": 5.375,
+      "learning_rate": 4.549233931240658e-05,
+      "loss": 1.2643,
+      "step": 10615
+    },
+    {
+      "epoch": 0.7215654300856095,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.549021606196494e-05,
+      "loss": 1.3287,
+      "step": 10620
+    },
+    {
+      "epoch": 0.7219051501562712,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.548809281152331e-05,
+      "loss": 1.3351,
+      "step": 10625
+    },
+    {
+      "epoch": 0.722244870226933,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.548596956108167e-05,
+      "loss": 1.3092,
+      "step": 10630
+    },
+    {
+      "epoch": 0.7225845902975948,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5483846310640035e-05,
+      "loss": 1.2805,
+      "step": 10635
+    },
+    {
+      "epoch": 0.7229243103682566,
+      "grad_norm": 5.25,
+      "learning_rate": 4.54817230601984e-05,
+      "loss": 1.2075,
+      "step": 10640
+    },
+    {
+      "epoch": 0.7232640304389183,
+      "grad_norm": 5.625,
+      "learning_rate": 4.547959980975676e-05,
+      "loss": 1.211,
+      "step": 10645
+    },
+    {
+      "epoch": 0.7236037505095801,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.547747655931513e-05,
+      "loss": 1.231,
+      "step": 10650
+    },
+    {
+      "epoch": 0.7239434705802419,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.547535330887349e-05,
+      "loss": 1.2605,
+      "step": 10655
+    },
+    {
+      "epoch": 0.7242831906509036,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5473230058431855e-05,
+      "loss": 1.2751,
+      "step": 10660
+    },
+    {
+      "epoch": 0.7246229107215655,
+      "grad_norm": 5.75,
+      "learning_rate": 4.547110680799021e-05,
+      "loss": 1.3348,
+      "step": 10665
+    },
+    {
+      "epoch": 0.7249626307922272,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.546898355754858e-05,
+      "loss": 1.2613,
+      "step": 10670
+    },
+    {
+      "epoch": 0.725302350862889,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.546686030710695e-05,
+      "loss": 1.3,
+      "step": 10675
+    },
+    {
+      "epoch": 0.7256420709335507,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.5464737056665305e-05,
+      "loss": 1.14,
+      "step": 10680
+    },
+    {
+      "epoch": 0.7259817910042126,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5462613806223675e-05,
+      "loss": 1.189,
+      "step": 10685
+    },
+    {
+      "epoch": 0.7263215110748743,
+      "grad_norm": 5.25,
+      "learning_rate": 4.546049055578204e-05,
+      "loss": 1.2107,
+      "step": 10690
+    },
+    {
+      "epoch": 0.726661231145536,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.5458367305340397e-05,
+      "loss": 1.2643,
+      "step": 10695
+    },
+    {
+      "epoch": 0.7270009512161979,
+      "grad_norm": 5.125,
+      "learning_rate": 4.545624405489877e-05,
+      "loss": 1.3362,
+      "step": 10700
+    },
+    {
+      "epoch": 0.7273406712868596,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.545412080445713e-05,
+      "loss": 1.3114,
+      "step": 10705
+    },
+    {
+      "epoch": 0.7276803913575214,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.545199755401549e-05,
+      "loss": 1.2792,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7280201114281831,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.544987430357386e-05,
+      "loss": 1.2551,
+      "step": 10715
+    },
+    {
+      "epoch": 0.728359831498845,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.544775105313222e-05,
+      "loss": 1.2686,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7286995515695067,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.544562780269058e-05,
+      "loss": 1.2294,
+      "step": 10725
+    },
+    {
+      "epoch": 0.7290392716401685,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.544350455224895e-05,
+      "loss": 1.2285,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7293789917108303,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5441381301807315e-05,
+      "loss": 1.2146,
+      "step": 10735
+    },
+    {
+      "epoch": 0.7297187117814921,
+      "grad_norm": 5.125,
+      "learning_rate": 4.543925805136568e-05,
+      "loss": 1.344,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7300584318521538,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5437134800924043e-05,
+      "loss": 1.2754,
+      "step": 10745
+    },
+    {
+      "epoch": 0.7303981519228157,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.54350115504824e-05,
+      "loss": 1.1866,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7307378719934774,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.543288830004077e-05,
+      "loss": 1.3036,
+      "step": 10755
+    },
+    {
+      "epoch": 0.7310775920641391,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.5430765049599135e-05,
+      "loss": 1.2843,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7314173121348009,
+      "grad_norm": 5.5,
+      "learning_rate": 4.542864179915749e-05,
+      "loss": 1.2988,
+      "step": 10765
+    },
+    {
+      "epoch": 0.7317570322054627,
+      "grad_norm": 5.125,
+      "learning_rate": 4.5426518548715863e-05,
+      "loss": 1.1267,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7320967522761245,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.542439529827423e-05,
+      "loss": 1.3204,
+      "step": 10775
+    },
+    {
+      "epoch": 0.7324364723467862,
+      "grad_norm": 5.125,
+      "learning_rate": 4.5422272047832585e-05,
+      "loss": 1.3061,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7327761924174481,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.5420148797390955e-05,
+      "loss": 1.238,
+      "step": 10785
+    },
+    {
+      "epoch": 0.7331159124881098,
+      "grad_norm": 5.25,
+      "learning_rate": 4.541802554694932e-05,
+      "loss": 1.2602,
+      "step": 10790
+    },
+    {
+      "epoch": 0.7334556325587716,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.541590229650768e-05,
+      "loss": 1.2286,
+      "step": 10795
+    },
+    {
+      "epoch": 0.7337953526294333,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.541377904606605e-05,
+      "loss": 1.2007,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7341350727000951,
+      "grad_norm": 5.375,
+      "learning_rate": 4.541165579562441e-05,
+      "loss": 1.2334,
+      "step": 10805
+    },
+    {
+      "epoch": 0.7344747927707569,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.540953254518277e-05,
+      "loss": 1.2185,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7348145128414186,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.540740929474114e-05,
+      "loss": 1.2202,
+      "step": 10815
+    },
+    {
+      "epoch": 0.7351542329120805,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.54052860442995e-05,
+      "loss": 1.2947,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7354939529827422,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.540316279385786e-05,
+      "loss": 1.3582,
+      "step": 10825
+    },
+    {
+      "epoch": 0.735833673053404,
+      "grad_norm": 4.875,
+      "learning_rate": 4.540103954341623e-05,
+      "loss": 1.3189,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7361733931240658,
+      "grad_norm": 5.625,
+      "learning_rate": 4.539891629297459e-05,
+      "loss": 1.2361,
+      "step": 10835
+    },
+    {
+      "epoch": 0.7365131131947276,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.539679304253295e-05,
+      "loss": 1.2346,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7368528332653893,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.5394669792091323e-05,
+      "loss": 1.3087,
+      "step": 10845
+    },
+    {
+      "epoch": 0.737192553336051,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.539254654164968e-05,
+      "loss": 1.3473,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7375322734067129,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5390423291208045e-05,
+      "loss": 1.1827,
+      "step": 10855
+    },
+    {
+      "epoch": 0.7378719934773746,
+      "grad_norm": 5.0,
+      "learning_rate": 4.5388300040766415e-05,
+      "loss": 1.328,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7382117135480364,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.538617679032477e-05,
+      "loss": 1.2823,
+      "step": 10865
+    },
+    {
+      "epoch": 0.7385514336186982,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.538405353988314e-05,
+      "loss": 1.2959,
+      "step": 10870
+    },
+    {
+      "epoch": 0.73889115368936,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.538193028944151e-05,
+      "loss": 1.279,
+      "step": 10875
+    },
+    {
+      "epoch": 0.7392308737600217,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5379807038999865e-05,
+      "loss": 1.2989,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7395705938306835,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.537768378855823e-05,
+      "loss": 1.2924,
+      "step": 10885
+    },
+    {
+      "epoch": 0.7399103139013453,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.537556053811659e-05,
+      "loss": 1.2567,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7402500339720071,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.537343728767496e-05,
+      "loss": 1.286,
+      "step": 10895
+    },
+    {
+      "epoch": 0.7405897540426688,
+      "grad_norm": 5.0,
+      "learning_rate": 4.537131403723332e-05,
+      "loss": 1.2508,
+      "step": 10900
+    },
+    {
+      "epoch": 0.7409294741133307,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.5369190786791685e-05,
+      "loss": 1.1939,
+      "step": 10905
+    },
+    {
+      "epoch": 0.7412691941839924,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.536706753635005e-05,
+      "loss": 1.2093,
+      "step": 10910
+    },
+    {
+      "epoch": 0.7416089142546541,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.536494428590841e-05,
+      "loss": 1.2857,
+      "step": 10915
+    },
+    {
+      "epoch": 0.741948634325316,
+      "grad_norm": 4.625,
+      "learning_rate": 4.536282103546678e-05,
+      "loss": 1.2796,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7422883543959777,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.536069778502514e-05,
+      "loss": 1.3027,
+      "step": 10925
+    },
+    {
+      "epoch": 0.7426280744666395,
+      "grad_norm": 5.375,
+      "learning_rate": 4.5358574534583505e-05,
+      "loss": 1.2438,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7429677945373012,
+      "grad_norm": 5.875,
+      "learning_rate": 4.535645128414187e-05,
+      "loss": 1.2972,
+      "step": 10935
+    },
+    {
+      "epoch": 0.7433075146079631,
+      "grad_norm": 5.375,
+      "learning_rate": 4.535432803370023e-05,
+      "loss": 1.2008,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7436472346786248,
+      "grad_norm": 5.0,
+      "learning_rate": 4.53522047832586e-05,
+      "loss": 1.2311,
+      "step": 10945
+    },
+    {
+      "epoch": 0.7439869547492866,
+      "grad_norm": 5.5,
+      "learning_rate": 4.535008153281696e-05,
+      "loss": 1.3087,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7443266748199484,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.5347958282375325e-05,
+      "loss": 1.2732,
+      "step": 10955
+    },
+    {
+      "epoch": 0.7446663948906102,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.534583503193369e-05,
+      "loss": 1.2187,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7450061149612719,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.534371178149205e-05,
+      "loss": 1.2589,
+      "step": 10965
+    },
+    {
+      "epoch": 0.7453458350319336,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.534158853105042e-05,
+      "loss": 1.2722,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7456855551025955,
+      "grad_norm": 5.375,
+      "learning_rate": 4.533946528060878e-05,
+      "loss": 1.2165,
+      "step": 10975
+    },
+    {
+      "epoch": 0.7460252751732572,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.5337342030167145e-05,
+      "loss": 1.2464,
+      "step": 10980
+    },
+    {
+      "epoch": 0.746364995243919,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.533521877972551e-05,
+      "loss": 1.3308,
+      "step": 10985
+    },
+    {
+      "epoch": 0.7467047153145808,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.533309552928387e-05,
+      "loss": 1.2105,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7470444353852426,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.533097227884224e-05,
+      "loss": 1.1489,
+      "step": 10995
+    },
+    {
+      "epoch": 0.7473841554559043,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.53288490284006e-05,
+      "loss": 1.2282,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7477238755265662,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.5326725777958965e-05,
+      "loss": 1.1784,
+      "step": 11005
+    },
+    {
+      "epoch": 0.7480635955972279,
+      "grad_norm": 5.75,
+      "learning_rate": 4.532460252751733e-05,
+      "loss": 1.2551,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7484033156678896,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.532247927707569e-05,
+      "loss": 1.2036,
+      "step": 11015
+    },
+    {
+      "epoch": 0.7487430357385514,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.532035602663405e-05,
+      "loss": 1.2328,
+      "step": 11020
+    },
+    {
+      "epoch": 0.7490827558092132,
+      "grad_norm": 5.625,
+      "learning_rate": 4.531823277619242e-05,
+      "loss": 1.2258,
+      "step": 11025
+    },
+    {
+      "epoch": 0.749422475879875,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5316109525750785e-05,
+      "loss": 1.2262,
+      "step": 11030
+    },
+    {
+      "epoch": 0.7497621959505367,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.531398627530914e-05,
+      "loss": 1.271,
+      "step": 11035
+    },
+    {
+      "epoch": 0.7501019160211986,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.531186302486751e-05,
+      "loss": 1.3116,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7504416360918603,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.530973977442588e-05,
+      "loss": 1.1882,
+      "step": 11045
+    },
+    {
+      "epoch": 0.7507813561625221,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.5307616523984234e-05,
+      "loss": 1.3202,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7511210762331838,
+      "grad_norm": 6.375,
+      "learning_rate": 4.5305493273542605e-05,
+      "loss": 1.1813,
+      "step": 11055
+    },
+    {
+      "epoch": 0.7514607963038457,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.530337002310097e-05,
+      "loss": 1.3272,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7518005163745074,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.5301246772659326e-05,
+      "loss": 1.227,
+      "step": 11065
+    },
+    {
+      "epoch": 0.7521402364451691,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.52991235222177e-05,
+      "loss": 1.1405,
+      "step": 11070
+    },
+    {
+      "epoch": 0.752479956515831,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.529700027177606e-05,
+      "loss": 1.2829,
+      "step": 11075
+    },
+    {
+      "epoch": 0.7528196765864927,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.5294877021334425e-05,
+      "loss": 1.311,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7531593966571545,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.529275377089279e-05,
+      "loss": 1.2747,
+      "step": 11085
+    },
+    {
+      "epoch": 0.7534991167278163,
+      "grad_norm": 5.375,
+      "learning_rate": 4.5290630520451146e-05,
+      "loss": 1.253,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7538388367984781,
+      "grad_norm": 5.625,
+      "learning_rate": 4.528850727000952e-05,
+      "loss": 1.2439,
+      "step": 11095
+    },
+    {
+      "epoch": 0.7541785568691398,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.528638401956788e-05,
+      "loss": 1.2262,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7545182769398016,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.528426076912624e-05,
+      "loss": 1.3478,
+      "step": 11105
+    },
+    {
+      "epoch": 0.7548579970104634,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.528213751868461e-05,
+      "loss": 1.2636,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7551977170811252,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.528001426824297e-05,
+      "loss": 1.2325,
+      "step": 11115
+    },
+    {
+      "epoch": 0.7555374371517869,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.527789101780133e-05,
+      "loss": 1.2722,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7558771572224487,
+      "grad_norm": 5.25,
+      "learning_rate": 4.52757677673597e-05,
+      "loss": 1.3392,
+      "step": 11125
+    },
+    {
+      "epoch": 0.7562168772931105,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.5273644516918065e-05,
+      "loss": 1.233,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7565565973637722,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.527152126647642e-05,
+      "loss": 1.3927,
+      "step": 11135
+    },
+    {
+      "epoch": 0.756896317434434,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.526939801603479e-05,
+      "loss": 1.2597,
+      "step": 11140
+    },
+    {
+      "epoch": 0.7572360375050958,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.526727476559316e-05,
+      "loss": 1.2672,
+      "step": 11145
+    },
+    {
+      "epoch": 0.7575757575757576,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5265151515151514e-05,
+      "loss": 1.3221,
+      "step": 11150
+    },
+    {
+      "epoch": 0.7579154776464193,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5263028264709885e-05,
+      "loss": 1.1546,
+      "step": 11155
+    },
+    {
+      "epoch": 0.7582551977170812,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.526090501426825e-05,
+      "loss": 1.3006,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7585949177877429,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.5258781763826606e-05,
+      "loss": 1.1442,
+      "step": 11165
+    },
+    {
+      "epoch": 0.7589346378584046,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.525665851338498e-05,
+      "loss": 1.2234,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7592743579290665,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.5254535262943334e-05,
+      "loss": 1.3398,
+      "step": 11175
+    },
+    {
+      "epoch": 0.7596140779997282,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.52524120125017e-05,
+      "loss": 1.2498,
+      "step": 11180
+    },
+    {
+      "epoch": 0.75995379807039,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.525028876206007e-05,
+      "loss": 1.2407,
+      "step": 11185
+    },
+    {
+      "epoch": 0.7602935181410517,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5248165511618426e-05,
+      "loss": 1.2802,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7606332382117136,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.524604226117679e-05,
+      "loss": 1.2798,
+      "step": 11195
+    },
+    {
+      "epoch": 0.7609729582823753,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.524391901073516e-05,
+      "loss": 1.2624,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7613126783530371,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.524179576029352e-05,
+      "loss": 1.2057,
+      "step": 11205
+    },
+    {
+      "epoch": 0.7616523984236989,
+      "grad_norm": 4.21875,
+      "learning_rate": 4.523967250985188e-05,
+      "loss": 1.1508,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7619921184943607,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.523754925941025e-05,
+      "loss": 1.1641,
+      "step": 11215
+    },
+    {
+      "epoch": 0.7623318385650224,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.523542600896861e-05,
+      "loss": 1.251,
+      "step": 11220
+    },
+    {
+      "epoch": 0.7626715586356841,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.5233302758526974e-05,
+      "loss": 1.2754,
+      "step": 11225
+    },
+    {
+      "epoch": 0.763011278706346,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5231179508085345e-05,
+      "loss": 1.3425,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7633509987770077,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.52290562576437e-05,
+      "loss": 1.2914,
+      "step": 11235
+    },
+    {
+      "epoch": 0.7636907188476695,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5226933007202066e-05,
+      "loss": 1.3111,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7640304389183313,
+      "grad_norm": 5.375,
+      "learning_rate": 4.522480975676043e-05,
+      "loss": 1.1572,
+      "step": 11245
+    },
+    {
+      "epoch": 0.7643701589889931,
+      "grad_norm": 5.875,
+      "learning_rate": 4.5222686506318794e-05,
+      "loss": 1.2513,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7647098790596548,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.522056325587716e-05,
+      "loss": 1.2491,
+      "step": 11255
+    },
+    {
+      "epoch": 0.7650495991303167,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.521844000543552e-05,
+      "loss": 1.3054,
+      "step": 11260
+    },
+    {
+      "epoch": 0.7653893192009784,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.5216316754993886e-05,
+      "loss": 1.2984,
+      "step": 11265
+    },
+    {
+      "epoch": 0.7657290392716402,
+      "grad_norm": 5.125,
+      "learning_rate": 4.521419350455225e-05,
+      "loss": 1.1913,
+      "step": 11270
+    },
+    {
+      "epoch": 0.7660687593423019,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.5212070254110614e-05,
+      "loss": 1.1336,
+      "step": 11275
+    },
+    {
+      "epoch": 0.7664084794129638,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.520994700366898e-05,
+      "loss": 1.1976,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7667481994836255,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.520782375322734e-05,
+      "loss": 1.1948,
+      "step": 11285
+    },
+    {
+      "epoch": 0.7670879195542872,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.5205700502785706e-05,
+      "loss": 1.2936,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7674276396249491,
+      "grad_norm": 5.625,
+      "learning_rate": 4.520357725234407e-05,
+      "loss": 1.2334,
+      "step": 11295
+    },
+    {
+      "epoch": 0.7677673596956108,
+      "grad_norm": 5.0,
+      "learning_rate": 4.5201454001902434e-05,
+      "loss": 1.3122,
+      "step": 11300
+    },
+    {
+      "epoch": 0.7681070797662726,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.51993307514608e-05,
+      "loss": 1.2323,
+      "step": 11305
+    },
+    {
+      "epoch": 0.7684467998369343,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.519720750101916e-05,
+      "loss": 1.2576,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7687865199075962,
+      "grad_norm": 5.375,
+      "learning_rate": 4.5195084250577526e-05,
+      "loss": 1.1817,
+      "step": 11315
+    },
+    {
+      "epoch": 0.7691262399782579,
+      "grad_norm": 5.375,
+      "learning_rate": 4.519296100013589e-05,
+      "loss": 1.2567,
+      "step": 11320
+    },
+    {
+      "epoch": 0.7694659600489197,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.5190837749694254e-05,
+      "loss": 1.18,
+      "step": 11325
+    },
+    {
+      "epoch": 0.7698056801195815,
+      "grad_norm": 5.75,
+      "learning_rate": 4.518871449925262e-05,
+      "loss": 1.2245,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7701454001902432,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.518659124881098e-05,
+      "loss": 1.16,
+      "step": 11335
+    },
+    {
+      "epoch": 0.770485120260905,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5184467998369346e-05,
+      "loss": 1.247,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7708248403315668,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.518234474792771e-05,
+      "loss": 1.2486,
+      "step": 11345
+    },
+    {
+      "epoch": 0.7711645604022286,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.5180221497486074e-05,
+      "loss": 1.368,
+      "step": 11350
+    },
+    {
+      "epoch": 0.7715042804728903,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.517809824704444e-05,
+      "loss": 1.2716,
+      "step": 11355
+    },
+    {
+      "epoch": 0.7718440005435521,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.51759749966028e-05,
+      "loss": 1.2996,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7721837206142139,
+      "grad_norm": 5.375,
+      "learning_rate": 4.5173851746161166e-05,
+      "loss": 1.1986,
+      "step": 11365
+    },
+    {
+      "epoch": 0.7725234406848757,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.517172849571953e-05,
+      "loss": 1.3167,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7728631607555374,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.516960524527789e-05,
+      "loss": 1.1677,
+      "step": 11375
+    },
+    {
+      "epoch": 0.7732028808261993,
+      "grad_norm": 5.5,
+      "learning_rate": 4.516748199483626e-05,
+      "loss": 1.2659,
+      "step": 11380
+    },
+    {
+      "epoch": 0.773542600896861,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.516535874439462e-05,
+      "loss": 1.3473,
+      "step": 11385
+    },
+    {
+      "epoch": 0.7738823209675227,
+      "grad_norm": 5.5,
+      "learning_rate": 4.516323549395298e-05,
+      "loss": 1.211,
+      "step": 11390
+    },
+    {
+      "epoch": 0.7742220410381845,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.516111224351135e-05,
+      "loss": 1.2014,
+      "step": 11395
+    },
+    {
+      "epoch": 0.7745617611088463,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.5158988993069714e-05,
+      "loss": 1.2663,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7749014811795081,
+      "grad_norm": 6.0,
+      "learning_rate": 4.515686574262807e-05,
+      "loss": 1.3761,
+      "step": 11405
+    },
+    {
+      "epoch": 0.7752412012501698,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.515474249218644e-05,
+      "loss": 1.2638,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7755809213208317,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.5152619241744806e-05,
+      "loss": 1.2045,
+      "step": 11415
+    },
+    {
+      "epoch": 0.7759206413914934,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.515049599130317e-05,
+      "loss": 1.2793,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7762603614621552,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.5148372740861534e-05,
+      "loss": 1.3197,
+      "step": 11425
+    },
+    {
+      "epoch": 0.776600081532817,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.51462494904199e-05,
+      "loss": 1.2166,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7769398016034788,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.514412623997826e-05,
+      "loss": 1.2975,
+      "step": 11435
+    },
+    {
+      "epoch": 0.7772795216741405,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.5142002989536626e-05,
+      "loss": 1.3899,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7776192417448022,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.5139879739094984e-05,
+      "loss": 1.3539,
+      "step": 11445
+    },
+    {
+      "epoch": 0.7779589618154641,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5137756488653354e-05,
+      "loss": 1.2435,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7782986818861258,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.513563323821172e-05,
+      "loss": 1.5837,
+      "step": 11455
+    },
+    {
+      "epoch": 0.7786384019567876,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.5133509987770076e-05,
+      "loss": 1.2685,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7789781220274494,
+      "grad_norm": 5.25,
+      "learning_rate": 4.5131386737328446e-05,
+      "loss": 1.2956,
+      "step": 11465
+    },
+    {
+      "epoch": 0.7793178420981112,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.512926348688681e-05,
+      "loss": 1.2023,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7796575621687729,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.512714023644517e-05,
+      "loss": 1.2258,
+      "step": 11475
+    },
+    {
+      "epoch": 0.7799972822394347,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.512501698600354e-05,
+      "loss": 1.2845,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7803370023100965,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.51228937355619e-05,
+      "loss": 1.3197,
+      "step": 11485
+    },
+    {
+      "epoch": 0.7806767223807582,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.512077048512026e-05,
+      "loss": 1.2074,
+      "step": 11490
+    },
+    {
+      "epoch": 0.78101644245142,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.511864723467863e-05,
+      "loss": 1.2026,
+      "step": 11495
+    },
+    {
+      "epoch": 0.7813561625220818,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.5116523984236994e-05,
+      "loss": 1.2969,
+      "step": 11500
+    },
+    {
+      "epoch": 0.7816958825927436,
+      "grad_norm": 5.125,
+      "learning_rate": 4.511440073379535e-05,
+      "loss": 1.1754,
+      "step": 11505
+    },
+    {
+      "epoch": 0.7820356026634053,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.511227748335372e-05,
+      "loss": 1.2479,
+      "step": 11510
+    },
+    {
+      "epoch": 0.7823753227340672,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.511015423291208e-05,
+      "loss": 1.2452,
+      "step": 11515
+    },
+    {
+      "epoch": 0.7827150428047289,
+      "grad_norm": 5.5,
+      "learning_rate": 4.5108030982470444e-05,
+      "loss": 1.3881,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7830547628753907,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.5105907732028814e-05,
+      "loss": 1.2439,
+      "step": 11525
+    },
+    {
+      "epoch": 0.7833944829460524,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.510378448158717e-05,
+      "loss": 1.2174,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7837342030167143,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.5101661231145536e-05,
+      "loss": 1.27,
+      "step": 11535
+    },
+    {
+      "epoch": 0.784073923087376,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.5099537980703906e-05,
+      "loss": 1.3076,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7844136431580377,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.5097414730262264e-05,
+      "loss": 1.3102,
+      "step": 11545
+    },
+    {
+      "epoch": 0.7847533632286996,
+      "grad_norm": 5.375,
+      "learning_rate": 4.509529147982063e-05,
+      "loss": 1.2224,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7850930832993613,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.5093168229379e-05,
+      "loss": 1.3319,
+      "step": 11555
+    },
+    {
+      "epoch": 0.7854328033700231,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.5091044978937356e-05,
+      "loss": 1.2765,
+      "step": 11560
+    },
+    {
+      "epoch": 0.7857725234406848,
+      "grad_norm": 6.375,
+      "learning_rate": 4.508892172849572e-05,
+      "loss": 1.3279,
+      "step": 11565
+    },
+    {
+      "epoch": 0.7861122435113467,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.508679847805409e-05,
+      "loss": 1.197,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7864519635820084,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.508467522761245e-05,
+      "loss": 1.109,
+      "step": 11575
+    },
+    {
+      "epoch": 0.7867916836526702,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.508255197717081e-05,
+      "loss": 1.2523,
+      "step": 11580
+    },
+    {
+      "epoch": 0.787131403723332,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.508042872672918e-05,
+      "loss": 1.2186,
+      "step": 11585
+    },
+    {
+      "epoch": 0.7874711237939938,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.507830547628754e-05,
+      "loss": 1.282,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7878108438646555,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.5076182225845904e-05,
+      "loss": 1.2227,
+      "step": 11595
+    },
+    {
+      "epoch": 0.7881505639353173,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.507405897540427e-05,
+      "loss": 1.2302,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7884902840059791,
+      "grad_norm": 5.375,
+      "learning_rate": 4.507193572496263e-05,
+      "loss": 1.3071,
+      "step": 11605
+    },
+    {
+      "epoch": 0.7888300040766408,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.5069812474520996e-05,
+      "loss": 1.297,
+      "step": 11610
+    },
+    {
+      "epoch": 0.7891697241473026,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.506768922407936e-05,
+      "loss": 1.2435,
+      "step": 11615
+    },
+    {
+      "epoch": 0.7895094442179644,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.5065565973637724e-05,
+      "loss": 1.2642,
+      "step": 11620
+    },
+    {
+      "epoch": 0.7898491642886262,
+      "grad_norm": 5.5,
+      "learning_rate": 4.506344272319609e-05,
+      "loss": 1.2469,
+      "step": 11625
+    },
+    {
+      "epoch": 0.7901888843592879,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.506131947275445e-05,
+      "loss": 1.3417,
+      "step": 11630
+    },
+    {
+      "epoch": 0.7905286044299498,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.5059196222312816e-05,
+      "loss": 1.2917,
+      "step": 11635
+    },
+    {
+      "epoch": 0.7908683245006115,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.505707297187118e-05,
+      "loss": 1.2929,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7912080445712733,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.5054949721429544e-05,
+      "loss": 1.2568,
+      "step": 11645
+    },
+    {
+      "epoch": 0.791547764641935,
+      "grad_norm": 5.25,
+      "learning_rate": 4.505282647098791e-05,
+      "loss": 1.2993,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7918874847125968,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.505070322054627e-05,
+      "loss": 1.2903,
+      "step": 11655
+    },
+    {
+      "epoch": 0.7922272047832586,
+      "grad_norm": 5.875,
+      "learning_rate": 4.5048579970104636e-05,
+      "loss": 1.2916,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7925669248539203,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.5046456719663e-05,
+      "loss": 1.28,
+      "step": 11665
+    },
+    {
+      "epoch": 0.7929066449245822,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5044333469221364e-05,
+      "loss": 1.2384,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7932463649952439,
+      "grad_norm": 5.25,
+      "learning_rate": 4.504221021877973e-05,
+      "loss": 1.2258,
+      "step": 11675
+    },
+    {
+      "epoch": 0.7935860850659057,
+      "grad_norm": 5.0,
+      "learning_rate": 4.504008696833809e-05,
+      "loss": 1.1969,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7939258051365675,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.5037963717896456e-05,
+      "loss": 1.3815,
+      "step": 11685
+    },
+    {
+      "epoch": 0.7942655252072293,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.503584046745482e-05,
+      "loss": 1.1683,
+      "step": 11690
+    },
+    {
+      "epoch": 0.794605245277891,
+      "grad_norm": 6.0,
+      "learning_rate": 4.5033717217013184e-05,
+      "loss": 1.3166,
+      "step": 11695
+    },
+    {
+      "epoch": 0.7949449653485527,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.503159396657155e-05,
+      "loss": 1.2469,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7952846854192146,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.502947071612991e-05,
+      "loss": 1.188,
+      "step": 11705
+    },
+    {
+      "epoch": 0.7956244054898763,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5027347465688276e-05,
+      "loss": 1.2091,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7959641255605381,
+      "grad_norm": 6.0,
+      "learning_rate": 4.502522421524663e-05,
+      "loss": 1.2976,
+      "step": 11715
+    },
+    {
+      "epoch": 0.7963038456311999,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.5023100964805004e-05,
+      "loss": 1.2079,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7966435657018617,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.502097771436337e-05,
+      "loss": 1.2253,
+      "step": 11725
+    },
+    {
+      "epoch": 0.7969832857725234,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.5018854463921725e-05,
+      "loss": 1.2013,
+      "step": 11730
+    },
+    {
+      "epoch": 0.7973230058431852,
+      "grad_norm": 5.625,
+      "learning_rate": 4.5016731213480096e-05,
+      "loss": 1.2502,
+      "step": 11735
+    },
+    {
+      "epoch": 0.797662725913847,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.501460796303846e-05,
+      "loss": 1.0879,
+      "step": 11740
+    },
+    {
+      "epoch": 0.7980024459845088,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.501248471259682e-05,
+      "loss": 1.147,
+      "step": 11745
+    },
+    {
+      "epoch": 0.7983421660551705,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.501036146215519e-05,
+      "loss": 1.2058,
+      "step": 11750
+    },
+    {
+      "epoch": 0.7986818861258324,
+      "grad_norm": 5.75,
+      "learning_rate": 4.500823821171355e-05,
+      "loss": 1.24,
+      "step": 11755
+    },
+    {
+      "epoch": 0.7990216061964941,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.5006114961271916e-05,
+      "loss": 1.2309,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7993613262671558,
+      "grad_norm": 5.625,
+      "learning_rate": 4.500399171083028e-05,
+      "loss": 1.2523,
+      "step": 11765
+    },
+    {
+      "epoch": 0.7997010463378177,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.5001868460388644e-05,
+      "loss": 1.4293,
+      "step": 11770
+    },
+    {
+      "epoch": 0.8000407664084794,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.499974520994701e-05,
+      "loss": 1.2898,
+      "step": 11775
+    },
+    {
+      "epoch": 0.8003804864791412,
+      "grad_norm": 6.0,
+      "learning_rate": 4.499762195950537e-05,
+      "loss": 1.2948,
+      "step": 11780
+    },
+    {
+      "epoch": 0.8007202065498029,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.4995498709063736e-05,
+      "loss": 1.3493,
+      "step": 11785
+    },
+    {
+      "epoch": 0.8010599266204648,
+      "grad_norm": 5.25,
+      "learning_rate": 4.49933754586221e-05,
+      "loss": 1.29,
+      "step": 11790
+    },
+    {
+      "epoch": 0.8013996466911265,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.4991252208180464e-05,
+      "loss": 1.3093,
+      "step": 11795
+    },
+    {
+      "epoch": 0.8017393667617883,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.498912895773882e-05,
+      "loss": 1.3259,
+      "step": 11800
+    },
+    {
+      "epoch": 0.8020790868324501,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.498700570729719e-05,
+      "loss": 1.2782,
+      "step": 11805
+    },
+    {
+      "epoch": 0.8024188069031118,
+      "grad_norm": 5.5,
+      "learning_rate": 4.4984882456855556e-05,
+      "loss": 1.2222,
+      "step": 11810
+    },
+    {
+      "epoch": 0.8027585269737736,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.498275920641391e-05,
+      "loss": 1.2267,
+      "step": 11815
+    },
+    {
+      "epoch": 0.8030982470444353,
+      "grad_norm": 5.25,
+      "learning_rate": 4.4980635955972284e-05,
+      "loss": 1.2774,
+      "step": 11820
+    },
+    {
+      "epoch": 0.8034379671150972,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.497851270553065e-05,
+      "loss": 1.3325,
+      "step": 11825
+    },
+    {
+      "epoch": 0.8037776871857589,
+      "grad_norm": 6.75,
+      "learning_rate": 4.4976389455089005e-05,
+      "loss": 1.2681,
+      "step": 11830
+    },
+    {
+      "epoch": 0.8041174072564207,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.4974266204647376e-05,
+      "loss": 1.3664,
+      "step": 11835
+    },
+    {
+      "epoch": 0.8044571273270825,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.497214295420574e-05,
+      "loss": 1.2766,
+      "step": 11840
+    },
+    {
+      "epoch": 0.8047968473977443,
+      "grad_norm": 5.375,
+      "learning_rate": 4.49700197037641e-05,
+      "loss": 1.2373,
+      "step": 11845
+    },
+    {
+      "epoch": 0.805136567468406,
+      "grad_norm": 5.625,
+      "learning_rate": 4.496789645332247e-05,
+      "loss": 1.2818,
+      "step": 11850
+    },
+    {
+      "epoch": 0.8054762875390679,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.496577320288083e-05,
+      "loss": 1.1808,
+      "step": 11855
+    },
+    {
+      "epoch": 0.8058160076097296,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.496364995243919e-05,
+      "loss": 1.2148,
+      "step": 11860
+    },
+    {
+      "epoch": 0.8061557276803913,
+      "grad_norm": 5.875,
+      "learning_rate": 4.496152670199756e-05,
+      "loss": 1.2162,
+      "step": 11865
+    },
+    {
+      "epoch": 0.8064954477510531,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.495940345155592e-05,
+      "loss": 1.2969,
+      "step": 11870
+    },
+    {
+      "epoch": 0.8068351678217149,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.495728020111428e-05,
+      "loss": 1.2378,
+      "step": 11875
+    },
+    {
+      "epoch": 0.8071748878923767,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.495515695067265e-05,
+      "loss": 1.2002,
+      "step": 11880
+    },
+    {
+      "epoch": 0.8075146079630384,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.495303370023101e-05,
+      "loss": 1.3147,
+      "step": 11885
+    },
+    {
+      "epoch": 0.8078543280337003,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.495091044978937e-05,
+      "loss": 1.2337,
+      "step": 11890
+    },
+    {
+      "epoch": 0.808194048104362,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4948787199347744e-05,
+      "loss": 1.2983,
+      "step": 11895
+    },
+    {
+      "epoch": 0.8085337681750238,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.49466639489061e-05,
+      "loss": 1.1854,
+      "step": 11900
+    },
+    {
+      "epoch": 0.8088734882456855,
+      "grad_norm": 5.625,
+      "learning_rate": 4.4944540698464465e-05,
+      "loss": 1.327,
+      "step": 11905
+    },
+    {
+      "epoch": 0.8092132083163474,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.4942417448022836e-05,
+      "loss": 1.4289,
+      "step": 11910
+    },
+    {
+      "epoch": 0.8095529283870091,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.494029419758119e-05,
+      "loss": 1.2582,
+      "step": 11915
+    },
+    {
+      "epoch": 0.8098926484576708,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.493817094713956e-05,
+      "loss": 1.1726,
+      "step": 11920
+    },
+    {
+      "epoch": 0.8102323685283327,
+      "grad_norm": 5.75,
+      "learning_rate": 4.493604769669793e-05,
+      "loss": 1.2834,
+      "step": 11925
+    },
+    {
+      "epoch": 0.8105720885989944,
+      "grad_norm": 5.5,
+      "learning_rate": 4.4933924446256285e-05,
+      "loss": 1.2696,
+      "step": 11930
+    },
+    {
+      "epoch": 0.8109118086696562,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.493180119581465e-05,
+      "loss": 1.236,
+      "step": 11935
+    },
+    {
+      "epoch": 0.811251528740318,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.492967794537301e-05,
+      "loss": 1.3259,
+      "step": 11940
+    },
+    {
+      "epoch": 0.8115912488109798,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.492755469493138e-05,
+      "loss": 1.2259,
+      "step": 11945
+    },
+    {
+      "epoch": 0.8119309688816415,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.492543144448974e-05,
+      "loss": 1.396,
+      "step": 11950
+    },
+    {
+      "epoch": 0.8122706889523033,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.4923308194048105e-05,
+      "loss": 1.178,
+      "step": 11955
+    },
+    {
+      "epoch": 0.8126104090229651,
+      "grad_norm": 5.5,
+      "learning_rate": 4.492118494360647e-05,
+      "loss": 1.2771,
+      "step": 11960
+    },
+    {
+      "epoch": 0.8129501290936268,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.491906169316483e-05,
+      "loss": 1.2521,
+      "step": 11965
+    },
+    {
+      "epoch": 0.8132898491642886,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.49169384427232e-05,
+      "loss": 1.3003,
+      "step": 11970
+    },
+    {
+      "epoch": 0.8136295692349504,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.491481519228156e-05,
+      "loss": 1.2402,
+      "step": 11975
+    },
+    {
+      "epoch": 0.8139692893056122,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.4912691941839925e-05,
+      "loss": 1.2504,
+      "step": 11980
+    },
+    {
+      "epoch": 0.8143090093762739,
+      "grad_norm": 5.625,
+      "learning_rate": 4.491056869139829e-05,
+      "loss": 1.2959,
+      "step": 11985
+    },
+    {
+      "epoch": 0.8146487294469357,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.490844544095665e-05,
+      "loss": 1.2857,
+      "step": 11990
+    },
+    {
+      "epoch": 0.8149884495175975,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.490632219051502e-05,
+      "loss": 1.1544,
+      "step": 11995
+    },
+    {
+      "epoch": 0.8153281695882593,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.490419894007338e-05,
+      "loss": 1.2865,
+      "step": 12000
+    },
+    {
+      "epoch": 0.815667889658921,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.4902075689631745e-05,
+      "loss": 1.2246,
+      "step": 12005
+    },
+    {
+      "epoch": 0.8160076097295829,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.489995243919011e-05,
+      "loss": 1.2683,
+      "step": 12010
+    },
+    {
+      "epoch": 0.8163473298002446,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.489782918874847e-05,
+      "loss": 1.2369,
+      "step": 12015
+    },
+    {
+      "epoch": 0.8166870498709063,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.489570593830684e-05,
+      "loss": 1.2592,
+      "step": 12020
+    },
+    {
+      "epoch": 0.8170267699415682,
+      "grad_norm": 5.25,
+      "learning_rate": 4.48935826878652e-05,
+      "loss": 1.2303,
+      "step": 12025
+    },
+    {
+      "epoch": 0.8173664900122299,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4891459437423565e-05,
+      "loss": 1.2561,
+      "step": 12030
+    },
+    {
+      "epoch": 0.8177062100828917,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.488933618698193e-05,
+      "loss": 1.2506,
+      "step": 12035
+    },
+    {
+      "epoch": 0.8180459301535534,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.488721293654029e-05,
+      "loss": 1.2234,
+      "step": 12040
+    },
+    {
+      "epoch": 0.8183856502242153,
+      "grad_norm": 5.5,
+      "learning_rate": 4.488508968609866e-05,
+      "loss": 1.3143,
+      "step": 12045
+    },
+    {
+      "epoch": 0.818725370294877,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.488296643565702e-05,
+      "loss": 1.2884,
+      "step": 12050
+    },
+    {
+      "epoch": 0.8190650903655388,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.4880843185215385e-05,
+      "loss": 1.2647,
+      "step": 12055
+    },
+    {
+      "epoch": 0.8194048104362006,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.487871993477375e-05,
+      "loss": 1.29,
+      "step": 12060
+    },
+    {
+      "epoch": 0.8197445305068624,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.487659668433211e-05,
+      "loss": 1.2782,
+      "step": 12065
+    },
+    {
+      "epoch": 0.8200842505775241,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.487447343389047e-05,
+      "loss": 1.231,
+      "step": 12070
+    },
+    {
+      "epoch": 0.8204239706481858,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.487235018344884e-05,
+      "loss": 1.3413,
+      "step": 12075
+    },
+    {
+      "epoch": 0.8207636907188477,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.4870226933007205e-05,
+      "loss": 1.2297,
+      "step": 12080
+    },
+    {
+      "epoch": 0.8211034107895094,
+      "grad_norm": 5.25,
+      "learning_rate": 4.486810368256556e-05,
+      "loss": 1.1765,
+      "step": 12085
+    },
+    {
+      "epoch": 0.8214431308601712,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.486598043212393e-05,
+      "loss": 1.2824,
+      "step": 12090
+    },
+    {
+      "epoch": 0.821782850930833,
+      "grad_norm": 5.5,
+      "learning_rate": 4.48638571816823e-05,
+      "loss": 1.293,
+      "step": 12095
+    },
+    {
+      "epoch": 0.8221225710014948,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.486173393124066e-05,
+      "loss": 1.2785,
+      "step": 12100
+    },
+    {
+      "epoch": 0.8224622910721565,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.4859610680799025e-05,
+      "loss": 1.221,
+      "step": 12105
+    },
+    {
+      "epoch": 0.8228020111428184,
+      "grad_norm": 5.0,
+      "learning_rate": 4.485748743035739e-05,
+      "loss": 1.1451,
+      "step": 12110
+    },
+    {
+      "epoch": 0.8231417312134801,
+      "grad_norm": 5.125,
+      "learning_rate": 4.485536417991575e-05,
+      "loss": 1.2306,
+      "step": 12115
+    },
+    {
+      "epoch": 0.8234814512841419,
+      "grad_norm": 5.375,
+      "learning_rate": 4.485324092947412e-05,
+      "loss": 1.2339,
+      "step": 12120
+    },
+    {
+      "epoch": 0.8238211713548036,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.485111767903248e-05,
+      "loss": 1.1951,
+      "step": 12125
+    },
+    {
+      "epoch": 0.8241608914254654,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.4848994428590845e-05,
+      "loss": 1.3258,
+      "step": 12130
+    },
+    {
+      "epoch": 0.8245006114961272,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.484687117814921e-05,
+      "loss": 1.3026,
+      "step": 12135
+    },
+    {
+      "epoch": 0.8248403315667889,
+      "grad_norm": 4.75,
+      "learning_rate": 4.4844747927707567e-05,
+      "loss": 1.2814,
+      "step": 12140
+    },
+    {
+      "epoch": 0.8251800516374508,
+      "grad_norm": 5.5,
+      "learning_rate": 4.484262467726594e-05,
+      "loss": 1.3148,
+      "step": 12145
+    },
+    {
+      "epoch": 0.8255197717081125,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.48405014268243e-05,
+      "loss": 1.2949,
+      "step": 12150
+    },
+    {
+      "epoch": 0.8258594917787743,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.483837817638266e-05,
+      "loss": 1.2996,
+      "step": 12155
+    },
+    {
+      "epoch": 0.826199211849436,
+      "grad_norm": 5.375,
+      "learning_rate": 4.483625492594103e-05,
+      "loss": 1.1728,
+      "step": 12160
+    },
+    {
+      "epoch": 0.8265389319200979,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.4834131675499393e-05,
+      "loss": 1.2374,
+      "step": 12165
+    },
+    {
+      "epoch": 0.8268786519907596,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.483200842505775e-05,
+      "loss": 1.2739,
+      "step": 12170
+    },
+    {
+      "epoch": 0.8272183720614213,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.482988517461612e-05,
+      "loss": 1.2354,
+      "step": 12175
+    },
+    {
+      "epoch": 0.8275580921320832,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.4827761924174485e-05,
+      "loss": 1.3377,
+      "step": 12180
+    },
+    {
+      "epoch": 0.8278978122027449,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.482563867373284e-05,
+      "loss": 1.2522,
+      "step": 12185
+    },
+    {
+      "epoch": 0.8282375322734067,
+      "grad_norm": 4.75,
+      "learning_rate": 4.4823515423291213e-05,
+      "loss": 1.1579,
+      "step": 12190
+    },
+    {
+      "epoch": 0.8285772523440685,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.482139217284958e-05,
+      "loss": 1.2312,
+      "step": 12195
+    },
+    {
+      "epoch": 0.8289169724147303,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.4819268922407935e-05,
+      "loss": 1.2395,
+      "step": 12200
+    },
+    {
+      "epoch": 0.829256692485392,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.4817145671966305e-05,
+      "loss": 1.3373,
+      "step": 12205
+    },
+    {
+      "epoch": 0.8295964125560538,
+      "grad_norm": 4.875,
+      "learning_rate": 4.481502242152467e-05,
+      "loss": 1.2378,
+      "step": 12210
+    },
+    {
+      "epoch": 0.8299361326267156,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.481289917108303e-05,
+      "loss": 1.2904,
+      "step": 12215
+    },
+    {
+      "epoch": 0.8302758526973774,
+      "grad_norm": 4.75,
+      "learning_rate": 4.48107759206414e-05,
+      "loss": 1.1542,
+      "step": 12220
+    },
+    {
+      "epoch": 0.8306155727680391,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.4808652670199755e-05,
+      "loss": 1.3236,
+      "step": 12225
+    },
+    {
+      "epoch": 0.830955292838701,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.480652941975812e-05,
+      "loss": 1.2724,
+      "step": 12230
+    },
+    {
+      "epoch": 0.8312950129093627,
+      "grad_norm": 5.25,
+      "learning_rate": 4.480440616931649e-05,
+      "loss": 1.1545,
+      "step": 12235
+    },
+    {
+      "epoch": 0.8316347329800244,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.480228291887485e-05,
+      "loss": 1.1943,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8319744530506862,
+      "grad_norm": 5.25,
+      "learning_rate": 4.480015966843321e-05,
+      "loss": 1.2343,
+      "step": 12245
+    },
+    {
+      "epoch": 0.832314173121348,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.479803641799158e-05,
+      "loss": 1.1997,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8326538931920098,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.479591316754994e-05,
+      "loss": 1.2188,
+      "step": 12255
+    },
+    {
+      "epoch": 0.8329936132626715,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.47937899171083e-05,
+      "loss": 1.178,
+      "step": 12260
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.4791666666666673e-05,
+      "loss": 1.2502,
+      "step": 12265
+    },
+    {
+      "epoch": 0.8336730534039951,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.478954341622503e-05,
+      "loss": 1.0661,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8340127734746569,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4787420165783395e-05,
+      "loss": 1.2805,
+      "step": 12275
+    },
+    {
+      "epoch": 0.8343524935453187,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4785296915341765e-05,
+      "loss": 1.2806,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8346922136159804,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.478317366490012e-05,
+      "loss": 1.3002,
+      "step": 12285
+    },
+    {
+      "epoch": 0.8350319336866422,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.478105041445849e-05,
+      "loss": 1.1878,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8353716537573039,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.477892716401685e-05,
+      "loss": 1.2997,
+      "step": 12295
+    },
+    {
+      "epoch": 0.8357113738279658,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.4776803913575215e-05,
+      "loss": 1.213,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8360510938986275,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.477468066313358e-05,
+      "loss": 1.2034,
+      "step": 12305
+    },
+    {
+      "epoch": 0.8363908139692893,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.477255741269194e-05,
+      "loss": 1.2257,
+      "step": 12310
+    },
+    {
+      "epoch": 0.8367305340399511,
+      "grad_norm": 5.125,
+      "learning_rate": 4.477043416225031e-05,
+      "loss": 1.1994,
+      "step": 12315
+    },
+    {
+      "epoch": 0.8370702541106129,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.476831091180867e-05,
+      "loss": 1.2797,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8374099741812746,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.4766187661367035e-05,
+      "loss": 1.2947,
+      "step": 12325
+    },
+    {
+      "epoch": 0.8377496942519363,
+      "grad_norm": 4.125,
+      "learning_rate": 4.47640644109254e-05,
+      "loss": 1.0797,
+      "step": 12330
+    },
+    {
+      "epoch": 0.8380894143225982,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.476194116048376e-05,
+      "loss": 1.3748,
+      "step": 12335
+    },
+    {
+      "epoch": 0.83842913439326,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.475981791004213e-05,
+      "loss": 1.2012,
+      "step": 12340
+    },
+    {
+      "epoch": 0.8387688544639217,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.475769465960049e-05,
+      "loss": 1.3005,
+      "step": 12345
+    },
+    {
+      "epoch": 0.8391085745345835,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.4755571409158855e-05,
+      "loss": 1.1471,
+      "step": 12350
+    },
+    {
+      "epoch": 0.8394482946052453,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.475344815871722e-05,
+      "loss": 1.2067,
+      "step": 12355
+    },
+    {
+      "epoch": 0.839788014675907,
+      "grad_norm": 5.0,
+      "learning_rate": 4.475132490827558e-05,
+      "loss": 1.285,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8401277347465689,
+      "grad_norm": 5.5,
+      "learning_rate": 4.474920165783395e-05,
+      "loss": 1.3297,
+      "step": 12365
+    },
+    {
+      "epoch": 0.8404674548172306,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.474707840739231e-05,
+      "loss": 1.1748,
+      "step": 12370
+    },
+    {
+      "epoch": 0.8408071748878924,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.4744955156950675e-05,
+      "loss": 1.289,
+      "step": 12375
+    },
+    {
+      "epoch": 0.8411468949585541,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.474283190650904e-05,
+      "loss": 1.2377,
+      "step": 12380
+    },
+    {
+      "epoch": 0.841486615029216,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.47407086560674e-05,
+      "loss": 1.1988,
+      "step": 12385
+    },
+    {
+      "epoch": 0.8418263350998777,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.473858540562577e-05,
+      "loss": 1.2646,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8421660551705394,
+      "grad_norm": 5.375,
+      "learning_rate": 4.473646215518413e-05,
+      "loss": 1.2216,
+      "step": 12395
+    },
+    {
+      "epoch": 0.8425057752412013,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.4734338904742495e-05,
+      "loss": 1.2302,
+      "step": 12400
+    },
+    {
+      "epoch": 0.842845495311863,
+      "grad_norm": 5.125,
+      "learning_rate": 4.473221565430086e-05,
+      "loss": 1.2591,
+      "step": 12405
+    },
+    {
+      "epoch": 0.8431852153825248,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.473009240385922e-05,
+      "loss": 1.2524,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8435249354531865,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.472796915341759e-05,
+      "loss": 1.1939,
+      "step": 12415
+    },
+    {
+      "epoch": 0.8438646555238484,
+      "grad_norm": 5.75,
+      "learning_rate": 4.472584590297595e-05,
+      "loss": 1.2297,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8442043755945101,
+      "grad_norm": 5.875,
+      "learning_rate": 4.472372265253431e-05,
+      "loss": 1.3014,
+      "step": 12425
+    },
+    {
+      "epoch": 0.8445440956651719,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.472159940209268e-05,
+      "loss": 1.29,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8448838157358337,
+      "grad_norm": 5.5,
+      "learning_rate": 4.471947615165104e-05,
+      "loss": 1.1882,
+      "step": 12435
+    },
+    {
+      "epoch": 0.8452235358064955,
+      "grad_norm": 5.25,
+      "learning_rate": 4.471735290120941e-05,
+      "loss": 1.1683,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8455632558771572,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.471522965076777e-05,
+      "loss": 1.2714,
+      "step": 12445
+    },
+    {
+      "epoch": 0.845902975947819,
+      "grad_norm": 4.40625,
+      "learning_rate": 4.4713106400326135e-05,
+      "loss": 1.3175,
+      "step": 12450
+    },
+    {
+      "epoch": 0.8462426960184808,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.47109831498845e-05,
+      "loss": 1.3038,
+      "step": 12455
+    },
+    {
+      "epoch": 0.8465824160891425,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.470885989944286e-05,
+      "loss": 1.1808,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8469221361598043,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.470673664900123e-05,
+      "loss": 1.2165,
+      "step": 12465
+    },
+    {
+      "epoch": 0.8472618562304661,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.470461339855959e-05,
+      "loss": 1.1485,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8476015763011279,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.4702490148117955e-05,
+      "loss": 1.2321,
+      "step": 12475
+    },
+    {
+      "epoch": 0.8479412963717896,
+      "grad_norm": 5.375,
+      "learning_rate": 4.470036689767632e-05,
+      "loss": 1.26,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8482810164424515,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.469824364723468e-05,
+      "loss": 1.2419,
+      "step": 12485
+    },
+    {
+      "epoch": 0.8486207365131132,
+      "grad_norm": 6.125,
+      "learning_rate": 4.469612039679305e-05,
+      "loss": 1.3123,
+      "step": 12490
+    },
+    {
+      "epoch": 0.848960456583775,
+      "grad_norm": 5.625,
+      "learning_rate": 4.4693997146351404e-05,
+      "loss": 1.2793,
+      "step": 12495
+    },
+    {
+      "epoch": 0.8493001766544367,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4691873895909775e-05,
+      "loss": 1.2815,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8496398967250985,
+      "grad_norm": 5.875,
+      "learning_rate": 4.468975064546814e-05,
+      "loss": 1.1688,
+      "step": 12505
+    },
+    {
+      "epoch": 0.8499796167957603,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.4687627395026496e-05,
+      "loss": 1.2404,
+      "step": 12510
+    },
+    {
+      "epoch": 0.850319336866422,
+      "grad_norm": 4.375,
+      "learning_rate": 4.468550414458487e-05,
+      "loss": 1.2699,
+      "step": 12515
+    },
+    {
+      "epoch": 0.8506590569370839,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.468338089414323e-05,
+      "loss": 1.189,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8509987770077456,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.468125764370159e-05,
+      "loss": 1.3235,
+      "step": 12525
+    },
+    {
+      "epoch": 0.8513384970784074,
+      "grad_norm": 6.0,
+      "learning_rate": 4.467913439325996e-05,
+      "loss": 1.2097,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8516782171490692,
+      "grad_norm": 5.5,
+      "learning_rate": 4.467701114281832e-05,
+      "loss": 1.2797,
+      "step": 12535
+    },
+    {
+      "epoch": 0.852017937219731,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.467488789237668e-05,
+      "loss": 1.2518,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8523576572903927,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.467276464193505e-05,
+      "loss": 1.1987,
+      "step": 12545
+    },
+    {
+      "epoch": 0.8526973773610544,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.4670641391493415e-05,
+      "loss": 1.2294,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8530370974317163,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.466851814105177e-05,
+      "loss": 1.2659,
+      "step": 12555
+    },
+    {
+      "epoch": 0.853376817502378,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.466639489061014e-05,
+      "loss": 1.2959,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8537165375730398,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.46642716401685e-05,
+      "loss": 1.2159,
+      "step": 12565
+    },
+    {
+      "epoch": 0.8540562576437016,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4662148389726864e-05,
+      "loss": 1.2973,
+      "step": 12570
+    },
+    {
+      "epoch": 0.8543959777143634,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.4660025139285235e-05,
+      "loss": 1.1996,
+      "step": 12575
+    },
+    {
+      "epoch": 0.8547356977850251,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.465790188884359e-05,
+      "loss": 1.0957,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8550754178556869,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.4655778638401956e-05,
+      "loss": 1.2327,
+      "step": 12585
+    },
+    {
+      "epoch": 0.8554151379263487,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.465365538796033e-05,
+      "loss": 1.2876,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8557548579970105,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4651532137518684e-05,
+      "loss": 1.3265,
+      "step": 12595
+    },
+    {
+      "epoch": 0.8560945780676722,
+      "grad_norm": 4.65625,
+      "learning_rate": 4.464940888707705e-05,
+      "loss": 1.2417,
+      "step": 12600
+    },
+    {
+      "epoch": 0.856434298138334,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.464728563663542e-05,
+      "loss": 1.2604,
+      "step": 12605
+    },
+    {
+      "epoch": 0.8567740182089958,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.4645162386193776e-05,
+      "loss": 1.2433,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8571137382796575,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.464303913575214e-05,
+      "loss": 1.1929,
+      "step": 12615
+    },
+    {
+      "epoch": 0.8574534583503194,
+      "grad_norm": 5.25,
+      "learning_rate": 4.464091588531051e-05,
+      "loss": 1.246,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8577931784209811,
+      "grad_norm": 5.875,
+      "learning_rate": 4.463879263486887e-05,
+      "loss": 1.1964,
+      "step": 12625
+    },
+    {
+      "epoch": 0.8581328984916429,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.463666938442723e-05,
+      "loss": 1.2758,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8584726185623046,
+      "grad_norm": 5.125,
+      "learning_rate": 4.46345461339856e-05,
+      "loss": 1.3579,
+      "step": 12635
+    },
+    {
+      "epoch": 0.8588123386329665,
+      "grad_norm": 5.875,
+      "learning_rate": 4.463242288354396e-05,
+      "loss": 1.2887,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8591520587036282,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4630299633102324e-05,
+      "loss": 1.3022,
+      "step": 12645
+    },
+    {
+      "epoch": 0.85949177877429,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.462817638266069e-05,
+      "loss": 1.2505,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8598314988449518,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.462605313221905e-05,
+      "loss": 1.2819,
+      "step": 12655
+    },
+    {
+      "epoch": 0.8601712189156135,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.4623929881777416e-05,
+      "loss": 1.2581,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8605109389862753,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.462180663133578e-05,
+      "loss": 1.272,
+      "step": 12665
+    },
+    {
+      "epoch": 0.860850659056937,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.4619683380894144e-05,
+      "loss": 1.1982,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8611903791275989,
+      "grad_norm": 5.25,
+      "learning_rate": 4.461756013045251e-05,
+      "loss": 1.343,
+      "step": 12675
+    },
+    {
+      "epoch": 0.8615300991982606,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.461543688001087e-05,
+      "loss": 1.2064,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8618698192689224,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4613313629569236e-05,
+      "loss": 1.3267,
+      "step": 12685
+    },
+    {
+      "epoch": 0.8622095393395842,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.46111903791276e-05,
+      "loss": 1.2587,
+      "step": 12690
+    },
+    {
+      "epoch": 0.862549259410246,
+      "grad_norm": 5.5,
+      "learning_rate": 4.4609067128685964e-05,
+      "loss": 1.1989,
+      "step": 12695
+    },
+    {
+      "epoch": 0.8628889794809077,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.460694387824433e-05,
+      "loss": 1.2488,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8632286995515696,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.460482062780269e-05,
+      "loss": 1.1906,
+      "step": 12705
+    },
+    {
+      "epoch": 0.8635684196222313,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.4602697377361056e-05,
+      "loss": 1.2815,
+      "step": 12710
+    },
+    {
+      "epoch": 0.863908139692893,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.460057412691942e-05,
+      "loss": 1.1913,
+      "step": 12715
+    },
+    {
+      "epoch": 0.8642478597635548,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.4598450876477784e-05,
+      "loss": 1.2978,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8645875798342166,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.459632762603615e-05,
+      "loss": 1.2793,
+      "step": 12725
+    },
+    {
+      "epoch": 0.8649272999048784,
+      "grad_norm": 5.125,
+      "learning_rate": 4.459420437559451e-05,
+      "loss": 1.3684,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8652670199755401,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.4592081125152876e-05,
+      "loss": 1.3004,
+      "step": 12735
+    },
+    {
+      "epoch": 0.865606740046202,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.458995787471124e-05,
+      "loss": 1.3597,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8659464601168637,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.4587834624269604e-05,
+      "loss": 1.282,
+      "step": 12745
+    },
+    {
+      "epoch": 0.8662861801875255,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.458571137382797e-05,
+      "loss": 1.2649,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8666259002581872,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.458358812338633e-05,
+      "loss": 1.2636,
+      "step": 12755
+    },
+    {
+      "epoch": 0.866965620328849,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.4581464872944696e-05,
+      "loss": 1.2665,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8673053403995108,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4579341622503054e-05,
+      "loss": 1.2138,
+      "step": 12765
+    },
+    {
+      "epoch": 0.8676450604701725,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.4577218372061424e-05,
+      "loss": 1.1862,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8679847805408344,
+      "grad_norm": 5.125,
+      "learning_rate": 4.457509512161979e-05,
+      "loss": 1.2443,
+      "step": 12775
+    },
+    {
+      "epoch": 0.8683245006114961,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.457297187117815e-05,
+      "loss": 1.2125,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8686642206821579,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.4570848620736516e-05,
+      "loss": 1.3261,
+      "step": 12785
+    },
+    {
+      "epoch": 0.8690039407528197,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.456872537029488e-05,
+      "loss": 1.1999,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8693436608234815,
+      "grad_norm": 5.0,
+      "learning_rate": 4.4566602119853244e-05,
+      "loss": 1.29,
+      "step": 12795
+    },
+    {
+      "epoch": 0.8696833808941432,
+      "grad_norm": 4.625,
+      "learning_rate": 4.456447886941161e-05,
+      "loss": 1.2854,
+      "step": 12800
+    },
+    {
+      "epoch": 0.870023100964805,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.456235561896997e-05,
+      "loss": 1.2303,
+      "step": 12805
+    },
+    {
+      "epoch": 0.8703628210354668,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.4560232368528336e-05,
+      "loss": 1.3239,
+      "step": 12810
+    },
+    {
+      "epoch": 0.8707025411061285,
+      "grad_norm": 5.625,
+      "learning_rate": 4.45581091180867e-05,
+      "loss": 1.2973,
+      "step": 12815
+    },
+    {
+      "epoch": 0.8710422611767903,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.4555985867645064e-05,
+      "loss": 1.2355,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8713819812474521,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.455386261720343e-05,
+      "loss": 1.3573,
+      "step": 12825
+    },
+    {
+      "epoch": 0.8717217013181139,
+      "grad_norm": 5.625,
+      "learning_rate": 4.455173936676179e-05,
+      "loss": 1.1985,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8720614213887756,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.4549616116320156e-05,
+      "loss": 1.2452,
+      "step": 12835
+    },
+    {
+      "epoch": 0.8724011414594374,
+      "grad_norm": 5.5,
+      "learning_rate": 4.454749286587852e-05,
+      "loss": 1.254,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8727408615300992,
+      "grad_norm": 6.25,
+      "learning_rate": 4.4545369615436884e-05,
+      "loss": 1.2212,
+      "step": 12845
+    },
+    {
+      "epoch": 0.873080581600761,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.454324636499524e-05,
+      "loss": 1.4058,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8734203016714227,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.454112311455361e-05,
+      "loss": 1.2833,
+      "step": 12855
+    },
+    {
+      "epoch": 0.8737600217420846,
+      "grad_norm": 5.375,
+      "learning_rate": 4.4538999864111976e-05,
+      "loss": 1.1626,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8740997418127463,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.4536876613670334e-05,
+      "loss": 1.1897,
+      "step": 12865
+    },
+    {
+      "epoch": 0.874439461883408,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.4534753363228704e-05,
+      "loss": 1.0323,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8747791819540699,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.453263011278707e-05,
+      "loss": 1.294,
+      "step": 12875
+    },
+    {
+      "epoch": 0.8751189020247316,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.4530506862345426e-05,
+      "loss": 1.2045,
+      "step": 12880
+    },
+    {
+      "epoch": 0.8754586220953934,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4528383611903796e-05,
+      "loss": 1.2634,
+      "step": 12885
+    },
+    {
+      "epoch": 0.8757983421660551,
+      "grad_norm": 5.75,
+      "learning_rate": 4.452626036146216e-05,
+      "loss": 1.2424,
+      "step": 12890
+    },
+    {
+      "epoch": 0.876138062236717,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.452413711102052e-05,
+      "loss": 1.2894,
+      "step": 12895
+    },
+    {
+      "epoch": 0.8764777823073787,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.452201386057889e-05,
+      "loss": 1.3223,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8768175023780405,
+      "grad_norm": 5.75,
+      "learning_rate": 4.451989061013725e-05,
+      "loss": 1.276,
+      "step": 12905
+    },
+    {
+      "epoch": 0.8771572224487023,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.451776735969561e-05,
+      "loss": 1.2894,
+      "step": 12910
+    },
+    {
+      "epoch": 0.877496942519364,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.451564410925398e-05,
+      "loss": 1.2134,
+      "step": 12915
+    },
+    {
+      "epoch": 0.8778366625900258,
+      "grad_norm": 5.375,
+      "learning_rate": 4.451352085881234e-05,
+      "loss": 1.2833,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8781763826606876,
+      "grad_norm": 5.375,
+      "learning_rate": 4.45113976083707e-05,
+      "loss": 1.195,
+      "step": 12925
+    },
+    {
+      "epoch": 0.8785161027313494,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.450927435792907e-05,
+      "loss": 1.1976,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8788558228020111,
+      "grad_norm": 5.5,
+      "learning_rate": 4.450715110748743e-05,
+      "loss": 1.2609,
+      "step": 12935
+    },
+    {
+      "epoch": 0.8791955428726729,
+      "grad_norm": 5.125,
+      "learning_rate": 4.4505027857045794e-05,
+      "loss": 1.1418,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8795352629433347,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.4502904606604164e-05,
+      "loss": 1.2554,
+      "step": 12945
+    },
+    {
+      "epoch": 0.8798749830139965,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.450078135616252e-05,
+      "loss": 1.1859,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8802147030846582,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.4498658105720886e-05,
+      "loss": 1.2192,
+      "step": 12955
+    },
+    {
+      "epoch": 0.8805544231553201,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.4496534855279256e-05,
+      "loss": 1.2866,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8808941432259818,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.4494411604837614e-05,
+      "loss": 1.2555,
+      "step": 12965
+    },
+    {
+      "epoch": 0.8812338632966435,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.449228835439598e-05,
+      "loss": 1.2438,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8815735833673053,
+      "grad_norm": 5.125,
+      "learning_rate": 4.449016510395435e-05,
+      "loss": 1.2568,
+      "step": 12975
+    },
+    {
+      "epoch": 0.8819133034379671,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.4488041853512706e-05,
+      "loss": 1.2555,
+      "step": 12980
+    },
+    {
+      "epoch": 0.8822530235086289,
+      "grad_norm": 5.75,
+      "learning_rate": 4.448591860307107e-05,
+      "loss": 1.2326,
+      "step": 12985
+    },
+    {
+      "epoch": 0.8825927435792906,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.448379535262944e-05,
+      "loss": 1.3521,
+      "step": 12990
+    },
+    {
+      "epoch": 0.8829324636499525,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.44816721021878e-05,
+      "loss": 1.2178,
+      "step": 12995
+    },
+    {
+      "epoch": 0.8832721837206142,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.447954885174616e-05,
+      "loss": 1.2629,
+      "step": 13000
+    },
+    {
+      "epoch": 0.883611903791276,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4477425601304526e-05,
+      "loss": 1.3398,
+      "step": 13005
+    },
+    {
+      "epoch": 0.8839516238619378,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.447530235086289e-05,
+      "loss": 1.2208,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8842913439325996,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.4473179100421254e-05,
+      "loss": 1.3158,
+      "step": 13015
+    },
+    {
+      "epoch": 0.8846310640032613,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.447105584997962e-05,
+      "loss": 1.2954,
+      "step": 13020
+    },
+    {
+      "epoch": 0.884970784073923,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.446893259953798e-05,
+      "loss": 1.1522,
+      "step": 13025
+    },
+    {
+      "epoch": 0.8853105041445849,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.4466809349096346e-05,
+      "loss": 1.2627,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8856502242152466,
+      "grad_norm": 5.375,
+      "learning_rate": 4.446468609865471e-05,
+      "loss": 1.2615,
+      "step": 13035
+    },
+    {
+      "epoch": 0.8859899442859084,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.4462562848213074e-05,
+      "loss": 1.2648,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8863296643565702,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.446043959777144e-05,
+      "loss": 1.3051,
+      "step": 13045
+    },
+    {
+      "epoch": 0.886669384427232,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.44583163473298e-05,
+      "loss": 1.2887,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8870091044978937,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.4456193096888166e-05,
+      "loss": 1.2525,
+      "step": 13055
+    },
+    {
+      "epoch": 0.8873488245685555,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.445406984644653e-05,
+      "loss": 1.1736,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8876885446392173,
+      "grad_norm": 5.25,
+      "learning_rate": 4.4451946596004894e-05,
+      "loss": 1.5529,
+      "step": 13065
+    },
+    {
+      "epoch": 0.8880282647098791,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.444982334556326e-05,
+      "loss": 1.2534,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8883679847805408,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.444770009512162e-05,
+      "loss": 1.3322,
+      "step": 13075
+    },
+    {
+      "epoch": 0.8887077048512027,
+      "grad_norm": 5.375,
+      "learning_rate": 4.4445576844679986e-05,
+      "loss": 1.3008,
+      "step": 13080
+    },
+    {
+      "epoch": 0.8890474249218644,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.444345359423835e-05,
+      "loss": 1.2862,
+      "step": 13085
+    },
+    {
+      "epoch": 0.8893871449925261,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4441330343796714e-05,
+      "loss": 1.2897,
+      "step": 13090
+    },
+    {
+      "epoch": 0.889726865063188,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.443920709335508e-05,
+      "loss": 1.3325,
+      "step": 13095
+    },
+    {
+      "epoch": 0.8900665851338497,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.443708384291344e-05,
+      "loss": 1.2685,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8904063052045115,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4434960592471806e-05,
+      "loss": 1.2563,
+      "step": 13105
+    },
+    {
+      "epoch": 0.8907460252751732,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.443283734203017e-05,
+      "loss": 1.2471,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8910857453458351,
+      "grad_norm": 5.625,
+      "learning_rate": 4.4430714091588534e-05,
+      "loss": 1.2012,
+      "step": 13115
+    },
+    {
+      "epoch": 0.8914254654164968,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.44285908411469e-05,
+      "loss": 1.1938,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8917651854871586,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.442646759070526e-05,
+      "loss": 1.1818,
+      "step": 13125
+    },
+    {
+      "epoch": 0.8921049055578204,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.4424344340263626e-05,
+      "loss": 1.214,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8924446256284821,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.442222108982199e-05,
+      "loss": 1.1964,
+      "step": 13135
+    },
+    {
+      "epoch": 0.8927843456991439,
+      "grad_norm": 5.25,
+      "learning_rate": 4.4420097839380354e-05,
+      "loss": 1.28,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8931240657698056,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.441797458893872e-05,
+      "loss": 1.2743,
+      "step": 13145
+    },
+    {
+      "epoch": 0.8934637858404675,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.441585133849708e-05,
+      "loss": 1.3167,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8938035059111292,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.4413728088055446e-05,
+      "loss": 1.3637,
+      "step": 13155
+    },
+    {
+      "epoch": 0.894143225981791,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.441160483761381e-05,
+      "loss": 1.2104,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8944829460524528,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.4409481587172174e-05,
+      "loss": 1.2077,
+      "step": 13165
+    },
+    {
+      "epoch": 0.8948226661231146,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.440735833673054e-05,
+      "loss": 1.2682,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8951623861937763,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.44052350862889e-05,
+      "loss": 1.3327,
+      "step": 13175
+    },
+    {
+      "epoch": 0.8955021062644382,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.4403111835847266e-05,
+      "loss": 1.2627,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8958418263350999,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.440098858540563e-05,
+      "loss": 1.3023,
+      "step": 13185
+    },
+    {
+      "epoch": 0.8961815464057616,
+      "grad_norm": 5.75,
+      "learning_rate": 4.4398865334963994e-05,
+      "loss": 1.2094,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8965212664764234,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.439674208452236e-05,
+      "loss": 1.2525,
+      "step": 13195
+    },
+    {
+      "epoch": 0.8968609865470852,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.439461883408072e-05,
+      "loss": 1.258,
+      "step": 13200
+    },
+    {
+      "epoch": 0.897200706617747,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.439249558363908e-05,
+      "loss": 1.2255,
+      "step": 13205
+    },
+    {
+      "epoch": 0.8975404266884087,
+      "grad_norm": 5.125,
+      "learning_rate": 4.439037233319745e-05,
+      "loss": 1.1805,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8978801467590706,
+      "grad_norm": 5.125,
+      "learning_rate": 4.4388249082755814e-05,
+      "loss": 1.1722,
+      "step": 13215
+    },
+    {
+      "epoch": 0.8982198668297323,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.438612583231417e-05,
+      "loss": 1.2303,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8985595869003941,
+      "grad_norm": 5.0,
+      "learning_rate": 4.438400258187254e-05,
+      "loss": 1.2454,
+      "step": 13225
+    },
+    {
+      "epoch": 0.8988993069710558,
+      "grad_norm": 5.125,
+      "learning_rate": 4.4381879331430906e-05,
+      "loss": 1.2615,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8992390270417177,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.437975608098926e-05,
+      "loss": 1.2604,
+      "step": 13235
+    },
+    {
+      "epoch": 0.8995787471123794,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4377632830547634e-05,
+      "loss": 1.2461,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8999184671830411,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.4375509580106e-05,
+      "loss": 1.2228,
+      "step": 13245
+    },
+    {
+      "epoch": 0.900258187253703,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.4373386329664355e-05,
+      "loss": 1.27,
+      "step": 13250
+    },
+    {
+      "epoch": 0.9005979073243647,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.4371263079222726e-05,
+      "loss": 1.1938,
+      "step": 13255
+    },
+    {
+      "epoch": 0.9009376273950265,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.436913982878109e-05,
+      "loss": 1.2477,
+      "step": 13260
+    },
+    {
+      "epoch": 0.9012773474656883,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.436701657833945e-05,
+      "loss": 1.2043,
+      "step": 13265
+    },
+    {
+      "epoch": 0.9016170675363501,
+      "grad_norm": 5.25,
+      "learning_rate": 4.436489332789782e-05,
+      "loss": 1.1675,
+      "step": 13270
+    },
+    {
+      "epoch": 0.9019567876070118,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.4362770077456175e-05,
+      "loss": 1.2612,
+      "step": 13275
+    },
+    {
+      "epoch": 0.9022965076776736,
+      "grad_norm": 6.25,
+      "learning_rate": 4.436064682701454e-05,
+      "loss": 1.2117,
+      "step": 13280
+    },
+    {
+      "epoch": 0.9026362277483354,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.435852357657291e-05,
+      "loss": 1.301,
+      "step": 13285
+    },
+    {
+      "epoch": 0.9029759478189971,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.435640032613127e-05,
+      "loss": 1.3157,
+      "step": 13290
+    },
+    {
+      "epoch": 0.9033156678896589,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.435427707568963e-05,
+      "loss": 1.2373,
+      "step": 13295
+    },
+    {
+      "epoch": 0.9036553879603207,
+      "grad_norm": 5.625,
+      "learning_rate": 4.4352153825248e-05,
+      "loss": 1.3285,
+      "step": 13300
+    },
+    {
+      "epoch": 0.9039951080309825,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.435003057480636e-05,
+      "loss": 1.2784,
+      "step": 13305
+    },
+    {
+      "epoch": 0.9043348281016442,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.434790732436472e-05,
+      "loss": 1.3117,
+      "step": 13310
+    },
+    {
+      "epoch": 0.904674548172306,
+      "grad_norm": 5.875,
+      "learning_rate": 4.4345784073923094e-05,
+      "loss": 1.3783,
+      "step": 13315
+    },
+    {
+      "epoch": 0.9050142682429678,
+      "grad_norm": 5.375,
+      "learning_rate": 4.434366082348145e-05,
+      "loss": 1.2213,
+      "step": 13320
+    },
+    {
+      "epoch": 0.9053539883136296,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.4341537573039815e-05,
+      "loss": 1.1944,
+      "step": 13325
+    },
+    {
+      "epoch": 0.9056937083842913,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.4339414322598186e-05,
+      "loss": 1.2261,
+      "step": 13330
+    },
+    {
+      "epoch": 0.9060334284549532,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.433729107215654e-05,
+      "loss": 1.2055,
+      "step": 13335
+    },
+    {
+      "epoch": 0.9063731485256149,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.433516782171491e-05,
+      "loss": 1.2381,
+      "step": 13340
+    },
+    {
+      "epoch": 0.9067128685962766,
+      "grad_norm": 6.125,
+      "learning_rate": 4.433304457127327e-05,
+      "loss": 1.3519,
+      "step": 13345
+    },
+    {
+      "epoch": 0.9070525886669385,
+      "grad_norm": 5.375,
+      "learning_rate": 4.4330921320831635e-05,
+      "loss": 1.175,
+      "step": 13350
+    },
+    {
+      "epoch": 0.9073923087376002,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.432879807039e-05,
+      "loss": 1.2786,
+      "step": 13355
+    },
+    {
+      "epoch": 0.907732028808262,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.432667481994836e-05,
+      "loss": 1.3154,
+      "step": 13360
+    },
+    {
+      "epoch": 0.9080717488789237,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.432455156950673e-05,
+      "loss": 1.2406,
+      "step": 13365
+    },
+    {
+      "epoch": 0.9084114689495856,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.432242831906509e-05,
+      "loss": 1.3085,
+      "step": 13370
+    },
+    {
+      "epoch": 0.9087511890202473,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.4320305068623455e-05,
+      "loss": 1.1898,
+      "step": 13375
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.431818181818182e-05,
+      "loss": 1.3497,
+      "step": 13380
+    },
+    {
+      "epoch": 0.9094306291615709,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.431605856774018e-05,
+      "loss": 1.2525,
+      "step": 13385
+    },
+    {
+      "epoch": 0.9097703492322327,
+      "grad_norm": 5.75,
+      "learning_rate": 4.431393531729855e-05,
+      "loss": 1.0727,
+      "step": 13390
+    },
+    {
+      "epoch": 0.9101100693028944,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.431181206685691e-05,
+      "loss": 1.2205,
+      "step": 13395
+    },
+    {
+      "epoch": 0.9104497893735561,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.4309688816415275e-05,
+      "loss": 1.2034,
+      "step": 13400
+    },
+    {
+      "epoch": 0.910789509444218,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.430756556597364e-05,
+      "loss": 1.2803,
+      "step": 13405
+    },
+    {
+      "epoch": 0.9111292295148797,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.4305442315532e-05,
+      "loss": 1.1764,
+      "step": 13410
+    },
+    {
+      "epoch": 0.9114689495855415,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.430331906509037e-05,
+      "loss": 1.2458,
+      "step": 13415
+    },
+    {
+      "epoch": 0.9118086696562033,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.430119581464873e-05,
+      "loss": 1.1677,
+      "step": 13420
+    },
+    {
+      "epoch": 0.9121483897268651,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.4299072564207095e-05,
+      "loss": 1.294,
+      "step": 13425
+    },
+    {
+      "epoch": 0.9124881097975268,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.429694931376546e-05,
+      "loss": 1.1266,
+      "step": 13430
+    },
+    {
+      "epoch": 0.9128278298681887,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.429482606332382e-05,
+      "loss": 1.2967,
+      "step": 13435
+    },
+    {
+      "epoch": 0.9131675499388504,
+      "grad_norm": 4.625,
+      "learning_rate": 4.429270281288219e-05,
+      "loss": 1.2495,
+      "step": 13440
+    },
+    {
+      "epoch": 0.9135072700095122,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.429057956244055e-05,
+      "loss": 1.15,
+      "step": 13445
+    },
+    {
+      "epoch": 0.9138469900801739,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4288456311998915e-05,
+      "loss": 1.2797,
+      "step": 13450
+    },
+    {
+      "epoch": 0.9141867101508357,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.428633306155728e-05,
+      "loss": 1.2934,
+      "step": 13455
+    },
+    {
+      "epoch": 0.9145264302214975,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.428420981111564e-05,
+      "loss": 1.294,
+      "step": 13460
+    },
+    {
+      "epoch": 0.9148661502921592,
+      "grad_norm": 5.125,
+      "learning_rate": 4.428208656067401e-05,
+      "loss": 1.2653,
+      "step": 13465
+    },
+    {
+      "epoch": 0.9152058703628211,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.427996331023237e-05,
+      "loss": 1.1919,
+      "step": 13470
+    },
+    {
+      "epoch": 0.9155455904334828,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.4277840059790735e-05,
+      "loss": 1.1943,
+      "step": 13475
+    },
+    {
+      "epoch": 0.9158853105041446,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.42757168093491e-05,
+      "loss": 1.3182,
+      "step": 13480
+    },
+    {
+      "epoch": 0.9162250305748063,
+      "grad_norm": 5.625,
+      "learning_rate": 4.427359355890746e-05,
+      "loss": 1.1974,
+      "step": 13485
+    },
+    {
+      "epoch": 0.9165647506454682,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.427147030846583e-05,
+      "loss": 1.16,
+      "step": 13490
+    },
+    {
+      "epoch": 0.9169044707161299,
+      "grad_norm": 5.875,
+      "learning_rate": 4.426934705802419e-05,
+      "loss": 1.1946,
+      "step": 13495
+    },
+    {
+      "epoch": 0.9172441907867916,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.4267223807582555e-05,
+      "loss": 1.236,
+      "step": 13500
+    },
+    {
+      "epoch": 0.9175839108574535,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.426510055714092e-05,
+      "loss": 1.2421,
+      "step": 13505
+    },
+    {
+      "epoch": 0.9179236309281152,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.426297730669928e-05,
+      "loss": 1.3343,
+      "step": 13510
+    },
+    {
+      "epoch": 0.918263350998777,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.426085405625765e-05,
+      "loss": 1.2474,
+      "step": 13515
+    },
+    {
+      "epoch": 0.9186030710694388,
+      "grad_norm": 5.0,
+      "learning_rate": 4.425873080581601e-05,
+      "loss": 1.1175,
+      "step": 13520
+    },
+    {
+      "epoch": 0.9189427911401006,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.4256607555374375e-05,
+      "loss": 1.2716,
+      "step": 13525
+    },
+    {
+      "epoch": 0.9192825112107623,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.425448430493274e-05,
+      "loss": 1.1729,
+      "step": 13530
+    },
+    {
+      "epoch": 0.9196222312814241,
+      "grad_norm": 5.25,
+      "learning_rate": 4.42523610544911e-05,
+      "loss": 1.2879,
+      "step": 13535
+    },
+    {
+      "epoch": 0.9199619513520859,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.425023780404947e-05,
+      "loss": 1.2918,
+      "step": 13540
+    },
+    {
+      "epoch": 0.9203016714227477,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.4248114553607825e-05,
+      "loss": 1.3631,
+      "step": 13545
+    },
+    {
+      "epoch": 0.9206413914934094,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.4245991303166195e-05,
+      "loss": 1.3704,
+      "step": 13550
+    },
+    {
+      "epoch": 0.9209811115640713,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.424386805272456e-05,
+      "loss": 1.1724,
+      "step": 13555
+    },
+    {
+      "epoch": 0.921320831634733,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.4241744802282917e-05,
+      "loss": 1.2721,
+      "step": 13560
+    },
+    {
+      "epoch": 0.9216605517053947,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.423962155184129e-05,
+      "loss": 1.2365,
+      "step": 13565
+    },
+    {
+      "epoch": 0.9220002717760565,
+      "grad_norm": 5.625,
+      "learning_rate": 4.423749830139965e-05,
+      "loss": 1.2856,
+      "step": 13570
+    },
+    {
+      "epoch": 0.9223399918467183,
+      "grad_norm": 5.75,
+      "learning_rate": 4.423537505095801e-05,
+      "loss": 1.2041,
+      "step": 13575
+    },
+    {
+      "epoch": 0.9226797119173801,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.423325180051638e-05,
+      "loss": 1.2096,
+      "step": 13580
+    },
+    {
+      "epoch": 0.9230194319880418,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.423112855007474e-05,
+      "loss": 1.196,
+      "step": 13585
+    },
+    {
+      "epoch": 0.9233591520587037,
+      "grad_norm": 4.875,
+      "learning_rate": 4.42290052996331e-05,
+      "loss": 1.2142,
+      "step": 13590
+    },
+    {
+      "epoch": 0.9236988721293654,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.422688204919147e-05,
+      "loss": 1.2446,
+      "step": 13595
+    },
+    {
+      "epoch": 0.9240385922000272,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.4224758798749835e-05,
+      "loss": 1.1995,
+      "step": 13600
+    },
+    {
+      "epoch": 0.924378312270689,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.422263554830819e-05,
+      "loss": 1.3496,
+      "step": 13605
+    },
+    {
+      "epoch": 0.9247180323413507,
+      "grad_norm": 5.0,
+      "learning_rate": 4.4220512297866563e-05,
+      "loss": 1.2484,
+      "step": 13610
+    },
+    {
+      "epoch": 0.9250577524120125,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.421838904742493e-05,
+      "loss": 1.2143,
+      "step": 13615
+    },
+    {
+      "epoch": 0.9253974724826742,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.4216265796983285e-05,
+      "loss": 1.2765,
+      "step": 13620
+    },
+    {
+      "epoch": 0.9257371925533361,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.4214142546541655e-05,
+      "loss": 1.2259,
+      "step": 13625
+    },
+    {
+      "epoch": 0.9260769126239978,
+      "grad_norm": 5.625,
+      "learning_rate": 4.421201929610001e-05,
+      "loss": 1.1946,
+      "step": 13630
+    },
+    {
+      "epoch": 0.9264166326946596,
+      "grad_norm": 5.25,
+      "learning_rate": 4.420989604565838e-05,
+      "loss": 1.2091,
+      "step": 13635
+    },
+    {
+      "epoch": 0.9267563527653214,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.420777279521675e-05,
+      "loss": 1.2065,
+      "step": 13640
+    },
+    {
+      "epoch": 0.9270960728359832,
+      "grad_norm": 4.34375,
+      "learning_rate": 4.4205649544775105e-05,
+      "loss": 1.1235,
+      "step": 13645
+    },
+    {
+      "epoch": 0.9274357929066449,
+      "grad_norm": 6.125,
+      "learning_rate": 4.420352629433347e-05,
+      "loss": 1.186,
+      "step": 13650
+    },
+    {
+      "epoch": 0.9277755129773066,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.420140304389184e-05,
+      "loss": 1.2592,
+      "step": 13655
+    },
+    {
+      "epoch": 0.9281152330479685,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.41992797934502e-05,
+      "loss": 1.3153,
+      "step": 13660
+    },
+    {
+      "epoch": 0.9284549531186302,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.419715654300856e-05,
+      "loss": 1.2273,
+      "step": 13665
+    },
+    {
+      "epoch": 0.928794673189292,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.419503329256693e-05,
+      "loss": 1.2094,
+      "step": 13670
+    },
+    {
+      "epoch": 0.9291343932599538,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.419291004212529e-05,
+      "loss": 1.1768,
+      "step": 13675
+    },
+    {
+      "epoch": 0.9294741133306156,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.419078679168365e-05,
+      "loss": 1.2256,
+      "step": 13680
+    },
+    {
+      "epoch": 0.9298138334012773,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.4188663541242023e-05,
+      "loss": 1.1774,
+      "step": 13685
+    },
+    {
+      "epoch": 0.9301535534719392,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.418654029080038e-05,
+      "loss": 1.2136,
+      "step": 13690
+    },
+    {
+      "epoch": 0.9304932735426009,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.4184417040358745e-05,
+      "loss": 1.2401,
+      "step": 13695
+    },
+    {
+      "epoch": 0.9308329936132627,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.418229378991711e-05,
+      "loss": 1.1958,
+      "step": 13700
+    },
+    {
+      "epoch": 0.9311727136839244,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.418017053947547e-05,
+      "loss": 1.1763,
+      "step": 13705
+    },
+    {
+      "epoch": 0.9315124337545863,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.417804728903384e-05,
+      "loss": 1.2409,
+      "step": 13710
+    },
+    {
+      "epoch": 0.931852153825248,
+      "grad_norm": 6.0,
+      "learning_rate": 4.41759240385922e-05,
+      "loss": 1.3927,
+      "step": 13715
+    },
+    {
+      "epoch": 0.9321918738959097,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.4173800788150565e-05,
+      "loss": 1.2731,
+      "step": 13720
+    },
+    {
+      "epoch": 0.9325315939665716,
+      "grad_norm": 4.625,
+      "learning_rate": 4.417167753770893e-05,
+      "loss": 1.1963,
+      "step": 13725
+    },
+    {
+      "epoch": 0.9328713140372333,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.416955428726729e-05,
+      "loss": 1.2674,
+      "step": 13730
+    },
+    {
+      "epoch": 0.9332110341078951,
+      "grad_norm": 5.75,
+      "learning_rate": 4.416743103682566e-05,
+      "loss": 1.3261,
+      "step": 13735
+    },
+    {
+      "epoch": 0.9335507541785568,
+      "grad_norm": 5.875,
+      "learning_rate": 4.416530778638402e-05,
+      "loss": 1.259,
+      "step": 13740
+    },
+    {
+      "epoch": 0.9338904742492187,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.4163184535942385e-05,
+      "loss": 1.2383,
+      "step": 13745
+    },
+    {
+      "epoch": 0.9342301943198804,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.416106128550075e-05,
+      "loss": 1.1381,
+      "step": 13750
+    },
+    {
+      "epoch": 0.9345699143905422,
+      "grad_norm": 5.375,
+      "learning_rate": 4.415893803505911e-05,
+      "loss": 1.2698,
+      "step": 13755
+    },
+    {
+      "epoch": 0.934909634461204,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.415681478461748e-05,
+      "loss": 1.2736,
+      "step": 13760
+    },
+    {
+      "epoch": 0.9352493545318658,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.415469153417584e-05,
+      "loss": 1.221,
+      "step": 13765
+    },
+    {
+      "epoch": 0.9355890746025275,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.4152568283734205e-05,
+      "loss": 1.278,
+      "step": 13770
+    },
+    {
+      "epoch": 0.9359287946731893,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.415044503329257e-05,
+      "loss": 1.1842,
+      "step": 13775
+    },
+    {
+      "epoch": 0.9362685147438511,
+      "grad_norm": 5.5,
+      "learning_rate": 4.414832178285093e-05,
+      "loss": 1.2431,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9366082348145128,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.41461985324093e-05,
+      "loss": 1.2258,
+      "step": 13785
+    },
+    {
+      "epoch": 0.9369479548851746,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.414407528196766e-05,
+      "loss": 1.2248,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9372876749558364,
+      "grad_norm": 5.75,
+      "learning_rate": 4.4141952031526025e-05,
+      "loss": 1.3245,
+      "step": 13795
+    },
+    {
+      "epoch": 0.9376273950264982,
+      "grad_norm": 5.625,
+      "learning_rate": 4.413982878108439e-05,
+      "loss": 1.2212,
+      "step": 13800
+    },
+    {
+      "epoch": 0.9379671150971599,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.413770553064275e-05,
+      "loss": 1.2623,
+      "step": 13805
+    },
+    {
+      "epoch": 0.9383068351678218,
+      "grad_norm": 5.5,
+      "learning_rate": 4.413558228020112e-05,
+      "loss": 1.1033,
+      "step": 13810
+    },
+    {
+      "epoch": 0.9386465552384835,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.413345902975948e-05,
+      "loss": 1.2978,
+      "step": 13815
+    },
+    {
+      "epoch": 0.9389862753091452,
+      "grad_norm": 5.875,
+      "learning_rate": 4.4131335779317845e-05,
+      "loss": 1.1868,
+      "step": 13820
+    },
+    {
+      "epoch": 0.939325995379807,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.412921252887621e-05,
+      "loss": 1.1883,
+      "step": 13825
+    },
+    {
+      "epoch": 0.9396657154504688,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.412708927843457e-05,
+      "loss": 1.1975,
+      "step": 13830
+    },
+    {
+      "epoch": 0.9400054355211306,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.412496602799294e-05,
+      "loss": 1.3088,
+      "step": 13835
+    },
+    {
+      "epoch": 0.9403451555917923,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.41228427775513e-05,
+      "loss": 1.288,
+      "step": 13840
+    },
+    {
+      "epoch": 0.9406848756624542,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.4120719527109665e-05,
+      "loss": 1.1935,
+      "step": 13845
+    },
+    {
+      "epoch": 0.9410245957331159,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.411859627666803e-05,
+      "loss": 1.2231,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9413643158037777,
+      "grad_norm": 5.25,
+      "learning_rate": 4.411647302622639e-05,
+      "loss": 1.4583,
+      "step": 13855
+    },
+    {
+      "epoch": 0.9417040358744395,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.411434977578476e-05,
+      "loss": 1.4051,
+      "step": 13860
+    },
+    {
+      "epoch": 0.9420437559451013,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.411222652534312e-05,
+      "loss": 1.224,
+      "step": 13865
+    },
+    {
+      "epoch": 0.942383476015763,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.4110103274901485e-05,
+      "loss": 1.2394,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9427231960864247,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.410798002445985e-05,
+      "loss": 1.2492,
+      "step": 13875
+    },
+    {
+      "epoch": 0.9430629161570866,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.410585677401821e-05,
+      "loss": 1.309,
+      "step": 13880
+    },
+    {
+      "epoch": 0.9434026362277483,
+      "grad_norm": 4.875,
+      "learning_rate": 4.410373352357658e-05,
+      "loss": 1.1222,
+      "step": 13885
+    },
+    {
+      "epoch": 0.9437423562984101,
+      "grad_norm": 5.5,
+      "learning_rate": 4.410161027313494e-05,
+      "loss": 1.3475,
+      "step": 13890
+    },
+    {
+      "epoch": 0.9440820763690719,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.4099487022693305e-05,
+      "loss": 1.1974,
+      "step": 13895
+    },
+    {
+      "epoch": 0.9444217964397337,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.409736377225166e-05,
+      "loss": 1.1422,
+      "step": 13900
+    },
+    {
+      "epoch": 0.9447615165103954,
+      "grad_norm": 5.75,
+      "learning_rate": 4.409524052181003e-05,
+      "loss": 1.2031,
+      "step": 13905
+    },
+    {
+      "epoch": 0.9451012365810572,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.40931172713684e-05,
+      "loss": 1.341,
+      "step": 13910
+    },
+    {
+      "epoch": 0.945440956651719,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.4090994020926754e-05,
+      "loss": 1.2568,
+      "step": 13915
+    },
+    {
+      "epoch": 0.9457806767223808,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.4088870770485125e-05,
+      "loss": 1.1755,
+      "step": 13920
+    },
+    {
+      "epoch": 0.9461203967930425,
+      "grad_norm": 4.875,
+      "learning_rate": 4.408674752004349e-05,
+      "loss": 1.2094,
+      "step": 13925
+    },
+    {
+      "epoch": 0.9464601168637043,
+      "grad_norm": 5.625,
+      "learning_rate": 4.4084624269601846e-05,
+      "loss": 1.1937,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9467998369343661,
+      "grad_norm": 5.625,
+      "learning_rate": 4.408250101916022e-05,
+      "loss": 1.2302,
+      "step": 13935
+    },
+    {
+      "epoch": 0.9471395570050278,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.408037776871858e-05,
+      "loss": 1.3625,
+      "step": 13940
+    },
+    {
+      "epoch": 0.9474792770756897,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.407825451827694e-05,
+      "loss": 1.2166,
+      "step": 13945
+    },
+    {
+      "epoch": 0.9478189971463514,
+      "grad_norm": 5.0,
+      "learning_rate": 4.407613126783531e-05,
+      "loss": 1.198,
+      "step": 13950
+    },
+    {
+      "epoch": 0.9481587172170132,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.407400801739367e-05,
+      "loss": 1.2046,
+      "step": 13955
+    },
+    {
+      "epoch": 0.9484984372876749,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.407188476695203e-05,
+      "loss": 1.2691,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9488381573583368,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.40697615165104e-05,
+      "loss": 1.2282,
+      "step": 13965
+    },
+    {
+      "epoch": 0.9491778774289985,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.406763826606876e-05,
+      "loss": 1.1983,
+      "step": 13970
+    },
+    {
+      "epoch": 0.9495175974996602,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.406551501562712e-05,
+      "loss": 1.3094,
+      "step": 13975
+    },
+    {
+      "epoch": 0.9498573175703221,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.406339176518549e-05,
+      "loss": 1.4637,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9501970376409838,
+      "grad_norm": 4.875,
+      "learning_rate": 4.406126851474385e-05,
+      "loss": 1.1431,
+      "step": 13985
+    },
+    {
+      "epoch": 0.9505367577116456,
+      "grad_norm": 5.375,
+      "learning_rate": 4.4059145264302214e-05,
+      "loss": 1.153,
+      "step": 13990
+    },
+    {
+      "epoch": 0.9508764777823073,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.4057022013860585e-05,
+      "loss": 1.1679,
+      "step": 13995
+    },
+    {
+      "epoch": 0.9512161978529692,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.405489876341894e-05,
+      "loss": 1.1743,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9515559179236309,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.4052775512977306e-05,
+      "loss": 1.2426,
+      "step": 14005
+    },
+    {
+      "epoch": 0.9518956379942927,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.405065226253568e-05,
+      "loss": 1.1292,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9522353580649545,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.4048529012094034e-05,
+      "loss": 1.2389,
+      "step": 14015
+    },
+    {
+      "epoch": 0.9525750781356163,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.40464057616524e-05,
+      "loss": 1.2496,
+      "step": 14020
+    },
+    {
+      "epoch": 0.952914798206278,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.404428251121077e-05,
+      "loss": 1.2987,
+      "step": 14025
+    },
+    {
+      "epoch": 0.9532545182769399,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.4042159260769126e-05,
+      "loss": 1.1575,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9535942383476016,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.404003601032749e-05,
+      "loss": 1.0946,
+      "step": 14035
+    },
+    {
+      "epoch": 0.9539339584182633,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.403791275988586e-05,
+      "loss": 1.2572,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9542736784889251,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.403578950944422e-05,
+      "loss": 1.2961,
+      "step": 14045
+    },
+    {
+      "epoch": 0.9546133985595869,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.403366625900258e-05,
+      "loss": 1.2009,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9549531186302487,
+      "grad_norm": 5.625,
+      "learning_rate": 4.4031543008560946e-05,
+      "loss": 1.2586,
+      "step": 14055
+    },
+    {
+      "epoch": 0.9552928387009104,
+      "grad_norm": 5.375,
+      "learning_rate": 4.402941975811931e-05,
+      "loss": 1.2381,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9556325587715723,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.4027296507677674e-05,
+      "loss": 1.2619,
+      "step": 14065
+    },
+    {
+      "epoch": 0.955972278842234,
+      "grad_norm": 4.78125,
+      "learning_rate": 4.402517325723604e-05,
+      "loss": 1.2143,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9563119989128958,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.40230500067944e-05,
+      "loss": 1.24,
+      "step": 14075
+    },
+    {
+      "epoch": 0.9566517189835575,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.4020926756352766e-05,
+      "loss": 1.236,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9569914390542194,
+      "grad_norm": 5.125,
+      "learning_rate": 4.401880350591113e-05,
+      "loss": 1.2041,
+      "step": 14085
+    },
+    {
+      "epoch": 0.9573311591248811,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.4016680255469494e-05,
+      "loss": 1.1573,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9576708791955428,
+      "grad_norm": 5.375,
+      "learning_rate": 4.401455700502786e-05,
+      "loss": 1.2067,
+      "step": 14095
+    },
+    {
+      "epoch": 0.9580105992662047,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.401243375458622e-05,
+      "loss": 1.1755,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9583503193368664,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.4010310504144586e-05,
+      "loss": 1.1876,
+      "step": 14105
+    },
+    {
+      "epoch": 0.9586900394075282,
+      "grad_norm": 5.625,
+      "learning_rate": 4.400818725370295e-05,
+      "loss": 1.1633,
+      "step": 14110
+    },
+    {
+      "epoch": 0.95902975947819,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.4006064003261314e-05,
+      "loss": 1.2045,
+      "step": 14115
+    },
+    {
+      "epoch": 0.9593694795488518,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.400394075281968e-05,
+      "loss": 1.316,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9597091996195135,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.400181750237804e-05,
+      "loss": 1.273,
+      "step": 14125
+    },
+    {
+      "epoch": 0.9600489196901753,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.3999694251936406e-05,
+      "loss": 1.2777,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9603886397608371,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.399757100149477e-05,
+      "loss": 1.2828,
+      "step": 14135
+    },
+    {
+      "epoch": 0.9607283598314988,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.3995447751053134e-05,
+      "loss": 1.244,
+      "step": 14140
+    },
+    {
+      "epoch": 0.9610680799021606,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.39933245006115e-05,
+      "loss": 1.1513,
+      "step": 14145
+    },
+    {
+      "epoch": 0.9614077999728224,
+      "grad_norm": 5.25,
+      "learning_rate": 4.399120125016986e-05,
+      "loss": 1.3525,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9617475200434842,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.3989077999728226e-05,
+      "loss": 1.1456,
+      "step": 14155
+    },
+    {
+      "epoch": 0.9620872401141459,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.398695474928659e-05,
+      "loss": 1.2653,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9624269601848077,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.3984831498844954e-05,
+      "loss": 1.2754,
+      "step": 14165
+    },
+    {
+      "epoch": 0.9627666802554695,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.398270824840332e-05,
+      "loss": 1.2533,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9631064003261313,
+      "grad_norm": 5.875,
+      "learning_rate": 4.398058499796168e-05,
+      "loss": 1.2132,
+      "step": 14175
+    },
+    {
+      "epoch": 0.963446120396793,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.3978461747520046e-05,
+      "loss": 1.2698,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9637858404674549,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.397633849707841e-05,
+      "loss": 1.3084,
+      "step": 14185
+    },
+    {
+      "epoch": 0.9641255605381166,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.3974215246636774e-05,
+      "loss": 1.2281,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9644652806087783,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.397209199619514e-05,
+      "loss": 1.2369,
+      "step": 14195
+    },
+    {
+      "epoch": 0.9648050006794402,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.39699687457535e-05,
+      "loss": 1.2671,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9651447207501019,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.3967845495311866e-05,
+      "loss": 1.2596,
+      "step": 14205
+    },
+    {
+      "epoch": 0.9654844408207637,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.396572224487023e-05,
+      "loss": 1.252,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9658241608914254,
+      "grad_norm": 6.0,
+      "learning_rate": 4.3963598994428594e-05,
+      "loss": 1.2306,
+      "step": 14215
+    },
+    {
+      "epoch": 0.9661638809620873,
+      "grad_norm": 5.5,
+      "learning_rate": 4.396147574398696e-05,
+      "loss": 1.2353,
+      "step": 14220
+    },
+    {
+      "epoch": 0.966503601032749,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.395935249354532e-05,
+      "loss": 1.2106,
+      "step": 14225
+    },
+    {
+      "epoch": 0.9668433211034108,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.3957229243103686e-05,
+      "loss": 1.2212,
+      "step": 14230
+    },
+    {
+      "epoch": 0.9671830411740726,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.395510599266205e-05,
+      "loss": 1.283,
+      "step": 14235
+    },
+    {
+      "epoch": 0.9675227612447344,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3952982742220414e-05,
+      "loss": 1.1931,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9678624813153961,
+      "grad_norm": 5.625,
+      "learning_rate": 4.395085949177878e-05,
+      "loss": 1.241,
+      "step": 14245
+    },
+    {
+      "epoch": 0.9682022013860578,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.394873624133714e-05,
+      "loss": 1.2298,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9685419214567197,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.39466129908955e-05,
+      "loss": 1.2071,
+      "step": 14255
+    },
+    {
+      "epoch": 0.9688816415273814,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.394448974045387e-05,
+      "loss": 1.2242,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9692213615980432,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3942366490012234e-05,
+      "loss": 1.286,
+      "step": 14265
+    },
+    {
+      "epoch": 0.969561081668705,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.394024323957059e-05,
+      "loss": 1.1213,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9699008017393668,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.393811998912896e-05,
+      "loss": 1.2558,
+      "step": 14275
+    },
+    {
+      "epoch": 0.9702405218100285,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.3935996738687326e-05,
+      "loss": 1.1867,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9705802418806904,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.3933873488245684e-05,
+      "loss": 1.2104,
+      "step": 14285
+    },
+    {
+      "epoch": 0.9709199619513521,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.3931750237804054e-05,
+      "loss": 1.256,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9712596820220138,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.392962698736242e-05,
+      "loss": 1.2292,
+      "step": 14295
+    },
+    {
+      "epoch": 0.9715994020926756,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.3927503736920776e-05,
+      "loss": 1.2692,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9719391221633374,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.3925380486479146e-05,
+      "loss": 1.1954,
+      "step": 14305
+    },
+    {
+      "epoch": 0.9722788422339992,
+      "grad_norm": 4.5,
+      "learning_rate": 4.392325723603751e-05,
+      "loss": 1.2535,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9726185623046609,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.392113398559587e-05,
+      "loss": 1.194,
+      "step": 14315
+    },
+    {
+      "epoch": 0.9729582823753228,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.391901073515424e-05,
+      "loss": 1.3252,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9732980024459845,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3916887484712596e-05,
+      "loss": 1.1939,
+      "step": 14325
+    },
+    {
+      "epoch": 0.9736377225166463,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.391476423427096e-05,
+      "loss": 1.3529,
+      "step": 14330
+    },
+    {
+      "epoch": 0.973977442587308,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.391264098382933e-05,
+      "loss": 1.3696,
+      "step": 14335
+    },
+    {
+      "epoch": 0.9743171626579699,
+      "grad_norm": 6.375,
+      "learning_rate": 4.391051773338769e-05,
+      "loss": 1.193,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9746568827286316,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.390839448294605e-05,
+      "loss": 1.2367,
+      "step": 14345
+    },
+    {
+      "epoch": 0.9749966027992933,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.390627123250442e-05,
+      "loss": 1.2524,
+      "step": 14350
+    },
+    {
+      "epoch": 0.9753363228699552,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.390414798206278e-05,
+      "loss": 1.2,
+      "step": 14355
+    },
+    {
+      "epoch": 0.9756760429406169,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.3902024731621144e-05,
+      "loss": 1.2678,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9760157630112787,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.3899901481179514e-05,
+      "loss": 1.3105,
+      "step": 14365
+    },
+    {
+      "epoch": 0.9763554830819405,
+      "grad_norm": 5.5,
+      "learning_rate": 4.389777823073787e-05,
+      "loss": 1.2434,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9766952031526023,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.3895654980296236e-05,
+      "loss": 1.2963,
+      "step": 14375
+    },
+    {
+      "epoch": 0.977034923223264,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.3893531729854606e-05,
+      "loss": 1.2457,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9773746432939258,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.3891408479412964e-05,
+      "loss": 1.2878,
+      "step": 14385
+    },
+    {
+      "epoch": 0.9777143633645876,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.388928522897133e-05,
+      "loss": 1.1779,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9780540834352494,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.388716197852969e-05,
+      "loss": 1.2354,
+      "step": 14395
+    },
+    {
+      "epoch": 0.9783938035059111,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.3885038728088056e-05,
+      "loss": 1.1951,
+      "step": 14400
+    },
+    {
+      "epoch": 0.978733523576573,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.388291547764642e-05,
+      "loss": 1.2518,
+      "step": 14405
+    },
+    {
+      "epoch": 0.9790732436472347,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.3880792227204784e-05,
+      "loss": 1.1459,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9794129637178964,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.387866897676315e-05,
+      "loss": 1.1994,
+      "step": 14415
+    },
+    {
+      "epoch": 0.9797526837885582,
+      "grad_norm": 6.0,
+      "learning_rate": 4.387654572632151e-05,
+      "loss": 1.1468,
+      "step": 14420
+    },
+    {
+      "epoch": 0.98009240385922,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.3874422475879876e-05,
+      "loss": 1.3278,
+      "step": 14425
+    },
+    {
+      "epoch": 0.9804321239298818,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.387229922543824e-05,
+      "loss": 1.3339,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9807718440005435,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3870175974996604e-05,
+      "loss": 1.2362,
+      "step": 14435
+    },
+    {
+      "epoch": 0.9811115640712054,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.386805272455497e-05,
+      "loss": 1.3342,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9814512841418671,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.386592947411333e-05,
+      "loss": 1.2233,
+      "step": 14445
+    },
+    {
+      "epoch": 0.9817910042125289,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3863806223671696e-05,
+      "loss": 1.2085,
+      "step": 14450
+    },
+    {
+      "epoch": 0.9821307242831907,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.386168297323006e-05,
+      "loss": 1.209,
+      "step": 14455
+    },
+    {
+      "epoch": 0.9824704443538524,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.3859559722788424e-05,
+      "loss": 1.1824,
+      "step": 14460
+    },
+    {
+      "epoch": 0.9828101644245142,
+      "grad_norm": 6.0,
+      "learning_rate": 4.385743647234679e-05,
+      "loss": 1.2347,
+      "step": 14465
+    },
+    {
+      "epoch": 0.9831498844951759,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.385531322190515e-05,
+      "loss": 1.1067,
+      "step": 14470
+    },
+    {
+      "epoch": 0.9834896045658378,
+      "grad_norm": 5.5,
+      "learning_rate": 4.3853189971463516e-05,
+      "loss": 1.1852,
+      "step": 14475
+    },
+    {
+      "epoch": 0.9838293246364995,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.385106672102188e-05,
+      "loss": 1.2081,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9841690447071613,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.3848943470580244e-05,
+      "loss": 1.2885,
+      "step": 14485
+    },
+    {
+      "epoch": 0.9845087647778231,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.384682022013861e-05,
+      "loss": 1.1934,
+      "step": 14490
+    },
+    {
+      "epoch": 0.9848484848484849,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.384469696969697e-05,
+      "loss": 1.3078,
+      "step": 14495
+    },
+    {
+      "epoch": 0.9851882049191466,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.3842573719255336e-05,
+      "loss": 1.2572,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9855279249898083,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.38404504688137e-05,
+      "loss": 1.2187,
+      "step": 14505
+    },
+    {
+      "epoch": 0.9858676450604702,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.3838327218372064e-05,
+      "loss": 1.1919,
+      "step": 14510
+    },
+    {
+      "epoch": 0.9862073651311319,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.383620396793043e-05,
+      "loss": 1.2956,
+      "step": 14515
+    },
+    {
+      "epoch": 0.9865470852017937,
+      "grad_norm": 5.25,
+      "learning_rate": 4.383408071748879e-05,
+      "loss": 1.2752,
+      "step": 14520
+    },
+    {
+      "epoch": 0.9868868052724555,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.3831957467047156e-05,
+      "loss": 1.2119,
+      "step": 14525
+    },
+    {
+      "epoch": 0.9872265253431173,
+      "grad_norm": 5.125,
+      "learning_rate": 4.382983421660552e-05,
+      "loss": 1.2322,
+      "step": 14530
+    },
+    {
+      "epoch": 0.987566245413779,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.3827710966163884e-05,
+      "loss": 1.2043,
+      "step": 14535
+    },
+    {
+      "epoch": 0.9879059654844409,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.382558771572225e-05,
+      "loss": 1.2098,
+      "step": 14540
+    },
+    {
+      "epoch": 0.9882456855551026,
+      "grad_norm": 5.625,
+      "learning_rate": 4.382346446528061e-05,
+      "loss": 1.243,
+      "step": 14545
+    },
+    {
+      "epoch": 0.9885854056257644,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.3821341214838976e-05,
+      "loss": 1.301,
+      "step": 14550
+    },
+    {
+      "epoch": 0.9889251256964261,
+      "grad_norm": 5.375,
+      "learning_rate": 4.381921796439734e-05,
+      "loss": 1.2112,
+      "step": 14555
+    },
+    {
+      "epoch": 0.989264845767088,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.3817094713955704e-05,
+      "loss": 1.242,
+      "step": 14560
+    },
+    {
+      "epoch": 0.9896045658377497,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.381497146351407e-05,
+      "loss": 1.236,
+      "step": 14565
+    },
+    {
+      "epoch": 0.9899442859084114,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.381284821307243e-05,
+      "loss": 1.2841,
+      "step": 14570
+    },
+    {
+      "epoch": 0.9902840059790733,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.3810724962630796e-05,
+      "loss": 1.2192,
+      "step": 14575
+    },
+    {
+      "epoch": 0.990623726049735,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.380860171218916e-05,
+      "loss": 1.3014,
+      "step": 14580
+    },
+    {
+      "epoch": 0.9909634461203968,
+      "grad_norm": 5.25,
+      "learning_rate": 4.3806478461747524e-05,
+      "loss": 1.3244,
+      "step": 14585
+    },
+    {
+      "epoch": 0.9913031661910585,
+      "grad_norm": 5.375,
+      "learning_rate": 4.380435521130589e-05,
+      "loss": 1.2924,
+      "step": 14590
+    },
+    {
+      "epoch": 0.9916428862617204,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3802231960864245e-05,
+      "loss": 1.1944,
+      "step": 14595
+    },
+    {
+      "epoch": 0.9919826063323821,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.3800108710422616e-05,
+      "loss": 1.2561,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9923223264030439,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.379798545998098e-05,
+      "loss": 1.2697,
+      "step": 14605
+    },
+    {
+      "epoch": 0.9926620464737057,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.379586220953934e-05,
+      "loss": 1.2632,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9930017665443674,
+      "grad_norm": 5.25,
+      "learning_rate": 4.379373895909771e-05,
+      "loss": 1.2719,
+      "step": 14615
+    },
+    {
+      "epoch": 0.9933414866150292,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.379161570865607e-05,
+      "loss": 1.1847,
+      "step": 14620
+    },
+    {
+      "epoch": 0.993681206685691,
+      "grad_norm": 6.0,
+      "learning_rate": 4.378949245821443e-05,
+      "loss": 1.1708,
+      "step": 14625
+    },
+    {
+      "epoch": 0.9940209267563528,
+      "grad_norm": 5.5,
+      "learning_rate": 4.37873692077728e-05,
+      "loss": 1.2978,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9943606468270145,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.3785245957331164e-05,
+      "loss": 1.2321,
+      "step": 14635
+    },
+    {
+      "epoch": 0.9947003668976763,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.378312270688952e-05,
+      "loss": 1.2089,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9950400869683381,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.378099945644789e-05,
+      "loss": 1.1395,
+      "step": 14645
+    },
+    {
+      "epoch": 0.9953798070389999,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.3778876206006256e-05,
+      "loss": 1.2677,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9957195271096616,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.377675295556461e-05,
+      "loss": 1.2177,
+      "step": 14655
+    },
+    {
+      "epoch": 0.9960592471803235,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.3774629705122984e-05,
+      "loss": 1.3291,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9963989672509852,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.377250645468135e-05,
+      "loss": 1.2722,
+      "step": 14665
+    },
+    {
+      "epoch": 0.9967386873216469,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.3770383204239705e-05,
+      "loss": 1.2774,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9970784073923087,
+      "grad_norm": 6.125,
+      "learning_rate": 4.3768259953798076e-05,
+      "loss": 1.2961,
+      "step": 14675
+    },
+    {
+      "epoch": 0.9974181274629705,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.376613670335643e-05,
+      "loss": 1.1776,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9977578475336323,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.37640134529148e-05,
+      "loss": 1.1692,
+      "step": 14685
+    },
+    {
+      "epoch": 0.998097567604294,
+      "grad_norm": 5.25,
+      "learning_rate": 4.376189020247317e-05,
+      "loss": 1.3158,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9984372876749559,
+      "grad_norm": 5.875,
+      "learning_rate": 4.3759766952031525e-05,
+      "loss": 1.3805,
+      "step": 14695
+    },
+    {
+      "epoch": 0.9987770077456176,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.375764370158989e-05,
+      "loss": 1.2939,
+      "step": 14700
+    },
+    {
+      "epoch": 0.9991167278162794,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.375552045114826e-05,
+      "loss": 1.2312,
+      "step": 14705
+    },
+    {
+      "epoch": 0.9994564478869412,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.375339720070662e-05,
+      "loss": 1.2717,
+      "step": 14710
+    },
+    {
+      "epoch": 0.999796167957603,
+      "grad_norm": 5.0,
+      "learning_rate": 4.375127395026498e-05,
+      "loss": 1.2698,
+      "step": 14715
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.5637719631195068,
+      "eval_runtime": 146.4458,
+      "eval_samples_per_second": 70.463,
+      "eval_steps_per_second": 8.809,
+      "step": 14718
+    },
+    {
+      "epoch": 1.0001358880282647,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.374915069982335e-05,
+      "loss": 1.2018,
+      "step": 14720
+    },
+    {
+      "epoch": 1.0004756080989265,
+      "grad_norm": 4.59375,
+      "learning_rate": 4.374702744938171e-05,
+      "loss": 1.0695,
+      "step": 14725
+    },
+    {
+      "epoch": 1.0008153281695882,
+      "grad_norm": 4.4375,
+      "learning_rate": 4.374490419894007e-05,
+      "loss": 0.9714,
+      "step": 14730
+    },
+    {
+      "epoch": 1.00115504824025,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3742780948498444e-05,
+      "loss": 1.1104,
+      "step": 14735
+    },
+    {
+      "epoch": 1.0014947683109119,
+      "grad_norm": 5.875,
+      "learning_rate": 4.37406576980568e-05,
+      "loss": 1.0126,
+      "step": 14740
+    },
+    {
+      "epoch": 1.0018344883815735,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.3738534447615165e-05,
+      "loss": 1.0422,
+      "step": 14745
+    },
+    {
+      "epoch": 1.0021742084522354,
+      "grad_norm": 5.625,
+      "learning_rate": 4.373641119717353e-05,
+      "loss": 1.1246,
+      "step": 14750
+    },
+    {
+      "epoch": 1.0025139285228972,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.373428794673189e-05,
+      "loss": 1.0281,
+      "step": 14755
+    },
+    {
+      "epoch": 1.0028536485935589,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.373216469629026e-05,
+      "loss": 1.081,
+      "step": 14760
+    },
+    {
+      "epoch": 1.0031933686642207,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.373004144584862e-05,
+      "loss": 1.1052,
+      "step": 14765
+    },
+    {
+      "epoch": 1.0035330887348826,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.3727918195406985e-05,
+      "loss": 1.0656,
+      "step": 14770
+    },
+    {
+      "epoch": 1.0038728088055442,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.372579494496535e-05,
+      "loss": 1.1016,
+      "step": 14775
+    },
+    {
+      "epoch": 1.004212528876206,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.372367169452371e-05,
+      "loss": 1.1848,
+      "step": 14780
+    },
+    {
+      "epoch": 1.0045522489468677,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.372154844408208e-05,
+      "loss": 1.0373,
+      "step": 14785
+    },
+    {
+      "epoch": 1.0048919690175295,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.371942519364044e-05,
+      "loss": 1.1001,
+      "step": 14790
+    },
+    {
+      "epoch": 1.0052316890881914,
+      "grad_norm": 5.875,
+      "learning_rate": 4.3717301943198805e-05,
+      "loss": 1.1244,
+      "step": 14795
+    },
+    {
+      "epoch": 1.005571409158853,
+      "grad_norm": 5.125,
+      "learning_rate": 4.371517869275717e-05,
+      "loss": 0.9535,
+      "step": 14800
+    },
+    {
+      "epoch": 1.0059111292295149,
+      "grad_norm": 5.75,
+      "learning_rate": 4.371305544231553e-05,
+      "loss": 1.0385,
+      "step": 14805
+    },
+    {
+      "epoch": 1.0062508493001767,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.37109321918739e-05,
+      "loss": 1.0742,
+      "step": 14810
+    },
+    {
+      "epoch": 1.0065905693708384,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.370880894143226e-05,
+      "loss": 1.0783,
+      "step": 14815
+    },
+    {
+      "epoch": 1.0069302894415002,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.3706685690990625e-05,
+      "loss": 1.1003,
+      "step": 14820
+    },
+    {
+      "epoch": 1.007270009512162,
+      "grad_norm": 5.0,
+      "learning_rate": 4.370456244054899e-05,
+      "loss": 1.0418,
+      "step": 14825
+    },
+    {
+      "epoch": 1.0076097295828237,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.370243919010735e-05,
+      "loss": 1.0058,
+      "step": 14830
+    },
+    {
+      "epoch": 1.0079494496534855,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.370031593966572e-05,
+      "loss": 1.0785,
+      "step": 14835
+    },
+    {
+      "epoch": 1.0082891697241474,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.369819268922408e-05,
+      "loss": 1.1119,
+      "step": 14840
+    },
+    {
+      "epoch": 1.008628889794809,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.3696069438782445e-05,
+      "loss": 1.1512,
+      "step": 14845
+    },
+    {
+      "epoch": 1.0089686098654709,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.369394618834081e-05,
+      "loss": 1.1107,
+      "step": 14850
+    },
+    {
+      "epoch": 1.0093083299361327,
+      "grad_norm": 5.25,
+      "learning_rate": 4.369182293789917e-05,
+      "loss": 1.1896,
+      "step": 14855
+    },
+    {
+      "epoch": 1.0096480500067944,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.368969968745754e-05,
+      "loss": 1.1065,
+      "step": 14860
+    },
+    {
+      "epoch": 1.0099877700774562,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.36875764370159e-05,
+      "loss": 1.0064,
+      "step": 14865
+    },
+    {
+      "epoch": 1.0103274901481178,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.3685453186574265e-05,
+      "loss": 1.0494,
+      "step": 14870
+    },
+    {
+      "epoch": 1.0106672102187797,
+      "grad_norm": 5.25,
+      "learning_rate": 4.368332993613263e-05,
+      "loss": 1.3135,
+      "step": 14875
+    },
+    {
+      "epoch": 1.0110069302894416,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.368120668569099e-05,
+      "loss": 1.041,
+      "step": 14880
+    },
+    {
+      "epoch": 1.0113466503601032,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.367908343524936e-05,
+      "loss": 1.0158,
+      "step": 14885
+    },
+    {
+      "epoch": 1.011686370430765,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.367696018480772e-05,
+      "loss": 1.0098,
+      "step": 14890
+    },
+    {
+      "epoch": 1.012026090501427,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.3674836934366085e-05,
+      "loss": 1.0721,
+      "step": 14895
+    },
+    {
+      "epoch": 1.0123658105720885,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.367271368392445e-05,
+      "loss": 1.03,
+      "step": 14900
+    },
+    {
+      "epoch": 1.0127055306427504,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.367059043348281e-05,
+      "loss": 1.103,
+      "step": 14905
+    },
+    {
+      "epoch": 1.0130452507134122,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.366846718304118e-05,
+      "loss": 1.0943,
+      "step": 14910
+    },
+    {
+      "epoch": 1.0133849707840739,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.366634393259954e-05,
+      "loss": 1.0897,
+      "step": 14915
+    },
+    {
+      "epoch": 1.0137246908547357,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.3664220682157905e-05,
+      "loss": 1.0605,
+      "step": 14920
+    },
+    {
+      "epoch": 1.0140644109253976,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.366209743171627e-05,
+      "loss": 1.0348,
+      "step": 14925
+    },
+    {
+      "epoch": 1.0144041309960592,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.365997418127463e-05,
+      "loss": 1.0015,
+      "step": 14930
+    },
+    {
+      "epoch": 1.014743851066721,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.3657850930833e-05,
+      "loss": 1.0615,
+      "step": 14935
+    },
+    {
+      "epoch": 1.015083571137383,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.365572768039136e-05,
+      "loss": 0.9837,
+      "step": 14940
+    },
+    {
+      "epoch": 1.0154232912080445,
+      "grad_norm": 5.5,
+      "learning_rate": 4.3653604429949725e-05,
+      "loss": 1.117,
+      "step": 14945
+    },
+    {
+      "epoch": 1.0157630112787064,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.365148117950808e-05,
+      "loss": 1.1029,
+      "step": 14950
+    },
+    {
+      "epoch": 1.016102731349368,
+      "grad_norm": 5.25,
+      "learning_rate": 4.364935792906645e-05,
+      "loss": 1.0228,
+      "step": 14955
+    },
+    {
+      "epoch": 1.0164424514200299,
+      "grad_norm": 5.0,
+      "learning_rate": 4.364723467862482e-05,
+      "loss": 1.1051,
+      "step": 14960
+    },
+    {
+      "epoch": 1.0167821714906917,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3645111428183175e-05,
+      "loss": 0.9864,
+      "step": 14965
+    },
+    {
+      "epoch": 1.0171218915613534,
+      "grad_norm": 4.03125,
+      "learning_rate": 4.3642988177741545e-05,
+      "loss": 1.0318,
+      "step": 14970
+    },
+    {
+      "epoch": 1.0174616116320152,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.364086492729991e-05,
+      "loss": 1.0617,
+      "step": 14975
+    },
+    {
+      "epoch": 1.017801331702677,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3638741676858267e-05,
+      "loss": 1.0564,
+      "step": 14980
+    },
+    {
+      "epoch": 1.0181410517733387,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.363661842641664e-05,
+      "loss": 1.1503,
+      "step": 14985
+    },
+    {
+      "epoch": 1.0184807718440005,
+      "grad_norm": 5.125,
+      "learning_rate": 4.3634495175975e-05,
+      "loss": 1.0823,
+      "step": 14990
+    },
+    {
+      "epoch": 1.0188204919146624,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.363237192553336e-05,
+      "loss": 1.0148,
+      "step": 14995
+    },
+    {
+      "epoch": 1.019160211985324,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.363024867509173e-05,
+      "loss": 1.0648,
+      "step": 15000
+    },
+    {
+      "epoch": 1.0194999320559859,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.362812542465009e-05,
+      "loss": 1.1015,
+      "step": 15005
+    },
+    {
+      "epoch": 1.0198396521266477,
+      "grad_norm": 5.75,
+      "learning_rate": 4.362600217420845e-05,
+      "loss": 1.0914,
+      "step": 15010
+    },
+    {
+      "epoch": 1.0201793721973094,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.362387892376682e-05,
+      "loss": 1.079,
+      "step": 15015
+    },
+    {
+      "epoch": 1.0205190922679712,
+      "grad_norm": 5.75,
+      "learning_rate": 4.362175567332518e-05,
+      "loss": 1.0883,
+      "step": 15020
+    },
+    {
+      "epoch": 1.020858812338633,
+      "grad_norm": 5.375,
+      "learning_rate": 4.361963242288354e-05,
+      "loss": 0.9958,
+      "step": 15025
+    },
+    {
+      "epoch": 1.0211985324092947,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.3617509172441913e-05,
+      "loss": 1.1151,
+      "step": 15030
+    },
+    {
+      "epoch": 1.0215382524799566,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.361538592200027e-05,
+      "loss": 1.1202,
+      "step": 15035
+    },
+    {
+      "epoch": 1.0218779725506182,
+      "grad_norm": 6.0,
+      "learning_rate": 4.3613262671558635e-05,
+      "loss": 1.034,
+      "step": 15040
+    },
+    {
+      "epoch": 1.02221769262128,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.3611139421117005e-05,
+      "loss": 1.1466,
+      "step": 15045
+    },
+    {
+      "epoch": 1.022557412691942,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.360901617067536e-05,
+      "loss": 1.1348,
+      "step": 15050
+    },
+    {
+      "epoch": 1.0228971327626035,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.360689292023373e-05,
+      "loss": 1.0874,
+      "step": 15055
+    },
+    {
+      "epoch": 1.0232368528332654,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.36047696697921e-05,
+      "loss": 1.1275,
+      "step": 15060
+    },
+    {
+      "epoch": 1.0235765729039272,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.3602646419350455e-05,
+      "loss": 1.0494,
+      "step": 15065
+    },
+    {
+      "epoch": 1.0239162929745889,
+      "grad_norm": 5.875,
+      "learning_rate": 4.360052316890882e-05,
+      "loss": 1.061,
+      "step": 15070
+    },
+    {
+      "epoch": 1.0242560130452507,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.359839991846719e-05,
+      "loss": 1.1376,
+      "step": 15075
+    },
+    {
+      "epoch": 1.0245957331159126,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.359627666802555e-05,
+      "loss": 1.1103,
+      "step": 15080
+    },
+    {
+      "epoch": 1.0249354531865742,
+      "grad_norm": 5.0,
+      "learning_rate": 4.359415341758391e-05,
+      "loss": 1.0628,
+      "step": 15085
+    },
+    {
+      "epoch": 1.025275173257236,
+      "grad_norm": 5.625,
+      "learning_rate": 4.359203016714228e-05,
+      "loss": 1.1542,
+      "step": 15090
+    },
+    {
+      "epoch": 1.025614893327898,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.358990691670064e-05,
+      "loss": 1.2193,
+      "step": 15095
+    },
+    {
+      "epoch": 1.0259546133985595,
+      "grad_norm": 5.5,
+      "learning_rate": 4.3587783666259e-05,
+      "loss": 1.0549,
+      "step": 15100
+    },
+    {
+      "epoch": 1.0262943334692214,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.358566041581737e-05,
+      "loss": 1.0443,
+      "step": 15105
+    },
+    {
+      "epoch": 1.0266340535398832,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.358353716537573e-05,
+      "loss": 1.0359,
+      "step": 15110
+    },
+    {
+      "epoch": 1.0269737736105449,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.3581413914934095e-05,
+      "loss": 1.0201,
+      "step": 15115
+    },
+    {
+      "epoch": 1.0273134936812067,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.357929066449246e-05,
+      "loss": 1.185,
+      "step": 15120
+    },
+    {
+      "epoch": 1.0276532137518684,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.357716741405082e-05,
+      "loss": 1.08,
+      "step": 15125
+    },
+    {
+      "epoch": 1.0279929338225302,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.357504416360919e-05,
+      "loss": 1.0594,
+      "step": 15130
+    },
+    {
+      "epoch": 1.028332653893192,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.357292091316755e-05,
+      "loss": 0.9847,
+      "step": 15135
+    },
+    {
+      "epoch": 1.0286723739638537,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.3570797662725915e-05,
+      "loss": 1.0833,
+      "step": 15140
+    },
+    {
+      "epoch": 1.0290120940345155,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.356867441228428e-05,
+      "loss": 1.0672,
+      "step": 15145
+    },
+    {
+      "epoch": 1.0293518141051774,
+      "grad_norm": 5.125,
+      "learning_rate": 4.356655116184264e-05,
+      "loss": 1.0016,
+      "step": 15150
+    },
+    {
+      "epoch": 1.029691534175839,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.356442791140101e-05,
+      "loss": 1.1235,
+      "step": 15155
+    },
+    {
+      "epoch": 1.0300312542465009,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.356230466095937e-05,
+      "loss": 1.0815,
+      "step": 15160
+    },
+    {
+      "epoch": 1.0303709743171627,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.3560181410517735e-05,
+      "loss": 1.0143,
+      "step": 15165
+    },
+    {
+      "epoch": 1.0307106943878244,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.35580581600761e-05,
+      "loss": 1.0911,
+      "step": 15170
+    },
+    {
+      "epoch": 1.0310504144584862,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.355593490963446e-05,
+      "loss": 1.0943,
+      "step": 15175
+    },
+    {
+      "epoch": 1.031390134529148,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.355381165919283e-05,
+      "loss": 1.0582,
+      "step": 15180
+    },
+    {
+      "epoch": 1.0317298545998097,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.355168840875119e-05,
+      "loss": 1.1663,
+      "step": 15185
+    },
+    {
+      "epoch": 1.0320695746704716,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.3549565158309555e-05,
+      "loss": 1.0783,
+      "step": 15190
+    },
+    {
+      "epoch": 1.0324092947411334,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.354744190786792e-05,
+      "loss": 1.1416,
+      "step": 15195
+    },
+    {
+      "epoch": 1.032749014811795,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.354531865742628e-05,
+      "loss": 1.0399,
+      "step": 15200
+    },
+    {
+      "epoch": 1.033088734882457,
+      "grad_norm": 5.625,
+      "learning_rate": 4.354319540698465e-05,
+      "loss": 1.0956,
+      "step": 15205
+    },
+    {
+      "epoch": 1.0334284549531185,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.354107215654301e-05,
+      "loss": 1.0466,
+      "step": 15210
+    },
+    {
+      "epoch": 1.0337681750237804,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.3538948906101375e-05,
+      "loss": 1.084,
+      "step": 15215
+    },
+    {
+      "epoch": 1.0341078950944422,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.353682565565974e-05,
+      "loss": 1.0921,
+      "step": 15220
+    },
+    {
+      "epoch": 1.0344476151651039,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.35347024052181e-05,
+      "loss": 0.9853,
+      "step": 15225
+    },
+    {
+      "epoch": 1.0347873352357657,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.353257915477647e-05,
+      "loss": 0.9947,
+      "step": 15230
+    },
+    {
+      "epoch": 1.0351270553064276,
+      "grad_norm": 5.625,
+      "learning_rate": 4.353045590433483e-05,
+      "loss": 1.0958,
+      "step": 15235
+    },
+    {
+      "epoch": 1.0354667753770892,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.3528332653893195e-05,
+      "loss": 1.0595,
+      "step": 15240
+    },
+    {
+      "epoch": 1.035806495447751,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.352620940345156e-05,
+      "loss": 1.0722,
+      "step": 15245
+    },
+    {
+      "epoch": 1.036146215518413,
+      "grad_norm": 5.625,
+      "learning_rate": 4.352408615300992e-05,
+      "loss": 1.1831,
+      "step": 15250
+    },
+    {
+      "epoch": 1.0364859355890745,
+      "grad_norm": 5.625,
+      "learning_rate": 4.352196290256829e-05,
+      "loss": 1.112,
+      "step": 15255
+    },
+    {
+      "epoch": 1.0368256556597364,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.351983965212665e-05,
+      "loss": 1.0384,
+      "step": 15260
+    },
+    {
+      "epoch": 1.0371653757303982,
+      "grad_norm": 6.125,
+      "learning_rate": 4.3517716401685015e-05,
+      "loss": 1.1383,
+      "step": 15265
+    },
+    {
+      "epoch": 1.0375050958010599,
+      "grad_norm": 5.625,
+      "learning_rate": 4.351559315124338e-05,
+      "loss": 1.0749,
+      "step": 15270
+    },
+    {
+      "epoch": 1.0378448158717217,
+      "grad_norm": 4.5625,
+      "learning_rate": 4.351346990080174e-05,
+      "loss": 1.0581,
+      "step": 15275
+    },
+    {
+      "epoch": 1.0381845359423836,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.351134665036011e-05,
+      "loss": 1.0959,
+      "step": 15280
+    },
+    {
+      "epoch": 1.0385242560130452,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.350922339991847e-05,
+      "loss": 1.1106,
+      "step": 15285
+    },
+    {
+      "epoch": 1.038863976083707,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3507100149476835e-05,
+      "loss": 1.1439,
+      "step": 15290
+    },
+    {
+      "epoch": 1.0392036961543687,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.35049768990352e-05,
+      "loss": 1.0692,
+      "step": 15295
+    },
+    {
+      "epoch": 1.0395434162250305,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.350285364859356e-05,
+      "loss": 1.1076,
+      "step": 15300
+    },
+    {
+      "epoch": 1.0398831362956924,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.350073039815192e-05,
+      "loss": 0.9724,
+      "step": 15305
+    },
+    {
+      "epoch": 1.040222856366354,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.349860714771029e-05,
+      "loss": 0.9485,
+      "step": 15310
+    },
+    {
+      "epoch": 1.0405625764370159,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.3496483897268655e-05,
+      "loss": 1.0745,
+      "step": 15315
+    },
+    {
+      "epoch": 1.0409022965076777,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.349436064682701e-05,
+      "loss": 1.1008,
+      "step": 15320
+    },
+    {
+      "epoch": 1.0412420165783394,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.349223739638538e-05,
+      "loss": 1.1637,
+      "step": 15325
+    },
+    {
+      "epoch": 1.0415817366490012,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.349011414594375e-05,
+      "loss": 1.2398,
+      "step": 15330
+    },
+    {
+      "epoch": 1.041921456719663,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3487990895502104e-05,
+      "loss": 1.036,
+      "step": 15335
+    },
+    {
+      "epoch": 1.0422611767903247,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.3485867645060475e-05,
+      "loss": 1.077,
+      "step": 15340
+    },
+    {
+      "epoch": 1.0426008968609866,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.348374439461884e-05,
+      "loss": 1.086,
+      "step": 15345
+    },
+    {
+      "epoch": 1.0429406169316484,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.3481621144177196e-05,
+      "loss": 1.1431,
+      "step": 15350
+    },
+    {
+      "epoch": 1.04328033700231,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.347949789373557e-05,
+      "loss": 1.0246,
+      "step": 15355
+    },
+    {
+      "epoch": 1.043620057072972,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.347737464329393e-05,
+      "loss": 1.0849,
+      "step": 15360
+    },
+    {
+      "epoch": 1.0439597771436337,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.347525139285229e-05,
+      "loss": 1.0727,
+      "step": 15365
+    },
+    {
+      "epoch": 1.0442994972142954,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.347312814241066e-05,
+      "loss": 1.1079,
+      "step": 15370
+    },
+    {
+      "epoch": 1.0446392172849572,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.3471004891969016e-05,
+      "loss": 1.0908,
+      "step": 15375
+    },
+    {
+      "epoch": 1.0449789373556189,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.346888164152738e-05,
+      "loss": 1.0851,
+      "step": 15380
+    },
+    {
+      "epoch": 1.0453186574262807,
+      "grad_norm": 5.375,
+      "learning_rate": 4.346675839108575e-05,
+      "loss": 1.0485,
+      "step": 15385
+    },
+    {
+      "epoch": 1.0456583774969426,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.346463514064411e-05,
+      "loss": 1.1026,
+      "step": 15390
+    },
+    {
+      "epoch": 1.0459980975676042,
+      "grad_norm": 5.625,
+      "learning_rate": 4.346251189020247e-05,
+      "loss": 1.0686,
+      "step": 15395
+    },
+    {
+      "epoch": 1.046337817638266,
+      "grad_norm": 6.125,
+      "learning_rate": 4.346038863976084e-05,
+      "loss": 1.0882,
+      "step": 15400
+    },
+    {
+      "epoch": 1.046677537708928,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.34582653893192e-05,
+      "loss": 1.0103,
+      "step": 15405
+    },
+    {
+      "epoch": 1.0470172577795895,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.3456142138877564e-05,
+      "loss": 1.1118,
+      "step": 15410
+    },
+    {
+      "epoch": 1.0473569778502514,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.3454018888435935e-05,
+      "loss": 1.0198,
+      "step": 15415
+    },
+    {
+      "epoch": 1.0476966979209132,
+      "grad_norm": 7.28125,
+      "learning_rate": 4.345189563799429e-05,
+      "loss": 1.0701,
+      "step": 15420
+    },
+    {
+      "epoch": 1.0480364179915749,
+      "grad_norm": 5.75,
+      "learning_rate": 4.3449772387552656e-05,
+      "loss": 1.0667,
+      "step": 15425
+    },
+    {
+      "epoch": 1.0483761380622367,
+      "grad_norm": 5.75,
+      "learning_rate": 4.344764913711103e-05,
+      "loss": 1.1133,
+      "step": 15430
+    },
+    {
+      "epoch": 1.0487158581328986,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.3445525886669384e-05,
+      "loss": 1.0323,
+      "step": 15435
+    },
+    {
+      "epoch": 1.0490555782035602,
+      "grad_norm": 5.875,
+      "learning_rate": 4.344340263622775e-05,
+      "loss": 1.1263,
+      "step": 15440
+    },
+    {
+      "epoch": 1.049395298274222,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.344127938578612e-05,
+      "loss": 1.155,
+      "step": 15445
+    },
+    {
+      "epoch": 1.049735018344884,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.3439156135344476e-05,
+      "loss": 1.073,
+      "step": 15450
+    },
+    {
+      "epoch": 1.0500747384155455,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.343703288490284e-05,
+      "loss": 1.1137,
+      "step": 15455
+    },
+    {
+      "epoch": 1.0504144584862074,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3434909634461204e-05,
+      "loss": 1.0073,
+      "step": 15460
+    },
+    {
+      "epoch": 1.050754178556869,
+      "grad_norm": 5.5,
+      "learning_rate": 4.343278638401957e-05,
+      "loss": 1.0536,
+      "step": 15465
+    },
+    {
+      "epoch": 1.0510938986275309,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.343066313357793e-05,
+      "loss": 1.0837,
+      "step": 15470
+    },
+    {
+      "epoch": 1.0514336186981927,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.3428539883136296e-05,
+      "loss": 1.048,
+      "step": 15475
+    },
+    {
+      "epoch": 1.0517733387688544,
+      "grad_norm": 5.5,
+      "learning_rate": 4.342641663269466e-05,
+      "loss": 1.1072,
+      "step": 15480
+    },
+    {
+      "epoch": 1.0521130588395162,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.3424293382253024e-05,
+      "loss": 1.1468,
+      "step": 15485
+    },
+    {
+      "epoch": 1.052452778910178,
+      "grad_norm": 6.125,
+      "learning_rate": 4.342217013181139e-05,
+      "loss": 1.1762,
+      "step": 15490
+    },
+    {
+      "epoch": 1.0527924989808397,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.342004688136975e-05,
+      "loss": 0.9906,
+      "step": 15495
+    },
+    {
+      "epoch": 1.0531322190515016,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.3417923630928116e-05,
+      "loss": 1.0011,
+      "step": 15500
+    },
+    {
+      "epoch": 1.0534719391221634,
+      "grad_norm": 6.25,
+      "learning_rate": 4.341580038048648e-05,
+      "loss": 1.1283,
+      "step": 15505
+    },
+    {
+      "epoch": 1.053811659192825,
+      "grad_norm": 5.375,
+      "learning_rate": 4.3413677130044844e-05,
+      "loss": 1.1715,
+      "step": 15510
+    },
+    {
+      "epoch": 1.054151379263487,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3411553879603215e-05,
+      "loss": 0.9453,
+      "step": 15515
+    },
+    {
+      "epoch": 1.0544910993341488,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.340943062916157e-05,
+      "loss": 1.103,
+      "step": 15520
+    },
+    {
+      "epoch": 1.0548308194048104,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.3407307378719936e-05,
+      "loss": 1.0871,
+      "step": 15525
+    },
+    {
+      "epoch": 1.0551705394754722,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.34051841282783e-05,
+      "loss": 1.1215,
+      "step": 15530
+    },
+    {
+      "epoch": 1.055510259546134,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3403060877836664e-05,
+      "loss": 1.0171,
+      "step": 15535
+    },
+    {
+      "epoch": 1.0558499796167957,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.340093762739503e-05,
+      "loss": 1.0496,
+      "step": 15540
+    },
+    {
+      "epoch": 1.0561896996874576,
+      "grad_norm": 5.75,
+      "learning_rate": 4.339881437695339e-05,
+      "loss": 1.0907,
+      "step": 15545
+    },
+    {
+      "epoch": 1.0565294197581192,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.3396691126511756e-05,
+      "loss": 1.0981,
+      "step": 15550
+    },
+    {
+      "epoch": 1.056869139828781,
+      "grad_norm": 6.25,
+      "learning_rate": 4.339456787607012e-05,
+      "loss": 1.06,
+      "step": 15555
+    },
+    {
+      "epoch": 1.057208859899443,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.3392444625628484e-05,
+      "loss": 1.0644,
+      "step": 15560
+    },
+    {
+      "epoch": 1.0575485799701045,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.339032137518685e-05,
+      "loss": 1.1489,
+      "step": 15565
+    },
+    {
+      "epoch": 1.0578883000407664,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.338819812474521e-05,
+      "loss": 1.0082,
+      "step": 15570
+    },
+    {
+      "epoch": 1.0582280201114282,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.3386074874303576e-05,
+      "loss": 0.945,
+      "step": 15575
+    },
+    {
+      "epoch": 1.0585677401820899,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.338395162386194e-05,
+      "loss": 1.0499,
+      "step": 15580
+    },
+    {
+      "epoch": 1.0589074602527517,
+      "grad_norm": 5.75,
+      "learning_rate": 4.3381828373420304e-05,
+      "loss": 1.1149,
+      "step": 15585
+    },
+    {
+      "epoch": 1.0592471803234136,
+      "grad_norm": 5.625,
+      "learning_rate": 4.337970512297867e-05,
+      "loss": 1.0349,
+      "step": 15590
+    },
+    {
+      "epoch": 1.0595869003940752,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.337758187253703e-05,
+      "loss": 1.0907,
+      "step": 15595
+    },
+    {
+      "epoch": 1.059926620464737,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.3375458622095396e-05,
+      "loss": 1.1337,
+      "step": 15600
+    },
+    {
+      "epoch": 1.060266340535399,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.337333537165376e-05,
+      "loss": 1.0166,
+      "step": 15605
+    },
+    {
+      "epoch": 1.0606060606060606,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.3371212121212124e-05,
+      "loss": 1.0248,
+      "step": 15610
+    },
+    {
+      "epoch": 1.0609457806767224,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.336908887077049e-05,
+      "loss": 1.1507,
+      "step": 15615
+    },
+    {
+      "epoch": 1.0612855007473843,
+      "grad_norm": 6.375,
+      "learning_rate": 4.336696562032885e-05,
+      "loss": 1.1676,
+      "step": 15620
+    },
+    {
+      "epoch": 1.061625220818046,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.3364842369887216e-05,
+      "loss": 1.0579,
+      "step": 15625
+    },
+    {
+      "epoch": 1.0619649408887077,
+      "grad_norm": 5.75,
+      "learning_rate": 4.336271911944558e-05,
+      "loss": 1.069,
+      "step": 15630
+    },
+    {
+      "epoch": 1.0623046609593694,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3360595869003944e-05,
+      "loss": 1.1113,
+      "step": 15635
+    },
+    {
+      "epoch": 1.0626443810300312,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.335847261856231e-05,
+      "loss": 0.9171,
+      "step": 15640
+    },
+    {
+      "epoch": 1.062984101100693,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.335634936812067e-05,
+      "loss": 1.0631,
+      "step": 15645
+    },
+    {
+      "epoch": 1.0633238211713547,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.3354226117679036e-05,
+      "loss": 1.0274,
+      "step": 15650
+    },
+    {
+      "epoch": 1.0636635412420166,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.33521028672374e-05,
+      "loss": 1.1426,
+      "step": 15655
+    },
+    {
+      "epoch": 1.0640032613126784,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.334997961679576e-05,
+      "loss": 0.9993,
+      "step": 15660
+    },
+    {
+      "epoch": 1.06434298138334,
+      "grad_norm": 5.625,
+      "learning_rate": 4.334785636635413e-05,
+      "loss": 1.1099,
+      "step": 15665
+    },
+    {
+      "epoch": 1.064682701454002,
+      "grad_norm": 5.875,
+      "learning_rate": 4.334573311591249e-05,
+      "loss": 1.1467,
+      "step": 15670
+    },
+    {
+      "epoch": 1.0650224215246638,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.334360986547085e-05,
+      "loss": 1.0062,
+      "step": 15675
+    },
+    {
+      "epoch": 1.0653621415953254,
+      "grad_norm": 6.25,
+      "learning_rate": 4.334148661502922e-05,
+      "loss": 1.0709,
+      "step": 15680
+    },
+    {
+      "epoch": 1.0657018616659872,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.3339363364587584e-05,
+      "loss": 0.995,
+      "step": 15685
+    },
+    {
+      "epoch": 1.066041581736649,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.333724011414594e-05,
+      "loss": 1.041,
+      "step": 15690
+    },
+    {
+      "epoch": 1.0663813018073107,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.333511686370431e-05,
+      "loss": 0.9772,
+      "step": 15695
+    },
+    {
+      "epoch": 1.0667210218779726,
+      "grad_norm": 5.5,
+      "learning_rate": 4.3332993613262676e-05,
+      "loss": 0.955,
+      "step": 15700
+    },
+    {
+      "epoch": 1.0670607419486344,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.3330870362821034e-05,
+      "loss": 1.0609,
+      "step": 15705
+    },
+    {
+      "epoch": 1.067400462019296,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3328747112379404e-05,
+      "loss": 1.069,
+      "step": 15710
+    },
+    {
+      "epoch": 1.067740182089958,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.332662386193777e-05,
+      "loss": 1.077,
+      "step": 15715
+    },
+    {
+      "epoch": 1.0680799021606195,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.3324500611496126e-05,
+      "loss": 1.1053,
+      "step": 15720
+    },
+    {
+      "epoch": 1.0684196222312814,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.3322377361054496e-05,
+      "loss": 1.0784,
+      "step": 15725
+    },
+    {
+      "epoch": 1.0687593423019432,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.3320254110612854e-05,
+      "loss": 1.0195,
+      "step": 15730
+    },
+    {
+      "epoch": 1.0690990623726049,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.331813086017122e-05,
+      "loss": 1.1498,
+      "step": 15735
+    },
+    {
+      "epoch": 1.0694387824432667,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.331600760972959e-05,
+      "loss": 1.0687,
+      "step": 15740
+    },
+    {
+      "epoch": 1.0697785025139286,
+      "grad_norm": 5.875,
+      "learning_rate": 4.3313884359287946e-05,
+      "loss": 1.0205,
+      "step": 15745
+    },
+    {
+      "epoch": 1.0701182225845902,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.331176110884631e-05,
+      "loss": 1.0934,
+      "step": 15750
+    },
+    {
+      "epoch": 1.070457942655252,
+      "grad_norm": 5.875,
+      "learning_rate": 4.330963785840468e-05,
+      "loss": 1.2525,
+      "step": 15755
+    },
+    {
+      "epoch": 1.070797662725914,
+      "grad_norm": 5.375,
+      "learning_rate": 4.330751460796304e-05,
+      "loss": 1.0104,
+      "step": 15760
+    },
+    {
+      "epoch": 1.0711373827965756,
+      "grad_norm": 6.0,
+      "learning_rate": 4.33053913575214e-05,
+      "loss": 0.9806,
+      "step": 15765
+    },
+    {
+      "epoch": 1.0714771028672374,
+      "grad_norm": 4.875,
+      "learning_rate": 4.330326810707977e-05,
+      "loss": 0.9687,
+      "step": 15770
+    },
+    {
+      "epoch": 1.0718168229378993,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.330114485663813e-05,
+      "loss": 1.0886,
+      "step": 15775
+    },
+    {
+      "epoch": 1.072156543008561,
+      "grad_norm": 6.5,
+      "learning_rate": 4.3299021606196494e-05,
+      "loss": 1.1797,
+      "step": 15780
+    },
+    {
+      "epoch": 1.0724962630792227,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.3296898355754864e-05,
+      "loss": 1.0492,
+      "step": 15785
+    },
+    {
+      "epoch": 1.0728359831498846,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.329477510531322e-05,
+      "loss": 1.0666,
+      "step": 15790
+    },
+    {
+      "epoch": 1.0731757032205462,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.3292651854871586e-05,
+      "loss": 1.1107,
+      "step": 15795
+    },
+    {
+      "epoch": 1.073515423291208,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.329052860442995e-05,
+      "loss": 1.0436,
+      "step": 15800
+    },
+    {
+      "epoch": 1.07385514336187,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.3288405353988314e-05,
+      "loss": 1.1902,
+      "step": 15805
+    },
+    {
+      "epoch": 1.0741948634325316,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.328628210354668e-05,
+      "loss": 1.0783,
+      "step": 15810
+    },
+    {
+      "epoch": 1.0745345835031934,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.328415885310504e-05,
+      "loss": 1.0938,
+      "step": 15815
+    },
+    {
+      "epoch": 1.074874303573855,
+      "grad_norm": 6.0,
+      "learning_rate": 4.3282035602663406e-05,
+      "loss": 1.0507,
+      "step": 15820
+    },
+    {
+      "epoch": 1.075214023644517,
+      "grad_norm": 5.625,
+      "learning_rate": 4.327991235222177e-05,
+      "loss": 1.0352,
+      "step": 15825
+    },
+    {
+      "epoch": 1.0755537437151788,
+      "grad_norm": 4.53125,
+      "learning_rate": 4.3277789101780134e-05,
+      "loss": 1.0509,
+      "step": 15830
+    },
+    {
+      "epoch": 1.0758934637858404,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.32756658513385e-05,
+      "loss": 1.1408,
+      "step": 15835
+    },
+    {
+      "epoch": 1.0762331838565022,
+      "grad_norm": 5.875,
+      "learning_rate": 4.327354260089686e-05,
+      "loss": 1.0414,
+      "step": 15840
+    },
+    {
+      "epoch": 1.076572903927164,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3271419350455226e-05,
+      "loss": 0.9936,
+      "step": 15845
+    },
+    {
+      "epoch": 1.0769126239978257,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.326929610001359e-05,
+      "loss": 1.1018,
+      "step": 15850
+    },
+    {
+      "epoch": 1.0772523440684876,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.326717284957196e-05,
+      "loss": 1.0738,
+      "step": 15855
+    },
+    {
+      "epoch": 1.0775920641391494,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.326504959913032e-05,
+      "loss": 1.3743,
+      "step": 15860
+    },
+    {
+      "epoch": 1.077931784209811,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.326292634868868e-05,
+      "loss": 1.0602,
+      "step": 15865
+    },
+    {
+      "epoch": 1.078271504280473,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.326080309824705e-05,
+      "loss": 1.0845,
+      "step": 15870
+    },
+    {
+      "epoch": 1.0786112243511348,
+      "grad_norm": 5.75,
+      "learning_rate": 4.325867984780541e-05,
+      "loss": 1.0662,
+      "step": 15875
+    },
+    {
+      "epoch": 1.0789509444217964,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3256556597363774e-05,
+      "loss": 1.0749,
+      "step": 15880
+    },
+    {
+      "epoch": 1.0792906644924583,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.325443334692214e-05,
+      "loss": 1.1688,
+      "step": 15885
+    },
+    {
+      "epoch": 1.0796303845631199,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.32523100964805e-05,
+      "loss": 1.1148,
+      "step": 15890
+    },
+    {
+      "epoch": 1.0799701046337817,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.3250186846038866e-05,
+      "loss": 1.141,
+      "step": 15895
+    },
+    {
+      "epoch": 1.0803098247044436,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.324806359559723e-05,
+      "loss": 0.9438,
+      "step": 15900
+    },
+    {
+      "epoch": 1.0806495447751052,
+      "grad_norm": 5.75,
+      "learning_rate": 4.3245940345155594e-05,
+      "loss": 1.1055,
+      "step": 15905
+    },
+    {
+      "epoch": 1.080989264845767,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.324381709471396e-05,
+      "loss": 1.1885,
+      "step": 15910
+    },
+    {
+      "epoch": 1.081328984916429,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.324169384427232e-05,
+      "loss": 1.0298,
+      "step": 15915
+    },
+    {
+      "epoch": 1.0816687049870906,
+      "grad_norm": 6.375,
+      "learning_rate": 4.3239570593830686e-05,
+      "loss": 1.1562,
+      "step": 15920
+    },
+    {
+      "epoch": 1.0820084250577524,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.323744734338905e-05,
+      "loss": 1.0149,
+      "step": 15925
+    },
+    {
+      "epoch": 1.0823481451284143,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.3235324092947414e-05,
+      "loss": 1.0796,
+      "step": 15930
+    },
+    {
+      "epoch": 1.082687865199076,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.323320084250578e-05,
+      "loss": 1.0815,
+      "step": 15935
+    },
+    {
+      "epoch": 1.0830275852697377,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.323107759206414e-05,
+      "loss": 1.0361,
+      "step": 15940
+    },
+    {
+      "epoch": 1.0833673053403996,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3228954341622506e-05,
+      "loss": 1.0719,
+      "step": 15945
+    },
+    {
+      "epoch": 1.0837070254110612,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.322683109118087e-05,
+      "loss": 1.1098,
+      "step": 15950
+    },
+    {
+      "epoch": 1.084046745481723,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.3224707840739234e-05,
+      "loss": 1.043,
+      "step": 15955
+    },
+    {
+      "epoch": 1.084386465552385,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.32225845902976e-05,
+      "loss": 1.0297,
+      "step": 15960
+    },
+    {
+      "epoch": 1.0847261856230466,
+      "grad_norm": 6.375,
+      "learning_rate": 4.322046133985596e-05,
+      "loss": 1.0849,
+      "step": 15965
+    },
+    {
+      "epoch": 1.0850659056937084,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3218338089414326e-05,
+      "loss": 1.0146,
+      "step": 15970
+    },
+    {
+      "epoch": 1.0854056257643703,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.321621483897269e-05,
+      "loss": 1.0524,
+      "step": 15975
+    },
+    {
+      "epoch": 1.085745345835032,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.3214091588531054e-05,
+      "loss": 1.1207,
+      "step": 15980
+    },
+    {
+      "epoch": 1.0860850659056938,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.321196833808942e-05,
+      "loss": 1.0772,
+      "step": 15985
+    },
+    {
+      "epoch": 1.0864247859763554,
+      "grad_norm": 5.375,
+      "learning_rate": 4.320984508764778e-05,
+      "loss": 1.1092,
+      "step": 15990
+    },
+    {
+      "epoch": 1.0867645060470172,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3207721837206146e-05,
+      "loss": 1.1598,
+      "step": 15995
+    },
+    {
+      "epoch": 1.087104226117679,
+      "grad_norm": 5.375,
+      "learning_rate": 4.32055985867645e-05,
+      "loss": 1.0623,
+      "step": 16000
+    },
+    {
+      "epoch": 1.0874439461883407,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.3203475336322874e-05,
+      "loss": 1.0617,
+      "step": 16005
+    },
+    {
+      "epoch": 1.0877836662590026,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.320135208588124e-05,
+      "loss": 1.0048,
+      "step": 16010
+    },
+    {
+      "epoch": 1.0881233863296644,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.3199228835439595e-05,
+      "loss": 1.0717,
+      "step": 16015
+    },
+    {
+      "epoch": 1.088463106400326,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.3197105584997966e-05,
+      "loss": 1.0395,
+      "step": 16020
+    },
+    {
+      "epoch": 1.088802826470988,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.319498233455633e-05,
+      "loss": 1.0795,
+      "step": 16025
+    },
+    {
+      "epoch": 1.0891425465416498,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.319285908411469e-05,
+      "loss": 1.1311,
+      "step": 16030
+    },
+    {
+      "epoch": 1.0894822666123114,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.319073583367306e-05,
+      "loss": 1.0788,
+      "step": 16035
+    },
+    {
+      "epoch": 1.0898219866829733,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.318861258323142e-05,
+      "loss": 0.992,
+      "step": 16040
+    },
+    {
+      "epoch": 1.090161706753635,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.318648933278978e-05,
+      "loss": 1.0288,
+      "step": 16045
+    },
+    {
+      "epoch": 1.0905014268242967,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.318436608234815e-05,
+      "loss": 1.1326,
+      "step": 16050
+    },
+    {
+      "epoch": 1.0908411468949586,
+      "grad_norm": 6.5,
+      "learning_rate": 4.3182242831906514e-05,
+      "loss": 1.014,
+      "step": 16055
+    },
+    {
+      "epoch": 1.0911808669656202,
+      "grad_norm": 6.5,
+      "learning_rate": 4.318011958146487e-05,
+      "loss": 1.0673,
+      "step": 16060
+    },
+    {
+      "epoch": 1.091520587036282,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.317799633102324e-05,
+      "loss": 1.0886,
+      "step": 16065
+    },
+    {
+      "epoch": 1.091860307106944,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.3175873080581606e-05,
+      "loss": 1.0691,
+      "step": 16070
+    },
+    {
+      "epoch": 1.0922000271776056,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.317374983013996e-05,
+      "loss": 1.0633,
+      "step": 16075
+    },
+    {
+      "epoch": 1.0925397472482674,
+      "grad_norm": 6.25,
+      "learning_rate": 4.3171626579698334e-05,
+      "loss": 1.0715,
+      "step": 16080
+    },
+    {
+      "epoch": 1.0928794673189293,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.316950332925669e-05,
+      "loss": 1.3461,
+      "step": 16085
+    },
+    {
+      "epoch": 1.093219187389591,
+      "grad_norm": 5.75,
+      "learning_rate": 4.3167380078815055e-05,
+      "loss": 1.0611,
+      "step": 16090
+    },
+    {
+      "epoch": 1.0935589074602527,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.3165256828373426e-05,
+      "loss": 1.1093,
+      "step": 16095
+    },
+    {
+      "epoch": 1.0938986275309146,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.316313357793178e-05,
+      "loss": 1.138,
+      "step": 16100
+    },
+    {
+      "epoch": 1.0942383476015762,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.316101032749015e-05,
+      "loss": 1.0543,
+      "step": 16105
+    },
+    {
+      "epoch": 1.094578067672238,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.315888707704852e-05,
+      "loss": 1.0226,
+      "step": 16110
+    },
+    {
+      "epoch": 1.0949177877429,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.3156763826606875e-05,
+      "loss": 1.0819,
+      "step": 16115
+    },
+    {
+      "epoch": 1.0952575078135616,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.315464057616524e-05,
+      "loss": 1.0138,
+      "step": 16120
+    },
+    {
+      "epoch": 1.0955972278842234,
+      "grad_norm": 5.625,
+      "learning_rate": 4.315251732572361e-05,
+      "loss": 1.0524,
+      "step": 16125
+    },
+    {
+      "epoch": 1.0959369479548853,
+      "grad_norm": 5.625,
+      "learning_rate": 4.315039407528197e-05,
+      "loss": 1.0911,
+      "step": 16130
+    },
+    {
+      "epoch": 1.096276668025547,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.314827082484033e-05,
+      "loss": 1.1086,
+      "step": 16135
+    },
+    {
+      "epoch": 1.0966163880962088,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.31461475743987e-05,
+      "loss": 1.123,
+      "step": 16140
+    },
+    {
+      "epoch": 1.0969561081668706,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.314402432395706e-05,
+      "loss": 1.0112,
+      "step": 16145
+    },
+    {
+      "epoch": 1.0972958282375322,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.314190107351542e-05,
+      "loss": 1.0807,
+      "step": 16150
+    },
+    {
+      "epoch": 1.097635548308194,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.313977782307379e-05,
+      "loss": 1.075,
+      "step": 16155
+    },
+    {
+      "epoch": 1.0979752683788557,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.313765457263215e-05,
+      "loss": 1.0078,
+      "step": 16160
+    },
+    {
+      "epoch": 1.0983149884495176,
+      "grad_norm": 5.875,
+      "learning_rate": 4.3135531322190515e-05,
+      "loss": 1.026,
+      "step": 16165
+    },
+    {
+      "epoch": 1.0986547085201794,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.313340807174888e-05,
+      "loss": 1.1312,
+      "step": 16170
+    },
+    {
+      "epoch": 1.098994428590841,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.313128482130724e-05,
+      "loss": 1.1107,
+      "step": 16175
+    },
+    {
+      "epoch": 1.099334148661503,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.312916157086561e-05,
+      "loss": 1.0509,
+      "step": 16180
+    },
+    {
+      "epoch": 1.0996738687321648,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.312703832042397e-05,
+      "loss": 1.0618,
+      "step": 16185
+    },
+    {
+      "epoch": 1.1000135888028264,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.3124915069982335e-05,
+      "loss": 1.082,
+      "step": 16190
+    },
+    {
+      "epoch": 1.1003533088734883,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.3122791819540706e-05,
+      "loss": 1.0405,
+      "step": 16195
+    },
+    {
+      "epoch": 1.10069302894415,
+      "grad_norm": 6.25,
+      "learning_rate": 4.312066856909906e-05,
+      "loss": 1.1004,
+      "step": 16200
+    },
+    {
+      "epoch": 1.1010327490148117,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.311854531865743e-05,
+      "loss": 1.0712,
+      "step": 16205
+    },
+    {
+      "epoch": 1.1013724690854736,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.31164220682158e-05,
+      "loss": 0.9996,
+      "step": 16210
+    },
+    {
+      "epoch": 1.1017121891561354,
+      "grad_norm": 6.0,
+      "learning_rate": 4.3114298817774155e-05,
+      "loss": 0.9863,
+      "step": 16215
+    },
+    {
+      "epoch": 1.102051909226797,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.311217556733252e-05,
+      "loss": 1.1462,
+      "step": 16220
+    },
+    {
+      "epoch": 1.102391629297459,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.311005231689088e-05,
+      "loss": 1.137,
+      "step": 16225
+    },
+    {
+      "epoch": 1.1027313493681206,
+      "grad_norm": 4.71875,
+      "learning_rate": 4.310792906644925e-05,
+      "loss": 1.0238,
+      "step": 16230
+    },
+    {
+      "epoch": 1.1030710694387824,
+      "grad_norm": 6.375,
+      "learning_rate": 4.310580581600761e-05,
+      "loss": 1.1448,
+      "step": 16235
+    },
+    {
+      "epoch": 1.1034107895094443,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.3103682565565975e-05,
+      "loss": 1.0881,
+      "step": 16240
+    },
+    {
+      "epoch": 1.103750509580106,
+      "grad_norm": 5.375,
+      "learning_rate": 4.310155931512434e-05,
+      "loss": 1.1404,
+      "step": 16245
+    },
+    {
+      "epoch": 1.1040902296507678,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.30994360646827e-05,
+      "loss": 1.1161,
+      "step": 16250
+    },
+    {
+      "epoch": 1.1044299497214296,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.309731281424107e-05,
+      "loss": 1.1048,
+      "step": 16255
+    },
+    {
+      "epoch": 1.1047696697920912,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.309518956379943e-05,
+      "loss": 1.0785,
+      "step": 16260
+    },
+    {
+      "epoch": 1.105109389862753,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.3093066313357795e-05,
+      "loss": 1.1207,
+      "step": 16265
+    },
+    {
+      "epoch": 1.105449109933415,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.309094306291616e-05,
+      "loss": 1.0583,
+      "step": 16270
+    },
+    {
+      "epoch": 1.1057888300040766,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.308881981247452e-05,
+      "loss": 1.046,
+      "step": 16275
+    },
+    {
+      "epoch": 1.1061285500747384,
+      "grad_norm": 5.875,
+      "learning_rate": 4.308669656203289e-05,
+      "loss": 0.9996,
+      "step": 16280
+    },
+    {
+      "epoch": 1.1064682701454003,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.308457331159125e-05,
+      "loss": 1.1326,
+      "step": 16285
+    },
+    {
+      "epoch": 1.106807990216062,
+      "grad_norm": 6.125,
+      "learning_rate": 4.3082450061149615e-05,
+      "loss": 1.1101,
+      "step": 16290
+    },
+    {
+      "epoch": 1.1071477102867238,
+      "grad_norm": 6.25,
+      "learning_rate": 4.308032681070798e-05,
+      "loss": 1.046,
+      "step": 16295
+    },
+    {
+      "epoch": 1.1074874303573856,
+      "grad_norm": 4.875,
+      "learning_rate": 4.307820356026634e-05,
+      "loss": 1.0454,
+      "step": 16300
+    },
+    {
+      "epoch": 1.1078271504280472,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.307608030982471e-05,
+      "loss": 1.0997,
+      "step": 16305
+    },
+    {
+      "epoch": 1.108166870498709,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.307395705938307e-05,
+      "loss": 1.1872,
+      "step": 16310
+    },
+    {
+      "epoch": 1.108506590569371,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.3071833808941435e-05,
+      "loss": 1.0793,
+      "step": 16315
+    },
+    {
+      "epoch": 1.1088463106400326,
+      "grad_norm": 6.125,
+      "learning_rate": 4.30697105584998e-05,
+      "loss": 1.0976,
+      "step": 16320
+    },
+    {
+      "epoch": 1.1091860307106944,
+      "grad_norm": 5.5,
+      "learning_rate": 4.306758730805816e-05,
+      "loss": 1.0755,
+      "step": 16325
+    },
+    {
+      "epoch": 1.109525750781356,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.306546405761653e-05,
+      "loss": 1.0443,
+      "step": 16330
+    },
+    {
+      "epoch": 1.109865470852018,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.306334080717489e-05,
+      "loss": 1.0969,
+      "step": 16335
+    },
+    {
+      "epoch": 1.1102051909226798,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.3061217556733255e-05,
+      "loss": 1.0096,
+      "step": 16340
+    },
+    {
+      "epoch": 1.1105449109933414,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.305909430629162e-05,
+      "loss": 1.1137,
+      "step": 16345
+    },
+    {
+      "epoch": 1.1108846310640033,
+      "grad_norm": 5.875,
+      "learning_rate": 4.305697105584998e-05,
+      "loss": 1.1597,
+      "step": 16350
+    },
+    {
+      "epoch": 1.1112243511346651,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.305484780540834e-05,
+      "loss": 1.1256,
+      "step": 16355
+    },
+    {
+      "epoch": 1.1115640712053267,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.305272455496671e-05,
+      "loss": 1.0573,
+      "step": 16360
+    },
+    {
+      "epoch": 1.1119037912759886,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.3050601304525075e-05,
+      "loss": 1.097,
+      "step": 16365
+    },
+    {
+      "epoch": 1.1122435113466504,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.304847805408343e-05,
+      "loss": 1.0655,
+      "step": 16370
+    },
+    {
+      "epoch": 1.112583231417312,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.30463548036418e-05,
+      "loss": 1.1167,
+      "step": 16375
+    },
+    {
+      "epoch": 1.112922951487974,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.304423155320017e-05,
+      "loss": 1.1453,
+      "step": 16380
+    },
+    {
+      "epoch": 1.1132626715586358,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.3042108302758525e-05,
+      "loss": 1.1924,
+      "step": 16385
+    },
+    {
+      "epoch": 1.1136023916292974,
+      "grad_norm": 5.625,
+      "learning_rate": 4.3039985052316895e-05,
+      "loss": 1.0443,
+      "step": 16390
+    },
+    {
+      "epoch": 1.1139421116999593,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.303786180187526e-05,
+      "loss": 1.0033,
+      "step": 16395
+    },
+    {
+      "epoch": 1.114281831770621,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.3035738551433617e-05,
+      "loss": 1.049,
+      "step": 16400
+    },
+    {
+      "epoch": 1.1146215518412828,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.303361530099199e-05,
+      "loss": 1.0611,
+      "step": 16405
+    },
+    {
+      "epoch": 1.1149612719119446,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.303149205055035e-05,
+      "loss": 1.0544,
+      "step": 16410
+    },
+    {
+      "epoch": 1.1153009919826062,
+      "grad_norm": 5.625,
+      "learning_rate": 4.302936880010871e-05,
+      "loss": 1.0964,
+      "step": 16415
+    },
+    {
+      "epoch": 1.115640712053268,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.302724554966708e-05,
+      "loss": 1.1436,
+      "step": 16420
+    },
+    {
+      "epoch": 1.11598043212393,
+      "grad_norm": 5.375,
+      "learning_rate": 4.3025122299225437e-05,
+      "loss": 1.0553,
+      "step": 16425
+    },
+    {
+      "epoch": 1.1163201521945916,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.30229990487838e-05,
+      "loss": 1.0648,
+      "step": 16430
+    },
+    {
+      "epoch": 1.1166598722652534,
+      "grad_norm": 6.625,
+      "learning_rate": 4.302087579834217e-05,
+      "loss": 1.0321,
+      "step": 16435
+    },
+    {
+      "epoch": 1.1169995923359153,
+      "grad_norm": 6.5,
+      "learning_rate": 4.301875254790053e-05,
+      "loss": 1.1171,
+      "step": 16440
+    },
+    {
+      "epoch": 1.117339312406577,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.301662929745889e-05,
+      "loss": 1.0511,
+      "step": 16445
+    },
+    {
+      "epoch": 1.1176790324772388,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.301450604701726e-05,
+      "loss": 1.1738,
+      "step": 16450
+    },
+    {
+      "epoch": 1.1180187525479006,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.301238279657562e-05,
+      "loss": 1.0532,
+      "step": 16455
+    },
+    {
+      "epoch": 1.1183584726185622,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.3010259546133985e-05,
+      "loss": 1.1234,
+      "step": 16460
+    },
+    {
+      "epoch": 1.118698192689224,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.3008136295692355e-05,
+      "loss": 1.0855,
+      "step": 16465
+    },
+    {
+      "epoch": 1.119037912759886,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.300601304525071e-05,
+      "loss": 1.0154,
+      "step": 16470
+    },
+    {
+      "epoch": 1.1193776328305476,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.300388979480908e-05,
+      "loss": 1.1031,
+      "step": 16475
+    },
+    {
+      "epoch": 1.1197173529012094,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.300176654436745e-05,
+      "loss": 1.0579,
+      "step": 16480
+    },
+    {
+      "epoch": 1.1200570729718713,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.2999643293925805e-05,
+      "loss": 1.0665,
+      "step": 16485
+    },
+    {
+      "epoch": 1.120396793042533,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.299752004348417e-05,
+      "loss": 1.0765,
+      "step": 16490
+    },
+    {
+      "epoch": 1.1207365131131948,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.299539679304254e-05,
+      "loss": 1.1352,
+      "step": 16495
+    },
+    {
+      "epoch": 1.1210762331838564,
+      "grad_norm": 5.5,
+      "learning_rate": 4.29932735426009e-05,
+      "loss": 1.1427,
+      "step": 16500
+    },
+    {
+      "epoch": 1.1214159532545183,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.299115029215926e-05,
+      "loss": 1.0211,
+      "step": 16505
+    },
+    {
+      "epoch": 1.1217556733251801,
+      "grad_norm": 5.25,
+      "learning_rate": 4.2989027041717625e-05,
+      "loss": 1.0629,
+      "step": 16510
+    },
+    {
+      "epoch": 1.1220953933958417,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.298690379127599e-05,
+      "loss": 1.0697,
+      "step": 16515
+    },
+    {
+      "epoch": 1.1224351134665036,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.298478054083435e-05,
+      "loss": 1.0686,
+      "step": 16520
+    },
+    {
+      "epoch": 1.1227748335371655,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.298265729039272e-05,
+      "loss": 1.1032,
+      "step": 16525
+    },
+    {
+      "epoch": 1.123114553607827,
+      "grad_norm": 6.125,
+      "learning_rate": 4.298053403995108e-05,
+      "loss": 1.0514,
+      "step": 16530
+    },
+    {
+      "epoch": 1.123454273678489,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.297841078950945e-05,
+      "loss": 1.089,
+      "step": 16535
+    },
+    {
+      "epoch": 1.1237939937491508,
+      "grad_norm": 6.375,
+      "learning_rate": 4.297628753906781e-05,
+      "loss": 1.1378,
+      "step": 16540
+    },
+    {
+      "epoch": 1.1241337138198124,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.297416428862617e-05,
+      "loss": 1.2398,
+      "step": 16545
+    },
+    {
+      "epoch": 1.1244734338904743,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2972041038184543e-05,
+      "loss": 1.0917,
+      "step": 16550
+    },
+    {
+      "epoch": 1.1248131539611361,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.29699177877429e-05,
+      "loss": 1.1306,
+      "step": 16555
+    },
+    {
+      "epoch": 1.1251528740317978,
+      "grad_norm": 5.75,
+      "learning_rate": 4.2967794537301265e-05,
+      "loss": 1.0715,
+      "step": 16560
+    },
+    {
+      "epoch": 1.1254925941024596,
+      "grad_norm": 5.03125,
+      "learning_rate": 4.2965671286859635e-05,
+      "loss": 1.0405,
+      "step": 16565
+    },
+    {
+      "epoch": 1.1258323141731212,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.296354803641799e-05,
+      "loss": 1.0882,
+      "step": 16570
+    },
+    {
+      "epoch": 1.126172034243783,
+      "grad_norm": 5.625,
+      "learning_rate": 4.296142478597636e-05,
+      "loss": 1.168,
+      "step": 16575
+    },
+    {
+      "epoch": 1.126511754314445,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.295930153553472e-05,
+      "loss": 1.0176,
+      "step": 16580
+    },
+    {
+      "epoch": 1.1268514743851066,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.2957178285093085e-05,
+      "loss": 1.1081,
+      "step": 16585
+    },
+    {
+      "epoch": 1.1271911944557684,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.295505503465145e-05,
+      "loss": 1.1078,
+      "step": 16590
+    },
+    {
+      "epoch": 1.1275309145264303,
+      "grad_norm": 5.75,
+      "learning_rate": 4.295293178420981e-05,
+      "loss": 1.1251,
+      "step": 16595
+    },
+    {
+      "epoch": 1.127870634597092,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.295080853376818e-05,
+      "loss": 1.171,
+      "step": 16600
+    },
+    {
+      "epoch": 1.1282103546677538,
+      "grad_norm": 5.25,
+      "learning_rate": 4.294868528332654e-05,
+      "loss": 1.0039,
+      "step": 16605
+    },
+    {
+      "epoch": 1.1285500747384156,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.2946562032884905e-05,
+      "loss": 1.0426,
+      "step": 16610
+    },
+    {
+      "epoch": 1.1288897948090773,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.294443878244327e-05,
+      "loss": 1.0812,
+      "step": 16615
+    },
+    {
+      "epoch": 1.129229514879739,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.294231553200163e-05,
+      "loss": 1.1724,
+      "step": 16620
+    },
+    {
+      "epoch": 1.129569234950401,
+      "grad_norm": 5.75,
+      "learning_rate": 4.294019228156e-05,
+      "loss": 1.028,
+      "step": 16625
+    },
+    {
+      "epoch": 1.1299089550210626,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.293806903111836e-05,
+      "loss": 1.0855,
+      "step": 16630
+    },
+    {
+      "epoch": 1.1302486750917244,
+      "grad_norm": 5.5,
+      "learning_rate": 4.2935945780676725e-05,
+      "loss": 1.0907,
+      "step": 16635
+    },
+    {
+      "epoch": 1.1305883951623863,
+      "grad_norm": 5.625,
+      "learning_rate": 4.293382253023509e-05,
+      "loss": 1.0927,
+      "step": 16640
+    },
+    {
+      "epoch": 1.130928115233048,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.293169927979345e-05,
+      "loss": 1.0337,
+      "step": 16645
+    },
+    {
+      "epoch": 1.1312678353037098,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.292957602935182e-05,
+      "loss": 1.0889,
+      "step": 16650
+    },
+    {
+      "epoch": 1.1316075553743716,
+      "grad_norm": 5.875,
+      "learning_rate": 4.292745277891018e-05,
+      "loss": 1.0653,
+      "step": 16655
+    },
+    {
+      "epoch": 1.1319472754450333,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2925329528468545e-05,
+      "loss": 1.0865,
+      "step": 16660
+    },
+    {
+      "epoch": 1.1322869955156951,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.292320627802691e-05,
+      "loss": 1.1214,
+      "step": 16665
+    },
+    {
+      "epoch": 1.1326267155863567,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.292108302758527e-05,
+      "loss": 1.0585,
+      "step": 16670
+    },
+    {
+      "epoch": 1.1329664356570186,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.291895977714364e-05,
+      "loss": 1.1487,
+      "step": 16675
+    },
+    {
+      "epoch": 1.1333061557276805,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.2916836526702e-05,
+      "loss": 1.0854,
+      "step": 16680
+    },
+    {
+      "epoch": 1.133645875798342,
+      "grad_norm": 4.1875,
+      "learning_rate": 4.2914713276260365e-05,
+      "loss": 0.9938,
+      "step": 16685
+    },
+    {
+      "epoch": 1.133985595869004,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.291259002581873e-05,
+      "loss": 1.1422,
+      "step": 16690
+    },
+    {
+      "epoch": 1.1343253159396658,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.291046677537709e-05,
+      "loss": 1.1474,
+      "step": 16695
+    },
+    {
+      "epoch": 1.1346650360103274,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.290834352493546e-05,
+      "loss": 1.1085,
+      "step": 16700
+    },
+    {
+      "epoch": 1.1350047560809893,
+      "grad_norm": 6.0,
+      "learning_rate": 4.290622027449382e-05,
+      "loss": 0.9932,
+      "step": 16705
+    },
+    {
+      "epoch": 1.1353444761516511,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.290409702405218e-05,
+      "loss": 1.0426,
+      "step": 16710
+    },
+    {
+      "epoch": 1.1356841962223128,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.290197377361055e-05,
+      "loss": 1.1158,
+      "step": 16715
+    },
+    {
+      "epoch": 1.1360239162929746,
+      "grad_norm": 6.0,
+      "learning_rate": 4.289985052316891e-05,
+      "loss": 1.1291,
+      "step": 16720
+    },
+    {
+      "epoch": 1.1363636363636362,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.289772727272727e-05,
+      "loss": 1.0215,
+      "step": 16725
+    },
+    {
+      "epoch": 1.136703356434298,
+      "grad_norm": 5.625,
+      "learning_rate": 4.289560402228564e-05,
+      "loss": 1.0574,
+      "step": 16730
+    },
+    {
+      "epoch": 1.13704307650496,
+      "grad_norm": 6.125,
+      "learning_rate": 4.2893480771844005e-05,
+      "loss": 1.0307,
+      "step": 16735
+    },
+    {
+      "epoch": 1.1373827965756216,
+      "grad_norm": 5.875,
+      "learning_rate": 4.289135752140236e-05,
+      "loss": 1.0183,
+      "step": 16740
+    },
+    {
+      "epoch": 1.1377225166462834,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.288923427096073e-05,
+      "loss": 1.178,
+      "step": 16745
+    },
+    {
+      "epoch": 1.1380622367169453,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.28871110205191e-05,
+      "loss": 1.0995,
+      "step": 16750
+    },
+    {
+      "epoch": 1.138401956787607,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.2884987770077454e-05,
+      "loss": 1.0335,
+      "step": 16755
+    },
+    {
+      "epoch": 1.1387416768582688,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.2882864519635825e-05,
+      "loss": 1.0663,
+      "step": 16760
+    },
+    {
+      "epoch": 1.1390813969289306,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.288074126919419e-05,
+      "loss": 1.0609,
+      "step": 16765
+    },
+    {
+      "epoch": 1.1394211169995923,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.2878618018752546e-05,
+      "loss": 1.0916,
+      "step": 16770
+    },
+    {
+      "epoch": 1.139760837070254,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.287649476831092e-05,
+      "loss": 1.0737,
+      "step": 16775
+    },
+    {
+      "epoch": 1.140100557140916,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.2874371517869274e-05,
+      "loss": 0.9839,
+      "step": 16780
+    },
+    {
+      "epoch": 1.1404402772115776,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.287224826742764e-05,
+      "loss": 1.1291,
+      "step": 16785
+    },
+    {
+      "epoch": 1.1407799972822394,
+      "grad_norm": 5.625,
+      "learning_rate": 4.287012501698601e-05,
+      "loss": 1.1526,
+      "step": 16790
+    },
+    {
+      "epoch": 1.1411197173529013,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.2868001766544366e-05,
+      "loss": 0.9761,
+      "step": 16795
+    },
+    {
+      "epoch": 1.141459437423563,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.286587851610273e-05,
+      "loss": 1.0925,
+      "step": 16800
+    },
+    {
+      "epoch": 1.1417991574942248,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.28637552656611e-05,
+      "loss": 1.0695,
+      "step": 16805
+    },
+    {
+      "epoch": 1.1421388775648866,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.286163201521946e-05,
+      "loss": 1.1434,
+      "step": 16810
+    },
+    {
+      "epoch": 1.1424785976355483,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.285950876477782e-05,
+      "loss": 1.0417,
+      "step": 16815
+    },
+    {
+      "epoch": 1.1428183177062101,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.285738551433619e-05,
+      "loss": 1.0767,
+      "step": 16820
+    },
+    {
+      "epoch": 1.143158037776872,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.285526226389455e-05,
+      "loss": 0.9688,
+      "step": 16825
+    },
+    {
+      "epoch": 1.1434977578475336,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.2853139013452914e-05,
+      "loss": 1.0948,
+      "step": 16830
+    },
+    {
+      "epoch": 1.1438374779181955,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2851015763011285e-05,
+      "loss": 1.1676,
+      "step": 16835
+    },
+    {
+      "epoch": 1.144177197988857,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.284889251256964e-05,
+      "loss": 1.103,
+      "step": 16840
+    },
+    {
+      "epoch": 1.144516918059519,
+      "grad_norm": 5.375,
+      "learning_rate": 4.2846769262128006e-05,
+      "loss": 1.0493,
+      "step": 16845
+    },
+    {
+      "epoch": 1.1448566381301808,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.284464601168637e-05,
+      "loss": 1.1362,
+      "step": 16850
+    },
+    {
+      "epoch": 1.1451963582008424,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.2842522761244734e-05,
+      "loss": 1.027,
+      "step": 16855
+    },
+    {
+      "epoch": 1.1455360782715043,
+      "grad_norm": 5.625,
+      "learning_rate": 4.28403995108031e-05,
+      "loss": 1.1299,
+      "step": 16860
+    },
+    {
+      "epoch": 1.1458757983421661,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.283827626036146e-05,
+      "loss": 1.0996,
+      "step": 16865
+    },
+    {
+      "epoch": 1.1462155184128278,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2836153009919826e-05,
+      "loss": 1.0275,
+      "step": 16870
+    },
+    {
+      "epoch": 1.1465552384834896,
+      "grad_norm": 5.625,
+      "learning_rate": 4.28340297594782e-05,
+      "loss": 1.1401,
+      "step": 16875
+    },
+    {
+      "epoch": 1.1468949585541515,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.2831906509036554e-05,
+      "loss": 1.0714,
+      "step": 16880
+    },
+    {
+      "epoch": 1.147234678624813,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.282978325859492e-05,
+      "loss": 1.1299,
+      "step": 16885
+    },
+    {
+      "epoch": 1.147574398695475,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.282766000815329e-05,
+      "loss": 1.0808,
+      "step": 16890
+    },
+    {
+      "epoch": 1.1479141187661366,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2825536757711646e-05,
+      "loss": 1.1151,
+      "step": 16895
+    },
+    {
+      "epoch": 1.1482538388367984,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.282341350727001e-05,
+      "loss": 1.092,
+      "step": 16900
+    },
+    {
+      "epoch": 1.1485935589074603,
+      "grad_norm": 5.375,
+      "learning_rate": 4.282129025682838e-05,
+      "loss": 1.0754,
+      "step": 16905
+    },
+    {
+      "epoch": 1.148933278978122,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.281916700638674e-05,
+      "loss": 1.0111,
+      "step": 16910
+    },
+    {
+      "epoch": 1.1492729990487838,
+      "grad_norm": 4.8125,
+      "learning_rate": 4.28170437559451e-05,
+      "loss": 1.0714,
+      "step": 16915
+    },
+    {
+      "epoch": 1.1496127191194456,
+      "grad_norm": 5.75,
+      "learning_rate": 4.281492050550347e-05,
+      "loss": 1.0351,
+      "step": 16920
+    },
+    {
+      "epoch": 1.1499524391901073,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.281279725506183e-05,
+      "loss": 1.0964,
+      "step": 16925
+    },
+    {
+      "epoch": 1.150292159260769,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2810674004620194e-05,
+      "loss": 1.092,
+      "step": 16930
+    },
+    {
+      "epoch": 1.150631879331431,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.280855075417856e-05,
+      "loss": 0.9947,
+      "step": 16935
+    },
+    {
+      "epoch": 1.1509715994020926,
+      "grad_norm": 6.25,
+      "learning_rate": 4.280642750373692e-05,
+      "loss": 1.1961,
+      "step": 16940
+    },
+    {
+      "epoch": 1.1513113194727544,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.2804304253295286e-05,
+      "loss": 1.1086,
+      "step": 16945
+    },
+    {
+      "epoch": 1.1516510395434163,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.280218100285365e-05,
+      "loss": 1.1818,
+      "step": 16950
+    },
+    {
+      "epoch": 1.151990759614078,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.2800057752412014e-05,
+      "loss": 1.0331,
+      "step": 16955
+    },
+    {
+      "epoch": 1.1523304796847398,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.279793450197038e-05,
+      "loss": 1.069,
+      "step": 16960
+    },
+    {
+      "epoch": 1.1526701997554016,
+      "grad_norm": 5.375,
+      "learning_rate": 4.279581125152874e-05,
+      "loss": 1.0716,
+      "step": 16965
+    },
+    {
+      "epoch": 1.1530099198260633,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2793688001087106e-05,
+      "loss": 1.0465,
+      "step": 16970
+    },
+    {
+      "epoch": 1.1533496398967251,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.279156475064547e-05,
+      "loss": 1.0513,
+      "step": 16975
+    },
+    {
+      "epoch": 1.153689359967387,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.2789441500203834e-05,
+      "loss": 1.0232,
+      "step": 16980
+    },
+    {
+      "epoch": 1.1540290800380486,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.27873182497622e-05,
+      "loss": 1.059,
+      "step": 16985
+    },
+    {
+      "epoch": 1.1543688001087105,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.278519499932056e-05,
+      "loss": 1.0229,
+      "step": 16990
+    },
+    {
+      "epoch": 1.1547085201793723,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2783071748878926e-05,
+      "loss": 1.0745,
+      "step": 16995
+    },
+    {
+      "epoch": 1.155048240250034,
+      "grad_norm": 6.25,
+      "learning_rate": 4.278094849843729e-05,
+      "loss": 1.1176,
+      "step": 17000
+    },
+    {
+      "epoch": 1.1553879603206958,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.2778825247995654e-05,
+      "loss": 1.0296,
+      "step": 17005
+    },
+    {
+      "epoch": 1.1557276803913574,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.277670199755402e-05,
+      "loss": 1.0527,
+      "step": 17010
+    },
+    {
+      "epoch": 1.1560674004620193,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.277457874711238e-05,
+      "loss": 1.1065,
+      "step": 17015
+    },
+    {
+      "epoch": 1.1564071205326811,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.2772455496670746e-05,
+      "loss": 1.1823,
+      "step": 17020
+    },
+    {
+      "epoch": 1.1567468406033428,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.277033224622911e-05,
+      "loss": 1.1978,
+      "step": 17025
+    },
+    {
+      "epoch": 1.1570865606740046,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.2768208995787474e-05,
+      "loss": 1.0382,
+      "step": 17030
+    },
+    {
+      "epoch": 1.1574262807446665,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.276608574534584e-05,
+      "loss": 1.0739,
+      "step": 17035
+    },
+    {
+      "epoch": 1.157766000815328,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.27639624949042e-05,
+      "loss": 0.9326,
+      "step": 17040
+    },
+    {
+      "epoch": 1.15810572088599,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.2761839244462566e-05,
+      "loss": 1.1478,
+      "step": 17045
+    },
+    {
+      "epoch": 1.1584454409566518,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.2759715994020924e-05,
+      "loss": 1.0475,
+      "step": 17050
+    },
+    {
+      "epoch": 1.1587851610273134,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.2757592743579294e-05,
+      "loss": 1.099,
+      "step": 17055
+    },
+    {
+      "epoch": 1.1591248810979753,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.275546949313766e-05,
+      "loss": 1.0672,
+      "step": 17060
+    },
+    {
+      "epoch": 1.159464601168637,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.2753346242696016e-05,
+      "loss": 1.1716,
+      "step": 17065
+    },
+    {
+      "epoch": 1.1598043212392988,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.2751222992254386e-05,
+      "loss": 1.1072,
+      "step": 17070
+    },
+    {
+      "epoch": 1.1601440413099606,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.274909974181275e-05,
+      "loss": 1.1017,
+      "step": 17075
+    },
+    {
+      "epoch": 1.1604837613806223,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.274697649137111e-05,
+      "loss": 1.0234,
+      "step": 17080
+    },
+    {
+      "epoch": 1.1608234814512841,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.274485324092948e-05,
+      "loss": 1.0528,
+      "step": 17085
+    },
+    {
+      "epoch": 1.161163201521946,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.274272999048784e-05,
+      "loss": 1.118,
+      "step": 17090
+    },
+    {
+      "epoch": 1.1615029215926076,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.27406067400462e-05,
+      "loss": 1.0882,
+      "step": 17095
+    },
+    {
+      "epoch": 1.1618426416632694,
+      "grad_norm": 5.5,
+      "learning_rate": 4.273848348960457e-05,
+      "loss": 0.9906,
+      "step": 17100
+    },
+    {
+      "epoch": 1.1621823617339313,
+      "grad_norm": 6.5,
+      "learning_rate": 4.2736360239162934e-05,
+      "loss": 1.1064,
+      "step": 17105
+    },
+    {
+      "epoch": 1.162522081804593,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.273423698872129e-05,
+      "loss": 1.1539,
+      "step": 17110
+    },
+    {
+      "epoch": 1.1628618018752548,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.273211373827966e-05,
+      "loss": 1.0482,
+      "step": 17115
+    },
+    {
+      "epoch": 1.1632015219459166,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.2729990487838026e-05,
+      "loss": 1.1102,
+      "step": 17120
+    },
+    {
+      "epoch": 1.1635412420165783,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2727867237396384e-05,
+      "loss": 1.1024,
+      "step": 17125
+    },
+    {
+      "epoch": 1.1638809620872401,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2725743986954754e-05,
+      "loss": 1.1696,
+      "step": 17130
+    },
+    {
+      "epoch": 1.164220682157902,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.272362073651311e-05,
+      "loss": 0.9943,
+      "step": 17135
+    },
+    {
+      "epoch": 1.1645604022285636,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2721497486071476e-05,
+      "loss": 1.0724,
+      "step": 17140
+    },
+    {
+      "epoch": 1.1649001222992255,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.2719374235629846e-05,
+      "loss": 1.1695,
+      "step": 17145
+    },
+    {
+      "epoch": 1.1652398423698873,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.2717250985188204e-05,
+      "loss": 1.0616,
+      "step": 17150
+    },
+    {
+      "epoch": 1.165579562440549,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.271512773474657e-05,
+      "loss": 1.0263,
+      "step": 17155
+    },
+    {
+      "epoch": 1.1659192825112108,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.271300448430494e-05,
+      "loss": 1.0568,
+      "step": 17160
+    },
+    {
+      "epoch": 1.1662590025818727,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.2710881233863296e-05,
+      "loss": 1.0376,
+      "step": 17165
+    },
+    {
+      "epoch": 1.1665987226525343,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.270875798342166e-05,
+      "loss": 1.0521,
+      "step": 17170
+    },
+    {
+      "epoch": 1.1669384427231961,
+      "grad_norm": 5.625,
+      "learning_rate": 4.270663473298003e-05,
+      "loss": 1.0717,
+      "step": 17175
+    },
+    {
+      "epoch": 1.1672781627938578,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.270451148253839e-05,
+      "loss": 1.1341,
+      "step": 17180
+    },
+    {
+      "epoch": 1.1676178828645196,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.270238823209675e-05,
+      "loss": 1.0399,
+      "step": 17185
+    },
+    {
+      "epoch": 1.1679576029351815,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.270026498165512e-05,
+      "loss": 1.0713,
+      "step": 17190
+    },
+    {
+      "epoch": 1.168297323005843,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.269814173121348e-05,
+      "loss": 1.0473,
+      "step": 17195
+    },
+    {
+      "epoch": 1.168637043076505,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.2696018480771844e-05,
+      "loss": 1.1385,
+      "step": 17200
+    },
+    {
+      "epoch": 1.1689767631471668,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.269389523033021e-05,
+      "loss": 1.0608,
+      "step": 17205
+    },
+    {
+      "epoch": 1.1693164832178284,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.269177197988857e-05,
+      "loss": 0.9888,
+      "step": 17210
+    },
+    {
+      "epoch": 1.1696562032884903,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.268964872944694e-05,
+      "loss": 1.0542,
+      "step": 17215
+    },
+    {
+      "epoch": 1.1699959233591521,
+      "grad_norm": 5.625,
+      "learning_rate": 4.26875254790053e-05,
+      "loss": 1.095,
+      "step": 17220
+    },
+    {
+      "epoch": 1.1703356434298138,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.2685402228563664e-05,
+      "loss": 1.0599,
+      "step": 17225
+    },
+    {
+      "epoch": 1.1706753635004756,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.2683278978122034e-05,
+      "loss": 1.0818,
+      "step": 17230
+    },
+    {
+      "epoch": 1.1710150835711373,
+      "grad_norm": 4.875,
+      "learning_rate": 4.268115572768039e-05,
+      "loss": 0.9651,
+      "step": 17235
+    },
+    {
+      "epoch": 1.1713548036417991,
+      "grad_norm": 5.375,
+      "learning_rate": 4.2679032477238756e-05,
+      "loss": 1.0216,
+      "step": 17240
+    },
+    {
+      "epoch": 1.171694523712461,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.2676909226797126e-05,
+      "loss": 1.0325,
+      "step": 17245
+    },
+    {
+      "epoch": 1.1720342437831226,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2674785976355484e-05,
+      "loss": 0.9881,
+      "step": 17250
+    },
+    {
+      "epoch": 1.1723739638537845,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.267266272591385e-05,
+      "loss": 1.1088,
+      "step": 17255
+    },
+    {
+      "epoch": 1.1727136839244463,
+      "grad_norm": 6.25,
+      "learning_rate": 4.267053947547222e-05,
+      "loss": 1.0971,
+      "step": 17260
+    },
+    {
+      "epoch": 1.173053403995108,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.2668416225030576e-05,
+      "loss": 1.091,
+      "step": 17265
+    },
+    {
+      "epoch": 1.1733931240657698,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.266629297458894e-05,
+      "loss": 1.0719,
+      "step": 17270
+    },
+    {
+      "epoch": 1.1737328441364316,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2664169724147304e-05,
+      "loss": 1.0203,
+      "step": 17275
+    },
+    {
+      "epoch": 1.1740725642070933,
+      "grad_norm": 5.625,
+      "learning_rate": 4.266204647370567e-05,
+      "loss": 1.1666,
+      "step": 17280
+    },
+    {
+      "epoch": 1.1744122842777551,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.265992322326403e-05,
+      "loss": 1.1522,
+      "step": 17285
+    },
+    {
+      "epoch": 1.174752004348417,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.2657799972822396e-05,
+      "loss": 1.0328,
+      "step": 17290
+    },
+    {
+      "epoch": 1.1750917244190786,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.265567672238076e-05,
+      "loss": 1.0656,
+      "step": 17295
+    },
+    {
+      "epoch": 1.1754314444897405,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.2653553471939124e-05,
+      "loss": 1.1164,
+      "step": 17300
+    },
+    {
+      "epoch": 1.1757711645604023,
+      "grad_norm": 6.0,
+      "learning_rate": 4.265143022149749e-05,
+      "loss": 1.111,
+      "step": 17305
+    },
+    {
+      "epoch": 1.176110884631064,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.264930697105585e-05,
+      "loss": 1.1709,
+      "step": 17310
+    },
+    {
+      "epoch": 1.1764506047017258,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.2647183720614216e-05,
+      "loss": 0.9545,
+      "step": 17315
+    },
+    {
+      "epoch": 1.1767903247723877,
+      "grad_norm": 5.5,
+      "learning_rate": 4.264506047017258e-05,
+      "loss": 1.1225,
+      "step": 17320
+    },
+    {
+      "epoch": 1.1771300448430493,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2642937219730944e-05,
+      "loss": 1.1087,
+      "step": 17325
+    },
+    {
+      "epoch": 1.1774697649137111,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.264081396928931e-05,
+      "loss": 1.1824,
+      "step": 17330
+    },
+    {
+      "epoch": 1.177809484984373,
+      "grad_norm": 5.625,
+      "learning_rate": 4.263869071884767e-05,
+      "loss": 1.1099,
+      "step": 17335
+    },
+    {
+      "epoch": 1.1781492050550346,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2636567468406036e-05,
+      "loss": 1.0937,
+      "step": 17340
+    },
+    {
+      "epoch": 1.1784889251256965,
+      "grad_norm": 5.5,
+      "learning_rate": 4.26344442179644e-05,
+      "loss": 0.9463,
+      "step": 17345
+    },
+    {
+      "epoch": 1.178828645196358,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2632320967522764e-05,
+      "loss": 1.1259,
+      "step": 17350
+    },
+    {
+      "epoch": 1.17916836526702,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.263019771708113e-05,
+      "loss": 1.0441,
+      "step": 17355
+    },
+    {
+      "epoch": 1.1795080853376818,
+      "grad_norm": 6.0,
+      "learning_rate": 4.262807446663949e-05,
+      "loss": 1.1213,
+      "step": 17360
+    },
+    {
+      "epoch": 1.1798478054083434,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.2625951216197856e-05,
+      "loss": 1.0776,
+      "step": 17365
+    },
+    {
+      "epoch": 1.1801875254790053,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.262382796575622e-05,
+      "loss": 1.0488,
+      "step": 17370
+    },
+    {
+      "epoch": 1.1805272455496671,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.2621704715314584e-05,
+      "loss": 1.022,
+      "step": 17375
+    },
+    {
+      "epoch": 1.1808669656203288,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.261958146487295e-05,
+      "loss": 1.136,
+      "step": 17380
+    },
+    {
+      "epoch": 1.1812066856909906,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.261745821443131e-05,
+      "loss": 1.0719,
+      "step": 17385
+    },
+    {
+      "epoch": 1.1815464057616525,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.2615334963989676e-05,
+      "loss": 1.1025,
+      "step": 17390
+    },
+    {
+      "epoch": 1.1818861258323141,
+      "grad_norm": 5.5,
+      "learning_rate": 4.261321171354804e-05,
+      "loss": 1.0324,
+      "step": 17395
+    },
+    {
+      "epoch": 1.182225845902976,
+      "grad_norm": 6.5,
+      "learning_rate": 4.2611088463106404e-05,
+      "loss": 1.0627,
+      "step": 17400
+    },
+    {
+      "epoch": 1.1825655659736376,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.260896521266476e-05,
+      "loss": 1.0796,
+      "step": 17405
+    },
+    {
+      "epoch": 1.1829052860442995,
+      "grad_norm": 5.25,
+      "learning_rate": 4.260684196222313e-05,
+      "loss": 1.0967,
+      "step": 17410
+    },
+    {
+      "epoch": 1.1832450061149613,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.2604718711781496e-05,
+      "loss": 1.0168,
+      "step": 17415
+    },
+    {
+      "epoch": 1.183584726185623,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.260259546133985e-05,
+      "loss": 1.0223,
+      "step": 17420
+    },
+    {
+      "epoch": 1.1839244462562848,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.2600472210898224e-05,
+      "loss": 1.0439,
+      "step": 17425
+    },
+    {
+      "epoch": 1.1842641663269466,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.259834896045659e-05,
+      "loss": 1.0322,
+      "step": 17430
+    },
+    {
+      "epoch": 1.1846038863976083,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.2596225710014945e-05,
+      "loss": 1.1459,
+      "step": 17435
+    },
+    {
+      "epoch": 1.1849436064682701,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.2594102459573316e-05,
+      "loss": 1.0166,
+      "step": 17440
+    },
+    {
+      "epoch": 1.185283326538932,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.259197920913168e-05,
+      "loss": 0.9659,
+      "step": 17445
+    },
+    {
+      "epoch": 1.1856230466095936,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.258985595869004e-05,
+      "loss": 1.0123,
+      "step": 17450
+    },
+    {
+      "epoch": 1.1859627666802555,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.258773270824841e-05,
+      "loss": 1.1254,
+      "step": 17455
+    },
+    {
+      "epoch": 1.1863024867509173,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.258560945780677e-05,
+      "loss": 1.1675,
+      "step": 17460
+    },
+    {
+      "epoch": 1.186642206821579,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.258348620736513e-05,
+      "loss": 1.0532,
+      "step": 17465
+    },
+    {
+      "epoch": 1.1869819268922408,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.25813629569235e-05,
+      "loss": 1.0862,
+      "step": 17470
+    },
+    {
+      "epoch": 1.1873216469629027,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.257923970648186e-05,
+      "loss": 1.0156,
+      "step": 17475
+    },
+    {
+      "epoch": 1.1876613670335643,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.257711645604022e-05,
+      "loss": 1.174,
+      "step": 17480
+    },
+    {
+      "epoch": 1.1880010871042261,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.257499320559859e-05,
+      "loss": 1.0052,
+      "step": 17485
+    },
+    {
+      "epoch": 1.188340807174888,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.257286995515695e-05,
+      "loss": 1.0131,
+      "step": 17490
+    },
+    {
+      "epoch": 1.1886805272455496,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.257074670471531e-05,
+      "loss": 1.0985,
+      "step": 17495
+    },
+    {
+      "epoch": 1.1890202473162115,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.2568623454273684e-05,
+      "loss": 1.0448,
+      "step": 17500
+    },
+    {
+      "epoch": 1.1893599673868733,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.256650020383204e-05,
+      "loss": 1.004,
+      "step": 17505
+    },
+    {
+      "epoch": 1.189699687457535,
+      "grad_norm": 5.5,
+      "learning_rate": 4.2564376953390405e-05,
+      "loss": 1.0405,
+      "step": 17510
+    },
+    {
+      "epoch": 1.1900394075281968,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.2562253702948776e-05,
+      "loss": 1.0271,
+      "step": 17515
+    },
+    {
+      "epoch": 1.1903791275988584,
+      "grad_norm": 5.875,
+      "learning_rate": 4.256013045250713e-05,
+      "loss": 1.0519,
+      "step": 17520
+    },
+    {
+      "epoch": 1.1907188476695203,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.25580072020655e-05,
+      "loss": 1.0596,
+      "step": 17525
+    },
+    {
+      "epoch": 1.1910585677401822,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.255588395162387e-05,
+      "loss": 1.0792,
+      "step": 17530
+    },
+    {
+      "epoch": 1.1913982878108438,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.2553760701182225e-05,
+      "loss": 1.0548,
+      "step": 17535
+    },
+    {
+      "epoch": 1.1917380078815056,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.255163745074059e-05,
+      "loss": 1.0053,
+      "step": 17540
+    },
+    {
+      "epoch": 1.1920777279521675,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.254951420029896e-05,
+      "loss": 1.0443,
+      "step": 17545
+    },
+    {
+      "epoch": 1.1924174480228291,
+      "grad_norm": 5.875,
+      "learning_rate": 4.254739094985732e-05,
+      "loss": 1.0077,
+      "step": 17550
+    },
+    {
+      "epoch": 1.192757168093491,
+      "grad_norm": 5.375,
+      "learning_rate": 4.254526769941569e-05,
+      "loss": 1.0889,
+      "step": 17555
+    },
+    {
+      "epoch": 1.1930968881641528,
+      "grad_norm": 5.5,
+      "learning_rate": 4.2543144448974045e-05,
+      "loss": 1.0675,
+      "step": 17560
+    },
+    {
+      "epoch": 1.1934366082348145,
+      "grad_norm": 4.875,
+      "learning_rate": 4.254102119853241e-05,
+      "loss": 1.1082,
+      "step": 17565
+    },
+    {
+      "epoch": 1.1937763283054763,
+      "grad_norm": 6.125,
+      "learning_rate": 4.253889794809078e-05,
+      "loss": 1.0541,
+      "step": 17570
+    },
+    {
+      "epoch": 1.194116048376138,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.253677469764914e-05,
+      "loss": 1.044,
+      "step": 17575
+    },
+    {
+      "epoch": 1.1944557684467998,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.25346514472075e-05,
+      "loss": 1.1779,
+      "step": 17580
+    },
+    {
+      "epoch": 1.1947954885174616,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.253252819676587e-05,
+      "loss": 1.1592,
+      "step": 17585
+    },
+    {
+      "epoch": 1.1951352085881233,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.253040494632423e-05,
+      "loss": 1.1354,
+      "step": 17590
+    },
+    {
+      "epoch": 1.1954749286587851,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.252828169588259e-05,
+      "loss": 1.1137,
+      "step": 17595
+    },
+    {
+      "epoch": 1.195814648729447,
+      "grad_norm": 5.375,
+      "learning_rate": 4.2526158445440964e-05,
+      "loss": 1.0481,
+      "step": 17600
+    },
+    {
+      "epoch": 1.1961543688001086,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.252403519499932e-05,
+      "loss": 1.0624,
+      "step": 17605
+    },
+    {
+      "epoch": 1.1964940888707705,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.2521911944557685e-05,
+      "loss": 1.0268,
+      "step": 17610
+    },
+    {
+      "epoch": 1.1968338089414323,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2519788694116056e-05,
+      "loss": 1.0572,
+      "step": 17615
+    },
+    {
+      "epoch": 1.197173529012094,
+      "grad_norm": 6.25,
+      "learning_rate": 4.251766544367441e-05,
+      "loss": 1.059,
+      "step": 17620
+    },
+    {
+      "epoch": 1.1975132490827558,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.251554219323278e-05,
+      "loss": 1.0602,
+      "step": 17625
+    },
+    {
+      "epoch": 1.1978529691534177,
+      "grad_norm": 5.5,
+      "learning_rate": 4.251341894279114e-05,
+      "loss": 1.0657,
+      "step": 17630
+    },
+    {
+      "epoch": 1.1981926892240793,
+      "grad_norm": 6.125,
+      "learning_rate": 4.2511295692349505e-05,
+      "loss": 1.1782,
+      "step": 17635
+    },
+    {
+      "epoch": 1.1985324092947411,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.250917244190787e-05,
+      "loss": 1.0015,
+      "step": 17640
+    },
+    {
+      "epoch": 1.198872129365403,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.250704919146623e-05,
+      "loss": 1.0873,
+      "step": 17645
+    },
+    {
+      "epoch": 1.1992118494360646,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.25049259410246e-05,
+      "loss": 1.0551,
+      "step": 17650
+    },
+    {
+      "epoch": 1.1995515695067265,
+      "grad_norm": 5.0,
+      "learning_rate": 4.250280269058296e-05,
+      "loss": 1.1149,
+      "step": 17655
+    },
+    {
+      "epoch": 1.1998912895773883,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.2500679440141325e-05,
+      "loss": 1.1334,
+      "step": 17660
+    },
+    {
+      "epoch": 1.20023100964805,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.249855618969969e-05,
+      "loss": 1.0865,
+      "step": 17665
+    },
+    {
+      "epoch": 1.2005707297187118,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.249643293925805e-05,
+      "loss": 1.0722,
+      "step": 17670
+    },
+    {
+      "epoch": 1.2009104497893737,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.249430968881642e-05,
+      "loss": 1.1032,
+      "step": 17675
+    },
+    {
+      "epoch": 1.2012501698600353,
+      "grad_norm": 6.0,
+      "learning_rate": 4.249218643837478e-05,
+      "loss": 1.0482,
+      "step": 17680
+    },
+    {
+      "epoch": 1.2015898899306972,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.2490063187933145e-05,
+      "loss": 1.1089,
+      "step": 17685
+    },
+    {
+      "epoch": 1.2019296100013588,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.248793993749151e-05,
+      "loss": 1.0736,
+      "step": 17690
+    },
+    {
+      "epoch": 1.2022693300720206,
+      "grad_norm": 6.125,
+      "learning_rate": 4.248581668704987e-05,
+      "loss": 1.0231,
+      "step": 17695
+    },
+    {
+      "epoch": 1.2026090501426825,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.248369343660824e-05,
+      "loss": 0.9201,
+      "step": 17700
+    },
+    {
+      "epoch": 1.2029487702133441,
+      "grad_norm": 5.625,
+      "learning_rate": 4.24815701861666e-05,
+      "loss": 1.0362,
+      "step": 17705
+    },
+    {
+      "epoch": 1.203288490284006,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.2479446935724965e-05,
+      "loss": 1.0839,
+      "step": 17710
+    },
+    {
+      "epoch": 1.2036282103546678,
+      "grad_norm": 6.25,
+      "learning_rate": 4.247732368528333e-05,
+      "loss": 1.0799,
+      "step": 17715
+    },
+    {
+      "epoch": 1.2039679304253295,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.247520043484169e-05,
+      "loss": 1.1492,
+      "step": 17720
+    },
+    {
+      "epoch": 1.2043076504959913,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.247307718440006e-05,
+      "loss": 1.0134,
+      "step": 17725
+    },
+    {
+      "epoch": 1.2046473705666532,
+      "grad_norm": 5.5,
+      "learning_rate": 4.247095393395842e-05,
+      "loss": 1.1067,
+      "step": 17730
+    },
+    {
+      "epoch": 1.2049870906373148,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.2468830683516785e-05,
+      "loss": 0.9907,
+      "step": 17735
+    },
+    {
+      "epoch": 1.2053268107079766,
+      "grad_norm": 5.625,
+      "learning_rate": 4.246670743307515e-05,
+      "loss": 1.0643,
+      "step": 17740
+    },
+    {
+      "epoch": 1.2056665307786383,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.246458418263351e-05,
+      "loss": 1.0712,
+      "step": 17745
+    },
+    {
+      "epoch": 1.2060062508493001,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.246246093219188e-05,
+      "loss": 1.1,
+      "step": 17750
+    },
+    {
+      "epoch": 1.206345970919962,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.246033768175024e-05,
+      "loss": 1.0344,
+      "step": 17755
+    },
+    {
+      "epoch": 1.2066856909906236,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.24582144313086e-05,
+      "loss": 1.1244,
+      "step": 17760
+    },
+    {
+      "epoch": 1.2070254110612855,
+      "grad_norm": 6.125,
+      "learning_rate": 4.245609118086697e-05,
+      "loss": 1.114,
+      "step": 17765
+    },
+    {
+      "epoch": 1.2073651311319473,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.245396793042533e-05,
+      "loss": 1.0646,
+      "step": 17770
+    },
+    {
+      "epoch": 1.207704851202609,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.245184467998369e-05,
+      "loss": 1.1322,
+      "step": 17775
+    },
+    {
+      "epoch": 1.2080445712732708,
+      "grad_norm": 5.75,
+      "learning_rate": 4.244972142954206e-05,
+      "loss": 1.0223,
+      "step": 17780
+    },
+    {
+      "epoch": 1.2083842913439327,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.2447598179100425e-05,
+      "loss": 1.0816,
+      "step": 17785
+    },
+    {
+      "epoch": 1.2087240114145943,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.244547492865878e-05,
+      "loss": 1.0783,
+      "step": 17790
+    },
+    {
+      "epoch": 1.2090637314852561,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.244335167821715e-05,
+      "loss": 1.1399,
+      "step": 17795
+    },
+    {
+      "epoch": 1.209403451555918,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.244122842777552e-05,
+      "loss": 1.1025,
+      "step": 17800
+    },
+    {
+      "epoch": 1.2097431716265796,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.2439105177333875e-05,
+      "loss": 1.0968,
+      "step": 17805
+    },
+    {
+      "epoch": 1.2100828916972415,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.2436981926892245e-05,
+      "loss": 0.9809,
+      "step": 17810
+    },
+    {
+      "epoch": 1.2104226117679033,
+      "grad_norm": 5.625,
+      "learning_rate": 4.243485867645061e-05,
+      "loss": 1.1101,
+      "step": 17815
+    },
+    {
+      "epoch": 1.210762331838565,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2432735426008967e-05,
+      "loss": 1.1136,
+      "step": 17820
+    },
+    {
+      "epoch": 1.2111020519092268,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.243061217556734e-05,
+      "loss": 1.0861,
+      "step": 17825
+    },
+    {
+      "epoch": 1.2114417719798887,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2428488925125695e-05,
+      "loss": 1.1342,
+      "step": 17830
+    },
+    {
+      "epoch": 1.2117814920505503,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.242636567468406e-05,
+      "loss": 1.0933,
+      "step": 17835
+    },
+    {
+      "epoch": 1.2121212121212122,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.242424242424243e-05,
+      "loss": 1.0588,
+      "step": 17840
+    },
+    {
+      "epoch": 1.212460932191874,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.2422119173800787e-05,
+      "loss": 1.1026,
+      "step": 17845
+    },
+    {
+      "epoch": 1.2128006522625356,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.241999592335915e-05,
+      "loss": 1.133,
+      "step": 17850
+    },
+    {
+      "epoch": 1.2131403723331975,
+      "grad_norm": 6.375,
+      "learning_rate": 4.241787267291752e-05,
+      "loss": 1.1146,
+      "step": 17855
+    },
+    {
+      "epoch": 1.2134800924038591,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.241574942247588e-05,
+      "loss": 1.173,
+      "step": 17860
+    },
+    {
+      "epoch": 1.213819812474521,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.241362617203424e-05,
+      "loss": 1.0402,
+      "step": 17865
+    },
+    {
+      "epoch": 1.2141595325451828,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.241150292159261e-05,
+      "loss": 1.0475,
+      "step": 17870
+    },
+    {
+      "epoch": 1.2144992526158445,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.240937967115097e-05,
+      "loss": 1.1633,
+      "step": 17875
+    },
+    {
+      "epoch": 1.2148389726865063,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.2407256420709335e-05,
+      "loss": 1.0465,
+      "step": 17880
+    },
+    {
+      "epoch": 1.2151786927571682,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2405133170267705e-05,
+      "loss": 1.0256,
+      "step": 17885
+    },
+    {
+      "epoch": 1.2155184128278298,
+      "grad_norm": 4.9375,
+      "learning_rate": 4.240300991982606e-05,
+      "loss": 1.093,
+      "step": 17890
+    },
+    {
+      "epoch": 1.2158581328984917,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2400886669384433e-05,
+      "loss": 1.1459,
+      "step": 17895
+    },
+    {
+      "epoch": 1.2161978529691535,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.23987634189428e-05,
+      "loss": 1.0985,
+      "step": 17900
+    },
+    {
+      "epoch": 1.2165375730398151,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2396640168501155e-05,
+      "loss": 1.1255,
+      "step": 17905
+    },
+    {
+      "epoch": 1.216877293110477,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.2394516918059525e-05,
+      "loss": 1.0501,
+      "step": 17910
+    },
+    {
+      "epoch": 1.2172170131811386,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.239239366761788e-05,
+      "loss": 1.0253,
+      "step": 17915
+    },
+    {
+      "epoch": 1.2175567332518005,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.239027041717625e-05,
+      "loss": 1.0621,
+      "step": 17920
+    },
+    {
+      "epoch": 1.2178964533224623,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.238814716673462e-05,
+      "loss": 1.0525,
+      "step": 17925
+    },
+    {
+      "epoch": 1.218236173393124,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.2386023916292975e-05,
+      "loss": 1.0092,
+      "step": 17930
+    },
+    {
+      "epoch": 1.2185758934637858,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.238390066585134e-05,
+      "loss": 1.0224,
+      "step": 17935
+    },
+    {
+      "epoch": 1.2189156135344477,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.238177741540971e-05,
+      "loss": 1.0073,
+      "step": 17940
+    },
+    {
+      "epoch": 1.2192553336051093,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.237965416496807e-05,
+      "loss": 1.2149,
+      "step": 17945
+    },
+    {
+      "epoch": 1.2195950536757711,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.237753091452643e-05,
+      "loss": 1.1402,
+      "step": 17950
+    },
+    {
+      "epoch": 1.219934773746433,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.23754076640848e-05,
+      "loss": 1.0431,
+      "step": 17955
+    },
+    {
+      "epoch": 1.2202744938170946,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.237328441364316e-05,
+      "loss": 1.1644,
+      "step": 17960
+    },
+    {
+      "epoch": 1.2206142138877565,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.237116116320152e-05,
+      "loss": 1.0492,
+      "step": 17965
+    },
+    {
+      "epoch": 1.2209539339584183,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.2369037912759893e-05,
+      "loss": 1.1579,
+      "step": 17970
+    },
+    {
+      "epoch": 1.22129365402908,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.236691466231825e-05,
+      "loss": 1.0688,
+      "step": 17975
+    },
+    {
+      "epoch": 1.2216333740997418,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.2364791411876615e-05,
+      "loss": 1.14,
+      "step": 17980
+    },
+    {
+      "epoch": 1.2219730941704037,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.236266816143498e-05,
+      "loss": 1.0041,
+      "step": 17985
+    },
+    {
+      "epoch": 1.2223128142410653,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.236054491099334e-05,
+      "loss": 1.0275,
+      "step": 17990
+    },
+    {
+      "epoch": 1.2226525343117272,
+      "grad_norm": 5.625,
+      "learning_rate": 4.235842166055171e-05,
+      "loss": 1.1028,
+      "step": 17995
+    },
+    {
+      "epoch": 1.222992254382389,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.235629841011007e-05,
+      "loss": 0.9871,
+      "step": 18000
+    },
+    {
+      "epoch": 1.2233319744530506,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.2354175159668435e-05,
+      "loss": 1.0625,
+      "step": 18005
+    },
+    {
+      "epoch": 1.2236716945237125,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.23520519092268e-05,
+      "loss": 1.0701,
+      "step": 18010
+    },
+    {
+      "epoch": 1.2240114145943743,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.234992865878516e-05,
+      "loss": 1.0535,
+      "step": 18015
+    },
+    {
+      "epoch": 1.224351134665036,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.234780540834353e-05,
+      "loss": 1.0818,
+      "step": 18020
+    },
+    {
+      "epoch": 1.2246908547356978,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.234568215790189e-05,
+      "loss": 1.0791,
+      "step": 18025
+    },
+    {
+      "epoch": 1.2250305748063597,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2343558907460255e-05,
+      "loss": 1.0048,
+      "step": 18030
+    },
+    {
+      "epoch": 1.2253702948770213,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.234143565701862e-05,
+      "loss": 1.103,
+      "step": 18035
+    },
+    {
+      "epoch": 1.2257100149476832,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.233931240657698e-05,
+      "loss": 1.0823,
+      "step": 18040
+    },
+    {
+      "epoch": 1.2260497350183448,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.233718915613535e-05,
+      "loss": 1.1175,
+      "step": 18045
+    },
+    {
+      "epoch": 1.2263894550890067,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.233506590569371e-05,
+      "loss": 1.02,
+      "step": 18050
+    },
+    {
+      "epoch": 1.2267291751596685,
+      "grad_norm": 6.625,
+      "learning_rate": 4.2332942655252075e-05,
+      "loss": 1.0931,
+      "step": 18055
+    },
+    {
+      "epoch": 1.2270688952303301,
+      "grad_norm": 5.625,
+      "learning_rate": 4.233081940481044e-05,
+      "loss": 0.9754,
+      "step": 18060
+    },
+    {
+      "epoch": 1.227408615300992,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.23286961543688e-05,
+      "loss": 1.0552,
+      "step": 18065
+    },
+    {
+      "epoch": 1.2277483353716538,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.232657290392717e-05,
+      "loss": 1.1444,
+      "step": 18070
+    },
+    {
+      "epoch": 1.2280880554423155,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.232444965348553e-05,
+      "loss": 1.085,
+      "step": 18075
+    },
+    {
+      "epoch": 1.2284277755129773,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.2322326403043895e-05,
+      "loss": 1.0076,
+      "step": 18080
+    },
+    {
+      "epoch": 1.228767495583639,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.232020315260226e-05,
+      "loss": 1.0541,
+      "step": 18085
+    },
+    {
+      "epoch": 1.2291072156543008,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.231807990216062e-05,
+      "loss": 1.0923,
+      "step": 18090
+    },
+    {
+      "epoch": 1.2294469357249627,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.231595665171899e-05,
+      "loss": 1.0601,
+      "step": 18095
+    },
+    {
+      "epoch": 1.2297866557956243,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.231383340127735e-05,
+      "loss": 1.1171,
+      "step": 18100
+    },
+    {
+      "epoch": 1.2301263758662861,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.2311710150835715e-05,
+      "loss": 1.0773,
+      "step": 18105
+    },
+    {
+      "epoch": 1.230466095936948,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.230958690039408e-05,
+      "loss": 1.112,
+      "step": 18110
+    },
+    {
+      "epoch": 1.2308058160076096,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2307463649952436e-05,
+      "loss": 1.0596,
+      "step": 18115
+    },
+    {
+      "epoch": 1.2311455360782715,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.230534039951081e-05,
+      "loss": 1.1157,
+      "step": 18120
+    },
+    {
+      "epoch": 1.2314852561489333,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.230321714906917e-05,
+      "loss": 1.0938,
+      "step": 18125
+    },
+    {
+      "epoch": 1.231824976219595,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.230109389862753e-05,
+      "loss": 1.0592,
+      "step": 18130
+    },
+    {
+      "epoch": 1.2321646962902568,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.22989706481859e-05,
+      "loss": 1.0709,
+      "step": 18135
+    },
+    {
+      "epoch": 1.2325044163609187,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.229684739774426e-05,
+      "loss": 1.0801,
+      "step": 18140
+    },
+    {
+      "epoch": 1.2328441364315803,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.229472414730262e-05,
+      "loss": 1.0299,
+      "step": 18145
+    },
+    {
+      "epoch": 1.2331838565022422,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.229260089686099e-05,
+      "loss": 1.0314,
+      "step": 18150
+    },
+    {
+      "epoch": 1.233523576572904,
+      "grad_norm": 6.25,
+      "learning_rate": 4.2290477646419355e-05,
+      "loss": 1.111,
+      "step": 18155
+    },
+    {
+      "epoch": 1.2338632966435656,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.228835439597771e-05,
+      "loss": 1.1332,
+      "step": 18160
+    },
+    {
+      "epoch": 1.2342030167142275,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.228623114553608e-05,
+      "loss": 1.0555,
+      "step": 18165
+    },
+    {
+      "epoch": 1.2345427367848893,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.228410789509445e-05,
+      "loss": 1.0472,
+      "step": 18170
+    },
+    {
+      "epoch": 1.234882456855551,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.2281984644652804e-05,
+      "loss": 1.0569,
+      "step": 18175
+    },
+    {
+      "epoch": 1.2352221769262128,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.2279861394211175e-05,
+      "loss": 1.0471,
+      "step": 18180
+    },
+    {
+      "epoch": 1.2355618969968747,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.227773814376953e-05,
+      "loss": 1.0626,
+      "step": 18185
+    },
+    {
+      "epoch": 1.2359016170675363,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2275614893327896e-05,
+      "loss": 1.1548,
+      "step": 18190
+    },
+    {
+      "epoch": 1.2362413371381982,
+      "grad_norm": 6.0,
+      "learning_rate": 4.227349164288627e-05,
+      "loss": 1.0704,
+      "step": 18195
+    },
+    {
+      "epoch": 1.23658105720886,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.2271368392444624e-05,
+      "loss": 1.0465,
+      "step": 18200
+    },
+    {
+      "epoch": 1.2369207772795217,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.226924514200299e-05,
+      "loss": 0.9629,
+      "step": 18205
+    },
+    {
+      "epoch": 1.2372604973501835,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.226712189156136e-05,
+      "loss": 1.085,
+      "step": 18210
+    },
+    {
+      "epoch": 1.2376002174208451,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.2264998641119716e-05,
+      "loss": 1.062,
+      "step": 18215
+    },
+    {
+      "epoch": 1.237939937491507,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.226287539067808e-05,
+      "loss": 1.0618,
+      "step": 18220
+    },
+    {
+      "epoch": 1.2382796575621688,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.226075214023645e-05,
+      "loss": 1.0647,
+      "step": 18225
+    },
+    {
+      "epoch": 1.2386193776328305,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.225862888979481e-05,
+      "loss": 1.0426,
+      "step": 18230
+    },
+    {
+      "epoch": 1.2389590977034923,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.225650563935318e-05,
+      "loss": 1.1081,
+      "step": 18235
+    },
+    {
+      "epoch": 1.2392988177741542,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.225438238891154e-05,
+      "loss": 1.1202,
+      "step": 18240
+    },
+    {
+      "epoch": 1.2396385378448158,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.22522591384699e-05,
+      "loss": 1.0233,
+      "step": 18245
+    },
+    {
+      "epoch": 1.2399782579154777,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.225013588802827e-05,
+      "loss": 1.1394,
+      "step": 18250
+    },
+    {
+      "epoch": 1.2403179779861393,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.224801263758663e-05,
+      "loss": 1.0447,
+      "step": 18255
+    },
+    {
+      "epoch": 1.2406576980568012,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.224588938714499e-05,
+      "loss": 0.9963,
+      "step": 18260
+    },
+    {
+      "epoch": 1.240997418127463,
+      "grad_norm": 6.0,
+      "learning_rate": 4.224376613670336e-05,
+      "loss": 1.1658,
+      "step": 18265
+    },
+    {
+      "epoch": 1.2413371381981246,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.224164288626172e-05,
+      "loss": 0.985,
+      "step": 18270
+    },
+    {
+      "epoch": 1.2416768582687865,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2239519635820084e-05,
+      "loss": 1.08,
+      "step": 18275
+    },
+    {
+      "epoch": 1.2420165783394483,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.2237396385378455e-05,
+      "loss": 0.987,
+      "step": 18280
+    },
+    {
+      "epoch": 1.24235629841011,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.223527313493681e-05,
+      "loss": 1.2064,
+      "step": 18285
+    },
+    {
+      "epoch": 1.2426960184807718,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2233149884495176e-05,
+      "loss": 1.1161,
+      "step": 18290
+    },
+    {
+      "epoch": 1.2430357385514337,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.223102663405355e-05,
+      "loss": 1.0646,
+      "step": 18295
+    },
+    {
+      "epoch": 1.2433754586220953,
+      "grad_norm": 6.25,
+      "learning_rate": 4.2228903383611904e-05,
+      "loss": 1.1437,
+      "step": 18300
+    },
+    {
+      "epoch": 1.2437151786927572,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.222678013317027e-05,
+      "loss": 1.1449,
+      "step": 18305
+    },
+    {
+      "epoch": 1.244054898763419,
+      "grad_norm": 6.0,
+      "learning_rate": 4.222465688272864e-05,
+      "loss": 1.0548,
+      "step": 18310
+    },
+    {
+      "epoch": 1.2443946188340806,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.2222533632286996e-05,
+      "loss": 1.1285,
+      "step": 18315
+    },
+    {
+      "epoch": 1.2447343389047425,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.222041038184536e-05,
+      "loss": 1.0564,
+      "step": 18320
+    },
+    {
+      "epoch": 1.2450740589754044,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.221828713140373e-05,
+      "loss": 1.0569,
+      "step": 18325
+    },
+    {
+      "epoch": 1.245413779046066,
+      "grad_norm": 6.25,
+      "learning_rate": 4.221616388096209e-05,
+      "loss": 1.0965,
+      "step": 18330
+    },
+    {
+      "epoch": 1.2457534991167278,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.221404063052045e-05,
+      "loss": 0.9743,
+      "step": 18335
+    },
+    {
+      "epoch": 1.2460932191873897,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.2211917380078816e-05,
+      "loss": 1.0699,
+      "step": 18340
+    },
+    {
+      "epoch": 1.2464329392580513,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.220979412963718e-05,
+      "loss": 1.0714,
+      "step": 18345
+    },
+    {
+      "epoch": 1.2467726593287132,
+      "grad_norm": 5.0,
+      "learning_rate": 4.2207670879195544e-05,
+      "loss": 1.0264,
+      "step": 18350
+    },
+    {
+      "epoch": 1.247112379399375,
+      "grad_norm": 6.0,
+      "learning_rate": 4.220554762875391e-05,
+      "loss": 1.0773,
+      "step": 18355
+    },
+    {
+      "epoch": 1.2474520994700367,
+      "grad_norm": 6.0,
+      "learning_rate": 4.220342437831227e-05,
+      "loss": 1.0896,
+      "step": 18360
+    },
+    {
+      "epoch": 1.2477918195406985,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.2201301127870636e-05,
+      "loss": 1.0284,
+      "step": 18365
+    },
+    {
+      "epoch": 1.2481315396113604,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.2199177877429e-05,
+      "loss": 1.0346,
+      "step": 18370
+    },
+    {
+      "epoch": 1.248471259682022,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.2197054626987364e-05,
+      "loss": 1.0708,
+      "step": 18375
+    },
+    {
+      "epoch": 1.2488109797526838,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.219493137654573e-05,
+      "loss": 0.993,
+      "step": 18380
+    },
+    {
+      "epoch": 1.2491506998233455,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.219280812610409e-05,
+      "loss": 1.0463,
+      "step": 18385
+    },
+    {
+      "epoch": 1.2494904198940073,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.2190684875662456e-05,
+      "loss": 1.0746,
+      "step": 18390
+    },
+    {
+      "epoch": 1.2498301399646692,
+      "grad_norm": 5.75,
+      "learning_rate": 4.218856162522082e-05,
+      "loss": 1.0639,
+      "step": 18395
+    },
+    {
+      "epoch": 1.2501698600353308,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2186438374779184e-05,
+      "loss": 1.0649,
+      "step": 18400
+    },
+    {
+      "epoch": 1.2505095801059927,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.218431512433755e-05,
+      "loss": 1.0627,
+      "step": 18405
+    },
+    {
+      "epoch": 1.2508493001766543,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.218219187389591e-05,
+      "loss": 1.0528,
+      "step": 18410
+    },
+    {
+      "epoch": 1.2511890202473162,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.2180068623454276e-05,
+      "loss": 1.1229,
+      "step": 18415
+    },
+    {
+      "epoch": 1.251528740317978,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.217794537301264e-05,
+      "loss": 1.0393,
+      "step": 18420
+    },
+    {
+      "epoch": 1.2518684603886396,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2175822122571004e-05,
+      "loss": 1.0497,
+      "step": 18425
+    },
+    {
+      "epoch": 1.2522081804593015,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.217369887212937e-05,
+      "loss": 1.0076,
+      "step": 18430
+    },
+    {
+      "epoch": 1.2525479005299633,
+      "grad_norm": 5.75,
+      "learning_rate": 4.217157562168773e-05,
+      "loss": 1.1064,
+      "step": 18435
+    },
+    {
+      "epoch": 1.252887620600625,
+      "grad_norm": 6.5,
+      "learning_rate": 4.2169452371246096e-05,
+      "loss": 1.0513,
+      "step": 18440
+    },
+    {
+      "epoch": 1.2532273406712868,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.216732912080446e-05,
+      "loss": 1.0248,
+      "step": 18445
+    },
+    {
+      "epoch": 1.2535670607419487,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.2165205870362824e-05,
+      "loss": 1.0873,
+      "step": 18450
+    },
+    {
+      "epoch": 1.2539067808126103,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.216308261992118e-05,
+      "loss": 0.9628,
+      "step": 18455
+    },
+    {
+      "epoch": 1.2542465008832722,
+      "grad_norm": 6.375,
+      "learning_rate": 4.216095936947955e-05,
+      "loss": 1.1113,
+      "step": 18460
+    },
+    {
+      "epoch": 1.254586220953934,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.2158836119037916e-05,
+      "loss": 1.0769,
+      "step": 18465
+    },
+    {
+      "epoch": 1.2549259410245956,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.2156712868596274e-05,
+      "loss": 1.1509,
+      "step": 18470
+    },
+    {
+      "epoch": 1.2552656610952575,
+      "grad_norm": 5.75,
+      "learning_rate": 4.2154589618154644e-05,
+      "loss": 1.1203,
+      "step": 18475
+    },
+    {
+      "epoch": 1.2556053811659194,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.215246636771301e-05,
+      "loss": 1.0871,
+      "step": 18480
+    },
+    {
+      "epoch": 1.255945101236581,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.2150343117271366e-05,
+      "loss": 1.0807,
+      "step": 18485
+    },
+    {
+      "epoch": 1.2562848213072428,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2148219866829736e-05,
+      "loss": 0.9896,
+      "step": 18490
+    },
+    {
+      "epoch": 1.2566245413779047,
+      "grad_norm": 6.375,
+      "learning_rate": 4.21460966163881e-05,
+      "loss": 1.1807,
+      "step": 18495
+    },
+    {
+      "epoch": 1.2569642614485663,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.214397336594646e-05,
+      "loss": 1.1291,
+      "step": 18500
+    },
+    {
+      "epoch": 1.2573039815192282,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.214185011550483e-05,
+      "loss": 1.0223,
+      "step": 18505
+    },
+    {
+      "epoch": 1.25764370158989,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.213972686506319e-05,
+      "loss": 1.049,
+      "step": 18510
+    },
+    {
+      "epoch": 1.2579834216605517,
+      "grad_norm": 6.0,
+      "learning_rate": 4.213760361462155e-05,
+      "loss": 1.1963,
+      "step": 18515
+    },
+    {
+      "epoch": 1.2583231417312135,
+      "grad_norm": 6.125,
+      "learning_rate": 4.213548036417992e-05,
+      "loss": 1.0551,
+      "step": 18520
+    },
+    {
+      "epoch": 1.2586628618018754,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.2133357113738284e-05,
+      "loss": 1.0683,
+      "step": 18525
+    },
+    {
+      "epoch": 1.259002581872537,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.213123386329664e-05,
+      "loss": 1.1267,
+      "step": 18530
+    },
+    {
+      "epoch": 1.2593423019431988,
+      "grad_norm": 6.25,
+      "learning_rate": 4.212911061285501e-05,
+      "loss": 1.1507,
+      "step": 18535
+    },
+    {
+      "epoch": 1.2596820220138607,
+      "grad_norm": 5.25,
+      "learning_rate": 4.212698736241337e-05,
+      "loss": 1.0415,
+      "step": 18540
+    },
+    {
+      "epoch": 1.2600217420845223,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.2124864111971734e-05,
+      "loss": 1.0868,
+      "step": 18545
+    },
+    {
+      "epoch": 1.2603614621551842,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.2122740861530104e-05,
+      "loss": 1.088,
+      "step": 18550
+    },
+    {
+      "epoch": 1.260701182225846,
+      "grad_norm": 6.0,
+      "learning_rate": 4.212061761108846e-05,
+      "loss": 1.0432,
+      "step": 18555
+    },
+    {
+      "epoch": 1.2610409022965077,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.2118494360646826e-05,
+      "loss": 1.1609,
+      "step": 18560
+    },
+    {
+      "epoch": 1.2613806223671695,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2116371110205196e-05,
+      "loss": 1.1719,
+      "step": 18565
+    },
+    {
+      "epoch": 1.2617203424378312,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2114247859763554e-05,
+      "loss": 1.0814,
+      "step": 18570
+    },
+    {
+      "epoch": 1.262060062508493,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.2112124609321924e-05,
+      "loss": 1.0525,
+      "step": 18575
+    },
+    {
+      "epoch": 1.2623997825791546,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.211000135888029e-05,
+      "loss": 1.0738,
+      "step": 18580
+    },
+    {
+      "epoch": 1.2627395026498165,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.2107878108438646e-05,
+      "loss": 1.0905,
+      "step": 18585
+    },
+    {
+      "epoch": 1.2630792227204783,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.2105754857997016e-05,
+      "loss": 1.0339,
+      "step": 18590
+    },
+    {
+      "epoch": 1.26341894279114,
+      "grad_norm": 5.875,
+      "learning_rate": 4.210363160755538e-05,
+      "loss": 1.026,
+      "step": 18595
+    },
+    {
+      "epoch": 1.2637586628618018,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.210150835711374e-05,
+      "loss": 1.0486,
+      "step": 18600
+    },
+    {
+      "epoch": 1.2640983829324637,
+      "grad_norm": 6.25,
+      "learning_rate": 4.209938510667211e-05,
+      "loss": 1.0741,
+      "step": 18605
+    },
+    {
+      "epoch": 1.2644381030031253,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.2097261856230466e-05,
+      "loss": 1.0476,
+      "step": 18610
+    },
+    {
+      "epoch": 1.2647778230737872,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.209513860578883e-05,
+      "loss": 1.1079,
+      "step": 18615
+    },
+    {
+      "epoch": 1.265117543144449,
+      "grad_norm": 5.75,
+      "learning_rate": 4.20930153553472e-05,
+      "loss": 1.0768,
+      "step": 18620
+    },
+    {
+      "epoch": 1.2654572632151107,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.209089210490556e-05,
+      "loss": 1.0426,
+      "step": 18625
+    },
+    {
+      "epoch": 1.2657969832857725,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.208876885446392e-05,
+      "loss": 1.0563,
+      "step": 18630
+    },
+    {
+      "epoch": 1.2661367033564344,
+      "grad_norm": 6.25,
+      "learning_rate": 4.208664560402229e-05,
+      "loss": 1.1121,
+      "step": 18635
+    },
+    {
+      "epoch": 1.266476423427096,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.208452235358065e-05,
+      "loss": 1.0042,
+      "step": 18640
+    },
+    {
+      "epoch": 1.2668161434977578,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.2082399103139014e-05,
+      "loss": 1.0064,
+      "step": 18645
+    },
+    {
+      "epoch": 1.2671558635684197,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.2080275852697384e-05,
+      "loss": 1.0611,
+      "step": 18650
+    },
+    {
+      "epoch": 1.2674955836390813,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.207815260225574e-05,
+      "loss": 1.1372,
+      "step": 18655
+    },
+    {
+      "epoch": 1.2678353037097432,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2076029351814106e-05,
+      "loss": 1.0297,
+      "step": 18660
+    },
+    {
+      "epoch": 1.268175023780405,
+      "grad_norm": 5.875,
+      "learning_rate": 4.2073906101372476e-05,
+      "loss": 1.1721,
+      "step": 18665
+    },
+    {
+      "epoch": 1.2685147438510667,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.2071782850930834e-05,
+      "loss": 1.0506,
+      "step": 18670
+    },
+    {
+      "epoch": 1.2688544639217285,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.20696596004892e-05,
+      "loss": 1.0892,
+      "step": 18675
+    },
+    {
+      "epoch": 1.2691941839923904,
+      "grad_norm": 5.75,
+      "learning_rate": 4.206753635004756e-05,
+      "loss": 1.1033,
+      "step": 18680
+    },
+    {
+      "epoch": 1.269533904063052,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.2065413099605926e-05,
+      "loss": 1.1195,
+      "step": 18685
+    },
+    {
+      "epoch": 1.2698736241337139,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.206328984916429e-05,
+      "loss": 1.0582,
+      "step": 18690
+    },
+    {
+      "epoch": 1.2702133442043757,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.2061166598722654e-05,
+      "loss": 1.0452,
+      "step": 18695
+    },
+    {
+      "epoch": 1.2705530642750373,
+      "grad_norm": 6.125,
+      "learning_rate": 4.205904334828102e-05,
+      "loss": 1.0935,
+      "step": 18700
+    },
+    {
+      "epoch": 1.2708927843456992,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.205692009783938e-05,
+      "loss": 1.1653,
+      "step": 18705
+    },
+    {
+      "epoch": 1.271232504416361,
+      "grad_norm": 5.625,
+      "learning_rate": 4.2054796847397746e-05,
+      "loss": 1.0455,
+      "step": 18710
+    },
+    {
+      "epoch": 1.2715722244870227,
+      "grad_norm": 5.625,
+      "learning_rate": 4.205267359695611e-05,
+      "loss": 0.9639,
+      "step": 18715
+    },
+    {
+      "epoch": 1.2719119445576845,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.2050550346514474e-05,
+      "loss": 1.0697,
+      "step": 18720
+    },
+    {
+      "epoch": 1.2722516646283464,
+      "grad_norm": 5.75,
+      "learning_rate": 4.204842709607284e-05,
+      "loss": 1.0744,
+      "step": 18725
+    },
+    {
+      "epoch": 1.272591384699008,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.20463038456312e-05,
+      "loss": 1.1316,
+      "step": 18730
+    },
+    {
+      "epoch": 1.2729311047696699,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.2044180595189566e-05,
+      "loss": 0.9685,
+      "step": 18735
+    },
+    {
+      "epoch": 1.2732708248403315,
+      "grad_norm": 6.5,
+      "learning_rate": 4.204205734474793e-05,
+      "loss": 1.0885,
+      "step": 18740
+    },
+    {
+      "epoch": 1.2736105449109933,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.2039934094306294e-05,
+      "loss": 1.1135,
+      "step": 18745
+    },
+    {
+      "epoch": 1.273950264981655,
+      "grad_norm": 5.75,
+      "learning_rate": 4.203781084386466e-05,
+      "loss": 1.038,
+      "step": 18750
+    },
+    {
+      "epoch": 1.2742899850523168,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.203568759342302e-05,
+      "loss": 1.0339,
+      "step": 18755
+    },
+    {
+      "epoch": 1.2746297051229787,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.2033564342981386e-05,
+      "loss": 1.1466,
+      "step": 18760
+    },
+    {
+      "epoch": 1.2749694251936403,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.203144109253975e-05,
+      "loss": 1.0345,
+      "step": 18765
+    },
+    {
+      "epoch": 1.2753091452643022,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.2029317842098114e-05,
+      "loss": 1.055,
+      "step": 18770
+    },
+    {
+      "epoch": 1.275648865334964,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.202719459165648e-05,
+      "loss": 1.0626,
+      "step": 18775
+    },
+    {
+      "epoch": 1.2759885854056257,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.202507134121484e-05,
+      "loss": 1.1649,
+      "step": 18780
+    },
+    {
+      "epoch": 1.2763283054762875,
+      "grad_norm": 5.5,
+      "learning_rate": 4.2022948090773206e-05,
+      "loss": 1.072,
+      "step": 18785
+    },
+    {
+      "epoch": 1.2766680255469494,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.202082484033157e-05,
+      "loss": 1.1182,
+      "step": 18790
+    },
+    {
+      "epoch": 1.277007745617611,
+      "grad_norm": 5.75,
+      "learning_rate": 4.2018701589889934e-05,
+      "loss": 1.0987,
+      "step": 18795
+    },
+    {
+      "epoch": 1.2773474656882728,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.20165783394483e-05,
+      "loss": 1.1108,
+      "step": 18800
+    },
+    {
+      "epoch": 1.2776871857589347,
+      "grad_norm": 5.125,
+      "learning_rate": 4.201445508900666e-05,
+      "loss": 1.0644,
+      "step": 18805
+    },
+    {
+      "epoch": 1.2780269058295963,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.201233183856502e-05,
+      "loss": 1.0909,
+      "step": 18810
+    },
+    {
+      "epoch": 1.2783666259002582,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.201020858812339e-05,
+      "loss": 1.1249,
+      "step": 18815
+    },
+    {
+      "epoch": 1.27870634597092,
+      "grad_norm": 6.0,
+      "learning_rate": 4.2008085337681754e-05,
+      "loss": 1.0588,
+      "step": 18820
+    },
+    {
+      "epoch": 1.2790460660415817,
+      "grad_norm": 6.0,
+      "learning_rate": 4.200596208724011e-05,
+      "loss": 1.1105,
+      "step": 18825
+    },
+    {
+      "epoch": 1.2793857861122435,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.200383883679848e-05,
+      "loss": 1.1216,
+      "step": 18830
+    },
+    {
+      "epoch": 1.2797255061829054,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.2001715586356846e-05,
+      "loss": 1.1314,
+      "step": 18835
+    },
+    {
+      "epoch": 1.280065226253567,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.19995923359152e-05,
+      "loss": 1.0717,
+      "step": 18840
+    },
+    {
+      "epoch": 1.2804049463242289,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.1997469085473574e-05,
+      "loss": 1.0329,
+      "step": 18845
+    },
+    {
+      "epoch": 1.2807446663948907,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.199534583503194e-05,
+      "loss": 1.0275,
+      "step": 18850
+    },
+    {
+      "epoch": 1.2810843864655523,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.1993222584590295e-05,
+      "loss": 1.1179,
+      "step": 18855
+    },
+    {
+      "epoch": 1.2814241065362142,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.1991099334148666e-05,
+      "loss": 0.9962,
+      "step": 18860
+    },
+    {
+      "epoch": 1.281763826606876,
+      "grad_norm": 5.625,
+      "learning_rate": 4.198897608370703e-05,
+      "loss": 1.0617,
+      "step": 18865
+    },
+    {
+      "epoch": 1.2821035466775377,
+      "grad_norm": 6.375,
+      "learning_rate": 4.198685283326539e-05,
+      "loss": 1.009,
+      "step": 18870
+    },
+    {
+      "epoch": 1.2824432667481995,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.198472958282376e-05,
+      "loss": 1.0887,
+      "step": 18875
+    },
+    {
+      "epoch": 1.2827829868188614,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.1982606332382115e-05,
+      "loss": 1.08,
+      "step": 18880
+    },
+    {
+      "epoch": 1.283122706889523,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.198048308194048e-05,
+      "loss": 1.0053,
+      "step": 18885
+    },
+    {
+      "epoch": 1.2834624269601849,
+      "grad_norm": 5.875,
+      "learning_rate": 4.197835983149885e-05,
+      "loss": 1.0596,
+      "step": 18890
+    },
+    {
+      "epoch": 1.2838021470308467,
+      "grad_norm": 6.0,
+      "learning_rate": 4.197623658105721e-05,
+      "loss": 1.1102,
+      "step": 18895
+    },
+    {
+      "epoch": 1.2841418671015083,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.197411333061557e-05,
+      "loss": 1.1531,
+      "step": 18900
+    },
+    {
+      "epoch": 1.2844815871721702,
+      "grad_norm": 6.0,
+      "learning_rate": 4.197199008017394e-05,
+      "loss": 1.0882,
+      "step": 18905
+    },
+    {
+      "epoch": 1.2848213072428318,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.19698668297323e-05,
+      "loss": 1.1009,
+      "step": 18910
+    },
+    {
+      "epoch": 1.2851610273134937,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.196774357929067e-05,
+      "loss": 1.0288,
+      "step": 18915
+    },
+    {
+      "epoch": 1.2855007473841553,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.1965620328849034e-05,
+      "loss": 1.0276,
+      "step": 18920
+    },
+    {
+      "epoch": 1.2858404674548172,
+      "grad_norm": 5.5,
+      "learning_rate": 4.196349707840739e-05,
+      "loss": 1.0437,
+      "step": 18925
+    },
+    {
+      "epoch": 1.286180187525479,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.196137382796576e-05,
+      "loss": 1.063,
+      "step": 18930
+    },
+    {
+      "epoch": 1.2865199075961407,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.1959250577524126e-05,
+      "loss": 0.9927,
+      "step": 18935
+    },
+    {
+      "epoch": 1.2868596276668025,
+      "grad_norm": 5.375,
+      "learning_rate": 4.195712732708248e-05,
+      "loss": 1.0419,
+      "step": 18940
+    },
+    {
+      "epoch": 1.2871993477374644,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.1955004076640854e-05,
+      "loss": 1.0335,
+      "step": 18945
+    },
+    {
+      "epoch": 1.287539067808126,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.195288082619922e-05,
+      "loss": 1.1721,
+      "step": 18950
+    },
+    {
+      "epoch": 1.2878787878787878,
+      "grad_norm": 4.75,
+      "learning_rate": 4.1950757575757575e-05,
+      "loss": 1.0429,
+      "step": 18955
+    },
+    {
+      "epoch": 1.2882185079494497,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.1948634325315946e-05,
+      "loss": 0.9859,
+      "step": 18960
+    },
+    {
+      "epoch": 1.2885582280201113,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.19465110748743e-05,
+      "loss": 1.0284,
+      "step": 18965
+    },
+    {
+      "epoch": 1.2888979480907732,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.194438782443267e-05,
+      "loss": 1.1237,
+      "step": 18970
+    },
+    {
+      "epoch": 1.289237668161435,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.194226457399104e-05,
+      "loss": 1.0636,
+      "step": 18975
+    },
+    {
+      "epoch": 1.2895773882320967,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.1940141323549395e-05,
+      "loss": 1.149,
+      "step": 18980
+    },
+    {
+      "epoch": 1.2899171083027585,
+      "grad_norm": 5.625,
+      "learning_rate": 4.193801807310776e-05,
+      "loss": 1.0562,
+      "step": 18985
+    },
+    {
+      "epoch": 1.2902568283734204,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.193589482266613e-05,
+      "loss": 1.0735,
+      "step": 18990
+    },
+    {
+      "epoch": 1.290596548444082,
+      "grad_norm": 5.5,
+      "learning_rate": 4.193377157222449e-05,
+      "loss": 1.196,
+      "step": 18995
+    },
+    {
+      "epoch": 1.2909362685147439,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.193164832178285e-05,
+      "loss": 1.0356,
+      "step": 19000
+    },
+    {
+      "epoch": 1.2912759885854057,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.192952507134122e-05,
+      "loss": 1.0942,
+      "step": 19005
+    },
+    {
+      "epoch": 1.2916157086560673,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.192740182089958e-05,
+      "loss": 1.0891,
+      "step": 19010
+    },
+    {
+      "epoch": 1.2919554287267292,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.192527857045794e-05,
+      "loss": 1.0356,
+      "step": 19015
+    },
+    {
+      "epoch": 1.292295148797391,
+      "grad_norm": 5.875,
+      "learning_rate": 4.1923155320016314e-05,
+      "loss": 1.0066,
+      "step": 19020
+    },
+    {
+      "epoch": 1.2926348688680527,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.192103206957467e-05,
+      "loss": 1.0583,
+      "step": 19025
+    },
+    {
+      "epoch": 1.2929745889387145,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.1918908819133035e-05,
+      "loss": 1.0861,
+      "step": 19030
+    },
+    {
+      "epoch": 1.2933143090093764,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.19167855686914e-05,
+      "loss": 1.1432,
+      "step": 19035
+    },
+    {
+      "epoch": 1.293654029080038,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.191466231824976e-05,
+      "loss": 1.0435,
+      "step": 19040
+    },
+    {
+      "epoch": 1.2939937491506999,
+      "grad_norm": 5.875,
+      "learning_rate": 4.191253906780813e-05,
+      "loss": 1.014,
+      "step": 19045
+    },
+    {
+      "epoch": 1.2943334692213617,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.191041581736649e-05,
+      "loss": 1.0798,
+      "step": 19050
+    },
+    {
+      "epoch": 1.2946731892920234,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.1908292566924855e-05,
+      "loss": 1.0441,
+      "step": 19055
+    },
+    {
+      "epoch": 1.2950129093626852,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.190616931648322e-05,
+      "loss": 1.1641,
+      "step": 19060
+    },
+    {
+      "epoch": 1.295352629433347,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.190404606604158e-05,
+      "loss": 1.0624,
+      "step": 19065
+    },
+    {
+      "epoch": 1.2956923495040087,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.190192281559995e-05,
+      "loss": 1.0041,
+      "step": 19070
+    },
+    {
+      "epoch": 1.2960320695746705,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.189979956515831e-05,
+      "loss": 0.9893,
+      "step": 19075
+    },
+    {
+      "epoch": 1.2963717896453322,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.1897676314716675e-05,
+      "loss": 1.0753,
+      "step": 19080
+    },
+    {
+      "epoch": 1.296711509715994,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.189555306427504e-05,
+      "loss": 1.0657,
+      "step": 19085
+    },
+    {
+      "epoch": 1.2970512297866557,
+      "grad_norm": 6.125,
+      "learning_rate": 4.18934298138334e-05,
+      "loss": 1.091,
+      "step": 19090
+    },
+    {
+      "epoch": 1.2973909498573175,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.189130656339177e-05,
+      "loss": 1.0433,
+      "step": 19095
+    },
+    {
+      "epoch": 1.2977306699279794,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.188918331295013e-05,
+      "loss": 1.0874,
+      "step": 19100
+    },
+    {
+      "epoch": 1.298070389998641,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.1887060062508495e-05,
+      "loss": 0.999,
+      "step": 19105
+    },
+    {
+      "epoch": 1.2984101100693028,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.188493681206686e-05,
+      "loss": 1.1192,
+      "step": 19110
+    },
+    {
+      "epoch": 1.2987498301399647,
+      "grad_norm": 6.25,
+      "learning_rate": 4.188281356162522e-05,
+      "loss": 1.0568,
+      "step": 19115
+    },
+    {
+      "epoch": 1.2990895502106263,
+      "grad_norm": 5.875,
+      "learning_rate": 4.188069031118359e-05,
+      "loss": 1.1151,
+      "step": 19120
+    },
+    {
+      "epoch": 1.2994292702812882,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.187856706074195e-05,
+      "loss": 1.07,
+      "step": 19125
+    },
+    {
+      "epoch": 1.29976899035195,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.1876443810300315e-05,
+      "loss": 1.0381,
+      "step": 19130
+    },
+    {
+      "epoch": 1.3001087104226117,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.187432055985868e-05,
+      "loss": 1.062,
+      "step": 19135
+    },
+    {
+      "epoch": 1.3004484304932735,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.187219730941704e-05,
+      "loss": 1.0362,
+      "step": 19140
+    },
+    {
+      "epoch": 1.3007881505639354,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.187007405897541e-05,
+      "loss": 1.0847,
+      "step": 19145
+    },
+    {
+      "epoch": 1.301127870634597,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.186795080853377e-05,
+      "loss": 1.1498,
+      "step": 19150
+    },
+    {
+      "epoch": 1.3014675907052589,
+      "grad_norm": 5.375,
+      "learning_rate": 4.1865827558092135e-05,
+      "loss": 1.1754,
+      "step": 19155
+    },
+    {
+      "epoch": 1.3018073107759207,
+      "grad_norm": 6.375,
+      "learning_rate": 4.18637043076505e-05,
+      "loss": 1.0563,
+      "step": 19160
+    },
+    {
+      "epoch": 1.3021470308465823,
+      "grad_norm": 5.375,
+      "learning_rate": 4.1861581057208856e-05,
+      "loss": 1.0652,
+      "step": 19165
+    },
+    {
+      "epoch": 1.3024867509172442,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.185945780676723e-05,
+      "loss": 1.1235,
+      "step": 19170
+    },
+    {
+      "epoch": 1.302826470987906,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.185733455632559e-05,
+      "loss": 1.079,
+      "step": 19175
+    },
+    {
+      "epoch": 1.3031661910585677,
+      "grad_norm": 5.625,
+      "learning_rate": 4.185521130588395e-05,
+      "loss": 0.9664,
+      "step": 19180
+    },
+    {
+      "epoch": 1.3035059111292295,
+      "grad_norm": 5.75,
+      "learning_rate": 4.185308805544232e-05,
+      "loss": 1.0188,
+      "step": 19185
+    },
+    {
+      "epoch": 1.3038456311998914,
+      "grad_norm": 5.875,
+      "learning_rate": 4.185096480500068e-05,
+      "loss": 1.1005,
+      "step": 19190
+    },
+    {
+      "epoch": 1.304185351270553,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.184884155455904e-05,
+      "loss": 1.0014,
+      "step": 19195
+    },
+    {
+      "epoch": 1.3045250713412149,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.184671830411741e-05,
+      "loss": 1.0677,
+      "step": 19200
+    },
+    {
+      "epoch": 1.3048647914118767,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.1844595053675775e-05,
+      "loss": 1.0667,
+      "step": 19205
+    },
+    {
+      "epoch": 1.3052045114825384,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.184247180323413e-05,
+      "loss": 1.1067,
+      "step": 19210
+    },
+    {
+      "epoch": 1.3055442315532002,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.18403485527925e-05,
+      "loss": 1.1476,
+      "step": 19215
+    },
+    {
+      "epoch": 1.305883951623862,
+      "grad_norm": 5.625,
+      "learning_rate": 4.183822530235087e-05,
+      "loss": 1.0886,
+      "step": 19220
+    },
+    {
+      "epoch": 1.3062236716945237,
+      "grad_norm": 5.875,
+      "learning_rate": 4.1836102051909225e-05,
+      "loss": 1.095,
+      "step": 19225
+    },
+    {
+      "epoch": 1.3065633917651855,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.1833978801467595e-05,
+      "loss": 1.0732,
+      "step": 19230
+    },
+    {
+      "epoch": 1.3069031118358474,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.183185555102595e-05,
+      "loss": 1.0439,
+      "step": 19235
+    },
+    {
+      "epoch": 1.307242831906509,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.1829732300584317e-05,
+      "loss": 1.0596,
+      "step": 19240
+    },
+    {
+      "epoch": 1.3075825519771709,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.182760905014269e-05,
+      "loss": 1.1321,
+      "step": 19245
+    },
+    {
+      "epoch": 1.3079222720478325,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.1825485799701045e-05,
+      "loss": 1.1109,
+      "step": 19250
+    },
+    {
+      "epoch": 1.3082619921184944,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.1823362549259415e-05,
+      "loss": 1.1357,
+      "step": 19255
+    },
+    {
+      "epoch": 1.308601712189156,
+      "grad_norm": 5.75,
+      "learning_rate": 4.182123929881778e-05,
+      "loss": 1.0808,
+      "step": 19260
+    },
+    {
+      "epoch": 1.3089414322598178,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1819116048376137e-05,
+      "loss": 1.0956,
+      "step": 19265
+    },
+    {
+      "epoch": 1.3092811523304797,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.181699279793451e-05,
+      "loss": 1.0125,
+      "step": 19270
+    },
+    {
+      "epoch": 1.3096208724011413,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.181486954749287e-05,
+      "loss": 1.1758,
+      "step": 19275
+    },
+    {
+      "epoch": 1.3099605924718032,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.181274629705123e-05,
+      "loss": 1.0089,
+      "step": 19280
+    },
+    {
+      "epoch": 1.310300312542465,
+      "grad_norm": 6.125,
+      "learning_rate": 4.18106230466096e-05,
+      "loss": 1.1099,
+      "step": 19285
+    },
+    {
+      "epoch": 1.3106400326131267,
+      "grad_norm": 6.5,
+      "learning_rate": 4.180849979616796e-05,
+      "loss": 1.087,
+      "step": 19290
+    },
+    {
+      "epoch": 1.3109797526837885,
+      "grad_norm": 6.125,
+      "learning_rate": 4.180637654572632e-05,
+      "loss": 1.0776,
+      "step": 19295
+    },
+    {
+      "epoch": 1.3113194727544504,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.180425329528469e-05,
+      "loss": 1.0523,
+      "step": 19300
+    },
+    {
+      "epoch": 1.311659192825112,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.180213004484305e-05,
+      "loss": 1.0746,
+      "step": 19305
+    },
+    {
+      "epoch": 1.3119989128957739,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.180000679440141e-05,
+      "loss": 1.0383,
+      "step": 19310
+    },
+    {
+      "epoch": 1.3123386329664357,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.1797883543959783e-05,
+      "loss": 1.0765,
+      "step": 19315
+    },
+    {
+      "epoch": 1.3126783530370973,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.179576029351814e-05,
+      "loss": 1.0897,
+      "step": 19320
+    },
+    {
+      "epoch": 1.3130180731077592,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.1793637043076505e-05,
+      "loss": 0.9803,
+      "step": 19325
+    },
+    {
+      "epoch": 1.313357793178421,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.1791513792634875e-05,
+      "loss": 1.0759,
+      "step": 19330
+    },
+    {
+      "epoch": 1.3136975132490827,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.178939054219323e-05,
+      "loss": 1.0006,
+      "step": 19335
+    },
+    {
+      "epoch": 1.3140372333197445,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.17872672917516e-05,
+      "loss": 1.1024,
+      "step": 19340
+    },
+    {
+      "epoch": 1.3143769533904064,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.178514404130997e-05,
+      "loss": 1.0696,
+      "step": 19345
+    },
+    {
+      "epoch": 1.314716673461068,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.1783020790868325e-05,
+      "loss": 0.9978,
+      "step": 19350
+    },
+    {
+      "epoch": 1.3150563935317299,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.178089754042669e-05,
+      "loss": 0.9479,
+      "step": 19355
+    },
+    {
+      "epoch": 1.3153961136023917,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.177877428998506e-05,
+      "loss": 1.0207,
+      "step": 19360
+    },
+    {
+      "epoch": 1.3157358336730534,
+      "grad_norm": 6.375,
+      "learning_rate": 4.177665103954342e-05,
+      "loss": 1.0114,
+      "step": 19365
+    },
+    {
+      "epoch": 1.3160755537437152,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.177452778910178e-05,
+      "loss": 1.0796,
+      "step": 19370
+    },
+    {
+      "epoch": 1.316415273814377,
+      "grad_norm": 6.0,
+      "learning_rate": 4.177240453866015e-05,
+      "loss": 1.133,
+      "step": 19375
+    },
+    {
+      "epoch": 1.3167549938850387,
+      "grad_norm": 5.75,
+      "learning_rate": 4.177028128821851e-05,
+      "loss": 1.0662,
+      "step": 19380
+    },
+    {
+      "epoch": 1.3170947139557005,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.176815803777687e-05,
+      "loss": 1.2278,
+      "step": 19385
+    },
+    {
+      "epoch": 1.3174344340263624,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.176603478733524e-05,
+      "loss": 1.0002,
+      "step": 19390
+    },
+    {
+      "epoch": 1.317774154097024,
+      "grad_norm": 5.75,
+      "learning_rate": 4.17639115368936e-05,
+      "loss": 0.9608,
+      "step": 19395
+    },
+    {
+      "epoch": 1.3181138741676859,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1761788286451965e-05,
+      "loss": 1.1575,
+      "step": 19400
+    },
+    {
+      "epoch": 1.3184535942383477,
+      "grad_norm": 5.375,
+      "learning_rate": 4.175966503601033e-05,
+      "loss": 1.0792,
+      "step": 19405
+    },
+    {
+      "epoch": 1.3187933143090094,
+      "grad_norm": 5.75,
+      "learning_rate": 4.175754178556869e-05,
+      "loss": 1.024,
+      "step": 19410
+    },
+    {
+      "epoch": 1.3191330343796712,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.175541853512706e-05,
+      "loss": 1.034,
+      "step": 19415
+    },
+    {
+      "epoch": 1.3194727544503329,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.175329528468542e-05,
+      "loss": 1.0409,
+      "step": 19420
+    },
+    {
+      "epoch": 1.3198124745209947,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1751172034243785e-05,
+      "loss": 1.0746,
+      "step": 19425
+    },
+    {
+      "epoch": 1.3201521945916566,
+      "grad_norm": 6.0,
+      "learning_rate": 4.174904878380215e-05,
+      "loss": 1.1099,
+      "step": 19430
+    },
+    {
+      "epoch": 1.3204919146623182,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.174692553336051e-05,
+      "loss": 1.1194,
+      "step": 19435
+    },
+    {
+      "epoch": 1.32083163473298,
+      "grad_norm": 5.75,
+      "learning_rate": 4.174480228291888e-05,
+      "loss": 1.0692,
+      "step": 19440
+    },
+    {
+      "epoch": 1.3211713548036417,
+      "grad_norm": 5.625,
+      "learning_rate": 4.174267903247724e-05,
+      "loss": 0.9552,
+      "step": 19445
+    },
+    {
+      "epoch": 1.3215110748743035,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1740555782035605e-05,
+      "loss": 1.0367,
+      "step": 19450
+    },
+    {
+      "epoch": 1.3218507949449654,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.173843253159397e-05,
+      "loss": 1.2458,
+      "step": 19455
+    },
+    {
+      "epoch": 1.322190515015627,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.173630928115233e-05,
+      "loss": 1.0353,
+      "step": 19460
+    },
+    {
+      "epoch": 1.3225302350862889,
+      "grad_norm": 5.875,
+      "learning_rate": 4.17341860307107e-05,
+      "loss": 1.0445,
+      "step": 19465
+    },
+    {
+      "epoch": 1.3228699551569507,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.173206278026906e-05,
+      "loss": 1.1306,
+      "step": 19470
+    },
+    {
+      "epoch": 1.3232096752276123,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.1729939529827425e-05,
+      "loss": 1.0814,
+      "step": 19475
+    },
+    {
+      "epoch": 1.3235493952982742,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.172781627938579e-05,
+      "loss": 1.0386,
+      "step": 19480
+    },
+    {
+      "epoch": 1.323889115368936,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.172569302894415e-05,
+      "loss": 1.0518,
+      "step": 19485
+    },
+    {
+      "epoch": 1.3242288354395977,
+      "grad_norm": 5.375,
+      "learning_rate": 4.172356977850252e-05,
+      "loss": 1.0185,
+      "step": 19490
+    },
+    {
+      "epoch": 1.3245685555102595,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.172144652806088e-05,
+      "loss": 1.0726,
+      "step": 19495
+    },
+    {
+      "epoch": 1.3249082755809214,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.1719323277619245e-05,
+      "loss": 1.0814,
+      "step": 19500
+    },
+    {
+      "epoch": 1.325247995651583,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.17172000271776e-05,
+      "loss": 1.0727,
+      "step": 19505
+    },
+    {
+      "epoch": 1.3255877157222449,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.171507677673597e-05,
+      "loss": 0.9957,
+      "step": 19510
+    },
+    {
+      "epoch": 1.3259274357929067,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.171295352629434e-05,
+      "loss": 1.1725,
+      "step": 19515
+    },
+    {
+      "epoch": 1.3262671558635684,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.1710830275852694e-05,
+      "loss": 1.0597,
+      "step": 19520
+    },
+    {
+      "epoch": 1.3266068759342302,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.1708707025411065e-05,
+      "loss": 1.1077,
+      "step": 19525
+    },
+    {
+      "epoch": 1.326946596004892,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.170658377496943e-05,
+      "loss": 1.0725,
+      "step": 19530
+    },
+    {
+      "epoch": 1.3272863160755537,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.1704460524527786e-05,
+      "loss": 0.9933,
+      "step": 19535
+    },
+    {
+      "epoch": 1.3276260361462155,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.170233727408616e-05,
+      "loss": 1.1299,
+      "step": 19540
+    },
+    {
+      "epoch": 1.3279657562168774,
+      "grad_norm": 4.875,
+      "learning_rate": 4.170021402364452e-05,
+      "loss": 0.97,
+      "step": 19545
+    },
+    {
+      "epoch": 1.328305476287539,
+      "grad_norm": 5.5,
+      "learning_rate": 4.169809077320288e-05,
+      "loss": 1.0458,
+      "step": 19550
+    },
+    {
+      "epoch": 1.3286451963582009,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.169596752276125e-05,
+      "loss": 1.1696,
+      "step": 19555
+    },
+    {
+      "epoch": 1.3289849164288627,
+      "grad_norm": 5.75,
+      "learning_rate": 4.169384427231961e-05,
+      "loss": 1.0605,
+      "step": 19560
+    },
+    {
+      "epoch": 1.3293246364995244,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.169172102187797e-05,
+      "loss": 1.1715,
+      "step": 19565
+    },
+    {
+      "epoch": 1.3296643565701862,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.168959777143634e-05,
+      "loss": 1.0343,
+      "step": 19570
+    },
+    {
+      "epoch": 1.330004076640848,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.1687474520994705e-05,
+      "loss": 1.0422,
+      "step": 19575
+    },
+    {
+      "epoch": 1.3303437967115097,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.168535127055306e-05,
+      "loss": 1.1319,
+      "step": 19580
+    },
+    {
+      "epoch": 1.3306835167821716,
+      "grad_norm": 6.25,
+      "learning_rate": 4.168322802011143e-05,
+      "loss": 1.0833,
+      "step": 19585
+    },
+    {
+      "epoch": 1.3310232368528332,
+      "grad_norm": 6.25,
+      "learning_rate": 4.168110476966979e-05,
+      "loss": 1.1398,
+      "step": 19590
+    },
+    {
+      "epoch": 1.331362956923495,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.167898151922816e-05,
+      "loss": 1.1267,
+      "step": 19595
+    },
+    {
+      "epoch": 1.331702676994157,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.1676858268786525e-05,
+      "loss": 1.0621,
+      "step": 19600
+    },
+    {
+      "epoch": 1.3320423970648185,
+      "grad_norm": 6.25,
+      "learning_rate": 4.167473501834488e-05,
+      "loss": 1.0313,
+      "step": 19605
+    },
+    {
+      "epoch": 1.3323821171354804,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.167261176790325e-05,
+      "loss": 1.0976,
+      "step": 19610
+    },
+    {
+      "epoch": 1.332721837206142,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.167048851746162e-05,
+      "loss": 1.0523,
+      "step": 19615
+    },
+    {
+      "epoch": 1.3330615572768039,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1668365267019974e-05,
+      "loss": 1.17,
+      "step": 19620
+    },
+    {
+      "epoch": 1.3334012773474657,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.1666242016578345e-05,
+      "loss": 1.1004,
+      "step": 19625
+    },
+    {
+      "epoch": 1.3337409974181273,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.166411876613671e-05,
+      "loss": 1.2353,
+      "step": 19630
+    },
+    {
+      "epoch": 1.3340807174887892,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1661995515695066e-05,
+      "loss": 1.1448,
+      "step": 19635
+    },
+    {
+      "epoch": 1.334420437559451,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.165987226525344e-05,
+      "loss": 1.0427,
+      "step": 19640
+    },
+    {
+      "epoch": 1.3347601576301127,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.16577490148118e-05,
+      "loss": 1.0479,
+      "step": 19645
+    },
+    {
+      "epoch": 1.3350998777007745,
+      "grad_norm": 4.375,
+      "learning_rate": 4.165562576437016e-05,
+      "loss": 1.0577,
+      "step": 19650
+    },
+    {
+      "epoch": 1.3354395977714364,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.165350251392853e-05,
+      "loss": 1.0676,
+      "step": 19655
+    },
+    {
+      "epoch": 1.335779317842098,
+      "grad_norm": 5.25,
+      "learning_rate": 4.1651379263486886e-05,
+      "loss": 1.1213,
+      "step": 19660
+    },
+    {
+      "epoch": 1.3361190379127599,
+      "grad_norm": 6.25,
+      "learning_rate": 4.164925601304525e-05,
+      "loss": 1.0261,
+      "step": 19665
+    },
+    {
+      "epoch": 1.3364587579834217,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.164713276260362e-05,
+      "loss": 1.1189,
+      "step": 19670
+    },
+    {
+      "epoch": 1.3367984780540834,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.164500951216198e-05,
+      "loss": 1.0159,
+      "step": 19675
+    },
+    {
+      "epoch": 1.3371381981247452,
+      "grad_norm": 5.625,
+      "learning_rate": 4.164288626172034e-05,
+      "loss": 0.9627,
+      "step": 19680
+    },
+    {
+      "epoch": 1.337477918195407,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.164076301127871e-05,
+      "loss": 1.1173,
+      "step": 19685
+    },
+    {
+      "epoch": 1.3378176382660687,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.163863976083707e-05,
+      "loss": 1.0821,
+      "step": 19690
+    },
+    {
+      "epoch": 1.3381573583367306,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.1636516510395434e-05,
+      "loss": 1.0622,
+      "step": 19695
+    },
+    {
+      "epoch": 1.3384970784073924,
+      "grad_norm": 5.625,
+      "learning_rate": 4.1634393259953805e-05,
+      "loss": 0.982,
+      "step": 19700
+    },
+    {
+      "epoch": 1.338836798478054,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.163227000951216e-05,
+      "loss": 1.1185,
+      "step": 19705
+    },
+    {
+      "epoch": 1.3391765185487159,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.1630146759070526e-05,
+      "loss": 1.0021,
+      "step": 19710
+    },
+    {
+      "epoch": 1.3395162386193777,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.16280235086289e-05,
+      "loss": 1.0281,
+      "step": 19715
+    },
+    {
+      "epoch": 1.3398559586900394,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.1625900258187254e-05,
+      "loss": 1.0815,
+      "step": 19720
+    },
+    {
+      "epoch": 1.3401956787607012,
+      "grad_norm": 6.5,
+      "learning_rate": 4.162377700774562e-05,
+      "loss": 1.0876,
+      "step": 19725
+    },
+    {
+      "epoch": 1.340535398831363,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.162165375730398e-05,
+      "loss": 1.074,
+      "step": 19730
+    },
+    {
+      "epoch": 1.3408751189020247,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.1619530506862346e-05,
+      "loss": 1.0761,
+      "step": 19735
+    },
+    {
+      "epoch": 1.3412148389726866,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.161740725642071e-05,
+      "loss": 1.0507,
+      "step": 19740
+    },
+    {
+      "epoch": 1.3415545590433484,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.1615284005979074e-05,
+      "loss": 1.1065,
+      "step": 19745
+    },
+    {
+      "epoch": 1.34189427911401,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.161316075553744e-05,
+      "loss": 0.997,
+      "step": 19750
+    },
+    {
+      "epoch": 1.342233999184672,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.16110375050958e-05,
+      "loss": 1.005,
+      "step": 19755
+    },
+    {
+      "epoch": 1.3425737192553335,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.1608914254654166e-05,
+      "loss": 1.1069,
+      "step": 19760
+    },
+    {
+      "epoch": 1.3429134393259954,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.160679100421253e-05,
+      "loss": 1.0818,
+      "step": 19765
+    },
+    {
+      "epoch": 1.3432531593966572,
+      "grad_norm": 6.25,
+      "learning_rate": 4.1604667753770894e-05,
+      "loss": 0.9913,
+      "step": 19770
+    },
+    {
+      "epoch": 1.3435928794673189,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.160254450332926e-05,
+      "loss": 1.0722,
+      "step": 19775
+    },
+    {
+      "epoch": 1.3439325995379807,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.160042125288762e-05,
+      "loss": 1.0755,
+      "step": 19780
+    },
+    {
+      "epoch": 1.3442723196086424,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.1598298002445986e-05,
+      "loss": 1.1182,
+      "step": 19785
+    },
+    {
+      "epoch": 1.3446120396793042,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.159617475200435e-05,
+      "loss": 1.1227,
+      "step": 19790
+    },
+    {
+      "epoch": 1.344951759749966,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.1594051501562714e-05,
+      "loss": 1.0307,
+      "step": 19795
+    },
+    {
+      "epoch": 1.3452914798206277,
+      "grad_norm": 5.625,
+      "learning_rate": 4.159192825112108e-05,
+      "loss": 1.027,
+      "step": 19800
+    },
+    {
+      "epoch": 1.3456311998912895,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.158980500067944e-05,
+      "loss": 1.1237,
+      "step": 19805
+    },
+    {
+      "epoch": 1.3459709199619514,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.1587681750237806e-05,
+      "loss": 1.1172,
+      "step": 19810
+    },
+    {
+      "epoch": 1.346310640032613,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.158555849979617e-05,
+      "loss": 1.1363,
+      "step": 19815
+    },
+    {
+      "epoch": 1.3466503601032749,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.1583435249354534e-05,
+      "loss": 1.1345,
+      "step": 19820
+    },
+    {
+      "epoch": 1.3469900801739367,
+      "grad_norm": 6.125,
+      "learning_rate": 4.15813119989129e-05,
+      "loss": 1.0605,
+      "step": 19825
+    },
+    {
+      "epoch": 1.3473298002445984,
+      "grad_norm": 5.0625,
+      "learning_rate": 4.157918874847126e-05,
+      "loss": 1.0724,
+      "step": 19830
+    },
+    {
+      "epoch": 1.3476695203152602,
+      "grad_norm": 6.0,
+      "learning_rate": 4.1577065498029626e-05,
+      "loss": 1.0899,
+      "step": 19835
+    },
+    {
+      "epoch": 1.348009240385922,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.157494224758799e-05,
+      "loss": 1.005,
+      "step": 19840
+    },
+    {
+      "epoch": 1.3483489604565837,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.1572818997146354e-05,
+      "loss": 1.0752,
+      "step": 19845
+    },
+    {
+      "epoch": 1.3486886805272456,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.157069574670472e-05,
+      "loss": 1.1133,
+      "step": 19850
+    },
+    {
+      "epoch": 1.3490284005979074,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.156857249626308e-05,
+      "loss": 1.1738,
+      "step": 19855
+    },
+    {
+      "epoch": 1.349368120668569,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.156644924582144e-05,
+      "loss": 1.0671,
+      "step": 19860
+    },
+    {
+      "epoch": 1.349707840739231,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.156432599537981e-05,
+      "loss": 1.0789,
+      "step": 19865
+    },
+    {
+      "epoch": 1.3500475608098927,
+      "grad_norm": 6.125,
+      "learning_rate": 4.1562202744938174e-05,
+      "loss": 1.1743,
+      "step": 19870
+    },
+    {
+      "epoch": 1.3503872808805544,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.156007949449653e-05,
+      "loss": 1.0227,
+      "step": 19875
+    },
+    {
+      "epoch": 1.3507270009512162,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.15579562440549e-05,
+      "loss": 1.1333,
+      "step": 19880
+    },
+    {
+      "epoch": 1.351066721021878,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.1555832993613266e-05,
+      "loss": 1.1313,
+      "step": 19885
+    },
+    {
+      "epoch": 1.3514064410925397,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.1553709743171624e-05,
+      "loss": 1.1199,
+      "step": 19890
+    },
+    {
+      "epoch": 1.3517461611632016,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.1551586492729994e-05,
+      "loss": 1.111,
+      "step": 19895
+    },
+    {
+      "epoch": 1.3520858812338634,
+      "grad_norm": 6.625,
+      "learning_rate": 4.154946324228836e-05,
+      "loss": 1.0765,
+      "step": 19900
+    },
+    {
+      "epoch": 1.352425601304525,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.1547339991846716e-05,
+      "loss": 1.0665,
+      "step": 19905
+    },
+    {
+      "epoch": 1.352765321375187,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1545216741405086e-05,
+      "loss": 1.1202,
+      "step": 19910
+    },
+    {
+      "epoch": 1.3531050414458488,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.154309349096345e-05,
+      "loss": 1.1453,
+      "step": 19915
+    },
+    {
+      "epoch": 1.3534447615165104,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.154097024052181e-05,
+      "loss": 1.1655,
+      "step": 19920
+    },
+    {
+      "epoch": 1.3537844815871722,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.153884699008018e-05,
+      "loss": 1.1875,
+      "step": 19925
+    },
+    {
+      "epoch": 1.3541242016578339,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.1536723739638536e-05,
+      "loss": 1.096,
+      "step": 19930
+    },
+    {
+      "epoch": 1.3544639217284957,
+      "grad_norm": 6.5,
+      "learning_rate": 4.1534600489196906e-05,
+      "loss": 1.1325,
+      "step": 19935
+    },
+    {
+      "epoch": 1.3548036417991576,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.153247723875527e-05,
+      "loss": 1.0687,
+      "step": 19940
+    },
+    {
+      "epoch": 1.3551433618698192,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.153035398831363e-05,
+      "loss": 1.0173,
+      "step": 19945
+    },
+    {
+      "epoch": 1.355483081940481,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.1528230737872e-05,
+      "loss": 1.1245,
+      "step": 19950
+    },
+    {
+      "epoch": 1.3558228020111427,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.152610748743036e-05,
+      "loss": 1.0453,
+      "step": 19955
+    },
+    {
+      "epoch": 1.3561625220818045,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.152398423698872e-05,
+      "loss": 1.1198,
+      "step": 19960
+    },
+    {
+      "epoch": 1.3565022421524664,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.152186098654709e-05,
+      "loss": 1.0594,
+      "step": 19965
+    },
+    {
+      "epoch": 1.356841962223128,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.1519737736105454e-05,
+      "loss": 1.0502,
+      "step": 19970
+    },
+    {
+      "epoch": 1.3571816822937899,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.151761448566381e-05,
+      "loss": 1.1166,
+      "step": 19975
+    },
+    {
+      "epoch": 1.3575214023644517,
+      "grad_norm": 6.125,
+      "learning_rate": 4.151549123522218e-05,
+      "loss": 1.1497,
+      "step": 19980
+    },
+    {
+      "epoch": 1.3578611224351134,
+      "grad_norm": 5.625,
+      "learning_rate": 4.1513367984780546e-05,
+      "loss": 1.0857,
+      "step": 19985
+    },
+    {
+      "epoch": 1.3582008425057752,
+      "grad_norm": 5.625,
+      "learning_rate": 4.1511244734338904e-05,
+      "loss": 1.0219,
+      "step": 19990
+    },
+    {
+      "epoch": 1.358540562576437,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.1509121483897274e-05,
+      "loss": 1.0543,
+      "step": 19995
+    },
+    {
+      "epoch": 1.3588802826470987,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.150699823345564e-05,
+      "loss": 1.0567,
+      "step": 20000
+    },
+    {
+      "epoch": 1.3592200027177606,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.1504874983013996e-05,
+      "loss": 1.0502,
+      "step": 20005
+    },
+    {
+      "epoch": 1.3595597227884224,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.1502751732572366e-05,
+      "loss": 1.0709,
+      "step": 20010
+    },
+    {
+      "epoch": 1.359899442859084,
+      "grad_norm": 5.75,
+      "learning_rate": 4.1500628482130724e-05,
+      "loss": 1.0524,
+      "step": 20015
+    },
+    {
+      "epoch": 1.360239162929746,
+      "grad_norm": 5.5,
+      "learning_rate": 4.149850523168909e-05,
+      "loss": 1.0261,
+      "step": 20020
+    },
+    {
+      "epoch": 1.3605788830004077,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.149638198124746e-05,
+      "loss": 1.0016,
+      "step": 20025
+    },
+    {
+      "epoch": 1.3609186030710694,
+      "grad_norm": 5.0,
+      "learning_rate": 4.1494258730805816e-05,
+      "loss": 1.0775,
+      "step": 20030
+    },
+    {
+      "epoch": 1.3612583231417312,
+      "grad_norm": 5.875,
+      "learning_rate": 4.149213548036418e-05,
+      "loss": 1.182,
+      "step": 20035
+    },
+    {
+      "epoch": 1.361598043212393,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.149001222992255e-05,
+      "loss": 1.0783,
+      "step": 20040
+    },
+    {
+      "epoch": 1.3619377632830547,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.148788897948091e-05,
+      "loss": 0.9962,
+      "step": 20045
+    },
+    {
+      "epoch": 1.3622774833537166,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.148576572903927e-05,
+      "loss": 1.0503,
+      "step": 20050
+    },
+    {
+      "epoch": 1.3626172034243784,
+      "grad_norm": 5.75,
+      "learning_rate": 4.148364247859764e-05,
+      "loss": 1.1158,
+      "step": 20055
+    },
+    {
+      "epoch": 1.36295692349504,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.1481519228156e-05,
+      "loss": 1.0761,
+      "step": 20060
+    },
+    {
+      "epoch": 1.363296643565702,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.1479395977714364e-05,
+      "loss": 1.1477,
+      "step": 20065
+    },
+    {
+      "epoch": 1.3636363636363638,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.1477272727272734e-05,
+      "loss": 1.0518,
+      "step": 20070
+    },
+    {
+      "epoch": 1.3639760837070254,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.147514947683109e-05,
+      "loss": 1.0831,
+      "step": 20075
+    },
+    {
+      "epoch": 1.3643158037776872,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.1473026226389456e-05,
+      "loss": 1.0802,
+      "step": 20080
+    },
+    {
+      "epoch": 1.364655523848349,
+      "grad_norm": 6.375,
+      "learning_rate": 4.147090297594782e-05,
+      "loss": 0.9903,
+      "step": 20085
+    },
+    {
+      "epoch": 1.3649952439190107,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.1468779725506184e-05,
+      "loss": 1.0064,
+      "step": 20090
+    },
+    {
+      "epoch": 1.3653349639896726,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.146665647506455e-05,
+      "loss": 1.0005,
+      "step": 20095
+    },
+    {
+      "epoch": 1.3656746840603342,
+      "grad_norm": 5.625,
+      "learning_rate": 4.146453322462291e-05,
+      "loss": 1.1578,
+      "step": 20100
+    },
+    {
+      "epoch": 1.366014404130996,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1462409974181276e-05,
+      "loss": 1.1136,
+      "step": 20105
+    },
+    {
+      "epoch": 1.366354124201658,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.146028672373964e-05,
+      "loss": 1.0984,
+      "step": 20110
+    },
+    {
+      "epoch": 1.3666938442723195,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1458163473298004e-05,
+      "loss": 1.0884,
+      "step": 20115
+    },
+    {
+      "epoch": 1.3670335643429814,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.145604022285637e-05,
+      "loss": 1.0508,
+      "step": 20120
+    },
+    {
+      "epoch": 1.367373284413643,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.145391697241473e-05,
+      "loss": 1.0593,
+      "step": 20125
+    },
+    {
+      "epoch": 1.3677130044843049,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.1451793721973096e-05,
+      "loss": 1.1046,
+      "step": 20130
+    },
+    {
+      "epoch": 1.3680527245549667,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.144967047153146e-05,
+      "loss": 1.0835,
+      "step": 20135
+    },
+    {
+      "epoch": 1.3683924446256284,
+      "grad_norm": 5.0,
+      "learning_rate": 4.1447547221089824e-05,
+      "loss": 0.9507,
+      "step": 20140
+    },
+    {
+      "epoch": 1.3687321646962902,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.144542397064819e-05,
+      "loss": 1.0357,
+      "step": 20145
+    },
+    {
+      "epoch": 1.369071884766952,
+      "grad_norm": 6.125,
+      "learning_rate": 4.144330072020655e-05,
+      "loss": 1.0085,
+      "step": 20150
+    },
+    {
+      "epoch": 1.3694116048376137,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1441177469764916e-05,
+      "loss": 1.1025,
+      "step": 20155
+    },
+    {
+      "epoch": 1.3697513249082756,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.143905421932328e-05,
+      "loss": 1.0148,
+      "step": 20160
+    },
+    {
+      "epoch": 1.3700910449789374,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.1436930968881644e-05,
+      "loss": 1.086,
+      "step": 20165
+    },
+    {
+      "epoch": 1.370430765049599,
+      "grad_norm": 6.125,
+      "learning_rate": 4.143480771844001e-05,
+      "loss": 1.0078,
+      "step": 20170
+    },
+    {
+      "epoch": 1.370770485120261,
+      "grad_norm": 5.75,
+      "learning_rate": 4.143268446799837e-05,
+      "loss": 1.0959,
+      "step": 20175
+    },
+    {
+      "epoch": 1.3711102051909227,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.1430561217556736e-05,
+      "loss": 1.0825,
+      "step": 20180
+    },
+    {
+      "epoch": 1.3714499252615844,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.14284379671151e-05,
+      "loss": 1.0204,
+      "step": 20185
+    },
+    {
+      "epoch": 1.3717896453322462,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.1426314716673464e-05,
+      "loss": 1.0553,
+      "step": 20190
+    },
+    {
+      "epoch": 1.372129365402908,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.142419146623183e-05,
+      "loss": 0.9895,
+      "step": 20195
+    },
+    {
+      "epoch": 1.3724690854735697,
+      "grad_norm": 6.125,
+      "learning_rate": 4.142206821579019e-05,
+      "loss": 1.0407,
+      "step": 20200
+    },
+    {
+      "epoch": 1.3728088055442316,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.1419944965348556e-05,
+      "loss": 1.0061,
+      "step": 20205
+    },
+    {
+      "epoch": 1.3731485256148934,
+      "grad_norm": 5.375,
+      "learning_rate": 4.141782171490692e-05,
+      "loss": 0.9545,
+      "step": 20210
+    },
+    {
+      "epoch": 1.373488245685555,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.141569846446528e-05,
+      "loss": 1.0646,
+      "step": 20215
+    },
+    {
+      "epoch": 1.373827965756217,
+      "grad_norm": 6.125,
+      "learning_rate": 4.141357521402365e-05,
+      "loss": 1.073,
+      "step": 20220
+    },
+    {
+      "epoch": 1.3741676858268788,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.141145196358201e-05,
+      "loss": 1.0837,
+      "step": 20225
+    },
+    {
+      "epoch": 1.3745074058975404,
+      "grad_norm": 5.625,
+      "learning_rate": 4.140932871314037e-05,
+      "loss": 1.0428,
+      "step": 20230
+    },
+    {
+      "epoch": 1.3748471259682022,
+      "grad_norm": 6.375,
+      "learning_rate": 4.140720546269874e-05,
+      "loss": 0.9973,
+      "step": 20235
+    },
+    {
+      "epoch": 1.375186846038864,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.1405082212257104e-05,
+      "loss": 1.1168,
+      "step": 20240
+    },
+    {
+      "epoch": 1.3755265661095257,
+      "grad_norm": 5.75,
+      "learning_rate": 4.140295896181546e-05,
+      "loss": 1.1606,
+      "step": 20245
+    },
+    {
+      "epoch": 1.3758662861801876,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.140083571137383e-05,
+      "loss": 1.0865,
+      "step": 20250
+    },
+    {
+      "epoch": 1.3762060062508494,
+      "grad_norm": 4.84375,
+      "learning_rate": 4.1398712460932196e-05,
+      "loss": 0.9928,
+      "step": 20255
+    },
+    {
+      "epoch": 1.376545726321511,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.139658921049055e-05,
+      "loss": 1.0835,
+      "step": 20260
+    },
+    {
+      "epoch": 1.376885446392173,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.1394465960048924e-05,
+      "loss": 1.0332,
+      "step": 20265
+    },
+    {
+      "epoch": 1.3772251664628345,
+      "grad_norm": 5.75,
+      "learning_rate": 4.139234270960729e-05,
+      "loss": 1.1557,
+      "step": 20270
+    },
+    {
+      "epoch": 1.3775648865334964,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.139021945916565e-05,
+      "loss": 1.1001,
+      "step": 20275
+    },
+    {
+      "epoch": 1.3779046066041583,
+      "grad_norm": 5.625,
+      "learning_rate": 4.1388096208724016e-05,
+      "loss": 1.1052,
+      "step": 20280
+    },
+    {
+      "epoch": 1.3782443266748199,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.138597295828237e-05,
+      "loss": 1.1148,
+      "step": 20285
+    },
+    {
+      "epoch": 1.3785840467454817,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.1383849707840744e-05,
+      "loss": 1.1151,
+      "step": 20290
+    },
+    {
+      "epoch": 1.3789237668161434,
+      "grad_norm": 5.75,
+      "learning_rate": 4.138172645739911e-05,
+      "loss": 1.1889,
+      "step": 20295
+    },
+    {
+      "epoch": 1.3792634868868052,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1379603206957465e-05,
+      "loss": 1.098,
+      "step": 20300
+    },
+    {
+      "epoch": 1.379603206957467,
+      "grad_norm": 6.25,
+      "learning_rate": 4.1377479956515836e-05,
+      "loss": 1.1179,
+      "step": 20305
+    },
+    {
+      "epoch": 1.3799429270281287,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.13753567060742e-05,
+      "loss": 1.1676,
+      "step": 20310
+    },
+    {
+      "epoch": 1.3802826470987906,
+      "grad_norm": 6.625,
+      "learning_rate": 4.137323345563256e-05,
+      "loss": 1.1561,
+      "step": 20315
+    },
+    {
+      "epoch": 1.3806223671694524,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.137111020519093e-05,
+      "loss": 1.0901,
+      "step": 20320
+    },
+    {
+      "epoch": 1.380962087240114,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.136898695474929e-05,
+      "loss": 1.0672,
+      "step": 20325
+    },
+    {
+      "epoch": 1.381301807310776,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.136686370430765e-05,
+      "loss": 1.0918,
+      "step": 20330
+    },
+    {
+      "epoch": 1.3816415273814378,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.136474045386602e-05,
+      "loss": 1.0753,
+      "step": 20335
+    },
+    {
+      "epoch": 1.3819812474520994,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.1362617203424384e-05,
+      "loss": 1.0669,
+      "step": 20340
+    },
+    {
+      "epoch": 1.3823209675227612,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.136049395298274e-05,
+      "loss": 1.1717,
+      "step": 20345
+    },
+    {
+      "epoch": 1.382660687593423,
+      "grad_norm": 6.125,
+      "learning_rate": 4.135837070254111e-05,
+      "loss": 1.0551,
+      "step": 20350
+    },
+    {
+      "epoch": 1.3830004076640847,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.1356247452099476e-05,
+      "loss": 0.9014,
+      "step": 20355
+    },
+    {
+      "epoch": 1.3833401277347466,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.135412420165783e-05,
+      "loss": 1.0314,
+      "step": 20360
+    },
+    {
+      "epoch": 1.3836798478054084,
+      "grad_norm": 6.125,
+      "learning_rate": 4.1352000951216204e-05,
+      "loss": 1.0958,
+      "step": 20365
+    },
+    {
+      "epoch": 1.38401956787607,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.134987770077456e-05,
+      "loss": 1.0852,
+      "step": 20370
+    },
+    {
+      "epoch": 1.384359287946732,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.1347754450332925e-05,
+      "loss": 1.0783,
+      "step": 20375
+    },
+    {
+      "epoch": 1.3846990080173938,
+      "grad_norm": 6.375,
+      "learning_rate": 4.1345631199891296e-05,
+      "loss": 1.1221,
+      "step": 20380
+    },
+    {
+      "epoch": 1.3850387280880554,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.134350794944965e-05,
+      "loss": 1.0729,
+      "step": 20385
+    },
+    {
+      "epoch": 1.3853784481587172,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.134138469900802e-05,
+      "loss": 1.0856,
+      "step": 20390
+    },
+    {
+      "epoch": 1.385718168229379,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.133926144856639e-05,
+      "loss": 0.991,
+      "step": 20395
+    },
+    {
+      "epoch": 1.3860578883000407,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.1337138198124745e-05,
+      "loss": 0.961,
+      "step": 20400
+    },
+    {
+      "epoch": 1.3863976083707026,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.133501494768311e-05,
+      "loss": 1.004,
+      "step": 20405
+    },
+    {
+      "epoch": 1.3867373284413644,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.133289169724148e-05,
+      "loss": 1.1103,
+      "step": 20410
+    },
+    {
+      "epoch": 1.387077048512026,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.133076844679984e-05,
+      "loss": 1.1119,
+      "step": 20415
+    },
+    {
+      "epoch": 1.387416768582688,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.13286451963582e-05,
+      "loss": 1.0455,
+      "step": 20420
+    },
+    {
+      "epoch": 1.3877564886533498,
+      "grad_norm": 6.0,
+      "learning_rate": 4.132652194591657e-05,
+      "loss": 1.0509,
+      "step": 20425
+    },
+    {
+      "epoch": 1.3880962087240114,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.132439869547493e-05,
+      "loss": 0.99,
+      "step": 20430
+    },
+    {
+      "epoch": 1.3884359287946733,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.132227544503329e-05,
+      "loss": 1.0745,
+      "step": 20435
+    },
+    {
+      "epoch": 1.3887756488653349,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.132015219459166e-05,
+      "loss": 1.0248,
+      "step": 20440
+    },
+    {
+      "epoch": 1.3891153689359967,
+      "grad_norm": 6.125,
+      "learning_rate": 4.131802894415002e-05,
+      "loss": 1.1355,
+      "step": 20445
+    },
+    {
+      "epoch": 1.3894550890066586,
+      "grad_norm": 6.625,
+      "learning_rate": 4.1315905693708385e-05,
+      "loss": 1.0293,
+      "step": 20450
+    },
+    {
+      "epoch": 1.3897948090773202,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.131378244326675e-05,
+      "loss": 0.9552,
+      "step": 20455
+    },
+    {
+      "epoch": 1.390134529147982,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.131165919282511e-05,
+      "loss": 1.175,
+      "step": 20460
+    },
+    {
+      "epoch": 1.3904742492186437,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.130953594238348e-05,
+      "loss": 1.041,
+      "step": 20465
+    },
+    {
+      "epoch": 1.3908139692893056,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.130741269194184e-05,
+      "loss": 1.0477,
+      "step": 20470
+    },
+    {
+      "epoch": 1.3911536893599674,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.1305289441500205e-05,
+      "loss": 1.0303,
+      "step": 20475
+    },
+    {
+      "epoch": 1.391493409430629,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.130316619105857e-05,
+      "loss": 1.0425,
+      "step": 20480
+    },
+    {
+      "epoch": 1.391833129501291,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.130104294061693e-05,
+      "loss": 1.0537,
+      "step": 20485
+    },
+    {
+      "epoch": 1.3921728495719528,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.12989196901753e-05,
+      "loss": 1.0398,
+      "step": 20490
+    },
+    {
+      "epoch": 1.3925125696426144,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.129679643973366e-05,
+      "loss": 1.0054,
+      "step": 20495
+    },
+    {
+      "epoch": 1.3928522897132762,
+      "grad_norm": 6.125,
+      "learning_rate": 4.1294673189292025e-05,
+      "loss": 0.9434,
+      "step": 20500
+    },
+    {
+      "epoch": 1.393192009783938,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.129254993885039e-05,
+      "loss": 1.0518,
+      "step": 20505
+    },
+    {
+      "epoch": 1.3935317298545997,
+      "grad_norm": 6.0,
+      "learning_rate": 4.129042668840875e-05,
+      "loss": 1.0386,
+      "step": 20510
+    },
+    {
+      "epoch": 1.3938714499252616,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.128830343796712e-05,
+      "loss": 1.0217,
+      "step": 20515
+    },
+    {
+      "epoch": 1.3942111699959234,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.128618018752548e-05,
+      "loss": 1.0802,
+      "step": 20520
+    },
+    {
+      "epoch": 1.394550890066585,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.1284056937083845e-05,
+      "loss": 1.0363,
+      "step": 20525
+    },
+    {
+      "epoch": 1.394890610137247,
+      "grad_norm": 6.75,
+      "learning_rate": 4.128193368664221e-05,
+      "loss": 1.0606,
+      "step": 20530
+    },
+    {
+      "epoch": 1.3952303302079088,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.127981043620057e-05,
+      "loss": 1.1236,
+      "step": 20535
+    },
+    {
+      "epoch": 1.3955700502785704,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.127768718575894e-05,
+      "loss": 1.1289,
+      "step": 20540
+    },
+    {
+      "epoch": 1.3959097703492322,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.12755639353173e-05,
+      "loss": 1.1033,
+      "step": 20545
+    },
+    {
+      "epoch": 1.396249490419894,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.1273440684875665e-05,
+      "loss": 1.0605,
+      "step": 20550
+    },
+    {
+      "epoch": 1.3965892104905557,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.127131743443403e-05,
+      "loss": 1.019,
+      "step": 20555
+    },
+    {
+      "epoch": 1.3969289305612176,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.126919418399239e-05,
+      "loss": 1.0582,
+      "step": 20560
+    },
+    {
+      "epoch": 1.3972686506318794,
+      "grad_norm": 4.875,
+      "learning_rate": 4.126707093355076e-05,
+      "loss": 1.0182,
+      "step": 20565
+    },
+    {
+      "epoch": 1.397608370702541,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.1264947683109114e-05,
+      "loss": 1.0308,
+      "step": 20570
+    },
+    {
+      "epoch": 1.397948090773203,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.1262824432667485e-05,
+      "loss": 1.1398,
+      "step": 20575
+    },
+    {
+      "epoch": 1.3982878108438648,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.126070118222585e-05,
+      "loss": 1.078,
+      "step": 20580
+    },
+    {
+      "epoch": 1.3986275309145264,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.1258577931784206e-05,
+      "loss": 1.0957,
+      "step": 20585
+    },
+    {
+      "epoch": 1.3989672509851883,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.125645468134258e-05,
+      "loss": 1.0467,
+      "step": 20590
+    },
+    {
+      "epoch": 1.3993069710558501,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.125433143090094e-05,
+      "loss": 1.0333,
+      "step": 20595
+    },
+    {
+      "epoch": 1.3996466911265117,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.12522081804593e-05,
+      "loss": 1.0905,
+      "step": 20600
+    },
+    {
+      "epoch": 1.3999864111971736,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.125008493001767e-05,
+      "loss": 1.0708,
+      "step": 20605
+    },
+    {
+      "epoch": 1.4003261312678352,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.124796167957603e-05,
+      "loss": 1.105,
+      "step": 20610
+    },
+    {
+      "epoch": 1.400665851338497,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.12458384291344e-05,
+      "loss": 1.1225,
+      "step": 20615
+    },
+    {
+      "epoch": 1.401005571409159,
+      "grad_norm": 5.875,
+      "learning_rate": 4.124371517869276e-05,
+      "loss": 0.9731,
+      "step": 20620
+    },
+    {
+      "epoch": 1.4013452914798206,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.1241591928251125e-05,
+      "loss": 1.0855,
+      "step": 20625
+    },
+    {
+      "epoch": 1.4016850115504824,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.123946867780949e-05,
+      "loss": 1.1941,
+      "step": 20630
+    },
+    {
+      "epoch": 1.402024731621144,
+      "grad_norm": 5.875,
+      "learning_rate": 4.123734542736785e-05,
+      "loss": 1.1076,
+      "step": 20635
+    },
+    {
+      "epoch": 1.402364451691806,
+      "grad_norm": 6.0,
+      "learning_rate": 4.123522217692621e-05,
+      "loss": 1.1231,
+      "step": 20640
+    },
+    {
+      "epoch": 1.4027041717624678,
+      "grad_norm": 5.875,
+      "learning_rate": 4.123309892648458e-05,
+      "loss": 1.0946,
+      "step": 20645
+    },
+    {
+      "epoch": 1.4030438918331294,
+      "grad_norm": 6.0,
+      "learning_rate": 4.1230975676042945e-05,
+      "loss": 1.0343,
+      "step": 20650
+    },
+    {
+      "epoch": 1.4033836119037912,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.12288524256013e-05,
+      "loss": 0.9753,
+      "step": 20655
+    },
+    {
+      "epoch": 1.403723331974453,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.122672917515967e-05,
+      "loss": 1.149,
+      "step": 20660
+    },
+    {
+      "epoch": 1.4040630520451147,
+      "grad_norm": 7.0,
+      "learning_rate": 4.122460592471804e-05,
+      "loss": 1.1202,
+      "step": 20665
+    },
+    {
+      "epoch": 1.4044027721157766,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.1222482674276395e-05,
+      "loss": 1.2117,
+      "step": 20670
+    },
+    {
+      "epoch": 1.4047424921864384,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.1220359423834765e-05,
+      "loss": 1.0371,
+      "step": 20675
+    },
+    {
+      "epoch": 1.4050822122571,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.121823617339313e-05,
+      "loss": 1.1287,
+      "step": 20680
+    },
+    {
+      "epoch": 1.405421932327762,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.1216112922951487e-05,
+      "loss": 1.127,
+      "step": 20685
+    },
+    {
+      "epoch": 1.4057616523984238,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.121398967250986e-05,
+      "loss": 1.0263,
+      "step": 20690
+    },
+    {
+      "epoch": 1.4061013724690854,
+      "grad_norm": 5.75,
+      "learning_rate": 4.121186642206822e-05,
+      "loss": 1.0545,
+      "step": 20695
+    },
+    {
+      "epoch": 1.4064410925397473,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.120974317162658e-05,
+      "loss": 1.0689,
+      "step": 20700
+    },
+    {
+      "epoch": 1.406780812610409,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.120761992118495e-05,
+      "loss": 1.144,
+      "step": 20705
+    },
+    {
+      "epoch": 1.4071205326810707,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.1205496670743307e-05,
+      "loss": 1.1467,
+      "step": 20710
+    },
+    {
+      "epoch": 1.4074602527517326,
+      "grad_norm": 5.875,
+      "learning_rate": 4.120337342030167e-05,
+      "loss": 1.0856,
+      "step": 20715
+    },
+    {
+      "epoch": 1.4077999728223944,
+      "grad_norm": 4.6875,
+      "learning_rate": 4.120125016986004e-05,
+      "loss": 1.0724,
+      "step": 20720
+    },
+    {
+      "epoch": 1.408139692893056,
+      "grad_norm": 6.125,
+      "learning_rate": 4.11991269194184e-05,
+      "loss": 1.0759,
+      "step": 20725
+    },
+    {
+      "epoch": 1.408479412963718,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.119700366897676e-05,
+      "loss": 1.0971,
+      "step": 20730
+    },
+    {
+      "epoch": 1.4088191330343798,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.119488041853513e-05,
+      "loss": 1.1328,
+      "step": 20735
+    },
+    {
+      "epoch": 1.4091588531050414,
+      "grad_norm": 5.375,
+      "learning_rate": 4.119275716809349e-05,
+      "loss": 1.079,
+      "step": 20740
+    },
+    {
+      "epoch": 1.4094985731757033,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.1190633917651855e-05,
+      "loss": 1.0989,
+      "step": 20745
+    },
+    {
+      "epoch": 1.4098382932463651,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.1188510667210225e-05,
+      "loss": 1.1141,
+      "step": 20750
+    },
+    {
+      "epoch": 1.4101780133170267,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.118638741676858e-05,
+      "loss": 1.0031,
+      "step": 20755
+    },
+    {
+      "epoch": 1.4105177333876886,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.118426416632695e-05,
+      "loss": 1.0683,
+      "step": 20760
+    },
+    {
+      "epoch": 1.4108574534583505,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.118214091588532e-05,
+      "loss": 1.0837,
+      "step": 20765
+    },
+    {
+      "epoch": 1.411197173529012,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.1180017665443675e-05,
+      "loss": 1.0079,
+      "step": 20770
+    },
+    {
+      "epoch": 1.411536893599674,
+      "grad_norm": 6.0,
+      "learning_rate": 4.117789441500204e-05,
+      "loss": 1.0416,
+      "step": 20775
+    },
+    {
+      "epoch": 1.4118766136703356,
+      "grad_norm": 5.5,
+      "learning_rate": 4.117577116456041e-05,
+      "loss": 1.1258,
+      "step": 20780
+    },
+    {
+      "epoch": 1.4122163337409974,
+      "grad_norm": 6.125,
+      "learning_rate": 4.117364791411877e-05,
+      "loss": 1.1053,
+      "step": 20785
+    },
+    {
+      "epoch": 1.4125560538116593,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.117152466367713e-05,
+      "loss": 1.0399,
+      "step": 20790
+    },
+    {
+      "epoch": 1.412895773882321,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.1169401413235495e-05,
+      "loss": 1.1556,
+      "step": 20795
+    },
+    {
+      "epoch": 1.4132354939529828,
+      "grad_norm": 6.125,
+      "learning_rate": 4.116727816279386e-05,
+      "loss": 1.0346,
+      "step": 20800
+    },
+    {
+      "epoch": 1.4135752140236444,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.116515491235222e-05,
+      "loss": 1.0955,
+      "step": 20805
+    },
+    {
+      "epoch": 1.4139149340943062,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.116303166191059e-05,
+      "loss": 1.1502,
+      "step": 20810
+    },
+    {
+      "epoch": 1.414254654164968,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.116090841146895e-05,
+      "loss": 1.0805,
+      "step": 20815
+    },
+    {
+      "epoch": 1.4145943742356297,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.1158785161027315e-05,
+      "loss": 1.0707,
+      "step": 20820
+    },
+    {
+      "epoch": 1.4149340943062916,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.115666191058568e-05,
+      "loss": 1.0699,
+      "step": 20825
+    },
+    {
+      "epoch": 1.4152738143769534,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.115453866014404e-05,
+      "loss": 1.0782,
+      "step": 20830
+    },
+    {
+      "epoch": 1.415613534447615,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.115241540970241e-05,
+      "loss": 1.0565,
+      "step": 20835
+    },
+    {
+      "epoch": 1.415953254518277,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.115029215926077e-05,
+      "loss": 1.0774,
+      "step": 20840
+    },
+    {
+      "epoch": 1.4162929745889388,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.1148168908819135e-05,
+      "loss": 1.0951,
+      "step": 20845
+    },
+    {
+      "epoch": 1.4166326946596004,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.11460456583775e-05,
+      "loss": 0.9972,
+      "step": 20850
+    },
+    {
+      "epoch": 1.4169724147302623,
+      "grad_norm": 5.5,
+      "learning_rate": 4.114392240793586e-05,
+      "loss": 1.0347,
+      "step": 20855
+    },
+    {
+      "epoch": 1.417312134800924,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.114179915749423e-05,
+      "loss": 1.0618,
+      "step": 20860
+    },
+    {
+      "epoch": 1.4176518548715857,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.113967590705259e-05,
+      "loss": 1.0574,
+      "step": 20865
+    },
+    {
+      "epoch": 1.4179915749422476,
+      "grad_norm": 6.25,
+      "learning_rate": 4.1137552656610955e-05,
+      "loss": 1.0802,
+      "step": 20870
+    },
+    {
+      "epoch": 1.4183312950129094,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.113542940616932e-05,
+      "loss": 1.1402,
+      "step": 20875
+    },
+    {
+      "epoch": 1.418671015083571,
+      "grad_norm": 5.5,
+      "learning_rate": 4.113330615572768e-05,
+      "loss": 1.0231,
+      "step": 20880
+    },
+    {
+      "epoch": 1.419010735154233,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.113118290528605e-05,
+      "loss": 1.1092,
+      "step": 20885
+    },
+    {
+      "epoch": 1.4193504552248948,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.112905965484441e-05,
+      "loss": 1.1317,
+      "step": 20890
+    },
+    {
+      "epoch": 1.4196901752955564,
+      "grad_norm": 6.0,
+      "learning_rate": 4.1126936404402775e-05,
+      "loss": 1.1497,
+      "step": 20895
+    },
+    {
+      "epoch": 1.4200298953662183,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.112481315396114e-05,
+      "loss": 1.0581,
+      "step": 20900
+    },
+    {
+      "epoch": 1.4203696154368801,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.11226899035195e-05,
+      "loss": 1.1229,
+      "step": 20905
+    },
+    {
+      "epoch": 1.4207093355075417,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.112056665307786e-05,
+      "loss": 1.0777,
+      "step": 20910
+    },
+    {
+      "epoch": 1.4210490555782036,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.111844340263623e-05,
+      "loss": 1.0659,
+      "step": 20915
+    },
+    {
+      "epoch": 1.4213887756488655,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.1116320152194595e-05,
+      "loss": 1.0933,
+      "step": 20920
+    },
+    {
+      "epoch": 1.421728495719527,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.111419690175295e-05,
+      "loss": 1.0383,
+      "step": 20925
+    },
+    {
+      "epoch": 1.422068215790189,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.111207365131132e-05,
+      "loss": 1.0364,
+      "step": 20930
+    },
+    {
+      "epoch": 1.4224079358608508,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.110995040086969e-05,
+      "loss": 1.0367,
+      "step": 20935
+    },
+    {
+      "epoch": 1.4227476559315124,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.1107827150428044e-05,
+      "loss": 1.0438,
+      "step": 20940
+    },
+    {
+      "epoch": 1.4230873760021743,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.1105703899986415e-05,
+      "loss": 1.007,
+      "step": 20945
+    },
+    {
+      "epoch": 1.423427096072836,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.110358064954478e-05,
+      "loss": 1.1161,
+      "step": 20950
+    },
+    {
+      "epoch": 1.4237668161434978,
+      "grad_norm": 5.75,
+      "learning_rate": 4.110145739910314e-05,
+      "loss": 1.1213,
+      "step": 20955
+    },
+    {
+      "epoch": 1.4241065362141596,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.109933414866151e-05,
+      "loss": 1.0587,
+      "step": 20960
+    },
+    {
+      "epoch": 1.4244462562848212,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.109721089821987e-05,
+      "loss": 1.1205,
+      "step": 20965
+    },
+    {
+      "epoch": 1.424785976355483,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.1095087647778235e-05,
+      "loss": 1.1249,
+      "step": 20970
+    },
+    {
+      "epoch": 1.4251256964261447,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.10929643973366e-05,
+      "loss": 1.1062,
+      "step": 20975
+    },
+    {
+      "epoch": 1.4254654164968066,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.109084114689496e-05,
+      "loss": 1.1544,
+      "step": 20980
+    },
+    {
+      "epoch": 1.4258051365674684,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.108871789645333e-05,
+      "loss": 1.0291,
+      "step": 20985
+    },
+    {
+      "epoch": 1.42614485663813,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.108659464601169e-05,
+      "loss": 1.0366,
+      "step": 20990
+    },
+    {
+      "epoch": 1.426484576708792,
+      "grad_norm": 5.625,
+      "learning_rate": 4.108447139557005e-05,
+      "loss": 0.9604,
+      "step": 20995
+    },
+    {
+      "epoch": 1.4268242967794538,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.108234814512842e-05,
+      "loss": 1.181,
+      "step": 21000
+    },
+    {
+      "epoch": 1.4271640168501154,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.108022489468678e-05,
+      "loss": 1.0121,
+      "step": 21005
+    },
+    {
+      "epoch": 1.4275037369207773,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.107810164424514e-05,
+      "loss": 1.0359,
+      "step": 21010
+    },
+    {
+      "epoch": 1.427843456991439,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.107597839380351e-05,
+      "loss": 1.0817,
+      "step": 21015
+    },
+    {
+      "epoch": 1.4281831770621007,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.1073855143361875e-05,
+      "loss": 1.0397,
+      "step": 21020
+    },
+    {
+      "epoch": 1.4285228971327626,
+      "grad_norm": 6.5,
+      "learning_rate": 4.107173189292023e-05,
+      "loss": 1.0363,
+      "step": 21025
+    },
+    {
+      "epoch": 1.4288626172034244,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.10696086424786e-05,
+      "loss": 1.0631,
+      "step": 21030
+    },
+    {
+      "epoch": 1.429202337274086,
+      "grad_norm": 5.875,
+      "learning_rate": 4.106748539203697e-05,
+      "loss": 1.0172,
+      "step": 21035
+    },
+    {
+      "epoch": 1.429542057344748,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.1065362141595324e-05,
+      "loss": 1.0604,
+      "step": 21040
+    },
+    {
+      "epoch": 1.4298817774154098,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.1063238891153695e-05,
+      "loss": 1.084,
+      "step": 21045
+    },
+    {
+      "epoch": 1.4302214974860714,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.106111564071206e-05,
+      "loss": 1.0977,
+      "step": 21050
+    },
+    {
+      "epoch": 1.4305612175567333,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.1058992390270416e-05,
+      "loss": 1.0949,
+      "step": 21055
+    },
+    {
+      "epoch": 1.4309009376273951,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.105686913982879e-05,
+      "loss": 1.1395,
+      "step": 21060
+    },
+    {
+      "epoch": 1.4312406576980568,
+      "grad_norm": 5.375,
+      "learning_rate": 4.1054745889387144e-05,
+      "loss": 1.1766,
+      "step": 21065
+    },
+    {
+      "epoch": 1.4315803777687186,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.105262263894551e-05,
+      "loss": 1.0741,
+      "step": 21070
+    },
+    {
+      "epoch": 1.4319200978393805,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.105049938850388e-05,
+      "loss": 1.0962,
+      "step": 21075
+    },
+    {
+      "epoch": 1.432259817910042,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.1048376138062236e-05,
+      "loss": 0.994,
+      "step": 21080
+    },
+    {
+      "epoch": 1.432599537980704,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.10462528876206e-05,
+      "loss": 1.104,
+      "step": 21085
+    },
+    {
+      "epoch": 1.4329392580513658,
+      "grad_norm": 6.0,
+      "learning_rate": 4.104412963717897e-05,
+      "loss": 1.0071,
+      "step": 21090
+    },
+    {
+      "epoch": 1.4332789781220274,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.104200638673733e-05,
+      "loss": 0.9769,
+      "step": 21095
+    },
+    {
+      "epoch": 1.4336186981926893,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.103988313629569e-05,
+      "loss": 0.9166,
+      "step": 21100
+    },
+    {
+      "epoch": 1.4339584182633511,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.103775988585406e-05,
+      "loss": 1.0518,
+      "step": 21105
+    },
+    {
+      "epoch": 1.4342981383340128,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.103563663541242e-05,
+      "loss": 1.0106,
+      "step": 21110
+    },
+    {
+      "epoch": 1.4346378584046746,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1033513384970784e-05,
+      "loss": 1.0348,
+      "step": 21115
+    },
+    {
+      "epoch": 1.4349775784753362,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.1031390134529155e-05,
+      "loss": 1.2119,
+      "step": 21120
+    },
+    {
+      "epoch": 1.435317298545998,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.102926688408751e-05,
+      "loss": 1.1382,
+      "step": 21125
+    },
+    {
+      "epoch": 1.43565701861666,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1027143633645876e-05,
+      "loss": 1.0436,
+      "step": 21130
+    },
+    {
+      "epoch": 1.4359967386873216,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.102502038320424e-05,
+      "loss": 1.0531,
+      "step": 21135
+    },
+    {
+      "epoch": 1.4363364587579834,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.1022897132762604e-05,
+      "loss": 1.1351,
+      "step": 21140
+    },
+    {
+      "epoch": 1.436676178828645,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.102077388232097e-05,
+      "loss": 1.1539,
+      "step": 21145
+    },
+    {
+      "epoch": 1.437015898899307,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.101865063187933e-05,
+      "loss": 1.0088,
+      "step": 21150
+    },
+    {
+      "epoch": 1.4373556189699688,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.1016527381437696e-05,
+      "loss": 1.1089,
+      "step": 21155
+    },
+    {
+      "epoch": 1.4376953390406304,
+      "grad_norm": 6.125,
+      "learning_rate": 4.101440413099606e-05,
+      "loss": 1.0397,
+      "step": 21160
+    },
+    {
+      "epoch": 1.4380350591112923,
+      "grad_norm": 5.5,
+      "learning_rate": 4.1012280880554424e-05,
+      "loss": 1.0386,
+      "step": 21165
+    },
+    {
+      "epoch": 1.438374779181954,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.101015763011279e-05,
+      "loss": 1.2252,
+      "step": 21170
+    },
+    {
+      "epoch": 1.4387144992526157,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.100803437967115e-05,
+      "loss": 1.1525,
+      "step": 21175
+    },
+    {
+      "epoch": 1.4390542193232776,
+      "grad_norm": 5.875,
+      "learning_rate": 4.1005911129229516e-05,
+      "loss": 1.1221,
+      "step": 21180
+    },
+    {
+      "epoch": 1.4393939393939394,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.100378787878788e-05,
+      "loss": 1.1163,
+      "step": 21185
+    },
+    {
+      "epoch": 1.439733659464601,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.1001664628346244e-05,
+      "loss": 1.0465,
+      "step": 21190
+    },
+    {
+      "epoch": 1.440073379535263,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.099954137790461e-05,
+      "loss": 1.0661,
+      "step": 21195
+    },
+    {
+      "epoch": 1.4404130996059248,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.099741812746297e-05,
+      "loss": 1.109,
+      "step": 21200
+    },
+    {
+      "epoch": 1.4407528196765864,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.0995294877021336e-05,
+      "loss": 1.1604,
+      "step": 21205
+    },
+    {
+      "epoch": 1.4410925397472483,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.09931716265797e-05,
+      "loss": 1.0412,
+      "step": 21210
+    },
+    {
+      "epoch": 1.4414322598179101,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.0991048376138064e-05,
+      "loss": 1.1418,
+      "step": 21215
+    },
+    {
+      "epoch": 1.4417719798885718,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.098892512569643e-05,
+      "loss": 1.1069,
+      "step": 21220
+    },
+    {
+      "epoch": 1.4421116999592336,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.098680187525479e-05,
+      "loss": 1.0435,
+      "step": 21225
+    },
+    {
+      "epoch": 1.4424514200298955,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0984678624813156e-05,
+      "loss": 1.1814,
+      "step": 21230
+    },
+    {
+      "epoch": 1.442791140100557,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.098255537437152e-05,
+      "loss": 1.0445,
+      "step": 21235
+    },
+    {
+      "epoch": 1.443130860171219,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.0980432123929884e-05,
+      "loss": 1.0387,
+      "step": 21240
+    },
+    {
+      "epoch": 1.4434705802418808,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.097830887348825e-05,
+      "loss": 0.9617,
+      "step": 21245
+    },
+    {
+      "epoch": 1.4438103003125424,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.097618562304661e-05,
+      "loss": 1.0098,
+      "step": 21250
+    },
+    {
+      "epoch": 1.4441500203832043,
+      "grad_norm": 5.75,
+      "learning_rate": 4.0974062372604976e-05,
+      "loss": 1.0811,
+      "step": 21255
+    },
+    {
+      "epoch": 1.4444897404538661,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.097193912216334e-05,
+      "loss": 1.008,
+      "step": 21260
+    },
+    {
+      "epoch": 1.4448294605245278,
+      "grad_norm": 5.875,
+      "learning_rate": 4.09698158717217e-05,
+      "loss": 1.0612,
+      "step": 21265
+    },
+    {
+      "epoch": 1.4451691805951896,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.096769262128007e-05,
+      "loss": 1.0546,
+      "step": 21270
+    },
+    {
+      "epoch": 1.4455089006658515,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.096556937083843e-05,
+      "loss": 0.9774,
+      "step": 21275
+    },
+    {
+      "epoch": 1.445848620736513,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.096344612039679e-05,
+      "loss": 0.988,
+      "step": 21280
+    },
+    {
+      "epoch": 1.446188340807175,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.096132286995516e-05,
+      "loss": 1.0301,
+      "step": 21285
+    },
+    {
+      "epoch": 1.4465280608778366,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.0959199619513524e-05,
+      "loss": 1.1463,
+      "step": 21290
+    },
+    {
+      "epoch": 1.4468677809484984,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.095707636907189e-05,
+      "loss": 0.9602,
+      "step": 21295
+    },
+    {
+      "epoch": 1.4472075010191603,
+      "grad_norm": 6.25,
+      "learning_rate": 4.095495311863025e-05,
+      "loss": 1.0843,
+      "step": 21300
+    },
+    {
+      "epoch": 1.447547221089822,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0952829868188616e-05,
+      "loss": 1.0195,
+      "step": 21305
+    },
+    {
+      "epoch": 1.4478869411604838,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.095070661774698e-05,
+      "loss": 1.0755,
+      "step": 21310
+    },
+    {
+      "epoch": 1.4482266612311454,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.0948583367305344e-05,
+      "loss": 1.0857,
+      "step": 21315
+    },
+    {
+      "epoch": 1.4485663813018073,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.094646011686371e-05,
+      "loss": 1.1218,
+      "step": 21320
+    },
+    {
+      "epoch": 1.4489061013724691,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.094433686642207e-05,
+      "loss": 1.1733,
+      "step": 21325
+    },
+    {
+      "epoch": 1.4492458214431307,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.0942213615980436e-05,
+      "loss": 1.0589,
+      "step": 21330
+    },
+    {
+      "epoch": 1.4495855415137926,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.0940090365538794e-05,
+      "loss": 1.039,
+      "step": 21335
+    },
+    {
+      "epoch": 1.4499252615844545,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.0937967115097164e-05,
+      "loss": 1.0684,
+      "step": 21340
+    },
+    {
+      "epoch": 1.450264981655116,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.093584386465553e-05,
+      "loss": 1.1362,
+      "step": 21345
+    },
+    {
+      "epoch": 1.450604701725778,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.0933720614213886e-05,
+      "loss": 1.0448,
+      "step": 21350
+    },
+    {
+      "epoch": 1.4509444217964398,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.0931597363772256e-05,
+      "loss": 1.0702,
+      "step": 21355
+    },
+    {
+      "epoch": 1.4512841418671014,
+      "grad_norm": 5.75,
+      "learning_rate": 4.092947411333062e-05,
+      "loss": 1.063,
+      "step": 21360
+    },
+    {
+      "epoch": 1.4516238619377633,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.092735086288898e-05,
+      "loss": 1.0706,
+      "step": 21365
+    },
+    {
+      "epoch": 1.4519635820084251,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.092522761244735e-05,
+      "loss": 1.0772,
+      "step": 21370
+    },
+    {
+      "epoch": 1.4523033020790868,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.092310436200571e-05,
+      "loss": 1.0863,
+      "step": 21375
+    },
+    {
+      "epoch": 1.4526430221497486,
+      "grad_norm": 6.0,
+      "learning_rate": 4.092098111156407e-05,
+      "loss": 1.0684,
+      "step": 21380
+    },
+    {
+      "epoch": 1.4529827422204105,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.091885786112244e-05,
+      "loss": 1.092,
+      "step": 21385
+    },
+    {
+      "epoch": 1.453322462291072,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.0916734610680804e-05,
+      "loss": 1.0781,
+      "step": 21390
+    },
+    {
+      "epoch": 1.453662182361734,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.091461136023916e-05,
+      "loss": 1.0512,
+      "step": 21395
+    },
+    {
+      "epoch": 1.4540019024323958,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.091248810979753e-05,
+      "loss": 1.0763,
+      "step": 21400
+    },
+    {
+      "epoch": 1.4543416225030574,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.0910364859355896e-05,
+      "loss": 1.1639,
+      "step": 21405
+    },
+    {
+      "epoch": 1.4546813425737193,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.0908241608914254e-05,
+      "loss": 1.0529,
+      "step": 21410
+    },
+    {
+      "epoch": 1.4550210626443811,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.0906118358472624e-05,
+      "loss": 1.1184,
+      "step": 21415
+    },
+    {
+      "epoch": 1.4553607827150428,
+      "grad_norm": 5.75,
+      "learning_rate": 4.090399510803098e-05,
+      "loss": 1.1344,
+      "step": 21420
+    },
+    {
+      "epoch": 1.4557005027857046,
+      "grad_norm": 6.5,
+      "learning_rate": 4.0901871857589346e-05,
+      "loss": 1.0793,
+      "step": 21425
+    },
+    {
+      "epoch": 1.4560402228563665,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.0899748607147716e-05,
+      "loss": 1.1047,
+      "step": 21430
+    },
+    {
+      "epoch": 1.456379942927028,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.0897625356706074e-05,
+      "loss": 1.0627,
+      "step": 21435
+    },
+    {
+      "epoch": 1.45671966299769,
+      "grad_norm": 6.0,
+      "learning_rate": 4.089550210626444e-05,
+      "loss": 1.0521,
+      "step": 21440
+    },
+    {
+      "epoch": 1.4570593830683518,
+      "grad_norm": 5.625,
+      "learning_rate": 4.089337885582281e-05,
+      "loss": 0.9856,
+      "step": 21445
+    },
+    {
+      "epoch": 1.4573991031390134,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.0891255605381166e-05,
+      "loss": 1.1857,
+      "step": 21450
+    },
+    {
+      "epoch": 1.4577388232096753,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.088913235493953e-05,
+      "loss": 1.0699,
+      "step": 21455
+    },
+    {
+      "epoch": 1.458078543280337,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.08870091044979e-05,
+      "loss": 1.0275,
+      "step": 21460
+    },
+    {
+      "epoch": 1.4584182633509988,
+      "grad_norm": 6.5,
+      "learning_rate": 4.088488585405626e-05,
+      "loss": 1.0824,
+      "step": 21465
+    },
+    {
+      "epoch": 1.4587579834216606,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.088276260361462e-05,
+      "loss": 1.1054,
+      "step": 21470
+    },
+    {
+      "epoch": 1.4590977034923223,
+      "grad_norm": 5.875,
+      "learning_rate": 4.088063935317299e-05,
+      "loss": 1.0492,
+      "step": 21475
+    },
+    {
+      "epoch": 1.4594374235629841,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.087851610273135e-05,
+      "loss": 1.0656,
+      "step": 21480
+    },
+    {
+      "epoch": 1.4597771436336457,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.0876392852289714e-05,
+      "loss": 1.0519,
+      "step": 21485
+    },
+    {
+      "epoch": 1.4601168637043076,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.087426960184808e-05,
+      "loss": 1.1065,
+      "step": 21490
+    },
+    {
+      "epoch": 1.4604565837749695,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.087214635140644e-05,
+      "loss": 1.1053,
+      "step": 21495
+    },
+    {
+      "epoch": 1.460796303845631,
+      "grad_norm": 6.0,
+      "learning_rate": 4.0870023100964806e-05,
+      "loss": 1.0254,
+      "step": 21500
+    },
+    {
+      "epoch": 1.461136023916293,
+      "grad_norm": 5.75,
+      "learning_rate": 4.086789985052317e-05,
+      "loss": 1.0201,
+      "step": 21505
+    },
+    {
+      "epoch": 1.4614757439869548,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.0865776600081534e-05,
+      "loss": 1.1604,
+      "step": 21510
+    },
+    {
+      "epoch": 1.4618154640576164,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.08636533496399e-05,
+      "loss": 1.0428,
+      "step": 21515
+    },
+    {
+      "epoch": 1.4621551841282783,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.086153009919826e-05,
+      "loss": 1.0556,
+      "step": 21520
+    },
+    {
+      "epoch": 1.4624949041989401,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.0859406848756626e-05,
+      "loss": 1.0699,
+      "step": 21525
+    },
+    {
+      "epoch": 1.4628346242696018,
+      "grad_norm": 5.625,
+      "learning_rate": 4.085728359831499e-05,
+      "loss": 1.1186,
+      "step": 21530
+    },
+    {
+      "epoch": 1.4631743443402636,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.0855160347873354e-05,
+      "loss": 0.9973,
+      "step": 21535
+    },
+    {
+      "epoch": 1.4635140644109255,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.085303709743172e-05,
+      "loss": 1.0787,
+      "step": 21540
+    },
+    {
+      "epoch": 1.463853784481587,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.085091384699008e-05,
+      "loss": 1.0339,
+      "step": 21545
+    },
+    {
+      "epoch": 1.464193504552249,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.0848790596548446e-05,
+      "loss": 1.1066,
+      "step": 21550
+    },
+    {
+      "epoch": 1.4645332246229108,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.084666734610681e-05,
+      "loss": 1.0549,
+      "step": 21555
+    },
+    {
+      "epoch": 1.4648729446935724,
+      "grad_norm": 6.5,
+      "learning_rate": 4.0844544095665174e-05,
+      "loss": 1.1114,
+      "step": 21560
+    },
+    {
+      "epoch": 1.4652126647642343,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.084242084522354e-05,
+      "loss": 1.2065,
+      "step": 21565
+    },
+    {
+      "epoch": 1.4655523848348961,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.08402975947819e-05,
+      "loss": 1.0968,
+      "step": 21570
+    },
+    {
+      "epoch": 1.4658921049055578,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.0838174344340266e-05,
+      "loss": 1.0709,
+      "step": 21575
+    },
+    {
+      "epoch": 1.4662318249762196,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.083605109389863e-05,
+      "loss": 1.1159,
+      "step": 21580
+    },
+    {
+      "epoch": 1.4665715450468815,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.0833927843456994e-05,
+      "loss": 0.9922,
+      "step": 21585
+    },
+    {
+      "epoch": 1.466911265117543,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.083180459301536e-05,
+      "loss": 1.2018,
+      "step": 21590
+    },
+    {
+      "epoch": 1.467250985188205,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.082968134257372e-05,
+      "loss": 1.004,
+      "step": 21595
+    },
+    {
+      "epoch": 1.4675907052588668,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.0827558092132086e-05,
+      "loss": 1.1001,
+      "step": 21600
+    },
+    {
+      "epoch": 1.4679304253295284,
+      "grad_norm": 6.125,
+      "learning_rate": 4.082543484169045e-05,
+      "loss": 1.1371,
+      "step": 21605
+    },
+    {
+      "epoch": 1.4682701454001903,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0823311591248814e-05,
+      "loss": 1.0931,
+      "step": 21610
+    },
+    {
+      "epoch": 1.4686098654708521,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.082118834080718e-05,
+      "loss": 1.122,
+      "step": 21615
+    },
+    {
+      "epoch": 1.4689495855415138,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0819065090365535e-05,
+      "loss": 1.0451,
+      "step": 21620
+    },
+    {
+      "epoch": 1.4692893056121756,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.0816941839923906e-05,
+      "loss": 0.9835,
+      "step": 21625
+    },
+    {
+      "epoch": 1.4696290256828373,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.081481858948227e-05,
+      "loss": 1.1115,
+      "step": 21630
+    },
+    {
+      "epoch": 1.4699687457534991,
+      "grad_norm": 6.375,
+      "learning_rate": 4.0812695339040634e-05,
+      "loss": 1.1199,
+      "step": 21635
+    },
+    {
+      "epoch": 1.470308465824161,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.0810572088599e-05,
+      "loss": 1.0776,
+      "step": 21640
+    },
+    {
+      "epoch": 1.4706481858948226,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.080844883815736e-05,
+      "loss": 1.0776,
+      "step": 21645
+    },
+    {
+      "epoch": 1.4709879059654845,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.0806325587715726e-05,
+      "loss": 1.0067,
+      "step": 21650
+    },
+    {
+      "epoch": 1.471327626036146,
+      "grad_norm": 5.375,
+      "learning_rate": 4.080420233727409e-05,
+      "loss": 1.0345,
+      "step": 21655
+    },
+    {
+      "epoch": 1.471667346106808,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0802079086832454e-05,
+      "loss": 1.2108,
+      "step": 21660
+    },
+    {
+      "epoch": 1.4720070661774698,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.079995583639082e-05,
+      "loss": 1.0844,
+      "step": 21665
+    },
+    {
+      "epoch": 1.4723467862481314,
+      "grad_norm": 5.875,
+      "learning_rate": 4.079783258594918e-05,
+      "loss": 1.0681,
+      "step": 21670
+    },
+    {
+      "epoch": 1.4726865063187933,
+      "grad_norm": 6.125,
+      "learning_rate": 4.0795709335507546e-05,
+      "loss": 1.0487,
+      "step": 21675
+    },
+    {
+      "epoch": 1.4730262263894551,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.079358608506591e-05,
+      "loss": 1.0422,
+      "step": 21680
+    },
+    {
+      "epoch": 1.4733659464601168,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0791462834624274e-05,
+      "loss": 1.0405,
+      "step": 21685
+    },
+    {
+      "epoch": 1.4737056665307786,
+      "grad_norm": 5.625,
+      "learning_rate": 4.078933958418263e-05,
+      "loss": 1.0731,
+      "step": 21690
+    },
+    {
+      "epoch": 1.4740453866014405,
+      "grad_norm": 5.5,
+      "learning_rate": 4.0787216333741e-05,
+      "loss": 1.0149,
+      "step": 21695
+    },
+    {
+      "epoch": 1.474385106672102,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.0785093083299366e-05,
+      "loss": 1.1252,
+      "step": 21700
+    },
+    {
+      "epoch": 1.474724826742764,
+      "grad_norm": 6.25,
+      "learning_rate": 4.078296983285772e-05,
+      "loss": 1.0468,
+      "step": 21705
+    },
+    {
+      "epoch": 1.4750645468134258,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.0780846582416094e-05,
+      "loss": 1.0785,
+      "step": 21710
+    },
+    {
+      "epoch": 1.4754042668840874,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.077872333197446e-05,
+      "loss": 1.2088,
+      "step": 21715
+    },
+    {
+      "epoch": 1.4757439869547493,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0776600081532815e-05,
+      "loss": 1.0286,
+      "step": 21720
+    },
+    {
+      "epoch": 1.4760837070254111,
+      "grad_norm": 6.375,
+      "learning_rate": 4.0774476831091186e-05,
+      "loss": 1.021,
+      "step": 21725
+    },
+    {
+      "epoch": 1.4764234270960728,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.077235358064955e-05,
+      "loss": 1.0509,
+      "step": 21730
+    },
+    {
+      "epoch": 1.4767631471667346,
+      "grad_norm": 4.28125,
+      "learning_rate": 4.077023033020791e-05,
+      "loss": 1.0082,
+      "step": 21735
+    },
+    {
+      "epoch": 1.4771028672373965,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.076810707976628e-05,
+      "loss": 1.0348,
+      "step": 21740
+    },
+    {
+      "epoch": 1.477442587308058,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.076598382932464e-05,
+      "loss": 1.1271,
+      "step": 21745
+    },
+    {
+      "epoch": 1.47778230737872,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.0763860578883e-05,
+      "loss": 1.1312,
+      "step": 21750
+    },
+    {
+      "epoch": 1.4781220274493818,
+      "grad_norm": 5.5,
+      "learning_rate": 4.076173732844137e-05,
+      "loss": 1.0934,
+      "step": 21755
+    },
+    {
+      "epoch": 1.4784617475200434,
+      "grad_norm": 5.875,
+      "learning_rate": 4.075961407799973e-05,
+      "loss": 1.0283,
+      "step": 21760
+    },
+    {
+      "epoch": 1.4788014675907053,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.075749082755809e-05,
+      "loss": 1.1653,
+      "step": 21765
+    },
+    {
+      "epoch": 1.4791411876613672,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.075536757711646e-05,
+      "loss": 1.2081,
+      "step": 21770
+    },
+    {
+      "epoch": 1.4794809077320288,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.075324432667482e-05,
+      "loss": 1.0555,
+      "step": 21775
+    },
+    {
+      "epoch": 1.4798206278026906,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.075112107623318e-05,
+      "loss": 1.0307,
+      "step": 21780
+    },
+    {
+      "epoch": 1.4801603478733525,
+      "grad_norm": 5.625,
+      "learning_rate": 4.0748997825791554e-05,
+      "loss": 1.0584,
+      "step": 21785
+    },
+    {
+      "epoch": 1.4805000679440141,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.074687457534991e-05,
+      "loss": 1.0051,
+      "step": 21790
+    },
+    {
+      "epoch": 1.480839788014676,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0744751324908275e-05,
+      "loss": 1.0444,
+      "step": 21795
+    },
+    {
+      "epoch": 1.4811795080853376,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0742628074466646e-05,
+      "loss": 1.0328,
+      "step": 21800
+    },
+    {
+      "epoch": 1.4815192281559995,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.0740504824025e-05,
+      "loss": 1.1915,
+      "step": 21805
+    },
+    {
+      "epoch": 1.4818589482266613,
+      "grad_norm": 5.625,
+      "learning_rate": 4.073838157358337e-05,
+      "loss": 1.2146,
+      "step": 21810
+    },
+    {
+      "epoch": 1.482198668297323,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.073625832314174e-05,
+      "loss": 1.1019,
+      "step": 21815
+    },
+    {
+      "epoch": 1.4825383883679848,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0734135072700095e-05,
+      "loss": 1.047,
+      "step": 21820
+    },
+    {
+      "epoch": 1.4828781084386464,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.073201182225846e-05,
+      "loss": 1.0448,
+      "step": 21825
+    },
+    {
+      "epoch": 1.4832178285093083,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.072988857181683e-05,
+      "loss": 1.0475,
+      "step": 21830
+    },
+    {
+      "epoch": 1.4835575485799701,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.072776532137519e-05,
+      "loss": 1.0327,
+      "step": 21835
+    },
+    {
+      "epoch": 1.4838972686506318,
+      "grad_norm": 5.5,
+      "learning_rate": 4.072564207093355e-05,
+      "loss": 1.0514,
+      "step": 21840
+    },
+    {
+      "epoch": 1.4842369887212936,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.0723518820491915e-05,
+      "loss": 0.9638,
+      "step": 21845
+    },
+    {
+      "epoch": 1.4845767087919555,
+      "grad_norm": 5.625,
+      "learning_rate": 4.072139557005028e-05,
+      "loss": 0.9578,
+      "step": 21850
+    },
+    {
+      "epoch": 1.484916428862617,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.071927231960864e-05,
+      "loss": 1.1156,
+      "step": 21855
+    },
+    {
+      "epoch": 1.485256148933279,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.071714906916701e-05,
+      "loss": 1.0635,
+      "step": 21860
+    },
+    {
+      "epoch": 1.4855958690039408,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.071502581872537e-05,
+      "loss": 1.2062,
+      "step": 21865
+    },
+    {
+      "epoch": 1.4859355890746024,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0712902568283735e-05,
+      "loss": 1.0455,
+      "step": 21870
+    },
+    {
+      "epoch": 1.4862753091452643,
+      "grad_norm": 5.375,
+      "learning_rate": 4.07107793178421e-05,
+      "loss": 1.1162,
+      "step": 21875
+    },
+    {
+      "epoch": 1.4866150292159261,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.070865606740046e-05,
+      "loss": 1.0588,
+      "step": 21880
+    },
+    {
+      "epoch": 1.4869547492865878,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.070653281695883e-05,
+      "loss": 1.0079,
+      "step": 21885
+    },
+    {
+      "epoch": 1.4872944693572496,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.070440956651719e-05,
+      "loss": 1.0638,
+      "step": 21890
+    },
+    {
+      "epoch": 1.4876341894279115,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.0702286316075555e-05,
+      "loss": 1.1191,
+      "step": 21895
+    },
+    {
+      "epoch": 1.487973909498573,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.070016306563392e-05,
+      "loss": 0.9871,
+      "step": 21900
+    },
+    {
+      "epoch": 1.488313629569235,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.069803981519228e-05,
+      "loss": 0.9769,
+      "step": 21905
+    },
+    {
+      "epoch": 1.4886533496398968,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.069591656475065e-05,
+      "loss": 0.9809,
+      "step": 21910
+    },
+    {
+      "epoch": 1.4889930697105584,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.069379331430901e-05,
+      "loss": 1.0456,
+      "step": 21915
+    },
+    {
+      "epoch": 1.4893327897812203,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.0691670063867375e-05,
+      "loss": 0.9801,
+      "step": 21920
+    },
+    {
+      "epoch": 1.4896725098518822,
+      "grad_norm": 6.71875,
+      "learning_rate": 4.068954681342574e-05,
+      "loss": 1.1298,
+      "step": 21925
+    },
+    {
+      "epoch": 1.4900122299225438,
+      "grad_norm": 5.75,
+      "learning_rate": 4.06874235629841e-05,
+      "loss": 0.9658,
+      "step": 21930
+    },
+    {
+      "epoch": 1.4903519499932056,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.068530031254247e-05,
+      "loss": 1.1428,
+      "step": 21935
+    },
+    {
+      "epoch": 1.4906916700638675,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.068317706210083e-05,
+      "loss": 0.9696,
+      "step": 21940
+    },
+    {
+      "epoch": 1.4910313901345291,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0681053811659195e-05,
+      "loss": 1.0177,
+      "step": 21945
+    },
+    {
+      "epoch": 1.491371110205191,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.067893056121756e-05,
+      "loss": 1.0979,
+      "step": 21950
+    },
+    {
+      "epoch": 1.4917108302758528,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.067680731077592e-05,
+      "loss": 0.9467,
+      "step": 21955
+    },
+    {
+      "epoch": 1.4920505503465145,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.067468406033428e-05,
+      "loss": 1.0915,
+      "step": 21960
+    },
+    {
+      "epoch": 1.4923902704171763,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.067256080989265e-05,
+      "loss": 1.1188,
+      "step": 21965
+    },
+    {
+      "epoch": 1.492729990487838,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.0670437559451015e-05,
+      "loss": 1.0164,
+      "step": 21970
+    },
+    {
+      "epoch": 1.4930697105584998,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.066831430900938e-05,
+      "loss": 1.1398,
+      "step": 21975
+    },
+    {
+      "epoch": 1.4934094306291616,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.066619105856774e-05,
+      "loss": 0.9905,
+      "step": 21980
+    },
+    {
+      "epoch": 1.4937491506998233,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.066406780812611e-05,
+      "loss": 1.0183,
+      "step": 21985
+    },
+    {
+      "epoch": 1.4940888707704851,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.066194455768447e-05,
+      "loss": 1.0264,
+      "step": 21990
+    },
+    {
+      "epoch": 1.4944285908411468,
+      "grad_norm": 5.09375,
+      "learning_rate": 4.0659821307242835e-05,
+      "loss": 0.9916,
+      "step": 21995
+    },
+    {
+      "epoch": 1.4947683109118086,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.06576980568012e-05,
+      "loss": 1.014,
+      "step": 22000
+    },
+    {
+      "epoch": 1.4951080309824705,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.065557480635956e-05,
+      "loss": 1.0928,
+      "step": 22005
+    },
+    {
+      "epoch": 1.495447751053132,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.065345155591793e-05,
+      "loss": 1.0351,
+      "step": 22010
+    },
+    {
+      "epoch": 1.495787471123794,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.065132830547629e-05,
+      "loss": 1.0631,
+      "step": 22015
+    },
+    {
+      "epoch": 1.4961271911944558,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.0649205055034655e-05,
+      "loss": 1.0662,
+      "step": 22020
+    },
+    {
+      "epoch": 1.4964669112651174,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.064708180459302e-05,
+      "loss": 1.0721,
+      "step": 22025
+    },
+    {
+      "epoch": 1.4968066313357793,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.064495855415138e-05,
+      "loss": 0.9444,
+      "step": 22030
+    },
+    {
+      "epoch": 1.4971463514064411,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.064283530370975e-05,
+      "loss": 1.0977,
+      "step": 22035
+    },
+    {
+      "epoch": 1.4974860714771028,
+      "grad_norm": 6.96875,
+      "learning_rate": 4.064071205326811e-05,
+      "loss": 1.1011,
+      "step": 22040
+    },
+    {
+      "epoch": 1.4978257915477646,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.063858880282647e-05,
+      "loss": 1.0566,
+      "step": 22045
+    },
+    {
+      "epoch": 1.4981655116184265,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.063646555238484e-05,
+      "loss": 0.9884,
+      "step": 22050
+    },
+    {
+      "epoch": 1.4985052316890881,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.06343423019432e-05,
+      "loss": 1.0347,
+      "step": 22055
+    },
+    {
+      "epoch": 1.49884495175975,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.063221905150156e-05,
+      "loss": 1.1012,
+      "step": 22060
+    },
+    {
+      "epoch": 1.4991846718304118,
+      "grad_norm": 6.25,
+      "learning_rate": 4.063009580105993e-05,
+      "loss": 1.1481,
+      "step": 22065
+    },
+    {
+      "epoch": 1.4995243919010735,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0627972550618295e-05,
+      "loss": 1.0622,
+      "step": 22070
+    },
+    {
+      "epoch": 1.4998641119717353,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.062584930017665e-05,
+      "loss": 1.0034,
+      "step": 22075
+    },
+    {
+      "epoch": 1.5002038320423972,
+      "grad_norm": 5.5,
+      "learning_rate": 4.062372604973502e-05,
+      "loss": 0.9799,
+      "step": 22080
+    },
+    {
+      "epoch": 1.5005435521130588,
+      "grad_norm": 6.0,
+      "learning_rate": 4.062160279929339e-05,
+      "loss": 1.0717,
+      "step": 22085
+    },
+    {
+      "epoch": 1.5008832721837206,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.0619479548851745e-05,
+      "loss": 1.0501,
+      "step": 22090
+    },
+    {
+      "epoch": 1.5012229922543825,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.0617356298410115e-05,
+      "loss": 1.0359,
+      "step": 22095
+    },
+    {
+      "epoch": 1.5015627123250441,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.061523304796848e-05,
+      "loss": 1.0306,
+      "step": 22100
+    },
+    {
+      "epoch": 1.501902432395706,
+      "grad_norm": 5.625,
+      "learning_rate": 4.0613109797526837e-05,
+      "loss": 1.0629,
+      "step": 22105
+    },
+    {
+      "epoch": 1.5022421524663678,
+      "grad_norm": 6.0,
+      "learning_rate": 4.061098654708521e-05,
+      "loss": 1.1261,
+      "step": 22110
+    },
+    {
+      "epoch": 1.5025818725370295,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.0608863296643565e-05,
+      "loss": 1.0338,
+      "step": 22115
+    },
+    {
+      "epoch": 1.5029215926076913,
+      "grad_norm": 6.0,
+      "learning_rate": 4.060674004620193e-05,
+      "loss": 1.0526,
+      "step": 22120
+    },
+    {
+      "epoch": 1.5032613126783532,
+      "grad_norm": 5.625,
+      "learning_rate": 4.06046167957603e-05,
+      "loss": 1.0981,
+      "step": 22125
+    },
+    {
+      "epoch": 1.5036010327490148,
+      "grad_norm": 5.125,
+      "learning_rate": 4.0602493545318657e-05,
+      "loss": 1.0729,
+      "step": 22130
+    },
+    {
+      "epoch": 1.5039407528196764,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.060037029487702e-05,
+      "loss": 1.0872,
+      "step": 22135
+    },
+    {
+      "epoch": 1.5042804728903385,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.059824704443539e-05,
+      "loss": 1.0102,
+      "step": 22140
+    },
+    {
+      "epoch": 1.5046201929610001,
+      "grad_norm": 4.96875,
+      "learning_rate": 4.059612379399375e-05,
+      "loss": 1.0704,
+      "step": 22145
+    },
+    {
+      "epoch": 1.5049599130316618,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.059400054355211e-05,
+      "loss": 1.0917,
+      "step": 22150
+    },
+    {
+      "epoch": 1.5052996331023238,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.059187729311048e-05,
+      "loss": 1.0267,
+      "step": 22155
+    },
+    {
+      "epoch": 1.5056393531729855,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.058975404266884e-05,
+      "loss": 1.0848,
+      "step": 22160
+    },
+    {
+      "epoch": 1.505979073243647,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.0587630792227205e-05,
+      "loss": 1.0981,
+      "step": 22165
+    },
+    {
+      "epoch": 1.506318793314309,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.0585507541785575e-05,
+      "loss": 1.106,
+      "step": 22170
+    },
+    {
+      "epoch": 1.5066585133849708,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.058338429134393e-05,
+      "loss": 1.02,
+      "step": 22175
+    },
+    {
+      "epoch": 1.5069982334556324,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.05812610409023e-05,
+      "loss": 1.0709,
+      "step": 22180
+    },
+    {
+      "epoch": 1.5073379535262943,
+      "grad_norm": 5.3125,
+      "learning_rate": 4.057913779046066e-05,
+      "loss": 1.0862,
+      "step": 22185
+    },
+    {
+      "epoch": 1.5076776735969561,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.0577014540019025e-05,
+      "loss": 0.9809,
+      "step": 22190
+    },
+    {
+      "epoch": 1.5080173936676178,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.057489128957739e-05,
+      "loss": 1.1324,
+      "step": 22195
+    },
+    {
+      "epoch": 1.5083571137382796,
+      "grad_norm": 6.125,
+      "learning_rate": 4.057276803913575e-05,
+      "loss": 1.1564,
+      "step": 22200
+    },
+    {
+      "epoch": 1.5086968338089415,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.057064478869412e-05,
+      "loss": 0.9468,
+      "step": 22205
+    },
+    {
+      "epoch": 1.5090365538796031,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.056852153825248e-05,
+      "loss": 0.9587,
+      "step": 22210
+    },
+    {
+      "epoch": 1.509376273950265,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0566398287810845e-05,
+      "loss": 1.0476,
+      "step": 22215
+    },
+    {
+      "epoch": 1.5097159940209268,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.056427503736921e-05,
+      "loss": 1.0149,
+      "step": 22220
+    },
+    {
+      "epoch": 1.5100557140915885,
+      "grad_norm": 6.125,
+      "learning_rate": 4.056215178692757e-05,
+      "loss": 1.0557,
+      "step": 22225
+    },
+    {
+      "epoch": 1.5103954341622503,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.056002853648594e-05,
+      "loss": 1.0208,
+      "step": 22230
+    },
+    {
+      "epoch": 1.5107351542329122,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.05579052860443e-05,
+      "loss": 1.0026,
+      "step": 22235
+    },
+    {
+      "epoch": 1.5110748743035738,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0555782035602665e-05,
+      "loss": 1.1202,
+      "step": 22240
+    },
+    {
+      "epoch": 1.5114145943742356,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.055365878516103e-05,
+      "loss": 0.971,
+      "step": 22245
+    },
+    {
+      "epoch": 1.5117543144448975,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.055153553471939e-05,
+      "loss": 1.0635,
+      "step": 22250
+    },
+    {
+      "epoch": 1.5120940345155591,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.054941228427776e-05,
+      "loss": 1.1372,
+      "step": 22255
+    },
+    {
+      "epoch": 1.512433754586221,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.054728903383612e-05,
+      "loss": 0.9957,
+      "step": 22260
+    },
+    {
+      "epoch": 1.5127734746568828,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.0545165783394485e-05,
+      "loss": 1.093,
+      "step": 22265
+    },
+    {
+      "epoch": 1.5131131947275445,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.054304253295285e-05,
+      "loss": 0.9984,
+      "step": 22270
+    },
+    {
+      "epoch": 1.5134529147982063,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.054091928251121e-05,
+      "loss": 1.0232,
+      "step": 22275
+    },
+    {
+      "epoch": 1.5137926348688682,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.053879603206958e-05,
+      "loss": 0.9818,
+      "step": 22280
+    },
+    {
+      "epoch": 1.5141323549395298,
+      "grad_norm": 6.25,
+      "learning_rate": 4.053667278162794e-05,
+      "loss": 1.1398,
+      "step": 22285
+    },
+    {
+      "epoch": 1.5144720750101917,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.0534549531186305e-05,
+      "loss": 1.0717,
+      "step": 22290
+    },
+    {
+      "epoch": 1.5148117950808535,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.053242628074467e-05,
+      "loss": 1.0823,
+      "step": 22295
+    },
+    {
+      "epoch": 1.5151515151515151,
+      "grad_norm": 6.125,
+      "learning_rate": 4.053030303030303e-05,
+      "loss": 1.0236,
+      "step": 22300
+    },
+    {
+      "epoch": 1.5154912352221768,
+      "grad_norm": 6.125,
+      "learning_rate": 4.05281797798614e-05,
+      "loss": 1.167,
+      "step": 22305
+    },
+    {
+      "epoch": 1.5158309552928388,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.052605652941976e-05,
+      "loss": 1.0267,
+      "step": 22310
+    },
+    {
+      "epoch": 1.5161706753635005,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.052393327897812e-05,
+      "loss": 1.093,
+      "step": 22315
+    },
+    {
+      "epoch": 1.516510395434162,
+      "grad_norm": 6.125,
+      "learning_rate": 4.052181002853649e-05,
+      "loss": 1.1339,
+      "step": 22320
+    },
+    {
+      "epoch": 1.5168501155048242,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.051968677809485e-05,
+      "loss": 1.1305,
+      "step": 22325
+    },
+    {
+      "epoch": 1.5171898355754858,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.051756352765322e-05,
+      "loss": 1.0901,
+      "step": 22330
+    },
+    {
+      "epoch": 1.5175295556461474,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.051544027721158e-05,
+      "loss": 1.1524,
+      "step": 22335
+    },
+    {
+      "epoch": 1.5178692757168093,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.0513317026769945e-05,
+      "loss": 1.0639,
+      "step": 22340
+    },
+    {
+      "epoch": 1.5182089957874711,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.051119377632831e-05,
+      "loss": 1.0964,
+      "step": 22345
+    },
+    {
+      "epoch": 1.5185487158581328,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.050907052588667e-05,
+      "loss": 1.0696,
+      "step": 22350
+    },
+    {
+      "epoch": 1.5188884359287946,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.050694727544504e-05,
+      "loss": 1.0922,
+      "step": 22355
+    },
+    {
+      "epoch": 1.5192281559994565,
+      "grad_norm": 5.25,
+      "learning_rate": 4.05048240250034e-05,
+      "loss": 1.0913,
+      "step": 22360
+    },
+    {
+      "epoch": 1.5195678760701181,
+      "grad_norm": 6.25,
+      "learning_rate": 4.0502700774561765e-05,
+      "loss": 1.0492,
+      "step": 22365
+    },
+    {
+      "epoch": 1.51990759614078,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.050057752412013e-05,
+      "loss": 0.9538,
+      "step": 22370
+    },
+    {
+      "epoch": 1.5202473162114418,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.049845427367849e-05,
+      "loss": 1.066,
+      "step": 22375
+    },
+    {
+      "epoch": 1.5205870362821035,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.049633102323686e-05,
+      "loss": 1.104,
+      "step": 22380
+    },
+    {
+      "epoch": 1.5209267563527653,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0494207772795214e-05,
+      "loss": 1.019,
+      "step": 22385
+    },
+    {
+      "epoch": 1.5212664764234272,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.0492084522353585e-05,
+      "loss": 0.9923,
+      "step": 22390
+    },
+    {
+      "epoch": 1.5216061964940888,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.048996127191195e-05,
+      "loss": 1.0706,
+      "step": 22395
+    },
+    {
+      "epoch": 1.5219459165647506,
+      "grad_norm": 6.25,
+      "learning_rate": 4.0487838021470306e-05,
+      "loss": 1.182,
+      "step": 22400
+    },
+    {
+      "epoch": 1.5222856366354125,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.048571477102868e-05,
+      "loss": 1.065,
+      "step": 22405
+    },
+    {
+      "epoch": 1.5226253567060741,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.048359152058704e-05,
+      "loss": 1.0583,
+      "step": 22410
+    },
+    {
+      "epoch": 1.522965076776736,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.04814682701454e-05,
+      "loss": 1.1055,
+      "step": 22415
+    },
+    {
+      "epoch": 1.5233047968473978,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.047934501970377e-05,
+      "loss": 1.0913,
+      "step": 22420
+    },
+    {
+      "epoch": 1.5236445169180595,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.047722176926213e-05,
+      "loss": 1.1327,
+      "step": 22425
+    },
+    {
+      "epoch": 1.5239842369887213,
+      "grad_norm": 5.75,
+      "learning_rate": 4.047509851882049e-05,
+      "loss": 0.9872,
+      "step": 22430
+    },
+    {
+      "epoch": 1.5243239570593832,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.047297526837886e-05,
+      "loss": 1.0722,
+      "step": 22435
+    },
+    {
+      "epoch": 1.5246636771300448,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.0470852017937225e-05,
+      "loss": 1.1019,
+      "step": 22440
+    },
+    {
+      "epoch": 1.5250033972007067,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.046872876749558e-05,
+      "loss": 1.0639,
+      "step": 22445
+    },
+    {
+      "epoch": 1.5253431172713685,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.046660551705395e-05,
+      "loss": 0.999,
+      "step": 22450
+    },
+    {
+      "epoch": 1.5256828373420301,
+      "grad_norm": 5.875,
+      "learning_rate": 4.046448226661232e-05,
+      "loss": 1.0966,
+      "step": 22455
+    },
+    {
+      "epoch": 1.526022557412692,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.0462359016170674e-05,
+      "loss": 1.1033,
+      "step": 22460
+    },
+    {
+      "epoch": 1.5263622774833538,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0460235765729045e-05,
+      "loss": 1.1605,
+      "step": 22465
+    },
+    {
+      "epoch": 1.5267019975540155,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.04581125152874e-05,
+      "loss": 1.0973,
+      "step": 22470
+    },
+    {
+      "epoch": 1.527041717624677,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.0455989264845766e-05,
+      "loss": 1.1132,
+      "step": 22475
+    },
+    {
+      "epoch": 1.5273814376953392,
+      "grad_norm": 4.90625,
+      "learning_rate": 4.045386601440414e-05,
+      "loss": 0.9988,
+      "step": 22480
+    },
+    {
+      "epoch": 1.5277211577660008,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.0451742763962494e-05,
+      "loss": 1.093,
+      "step": 22485
+    },
+    {
+      "epoch": 1.5280608778366624,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.044961951352086e-05,
+      "loss": 1.087,
+      "step": 22490
+    },
+    {
+      "epoch": 1.5284005979073245,
+      "grad_norm": 5.625,
+      "learning_rate": 4.044749626307923e-05,
+      "loss": 1.0064,
+      "step": 22495
+    },
+    {
+      "epoch": 1.5287403179779862,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.0445373012637586e-05,
+      "loss": 1.0473,
+      "step": 22500
+    },
+    {
+      "epoch": 1.5290800380486478,
+      "grad_norm": 6.25,
+      "learning_rate": 4.044324976219595e-05,
+      "loss": 1.048,
+      "step": 22505
+    },
+    {
+      "epoch": 1.5294197581193096,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.044112651175432e-05,
+      "loss": 1.0282,
+      "step": 22510
+    },
+    {
+      "epoch": 1.5297594781899715,
+      "grad_norm": 6.125,
+      "learning_rate": 4.043900326131268e-05,
+      "loss": 1.0696,
+      "step": 22515
+    },
+    {
+      "epoch": 1.5300991982606331,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.043688001087104e-05,
+      "loss": 1.0341,
+      "step": 22520
+    },
+    {
+      "epoch": 1.530438918331295,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.043475676042941e-05,
+      "loss": 1.1095,
+      "step": 22525
+    },
+    {
+      "epoch": 1.5307786384019568,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.043263350998777e-05,
+      "loss": 1.0642,
+      "step": 22530
+    },
+    {
+      "epoch": 1.5311183584726185,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.0430510259546134e-05,
+      "loss": 0.998,
+      "step": 22535
+    },
+    {
+      "epoch": 1.5314580785432803,
+      "grad_norm": 5.625,
+      "learning_rate": 4.04283870091045e-05,
+      "loss": 1.0276,
+      "step": 22540
+    },
+    {
+      "epoch": 1.5317977986139422,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.042626375866286e-05,
+      "loss": 1.0605,
+      "step": 22545
+    },
+    {
+      "epoch": 1.5321375186846038,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.0424140508221226e-05,
+      "loss": 1.12,
+      "step": 22550
+    },
+    {
+      "epoch": 1.5324772387552656,
+      "grad_norm": 5.625,
+      "learning_rate": 4.042201725777959e-05,
+      "loss": 1.1165,
+      "step": 22555
+    },
+    {
+      "epoch": 1.5328169588259275,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.0419894007337954e-05,
+      "loss": 1.0506,
+      "step": 22560
+    },
+    {
+      "epoch": 1.5331566788965891,
+      "grad_norm": 6.125,
+      "learning_rate": 4.041777075689632e-05,
+      "loss": 1.1184,
+      "step": 22565
+    },
+    {
+      "epoch": 1.533496398967251,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.041564750645468e-05,
+      "loss": 1.1385,
+      "step": 22570
+    },
+    {
+      "epoch": 1.5338361190379128,
+      "grad_norm": 5.34375,
+      "learning_rate": 4.0413524256013046e-05,
+      "loss": 1.0253,
+      "step": 22575
+    },
+    {
+      "epoch": 1.5341758391085745,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.041140100557141e-05,
+      "loss": 1.1151,
+      "step": 22580
+    },
+    {
+      "epoch": 1.5345155591792363,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.0409277755129774e-05,
+      "loss": 1.0705,
+      "step": 22585
+    },
+    {
+      "epoch": 1.5348552792498982,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.040715450468814e-05,
+      "loss": 1.1367,
+      "step": 22590
+    },
+    {
+      "epoch": 1.5351949993205598,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.04050312542465e-05,
+      "loss": 1.0538,
+      "step": 22595
+    },
+    {
+      "epoch": 1.5355347193912217,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.0402908003804866e-05,
+      "loss": 1.0696,
+      "step": 22600
+    },
+    {
+      "epoch": 1.5358744394618835,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.040078475336323e-05,
+      "loss": 1.0508,
+      "step": 22605
+    },
+    {
+      "epoch": 1.5362141595325451,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.0398661502921594e-05,
+      "loss": 0.9696,
+      "step": 22610
+    },
+    {
+      "epoch": 1.536553879603207,
+      "grad_norm": 5.875,
+      "learning_rate": 4.039653825247996e-05,
+      "loss": 1.1889,
+      "step": 22615
+    },
+    {
+      "epoch": 1.5368935996738688,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.039441500203832e-05,
+      "loss": 1.0919,
+      "step": 22620
+    },
+    {
+      "epoch": 1.5372333197445305,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.0392291751596686e-05,
+      "loss": 1.0846,
+      "step": 22625
+    },
+    {
+      "epoch": 1.5375730398151923,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.039016850115505e-05,
+      "loss": 1.0378,
+      "step": 22630
+    },
+    {
+      "epoch": 1.5379127598858542,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.0388045250713414e-05,
+      "loss": 1.1387,
+      "step": 22635
+    },
+    {
+      "epoch": 1.5382524799565158,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.038592200027178e-05,
+      "loss": 1.125,
+      "step": 22640
+    },
+    {
+      "epoch": 1.5385922000271774,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.038379874983014e-05,
+      "loss": 1.02,
+      "step": 22645
+    },
+    {
+      "epoch": 1.5389319200978395,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.0381675499388506e-05,
+      "loss": 1.076,
+      "step": 22650
+    },
+    {
+      "epoch": 1.5392716401685012,
+      "grad_norm": 5.875,
+      "learning_rate": 4.037955224894687e-05,
+      "loss": 1.0211,
+      "step": 22655
+    },
+    {
+      "epoch": 1.5396113602391628,
+      "grad_norm": 6.5,
+      "learning_rate": 4.0377428998505234e-05,
+      "loss": 1.0905,
+      "step": 22660
+    },
+    {
+      "epoch": 1.5399510803098249,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.03753057480636e-05,
+      "loss": 0.9308,
+      "step": 22665
+    },
+    {
+      "epoch": 1.5402908003804865,
+      "grad_norm": 5.75,
+      "learning_rate": 4.037318249762196e-05,
+      "loss": 1.0431,
+      "step": 22670
+    },
+    {
+      "epoch": 1.5406305204511481,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0371059247180326e-05,
+      "loss": 0.996,
+      "step": 22675
+    },
+    {
+      "epoch": 1.54097024052181,
+      "grad_norm": 6.0,
+      "learning_rate": 4.036893599673869e-05,
+      "loss": 0.958,
+      "step": 22680
+    },
+    {
+      "epoch": 1.5413099605924718,
+      "grad_norm": 6.0,
+      "learning_rate": 4.0366812746297054e-05,
+      "loss": 1.1046,
+      "step": 22685
+    },
+    {
+      "epoch": 1.5416496806631335,
+      "grad_norm": 6.75,
+      "learning_rate": 4.036468949585542e-05,
+      "loss": 1.0494,
+      "step": 22690
+    },
+    {
+      "epoch": 1.5419894007337953,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.036256624541378e-05,
+      "loss": 1.1065,
+      "step": 22695
+    },
+    {
+      "epoch": 1.5423291208044572,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.0360442994972146e-05,
+      "loss": 1.1474,
+      "step": 22700
+    },
+    {
+      "epoch": 1.5426688408751188,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.035831974453051e-05,
+      "loss": 1.0459,
+      "step": 22705
+    },
+    {
+      "epoch": 1.5430085609457806,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0356196494088874e-05,
+      "loss": 1.0529,
+      "step": 22710
+    },
+    {
+      "epoch": 1.5433482810164425,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.035407324364724e-05,
+      "loss": 1.0768,
+      "step": 22715
+    },
+    {
+      "epoch": 1.5436880010871041,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.03519499932056e-05,
+      "loss": 1.0483,
+      "step": 22720
+    },
+    {
+      "epoch": 1.544027721157766,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.0349826742763966e-05,
+      "loss": 1.0593,
+      "step": 22725
+    },
+    {
+      "epoch": 1.5443674412284278,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.034770349232233e-05,
+      "loss": 1.183,
+      "step": 22730
+    },
+    {
+      "epoch": 1.5447071612990895,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.0345580241880694e-05,
+      "loss": 1.13,
+      "step": 22735
+    },
+    {
+      "epoch": 1.5450468813697513,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.034345699143905e-05,
+      "loss": 1.0836,
+      "step": 22740
+    },
+    {
+      "epoch": 1.5453866014404132,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.034133374099742e-05,
+      "loss": 1.0984,
+      "step": 22745
+    },
+    {
+      "epoch": 1.5457263215110748,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.0339210490555786e-05,
+      "loss": 1.0366,
+      "step": 22750
+    },
+    {
+      "epoch": 1.5460660415817367,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.0337087240114144e-05,
+      "loss": 1.1933,
+      "step": 22755
+    },
+    {
+      "epoch": 1.5464057616523985,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.0334963989672514e-05,
+      "loss": 1.0479,
+      "step": 22760
+    },
+    {
+      "epoch": 1.5467454817230601,
+      "grad_norm": 6.625,
+      "learning_rate": 4.033284073923088e-05,
+      "loss": 1.0802,
+      "step": 22765
+    },
+    {
+      "epoch": 1.547085201793722,
+      "grad_norm": 5.25,
+      "learning_rate": 4.0330717488789236e-05,
+      "loss": 1.0147,
+      "step": 22770
+    },
+    {
+      "epoch": 1.5474249218643839,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.0328594238347606e-05,
+      "loss": 0.9643,
+      "step": 22775
+    },
+    {
+      "epoch": 1.5477646419350455,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.032647098790597e-05,
+      "loss": 1.0696,
+      "step": 22780
+    },
+    {
+      "epoch": 1.5481043620057073,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.032434773746433e-05,
+      "loss": 0.9584,
+      "step": 22785
+    },
+    {
+      "epoch": 1.5484440820763692,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.03222244870227e-05,
+      "loss": 1.1128,
+      "step": 22790
+    },
+    {
+      "epoch": 1.5487838021470308,
+      "grad_norm": 6.875,
+      "learning_rate": 4.032010123658106e-05,
+      "loss": 1.1715,
+      "step": 22795
+    },
+    {
+      "epoch": 1.5491235222176927,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.031797798613942e-05,
+      "loss": 0.9879,
+      "step": 22800
+    },
+    {
+      "epoch": 1.5494632422883545,
+      "grad_norm": 6.375,
+      "learning_rate": 4.031585473569779e-05,
+      "loss": 1.0054,
+      "step": 22805
+    },
+    {
+      "epoch": 1.5498029623590162,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.0313731485256154e-05,
+      "loss": 1.1095,
+      "step": 22810
+    },
+    {
+      "epoch": 1.5501426824296778,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.031160823481451e-05,
+      "loss": 1.0512,
+      "step": 22815
+    },
+    {
+      "epoch": 1.5504824025003399,
+      "grad_norm": 5.875,
+      "learning_rate": 4.030948498437288e-05,
+      "loss": 1.1007,
+      "step": 22820
+    },
+    {
+      "epoch": 1.5508221225710015,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.030736173393124e-05,
+      "loss": 1.0776,
+      "step": 22825
+    },
+    {
+      "epoch": 1.5511618426416631,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0305238483489604e-05,
+      "loss": 1.0544,
+      "step": 22830
+    },
+    {
+      "epoch": 1.5515015627123252,
+      "grad_norm": 5.75,
+      "learning_rate": 4.0303115233047974e-05,
+      "loss": 1.0435,
+      "step": 22835
+    },
+    {
+      "epoch": 1.5518412827829868,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.030099198260633e-05,
+      "loss": 0.995,
+      "step": 22840
+    },
+    {
+      "epoch": 1.5521810028536485,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.0298868732164696e-05,
+      "loss": 1.0983,
+      "step": 22845
+    },
+    {
+      "epoch": 1.5525207229243103,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.0296745481723066e-05,
+      "loss": 1.0437,
+      "step": 22850
+    },
+    {
+      "epoch": 1.5528604429949722,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0294622231281424e-05,
+      "loss": 0.9732,
+      "step": 22855
+    },
+    {
+      "epoch": 1.5532001630656338,
+      "grad_norm": 5.875,
+      "learning_rate": 4.029249898083979e-05,
+      "loss": 1.0963,
+      "step": 22860
+    },
+    {
+      "epoch": 1.5535398831362957,
+      "grad_norm": 5.875,
+      "learning_rate": 4.029037573039816e-05,
+      "loss": 0.9983,
+      "step": 22865
+    },
+    {
+      "epoch": 1.5538796032069575,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.0288252479956516e-05,
+      "loss": 1.0752,
+      "step": 22870
+    },
+    {
+      "epoch": 1.5542193232776191,
+      "grad_norm": 5.875,
+      "learning_rate": 4.028612922951488e-05,
+      "loss": 1.0414,
+      "step": 22875
+    },
+    {
+      "epoch": 1.554559043348281,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.028400597907325e-05,
+      "loss": 1.1613,
+      "step": 22880
+    },
+    {
+      "epoch": 1.5548987634189428,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.028188272863161e-05,
+      "loss": 1.0028,
+      "step": 22885
+    },
+    {
+      "epoch": 1.5552384834896045,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.027975947818997e-05,
+      "loss": 1.0575,
+      "step": 22890
+    },
+    {
+      "epoch": 1.5555782035602663,
+      "grad_norm": 5.5,
+      "learning_rate": 4.0277636227748336e-05,
+      "loss": 0.9816,
+      "step": 22895
+    },
+    {
+      "epoch": 1.5559179236309282,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.02755129773067e-05,
+      "loss": 1.0057,
+      "step": 22900
+    },
+    {
+      "epoch": 1.5562576437015898,
+      "grad_norm": 4.46875,
+      "learning_rate": 4.0273389726865064e-05,
+      "loss": 1.0028,
+      "step": 22905
+    },
+    {
+      "epoch": 1.5565973637722517,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.027126647642343e-05,
+      "loss": 1.0947,
+      "step": 22910
+    },
+    {
+      "epoch": 1.5569370838429135,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.026914322598179e-05,
+      "loss": 1.0666,
+      "step": 22915
+    },
+    {
+      "epoch": 1.5572768039135751,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.0267019975540156e-05,
+      "loss": 1.1618,
+      "step": 22920
+    },
+    {
+      "epoch": 1.557616523984237,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.026489672509852e-05,
+      "loss": 1.0968,
+      "step": 22925
+    },
+    {
+      "epoch": 1.5579562440548989,
+      "grad_norm": 6.65625,
+      "learning_rate": 4.0262773474656884e-05,
+      "loss": 1.0892,
+      "step": 22930
+    },
+    {
+      "epoch": 1.5582959641255605,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.026065022421525e-05,
+      "loss": 1.0514,
+      "step": 22935
+    },
+    {
+      "epoch": 1.5586356841962223,
+      "grad_norm": 6.125,
+      "learning_rate": 4.025852697377361e-05,
+      "loss": 1.0939,
+      "step": 22940
+    },
+    {
+      "epoch": 1.5589754042668842,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.0256403723331976e-05,
+      "loss": 1.1063,
+      "step": 22945
+    },
+    {
+      "epoch": 1.5593151243375458,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.025428047289034e-05,
+      "loss": 1.0234,
+      "step": 22950
+    },
+    {
+      "epoch": 1.5596548444082077,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.0252157222448704e-05,
+      "loss": 1.115,
+      "step": 22955
+    },
+    {
+      "epoch": 1.5599945644788695,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.025003397200707e-05,
+      "loss": 1.0808,
+      "step": 22960
+    },
+    {
+      "epoch": 1.5603342845495312,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.024791072156543e-05,
+      "loss": 1.0511,
+      "step": 22965
+    },
+    {
+      "epoch": 1.560674004620193,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.0245787471123796e-05,
+      "loss": 1.022,
+      "step": 22970
+    },
+    {
+      "epoch": 1.5610137246908549,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.024366422068216e-05,
+      "loss": 1.0248,
+      "step": 22975
+    },
+    {
+      "epoch": 1.5613534447615165,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0241540970240524e-05,
+      "loss": 1.007,
+      "step": 22980
+    },
+    {
+      "epoch": 1.5616931648321781,
+      "grad_norm": 5.75,
+      "learning_rate": 4.023941771979889e-05,
+      "loss": 1.0065,
+      "step": 22985
+    },
+    {
+      "epoch": 1.5620328849028402,
+      "grad_norm": 5.15625,
+      "learning_rate": 4.023729446935725e-05,
+      "loss": 1.0188,
+      "step": 22990
+    },
+    {
+      "epoch": 1.5623726049735018,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.0235171218915616e-05,
+      "loss": 1.0359,
+      "step": 22995
+    },
+    {
+      "epoch": 1.5627123250441635,
+      "grad_norm": 5.75,
+      "learning_rate": 4.023304796847398e-05,
+      "loss": 1.023,
+      "step": 23000
+    },
+    {
+      "epoch": 1.5630520451148255,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.0230924718032344e-05,
+      "loss": 0.9718,
+      "step": 23005
+    },
+    {
+      "epoch": 1.5633917651854872,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.022880146759071e-05,
+      "loss": 1.0704,
+      "step": 23010
+    },
+    {
+      "epoch": 1.5637314852561488,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.022667821714907e-05,
+      "loss": 1.134,
+      "step": 23015
+    },
+    {
+      "epoch": 1.5640712053268107,
+      "grad_norm": 5.46875,
+      "learning_rate": 4.0224554966707436e-05,
+      "loss": 1.1544,
+      "step": 23020
+    },
+    {
+      "epoch": 1.5644109253974725,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.02224317162658e-05,
+      "loss": 1.095,
+      "step": 23025
+    },
+    {
+      "epoch": 1.5647506454681341,
+      "grad_norm": 5.5,
+      "learning_rate": 4.0220308465824164e-05,
+      "loss": 1.0738,
+      "step": 23030
+    },
+    {
+      "epoch": 1.565090365538796,
+      "grad_norm": 5.875,
+      "learning_rate": 4.021818521538253e-05,
+      "loss": 1.0492,
+      "step": 23035
+    },
+    {
+      "epoch": 1.5654300856094578,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.021606196494089e-05,
+      "loss": 1.0575,
+      "step": 23040
+    },
+    {
+      "epoch": 1.5657698056801195,
+      "grad_norm": 6.375,
+      "learning_rate": 4.0213938714499256e-05,
+      "loss": 1.0859,
+      "step": 23045
+    },
+    {
+      "epoch": 1.5661095257507813,
+      "grad_norm": 6.375,
+      "learning_rate": 4.021181546405762e-05,
+      "loss": 1.0724,
+      "step": 23050
+    },
+    {
+      "epoch": 1.5664492458214432,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0209692213615984e-05,
+      "loss": 1.0909,
+      "step": 23055
+    },
+    {
+      "epoch": 1.5667889658921048,
+      "grad_norm": 6.375,
+      "learning_rate": 4.020756896317435e-05,
+      "loss": 1.1072,
+      "step": 23060
+    },
+    {
+      "epoch": 1.5671286859627667,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.020544571273271e-05,
+      "loss": 1.0933,
+      "step": 23065
+    },
+    {
+      "epoch": 1.5674684060334285,
+      "grad_norm": 6.0,
+      "learning_rate": 4.0203322462291076e-05,
+      "loss": 1.0749,
+      "step": 23070
+    },
+    {
+      "epoch": 1.5678081261040901,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.020119921184944e-05,
+      "loss": 1.1148,
+      "step": 23075
+    },
+    {
+      "epoch": 1.568147846174752,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.0199075961407804e-05,
+      "loss": 1.0648,
+      "step": 23080
+    },
+    {
+      "epoch": 1.5684875662454139,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.019695271096617e-05,
+      "loss": 1.057,
+      "step": 23085
+    },
+    {
+      "epoch": 1.5688272863160755,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.019482946052453e-05,
+      "loss": 1.0566,
+      "step": 23090
+    },
+    {
+      "epoch": 1.5691670063867373,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.019270621008289e-05,
+      "loss": 1.0226,
+      "step": 23095
+    },
+    {
+      "epoch": 1.5695067264573992,
+      "grad_norm": 6.53125,
+      "learning_rate": 4.019058295964126e-05,
+      "loss": 1.0643,
+      "step": 23100
+    },
+    {
+      "epoch": 1.5698464465280608,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.0188459709199624e-05,
+      "loss": 1.0186,
+      "step": 23105
+    },
+    {
+      "epoch": 1.5701861665987227,
+      "grad_norm": 6.4375,
+      "learning_rate": 4.018633645875798e-05,
+      "loss": 1.0236,
+      "step": 23110
+    },
+    {
+      "epoch": 1.5705258866693845,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.018421320831635e-05,
+      "loss": 1.1227,
+      "step": 23115
+    },
+    {
+      "epoch": 1.5708656067400462,
+      "grad_norm": 6.46875,
+      "learning_rate": 4.0182089957874716e-05,
+      "loss": 1.0494,
+      "step": 23120
+    },
+    {
+      "epoch": 1.571205326810708,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.017996670743307e-05,
+      "loss": 1.1868,
+      "step": 23125
+    },
+    {
+      "epoch": 1.5715450468813699,
+      "grad_norm": 6.5625,
+      "learning_rate": 4.0177843456991444e-05,
+      "loss": 1.0768,
+      "step": 23130
+    },
+    {
+      "epoch": 1.5718847669520315,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.017572020654981e-05,
+      "loss": 1.1231,
+      "step": 23135
+    },
+    {
+      "epoch": 1.5722244870226934,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.0173596956108165e-05,
+      "loss": 1.0915,
+      "step": 23140
+    },
+    {
+      "epoch": 1.5725642070933552,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.0171473705666536e-05,
+      "loss": 1.0284,
+      "step": 23145
+    },
+    {
+      "epoch": 1.5729039271640168,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.01693504552249e-05,
+      "loss": 1.1132,
+      "step": 23150
+    },
+    {
+      "epoch": 1.5732436472346785,
+      "grad_norm": 6.125,
+      "learning_rate": 4.016722720478326e-05,
+      "loss": 1.1042,
+      "step": 23155
+    },
+    {
+      "epoch": 1.5735833673053405,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.016510395434163e-05,
+      "loss": 1.0284,
+      "step": 23160
+    },
+    {
+      "epoch": 1.5739230873760022,
+      "grad_norm": 5.75,
+      "learning_rate": 4.0162980703899985e-05,
+      "loss": 1.1419,
+      "step": 23165
+    },
+    {
+      "epoch": 1.5742628074466638,
+      "grad_norm": 5.40625,
+      "learning_rate": 4.016085745345835e-05,
+      "loss": 1.0464,
+      "step": 23170
+    },
+    {
+      "epoch": 1.5746025275173259,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.015873420301672e-05,
+      "loss": 1.0337,
+      "step": 23175
+    },
+    {
+      "epoch": 1.5749422475879875,
+      "grad_norm": 6.25,
+      "learning_rate": 4.015661095257508e-05,
+      "loss": 1.1374,
+      "step": 23180
+    },
+    {
+      "epoch": 1.5752819676586491,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.015448770213344e-05,
+      "loss": 1.1288,
+      "step": 23185
+    },
+    {
+      "epoch": 1.575621687729311,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.015236445169181e-05,
+      "loss": 1.0463,
+      "step": 23190
+    },
+    {
+      "epoch": 1.5759614077999728,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.015024120125017e-05,
+      "loss": 1.122,
+      "step": 23195
+    },
+    {
+      "epoch": 1.5763011278706345,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.014811795080853e-05,
+      "loss": 1.1291,
+      "step": 23200
+    },
+    {
+      "epoch": 1.5766408479412963,
+      "grad_norm": 6.15625,
+      "learning_rate": 4.0145994700366904e-05,
+      "loss": 1.0846,
+      "step": 23205
+    },
+    {
+      "epoch": 1.5769805680119582,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.014387144992526e-05,
+      "loss": 0.9998,
+      "step": 23210
+    },
+    {
+      "epoch": 1.5773202880826198,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.0141748199483625e-05,
+      "loss": 1.0674,
+      "step": 23215
+    },
+    {
+      "epoch": 1.5776600081532817,
+      "grad_norm": 5.375,
+      "learning_rate": 4.0139624949041996e-05,
+      "loss": 1.05,
+      "step": 23220
+    },
+    {
+      "epoch": 1.5779997282239435,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.013750169860035e-05,
+      "loss": 1.0547,
+      "step": 23225
+    },
+    {
+      "epoch": 1.5783394482946052,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.013537844815872e-05,
+      "loss": 1.0491,
+      "step": 23230
+    },
+    {
+      "epoch": 1.578679168365267,
+      "grad_norm": 4.5,
+      "learning_rate": 4.013325519771709e-05,
+      "loss": 1.0576,
+      "step": 23235
+    },
+    {
+      "epoch": 1.5790188884359289,
+      "grad_norm": 5.9375,
+      "learning_rate": 4.0131131947275445e-05,
+      "loss": 1.1527,
+      "step": 23240
+    },
+    {
+      "epoch": 1.5793586085065905,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.012900869683381e-05,
+      "loss": 1.0232,
+      "step": 23245
+    },
+    {
+      "epoch": 1.5796983285772523,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.012688544639217e-05,
+      "loss": 1.0811,
+      "step": 23250
+    },
+    {
+      "epoch": 1.5800380486479142,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.012476219595054e-05,
+      "loss": 1.0517,
+      "step": 23255
+    },
+    {
+      "epoch": 1.5803777687185758,
+      "grad_norm": 5.53125,
+      "learning_rate": 4.01226389455089e-05,
+      "loss": 1.206,
+      "step": 23260
+    },
+    {
+      "epoch": 1.5807174887892377,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.0120515695067265e-05,
+      "loss": 1.1008,
+      "step": 23265
+    },
+    {
+      "epoch": 1.5810572088598995,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.011839244462563e-05,
+      "loss": 1.0872,
+      "step": 23270
+    },
+    {
+      "epoch": 1.5813969289305612,
+      "grad_norm": 5.875,
+      "learning_rate": 4.011626919418399e-05,
+      "loss": 1.1832,
+      "step": 23275
+    },
+    {
+      "epoch": 1.581736649001223,
+      "grad_norm": 5.875,
+      "learning_rate": 4.011414594374236e-05,
+      "loss": 1.1095,
+      "step": 23280
+    },
+    {
+      "epoch": 1.5820763690718849,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.011202269330072e-05,
+      "loss": 1.0913,
+      "step": 23285
+    },
+    {
+      "epoch": 1.5824160891425465,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.0109899442859085e-05,
+      "loss": 1.0793,
+      "step": 23290
+    },
+    {
+      "epoch": 1.5827558092132084,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.010777619241745e-05,
+      "loss": 1.0607,
+      "step": 23295
+    },
+    {
+      "epoch": 1.5830955292838702,
+      "grad_norm": 5.875,
+      "learning_rate": 4.010565294197581e-05,
+      "loss": 1.0723,
+      "step": 23300
+    },
+    {
+      "epoch": 1.5834352493545318,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.010352969153418e-05,
+      "loss": 1.0505,
+      "step": 23305
+    },
+    {
+      "epoch": 1.5837749694251937,
+      "grad_norm": 6.5,
+      "learning_rate": 4.010140644109254e-05,
+      "loss": 1.1147,
+      "step": 23310
+    },
+    {
+      "epoch": 1.5841146894958555,
+      "grad_norm": 6.25,
+      "learning_rate": 4.0099283190650905e-05,
+      "loss": 1.0216,
+      "step": 23315
+    },
+    {
+      "epoch": 1.5844544095665172,
+      "grad_norm": 6.125,
+      "learning_rate": 4.009715994020927e-05,
+      "loss": 1.0052,
+      "step": 23320
+    },
+    {
+      "epoch": 1.5847941296371788,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.009503668976763e-05,
+      "loss": 1.0468,
+      "step": 23325
+    },
+    {
+      "epoch": 1.5851338497078409,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0092913439326e-05,
+      "loss": 1.143,
+      "step": 23330
+    },
+    {
+      "epoch": 1.5854735697785025,
+      "grad_norm": 5.75,
+      "learning_rate": 4.009079018888436e-05,
+      "loss": 1.0629,
+      "step": 23335
+    },
+    {
+      "epoch": 1.5858132898491641,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.0088666938442725e-05,
+      "loss": 1.1483,
+      "step": 23340
+    },
+    {
+      "epoch": 1.5861530099198262,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.008654368800109e-05,
+      "loss": 1.0985,
+      "step": 23345
+    },
+    {
+      "epoch": 1.5864927299904878,
+      "grad_norm": 5.875,
+      "learning_rate": 4.008442043755945e-05,
+      "loss": 1.1129,
+      "step": 23350
+    },
+    {
+      "epoch": 1.5868324500611495,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.008229718711782e-05,
+      "loss": 1.1041,
+      "step": 23355
+    },
+    {
+      "epoch": 1.5871721701318113,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.008017393667618e-05,
+      "loss": 1.0268,
+      "step": 23360
+    },
+    {
+      "epoch": 1.5875118902024732,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.0078050686234545e-05,
+      "loss": 1.0815,
+      "step": 23365
+    },
+    {
+      "epoch": 1.5878516102731348,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.007592743579291e-05,
+      "loss": 1.0212,
+      "step": 23370
+    },
+    {
+      "epoch": 1.5881913303437967,
+      "grad_norm": 5.28125,
+      "learning_rate": 4.007380418535127e-05,
+      "loss": 0.9772,
+      "step": 23375
+    },
+    {
+      "epoch": 1.5885310504144585,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.007168093490964e-05,
+      "loss": 1.1891,
+      "step": 23380
+    },
+    {
+      "epoch": 1.5888707704851202,
+      "grad_norm": 5.5625,
+      "learning_rate": 4.0069557684468e-05,
+      "loss": 1.0343,
+      "step": 23385
+    },
+    {
+      "epoch": 1.589210490555782,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.0067434434026365e-05,
+      "loss": 1.0759,
+      "step": 23390
+    },
+    {
+      "epoch": 1.5895502106264439,
+      "grad_norm": 5.4375,
+      "learning_rate": 4.006531118358473e-05,
+      "loss": 1.1484,
+      "step": 23395
+    },
+    {
+      "epoch": 1.5898899306971055,
+      "grad_norm": 5.78125,
+      "learning_rate": 4.006318793314309e-05,
+      "loss": 0.9932,
+      "step": 23400
+    },
+    {
+      "epoch": 1.5902296507677673,
+      "grad_norm": 5.65625,
+      "learning_rate": 4.006106468270146e-05,
+      "loss": 1.1683,
+      "step": 23405
+    },
+    {
+      "epoch": 1.5905693708384292,
+      "grad_norm": 5.21875,
+      "learning_rate": 4.005894143225982e-05,
+      "loss": 1.1592,
+      "step": 23410
+    },
+    {
+      "epoch": 1.5909090909090908,
+      "grad_norm": 5.875,
+      "learning_rate": 4.0056818181818185e-05,
+      "loss": 1.0601,
+      "step": 23415
+    },
+    {
+      "epoch": 1.5912488109797527,
+      "grad_norm": 5.25,
+      "learning_rate": 4.005469493137655e-05,
+      "loss": 0.9489,
+      "step": 23420
+    },
+    {
+      "epoch": 1.5915885310504145,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.005257168093491e-05,
+      "loss": 1.1416,
+      "step": 23425
+    },
+    {
+      "epoch": 1.5919282511210762,
+      "grad_norm": 6.1875,
+      "learning_rate": 4.005044843049328e-05,
+      "loss": 1.1061,
+      "step": 23430
+    },
+    {
+      "epoch": 1.592267971191738,
+      "grad_norm": 6.0,
+      "learning_rate": 4.004832518005164e-05,
+      "loss": 1.0651,
+      "step": 23435
+    },
+    {
+      "epoch": 1.5926076912623999,
+      "grad_norm": 6.09375,
+      "learning_rate": 4.0046201929610005e-05,
+      "loss": 1.0324,
+      "step": 23440
+    },
+    {
+      "epoch": 1.5929474113330615,
+      "grad_norm": 6.34375,
+      "learning_rate": 4.004407867916837e-05,
+      "loss": 1.0726,
+      "step": 23445
+    },
+    {
+      "epoch": 1.5932871314037234,
+      "grad_norm": 6.0625,
+      "learning_rate": 4.0041955428726726e-05,
+      "loss": 1.0344,
+      "step": 23450
+    },
+    {
+      "epoch": 1.5936268514743852,
+      "grad_norm": 6.5,
+      "learning_rate": 4.00398321782851e-05,
+      "loss": 1.0126,
+      "step": 23455
+    },
+    {
+      "epoch": 1.5939665715450468,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.003770892784346e-05,
+      "loss": 1.2076,
+      "step": 23460
+    },
+    {
+      "epoch": 1.5943062916157087,
+      "grad_norm": 5.8125,
+      "learning_rate": 4.003558567740182e-05,
+      "loss": 1.0177,
+      "step": 23465
+    },
+    {
+      "epoch": 1.5946460116863705,
+      "grad_norm": 6.40625,
+      "learning_rate": 4.003346242696019e-05,
+      "loss": 1.0774,
+      "step": 23470
+    },
+    {
+      "epoch": 1.5949857317570322,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.003133917651855e-05,
+      "loss": 1.0305,
+      "step": 23475
+    },
+    {
+      "epoch": 1.595325451827694,
+      "grad_norm": 6.125,
+      "learning_rate": 4.002921592607691e-05,
+      "loss": 1.0348,
+      "step": 23480
+    },
+    {
+      "epoch": 1.5956651718983559,
+      "grad_norm": 5.96875,
+      "learning_rate": 4.002709267563528e-05,
+      "loss": 0.9967,
+      "step": 23485
+    },
+    {
+      "epoch": 1.5960048919690175,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.0024969425193645e-05,
+      "loss": 1.0523,
+      "step": 23490
+    },
+    {
+      "epoch": 1.5963446120396791,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.0022846174752e-05,
+      "loss": 1.0919,
+      "step": 23495
+    },
+    {
+      "epoch": 1.5966843321103412,
+      "grad_norm": 5.84375,
+      "learning_rate": 4.002072292431037e-05,
+      "loss": 0.9986,
+      "step": 23500
+    },
+    {
+      "epoch": 1.5970240521810029,
+      "grad_norm": 5.90625,
+      "learning_rate": 4.001859967386874e-05,
+      "loss": 1.0543,
+      "step": 23505
+    },
+    {
+      "epoch": 1.5973637722516645,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.0016476423427095e-05,
+      "loss": 1.1254,
+      "step": 23510
+    },
+    {
+      "epoch": 1.5977034923223266,
+      "grad_norm": 6.3125,
+      "learning_rate": 4.0014353172985465e-05,
+      "loss": 1.1755,
+      "step": 23515
+    },
+    {
+      "epoch": 1.5980432123929882,
+      "grad_norm": 5.6875,
+      "learning_rate": 4.001222992254382e-05,
+      "loss": 1.0935,
+      "step": 23520
+    },
+    {
+      "epoch": 1.5983829324636498,
+      "grad_norm": 5.1875,
+      "learning_rate": 4.0010106672102187e-05,
+      "loss": 1.0922,
+      "step": 23525
+    },
+    {
+      "epoch": 1.5987226525343117,
+      "grad_norm": 5.59375,
+      "learning_rate": 4.000798342166056e-05,
+      "loss": 1.05,
+      "step": 23530
+    },
+    {
+      "epoch": 1.5990623726049735,
+      "grad_norm": 6.03125,
+      "learning_rate": 4.0005860171218915e-05,
+      "loss": 0.95,
+      "step": 23535
+    },
+    {
+      "epoch": 1.5994020926756352,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.000373692077728e-05,
+      "loss": 1.0567,
+      "step": 23540
+    },
+    {
+      "epoch": 1.599741812746297,
+      "grad_norm": 5.71875,
+      "learning_rate": 4.000161367033565e-05,
+      "loss": 0.9719,
+      "step": 23545
+    },
+    {
+      "epoch": 1.6000815328169589,
+      "grad_norm": 6.25,
+      "learning_rate": 3.9999490419894007e-05,
+      "loss": 1.098,
+      "step": 23550
+    },
+    {
+      "epoch": 1.6004212528876205,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.999736716945237e-05,
+      "loss": 1.0216,
+      "step": 23555
+    },
+    {
+      "epoch": 1.6007609729582823,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.999524391901074e-05,
+      "loss": 1.0999,
+      "step": 23560
+    },
+    {
+      "epoch": 1.6011006930289442,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.99931206685691e-05,
+      "loss": 1.0548,
+      "step": 23565
+    },
+    {
+      "epoch": 1.6014404130996058,
+      "grad_norm": 5.875,
+      "learning_rate": 3.999099741812746e-05,
+      "loss": 1.0703,
+      "step": 23570
+    },
+    {
+      "epoch": 1.6017801331702677,
+      "grad_norm": 5.625,
+      "learning_rate": 3.998887416768583e-05,
+      "loss": 1.0424,
+      "step": 23575
+    },
+    {
+      "epoch": 1.6021198532409295,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.998675091724419e-05,
+      "loss": 1.0419,
+      "step": 23580
+    },
+    {
+      "epoch": 1.6024595733115912,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.9984627666802555e-05,
+      "loss": 1.0871,
+      "step": 23585
+    },
+    {
+      "epoch": 1.602799293382253,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.998250441636092e-05,
+      "loss": 1.0451,
+      "step": 23590
+    },
+    {
+      "epoch": 1.6031390134529149,
+      "grad_norm": 5.5,
+      "learning_rate": 3.998038116591928e-05,
+      "loss": 1.0896,
+      "step": 23595
+    },
+    {
+      "epoch": 1.6034787335235765,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.9978257915477647e-05,
+      "loss": 1.1678,
+      "step": 23600
+    },
+    {
+      "epoch": 1.6038184535942384,
+      "grad_norm": 5.875,
+      "learning_rate": 3.997613466503601e-05,
+      "loss": 1.0728,
+      "step": 23605
+    },
+    {
+      "epoch": 1.6041581736649002,
+      "grad_norm": 5.1875,
+      "learning_rate": 3.9974011414594375e-05,
+      "loss": 1.0404,
+      "step": 23610
+    },
+    {
+      "epoch": 1.6044978937355618,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.997188816415274e-05,
+      "loss": 1.0114,
+      "step": 23615
+    },
+    {
+      "epoch": 1.6048376138062237,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.99697649137111e-05,
+      "loss": 1.0093,
+      "step": 23620
+    },
+    {
+      "epoch": 1.6051773338768855,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.996764166326947e-05,
+      "loss": 1.0586,
+      "step": 23625
+    },
+    {
+      "epoch": 1.6055170539475472,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.996551841282783e-05,
+      "loss": 1.1131,
+      "step": 23630
+    },
+    {
+      "epoch": 1.605856774018209,
+      "grad_norm": 6.125,
+      "learning_rate": 3.9963395162386195e-05,
+      "loss": 0.9704,
+      "step": 23635
+    },
+    {
+      "epoch": 1.6061964940888709,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.996127191194456e-05,
+      "loss": 0.9937,
+      "step": 23640
+    },
+    {
+      "epoch": 1.6065362141595325,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.995914866150292e-05,
+      "loss": 1.0906,
+      "step": 23645
+    },
+    {
+      "epoch": 1.6068759342301944,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.995702541106129e-05,
+      "loss": 1.1002,
+      "step": 23650
+    },
+    {
+      "epoch": 1.6072156543008562,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.995490216061965e-05,
+      "loss": 1.0842,
+      "step": 23655
+    },
+    {
+      "epoch": 1.6075553743715179,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.9952778910178015e-05,
+      "loss": 1.2049,
+      "step": 23660
+    },
+    {
+      "epoch": 1.6078950944421795,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.995065565973638e-05,
+      "loss": 1.0314,
+      "step": 23665
+    },
+    {
+      "epoch": 1.6082348145128416,
+      "grad_norm": 5.625,
+      "learning_rate": 3.994853240929474e-05,
+      "loss": 1.0749,
+      "step": 23670
+    },
+    {
+      "epoch": 1.6085745345835032,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.994640915885311e-05,
+      "loss": 1.1234,
+      "step": 23675
+    },
+    {
+      "epoch": 1.6089142546541648,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.994428590841147e-05,
+      "loss": 1.0632,
+      "step": 23680
+    },
+    {
+      "epoch": 1.609253974724827,
+      "grad_norm": 4.84375,
+      "learning_rate": 3.9942162657969835e-05,
+      "loss": 1.0674,
+      "step": 23685
+    },
+    {
+      "epoch": 1.6095936947954885,
+      "grad_norm": 6.25,
+      "learning_rate": 3.99400394075282e-05,
+      "loss": 1.1429,
+      "step": 23690
+    },
+    {
+      "epoch": 1.6099334148661502,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.993791615708656e-05,
+      "loss": 0.963,
+      "step": 23695
+    },
+    {
+      "epoch": 1.610273134936812,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.993579290664493e-05,
+      "loss": 1.1099,
+      "step": 23700
+    },
+    {
+      "epoch": 1.6106128550074739,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.993366965620329e-05,
+      "loss": 1.0456,
+      "step": 23705
+    },
+    {
+      "epoch": 1.6109525750781355,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9931546405761655e-05,
+      "loss": 0.9764,
+      "step": 23710
+    },
+    {
+      "epoch": 1.6112922951487973,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.992942315532002e-05,
+      "loss": 1.0798,
+      "step": 23715
+    },
+    {
+      "epoch": 1.6116320152194592,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.992729990487838e-05,
+      "loss": 1.0763,
+      "step": 23720
+    },
+    {
+      "epoch": 1.6119717352901208,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.992517665443675e-05,
+      "loss": 1.0147,
+      "step": 23725
+    },
+    {
+      "epoch": 1.6123114553607827,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.992305340399511e-05,
+      "loss": 1.1362,
+      "step": 23730
+    },
+    {
+      "epoch": 1.6126511754314445,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9920930153553475e-05,
+      "loss": 1.0112,
+      "step": 23735
+    },
+    {
+      "epoch": 1.6129908955021062,
+      "grad_norm": 5.5,
+      "learning_rate": 3.991880690311184e-05,
+      "loss": 1.0078,
+      "step": 23740
+    },
+    {
+      "epoch": 1.613330615572768,
+      "grad_norm": 6.375,
+      "learning_rate": 3.99166836526702e-05,
+      "loss": 1.1305,
+      "step": 23745
+    },
+    {
+      "epoch": 1.6136703356434299,
+      "grad_norm": 5.5,
+      "learning_rate": 3.991456040222857e-05,
+      "loss": 1.0817,
+      "step": 23750
+    },
+    {
+      "epoch": 1.6140100557140915,
+      "grad_norm": 6.0,
+      "learning_rate": 3.991243715178693e-05,
+      "loss": 1.121,
+      "step": 23755
+    },
+    {
+      "epoch": 1.6143497757847534,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9910313901345295e-05,
+      "loss": 1.1003,
+      "step": 23760
+    },
+    {
+      "epoch": 1.6146894958554152,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.990819065090366e-05,
+      "loss": 1.018,
+      "step": 23765
+    },
+    {
+      "epoch": 1.6150292159260768,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.990606740046202e-05,
+      "loss": 1.0137,
+      "step": 23770
+    },
+    {
+      "epoch": 1.6153689359967387,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.990394415002039e-05,
+      "loss": 1.1709,
+      "step": 23775
+    },
+    {
+      "epoch": 1.6157086560674006,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.990182089957875e-05,
+      "loss": 1.0634,
+      "step": 23780
+    },
+    {
+      "epoch": 1.6160483761380622,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.9899697649137115e-05,
+      "loss": 1.0924,
+      "step": 23785
+    },
+    {
+      "epoch": 1.616388096208724,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.989757439869547e-05,
+      "loss": 1.0206,
+      "step": 23790
+    },
+    {
+      "epoch": 1.6167278162793859,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.989545114825384e-05,
+      "loss": 1.2023,
+      "step": 23795
+    },
+    {
+      "epoch": 1.6170675363500475,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.989332789781221e-05,
+      "loss": 1.0948,
+      "step": 23800
+    },
+    {
+      "epoch": 1.6174072564207094,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.9891204647370564e-05,
+      "loss": 0.9791,
+      "step": 23805
+    },
+    {
+      "epoch": 1.6177469764913712,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.9889081396928935e-05,
+      "loss": 1.0283,
+      "step": 23810
+    },
+    {
+      "epoch": 1.6180866965620329,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.98869581464873e-05,
+      "loss": 1.1011,
+      "step": 23815
+    },
+    {
+      "epoch": 1.6184264166326947,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.9884834896045656e-05,
+      "loss": 1.0196,
+      "step": 23820
+    },
+    {
+      "epoch": 1.6187661367033566,
+      "grad_norm": 6.125,
+      "learning_rate": 3.988271164560403e-05,
+      "loss": 1.1078,
+      "step": 23825
+    },
+    {
+      "epoch": 1.6191058567740182,
+      "grad_norm": 5.625,
+      "learning_rate": 3.988058839516239e-05,
+      "loss": 1.0178,
+      "step": 23830
+    },
+    {
+      "epoch": 1.6194455768446798,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.987846514472075e-05,
+      "loss": 1.051,
+      "step": 23835
+    },
+    {
+      "epoch": 1.619785296915342,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.987634189427912e-05,
+      "loss": 1.1202,
+      "step": 23840
+    },
+    {
+      "epoch": 1.6201250169860035,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.987421864383748e-05,
+      "loss": 1.0807,
+      "step": 23845
+    },
+    {
+      "epoch": 1.6204647370566652,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.987209539339584e-05,
+      "loss": 1.151,
+      "step": 23850
+    },
+    {
+      "epoch": 1.6208044571273272,
+      "grad_norm": 6.125,
+      "learning_rate": 3.986997214295421e-05,
+      "loss": 1.1105,
+      "step": 23855
+    },
+    {
+      "epoch": 1.6211441771979889,
+      "grad_norm": 5.5,
+      "learning_rate": 3.9867848892512575e-05,
+      "loss": 1.0045,
+      "step": 23860
+    },
+    {
+      "epoch": 1.6214838972686505,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.986572564207093e-05,
+      "loss": 1.1295,
+      "step": 23865
+    },
+    {
+      "epoch": 1.6218236173393124,
+      "grad_norm": 5.3125,
+      "learning_rate": 3.98636023916293e-05,
+      "loss": 1.0353,
+      "step": 23870
+    },
+    {
+      "epoch": 1.6221633374099742,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.986147914118766e-05,
+      "loss": 1.0258,
+      "step": 23875
+    },
+    {
+      "epoch": 1.6225030574806358,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.9859355890746024e-05,
+      "loss": 1.1404,
+      "step": 23880
+    },
+    {
+      "epoch": 1.6228427775512977,
+      "grad_norm": 6.625,
+      "learning_rate": 3.9857232640304395e-05,
+      "loss": 1.1692,
+      "step": 23885
+    },
+    {
+      "epoch": 1.6231824976219595,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.985510938986275e-05,
+      "loss": 1.0389,
+      "step": 23890
+    },
+    {
+      "epoch": 1.6235222176926212,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.9852986139421116e-05,
+      "loss": 1.0319,
+      "step": 23895
+    },
+    {
+      "epoch": 1.623861937763283,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.985086288897949e-05,
+      "loss": 1.1103,
+      "step": 23900
+    },
+    {
+      "epoch": 1.6242016578339449,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9848739638537844e-05,
+      "loss": 1.0074,
+      "step": 23905
+    },
+    {
+      "epoch": 1.6245413779046065,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.984661638809621e-05,
+      "loss": 1.0465,
+      "step": 23910
+    },
+    {
+      "epoch": 1.6248810979752684,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.984449313765458e-05,
+      "loss": 1.0771,
+      "step": 23915
+    },
+    {
+      "epoch": 1.6252208180459302,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.9842369887212936e-05,
+      "loss": 1.1113,
+      "step": 23920
+    },
+    {
+      "epoch": 1.6255605381165918,
+      "grad_norm": 6.125,
+      "learning_rate": 3.98402466367713e-05,
+      "loss": 1.1253,
+      "step": 23925
+    },
+    {
+      "epoch": 1.6259002581872537,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.983812338632967e-05,
+      "loss": 1.0748,
+      "step": 23930
+    },
+    {
+      "epoch": 1.6262399782579156,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.983600013588803e-05,
+      "loss": 1.0612,
+      "step": 23935
+    },
+    {
+      "epoch": 1.6265796983285772,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.983387688544639e-05,
+      "loss": 1.175,
+      "step": 23940
+    },
+    {
+      "epoch": 1.626919418399239,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.9831753635004756e-05,
+      "loss": 1.0721,
+      "step": 23945
+    },
+    {
+      "epoch": 1.627259138469901,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.982963038456312e-05,
+      "loss": 1.0405,
+      "step": 23950
+    },
+    {
+      "epoch": 1.6275988585405625,
+      "grad_norm": 6.25,
+      "learning_rate": 3.9827507134121484e-05,
+      "loss": 1.1142,
+      "step": 23955
+    },
+    {
+      "epoch": 1.6279385786112244,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.982538388367985e-05,
+      "loss": 1.2056,
+      "step": 23960
+    },
+    {
+      "epoch": 1.6282782986818862,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.982326063323821e-05,
+      "loss": 1.0407,
+      "step": 23965
+    },
+    {
+      "epoch": 1.6286180187525479,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.9821137382796576e-05,
+      "loss": 1.2332,
+      "step": 23970
+    },
+    {
+      "epoch": 1.6289577388232097,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.981901413235494e-05,
+      "loss": 1.1312,
+      "step": 23975
+    },
+    {
+      "epoch": 1.6292974588938716,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.9816890881913304e-05,
+      "loss": 1.0306,
+      "step": 23980
+    },
+    {
+      "epoch": 1.6296371789645332,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.981476763147167e-05,
+      "loss": 1.0861,
+      "step": 23985
+    },
+    {
+      "epoch": 1.629976899035195,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.981264438103003e-05,
+      "loss": 1.0585,
+      "step": 23990
+    },
+    {
+      "epoch": 1.630316619105857,
+      "grad_norm": 6.0,
+      "learning_rate": 3.9810521130588396e-05,
+      "loss": 1.1271,
+      "step": 23995
+    },
+    {
+      "epoch": 1.6306563391765185,
+      "grad_norm": 5.875,
+      "learning_rate": 3.980839788014676e-05,
+      "loss": 1.0168,
+      "step": 24000
+    },
+    {
+      "epoch": 1.6309960592471802,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9806274629705124e-05,
+      "loss": 1.0284,
+      "step": 24005
+    },
+    {
+      "epoch": 1.6313357793178422,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.980415137926349e-05,
+      "loss": 1.1205,
+      "step": 24010
+    },
+    {
+      "epoch": 1.6316754993885039,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.980202812882185e-05,
+      "loss": 1.0578,
+      "step": 24015
+    },
+    {
+      "epoch": 1.6320152194591655,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.9799904878380216e-05,
+      "loss": 1.0854,
+      "step": 24020
+    },
+    {
+      "epoch": 1.6323549395298276,
+      "grad_norm": 5.75,
+      "learning_rate": 3.979778162793858e-05,
+      "loss": 1.0088,
+      "step": 24025
+    },
+    {
+      "epoch": 1.6326946596004892,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.9795658377496944e-05,
+      "loss": 1.1161,
+      "step": 24030
+    },
+    {
+      "epoch": 1.6330343796711508,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.979353512705531e-05,
+      "loss": 1.1133,
+      "step": 24035
+    },
+    {
+      "epoch": 1.633374099741813,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.979141187661367e-05,
+      "loss": 1.0183,
+      "step": 24040
+    },
+    {
+      "epoch": 1.6337138198124745,
+      "grad_norm": 6.125,
+      "learning_rate": 3.9789288626172036e-05,
+      "loss": 1.0128,
+      "step": 24045
+    },
+    {
+      "epoch": 1.6340535398831362,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.97871653757304e-05,
+      "loss": 0.9533,
+      "step": 24050
+    },
+    {
+      "epoch": 1.634393259953798,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.9785042125288764e-05,
+      "loss": 0.9808,
+      "step": 24055
+    },
+    {
+      "epoch": 1.6347329800244599,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.978291887484713e-05,
+      "loss": 1.0309,
+      "step": 24060
+    },
+    {
+      "epoch": 1.6350727000951215,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.978079562440549e-05,
+      "loss": 1.0156,
+      "step": 24065
+    },
+    {
+      "epoch": 1.6354124201657834,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.9778672373963856e-05,
+      "loss": 1.1263,
+      "step": 24070
+    },
+    {
+      "epoch": 1.6357521402364452,
+      "grad_norm": 6.0,
+      "learning_rate": 3.977654912352222e-05,
+      "loss": 1.0943,
+      "step": 24075
+    },
+    {
+      "epoch": 1.6360918603071068,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9774425873080584e-05,
+      "loss": 1.0661,
+      "step": 24080
+    },
+    {
+      "epoch": 1.6364315803777687,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.977230262263895e-05,
+      "loss": 1.0377,
+      "step": 24085
+    },
+    {
+      "epoch": 1.6367713004484306,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.977017937219731e-05,
+      "loss": 1.0097,
+      "step": 24090
+    },
+    {
+      "epoch": 1.6371110205190922,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.9768056121755676e-05,
+      "loss": 1.0518,
+      "step": 24095
+    },
+    {
+      "epoch": 1.637450740589754,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.976593287131404e-05,
+      "loss": 1.096,
+      "step": 24100
+    },
+    {
+      "epoch": 1.637790460660416,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.9763809620872404e-05,
+      "loss": 1.0761,
+      "step": 24105
+    },
+    {
+      "epoch": 1.6381301807310775,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.976168637043077e-05,
+      "loss": 1.0505,
+      "step": 24110
+    },
+    {
+      "epoch": 1.6384699008017394,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.975956311998913e-05,
+      "loss": 1.0124,
+      "step": 24115
+    },
+    {
+      "epoch": 1.6388096208724012,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.9757439869547496e-05,
+      "loss": 0.95,
+      "step": 24120
+    },
+    {
+      "epoch": 1.6391493409430629,
+      "grad_norm": 5.25,
+      "learning_rate": 3.975531661910586e-05,
+      "loss": 1.0771,
+      "step": 24125
+    },
+    {
+      "epoch": 1.6394890610137247,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.9753193368664224e-05,
+      "loss": 1.044,
+      "step": 24130
+    },
+    {
+      "epoch": 1.6398287810843866,
+      "grad_norm": 5.5,
+      "learning_rate": 3.975107011822259e-05,
+      "loss": 1.0541,
+      "step": 24135
+    },
+    {
+      "epoch": 1.6401685011550482,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.974894686778095e-05,
+      "loss": 1.0283,
+      "step": 24140
+    },
+    {
+      "epoch": 1.64050822122571,
+      "grad_norm": 5.875,
+      "learning_rate": 3.974682361733931e-05,
+      "loss": 1.1541,
+      "step": 24145
+    },
+    {
+      "epoch": 1.640847941296372,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.974470036689768e-05,
+      "loss": 1.0719,
+      "step": 24150
+    },
+    {
+      "epoch": 1.6411876613670335,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.9742577116456044e-05,
+      "loss": 1.104,
+      "step": 24155
+    },
+    {
+      "epoch": 1.6415273814376954,
+      "grad_norm": 5.625,
+      "learning_rate": 3.97404538660144e-05,
+      "loss": 1.0474,
+      "step": 24160
+    },
+    {
+      "epoch": 1.6418671015083572,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.973833061557277e-05,
+      "loss": 1.0228,
+      "step": 24165
+    },
+    {
+      "epoch": 1.6422068215790189,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9736207365131136e-05,
+      "loss": 1.0052,
+      "step": 24170
+    },
+    {
+      "epoch": 1.6425465416496805,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.9734084114689493e-05,
+      "loss": 1.1098,
+      "step": 24175
+    },
+    {
+      "epoch": 1.6428862617203426,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.9731960864247864e-05,
+      "loss": 1.0717,
+      "step": 24180
+    },
+    {
+      "epoch": 1.6432259817910042,
+      "grad_norm": 6.0,
+      "learning_rate": 3.972983761380623e-05,
+      "loss": 1.1457,
+      "step": 24185
+    },
+    {
+      "epoch": 1.6435657018616658,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.9727714363364586e-05,
+      "loss": 1.0323,
+      "step": 24190
+    },
+    {
+      "epoch": 1.643905421932328,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.9725591112922956e-05,
+      "loss": 1.0971,
+      "step": 24195
+    },
+    {
+      "epoch": 1.6442451420029895,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.972346786248132e-05,
+      "loss": 1.1201,
+      "step": 24200
+    },
+    {
+      "epoch": 1.6445848620736512,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.972134461203968e-05,
+      "loss": 1.1236,
+      "step": 24205
+    },
+    {
+      "epoch": 1.6449245821443133,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.971922136159805e-05,
+      "loss": 1.0038,
+      "step": 24210
+    },
+    {
+      "epoch": 1.6452643022149749,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.9717098111156406e-05,
+      "loss": 1.0403,
+      "step": 24215
+    },
+    {
+      "epoch": 1.6456040222856365,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.971497486071477e-05,
+      "loss": 1.0695,
+      "step": 24220
+    },
+    {
+      "epoch": 1.6459437423562984,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.971285161027314e-05,
+      "loss": 1.1313,
+      "step": 24225
+    },
+    {
+      "epoch": 1.6462834624269602,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.97107283598315e-05,
+      "loss": 1.0807,
+      "step": 24230
+    },
+    {
+      "epoch": 1.6466231824976219,
+      "grad_norm": 5.09375,
+      "learning_rate": 3.970860510938986e-05,
+      "loss": 1.0473,
+      "step": 24235
+    },
+    {
+      "epoch": 1.6469629025682837,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.970648185894823e-05,
+      "loss": 1.045,
+      "step": 24240
+    },
+    {
+      "epoch": 1.6473026226389456,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.970435860850659e-05,
+      "loss": 1.0931,
+      "step": 24245
+    },
+    {
+      "epoch": 1.6476423427096072,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.9702235358064954e-05,
+      "loss": 1.1197,
+      "step": 24250
+    },
+    {
+      "epoch": 1.647982062780269,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9700112107623324e-05,
+      "loss": 1.0179,
+      "step": 24255
+    },
+    {
+      "epoch": 1.648321782850931,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.969798885718168e-05,
+      "loss": 1.0049,
+      "step": 24260
+    },
+    {
+      "epoch": 1.6486615029215925,
+      "grad_norm": 5.125,
+      "learning_rate": 3.9695865606740046e-05,
+      "loss": 1.0949,
+      "step": 24265
+    },
+    {
+      "epoch": 1.6490012229922544,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.9693742356298416e-05,
+      "loss": 1.0576,
+      "step": 24270
+    },
+    {
+      "epoch": 1.6493409430629162,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.9691619105856774e-05,
+      "loss": 1.1221,
+      "step": 24275
+    },
+    {
+      "epoch": 1.6496806631335779,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.968949585541514e-05,
+      "loss": 0.933,
+      "step": 24280
+    },
+    {
+      "epoch": 1.6500203832042397,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.968737260497351e-05,
+      "loss": 0.967,
+      "step": 24285
+    },
+    {
+      "epoch": 1.6503601032749016,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.9685249354531866e-05,
+      "loss": 1.0165,
+      "step": 24290
+    },
+    {
+      "epoch": 1.6506998233455632,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.968312610409023e-05,
+      "loss": 1.1347,
+      "step": 24295
+    },
+    {
+      "epoch": 1.651039543416225,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.9681002853648594e-05,
+      "loss": 1.0618,
+      "step": 24300
+    },
+    {
+      "epoch": 1.651379263486887,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.967887960320696e-05,
+      "loss": 1.1799,
+      "step": 24305
+    },
+    {
+      "epoch": 1.6517189835575485,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.967675635276532e-05,
+      "loss": 0.9987,
+      "step": 24310
+    },
+    {
+      "epoch": 1.6520587036282104,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.9674633102323686e-05,
+      "loss": 1.1284,
+      "step": 24315
+    },
+    {
+      "epoch": 1.6523984236988722,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.967250985188205e-05,
+      "loss": 1.0278,
+      "step": 24320
+    },
+    {
+      "epoch": 1.6527381437695339,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9670386601440414e-05,
+      "loss": 0.9907,
+      "step": 24325
+    },
+    {
+      "epoch": 1.6530778638401957,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.966826335099878e-05,
+      "loss": 1.0176,
+      "step": 24330
+    },
+    {
+      "epoch": 1.6534175839108576,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.966614010055714e-05,
+      "loss": 1.1222,
+      "step": 24335
+    },
+    {
+      "epoch": 1.6537573039815192,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.9664016850115506e-05,
+      "loss": 1.0244,
+      "step": 24340
+    },
+    {
+      "epoch": 1.6540970240521808,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.966189359967387e-05,
+      "loss": 1.0473,
+      "step": 24345
+    },
+    {
+      "epoch": 1.654436744122843,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.9659770349232234e-05,
+      "loss": 1.2032,
+      "step": 24350
+    },
+    {
+      "epoch": 1.6547764641935045,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.96576470987906e-05,
+      "loss": 1.1021,
+      "step": 24355
+    },
+    {
+      "epoch": 1.6551161842641662,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.965552384834896e-05,
+      "loss": 1.0287,
+      "step": 24360
+    },
+    {
+      "epoch": 1.6554559043348283,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.9653400597907326e-05,
+      "loss": 1.0514,
+      "step": 24365
+    },
+    {
+      "epoch": 1.6557956244054899,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.965127734746569e-05,
+      "loss": 1.112,
+      "step": 24370
+    },
+    {
+      "epoch": 1.6561353444761515,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.9649154097024054e-05,
+      "loss": 1.028,
+      "step": 24375
+    },
+    {
+      "epoch": 1.6564750645468136,
+      "grad_norm": 5.75,
+      "learning_rate": 3.964703084658242e-05,
+      "loss": 1.0302,
+      "step": 24380
+    },
+    {
+      "epoch": 1.6568147846174752,
+      "grad_norm": 7.0,
+      "learning_rate": 3.964490759614078e-05,
+      "loss": 1.098,
+      "step": 24385
+    },
+    {
+      "epoch": 1.6571545046881369,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.9642784345699146e-05,
+      "loss": 1.0446,
+      "step": 24390
+    },
+    {
+      "epoch": 1.6574942247587987,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.964066109525751e-05,
+      "loss": 1.0911,
+      "step": 24395
+    },
+    {
+      "epoch": 1.6578339448294606,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.9638537844815874e-05,
+      "loss": 1.1613,
+      "step": 24400
+    },
+    {
+      "epoch": 1.6581736649001222,
+      "grad_norm": 6.25,
+      "learning_rate": 3.963641459437424e-05,
+      "loss": 1.0016,
+      "step": 24405
+    },
+    {
+      "epoch": 1.658513384970784,
+      "grad_norm": 6.5,
+      "learning_rate": 3.96342913439326e-05,
+      "loss": 1.0999,
+      "step": 24410
+    },
+    {
+      "epoch": 1.658853105041446,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.9632168093490966e-05,
+      "loss": 0.9943,
+      "step": 24415
+    },
+    {
+      "epoch": 1.6591928251121075,
+      "grad_norm": 6.125,
+      "learning_rate": 3.963004484304933e-05,
+      "loss": 1.0218,
+      "step": 24420
+    },
+    {
+      "epoch": 1.6595325451827694,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.9627921592607694e-05,
+      "loss": 1.0726,
+      "step": 24425
+    },
+    {
+      "epoch": 1.6598722652534312,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.962579834216606e-05,
+      "loss": 1.0769,
+      "step": 24430
+    },
+    {
+      "epoch": 1.6602119853240929,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.962367509172442e-05,
+      "loss": 1.0224,
+      "step": 24435
+    },
+    {
+      "epoch": 1.6605517053947547,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.9621551841282786e-05,
+      "loss": 0.9925,
+      "step": 24440
+    },
+    {
+      "epoch": 1.6608914254654166,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.961942859084115e-05,
+      "loss": 1.0787,
+      "step": 24445
+    },
+    {
+      "epoch": 1.6612311455360782,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.9617305340399514e-05,
+      "loss": 1.1242,
+      "step": 24450
+    },
+    {
+      "epoch": 1.66157086560674,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.961518208995788e-05,
+      "loss": 1.0954,
+      "step": 24455
+    },
+    {
+      "epoch": 1.661910585677402,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.961305883951624e-05,
+      "loss": 1.1141,
+      "step": 24460
+    },
+    {
+      "epoch": 1.6622503057480635,
+      "grad_norm": 6.375,
+      "learning_rate": 3.9610935589074606e-05,
+      "loss": 1.035,
+      "step": 24465
+    },
+    {
+      "epoch": 1.6625900258187254,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.960881233863297e-05,
+      "loss": 1.0538,
+      "step": 24470
+    },
+    {
+      "epoch": 1.6629297458893872,
+      "grad_norm": 5.75,
+      "learning_rate": 3.9606689088191334e-05,
+      "loss": 1.0263,
+      "step": 24475
+    },
+    {
+      "epoch": 1.6632694659600489,
+      "grad_norm": 5.5,
+      "learning_rate": 3.96045658377497e-05,
+      "loss": 1.1107,
+      "step": 24480
+    },
+    {
+      "epoch": 1.6636091860307107,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.960244258730806e-05,
+      "loss": 1.0549,
+      "step": 24485
+    },
+    {
+      "epoch": 1.6639489061013726,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.9600319336866426e-05,
+      "loss": 1.0977,
+      "step": 24490
+    },
+    {
+      "epoch": 1.6642886261720342,
+      "grad_norm": 6.0,
+      "learning_rate": 3.959819608642479e-05,
+      "loss": 1.0895,
+      "step": 24495
+    },
+    {
+      "epoch": 1.664628346242696,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.959607283598315e-05,
+      "loss": 1.1109,
+      "step": 24500
+    },
+    {
+      "epoch": 1.664968066313358,
+      "grad_norm": 6.375,
+      "learning_rate": 3.959394958554152e-05,
+      "loss": 1.0205,
+      "step": 24505
+    },
+    {
+      "epoch": 1.6653077863840196,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.959182633509988e-05,
+      "loss": 1.0939,
+      "step": 24510
+    },
+    {
+      "epoch": 1.6656475064546812,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.958970308465824e-05,
+      "loss": 1.0466,
+      "step": 24515
+    },
+    {
+      "epoch": 1.6659872265253433,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.958757983421661e-05,
+      "loss": 0.9963,
+      "step": 24520
+    },
+    {
+      "epoch": 1.6663269465960049,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.9585456583774974e-05,
+      "loss": 1.0886,
+      "step": 24525
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.958333333333333e-05,
+      "loss": 1.1192,
+      "step": 24530
+    },
+    {
+      "epoch": 1.6670063867373286,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.95812100828917e-05,
+      "loss": 1.0985,
+      "step": 24535
+    },
+    {
+      "epoch": 1.6673461068079902,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.9579086832450066e-05,
+      "loss": 1.0439,
+      "step": 24540
+    },
+    {
+      "epoch": 1.6676858268786519,
+      "grad_norm": 7.25,
+      "learning_rate": 3.957696358200842e-05,
+      "loss": 1.0779,
+      "step": 24545
+    },
+    {
+      "epoch": 1.668025546949314,
+      "grad_norm": 6.875,
+      "learning_rate": 3.9574840331566794e-05,
+      "loss": 1.0447,
+      "step": 24550
+    },
+    {
+      "epoch": 1.6683652670199756,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.957271708112516e-05,
+      "loss": 1.1571,
+      "step": 24555
+    },
+    {
+      "epoch": 1.6687049870906372,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9570593830683515e-05,
+      "loss": 1.0423,
+      "step": 24560
+    },
+    {
+      "epoch": 1.669044707161299,
+      "grad_norm": 7.125,
+      "learning_rate": 3.9568470580241886e-05,
+      "loss": 1.0911,
+      "step": 24565
+    },
+    {
+      "epoch": 1.669384427231961,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.956634732980024e-05,
+      "loss": 1.119,
+      "step": 24570
+    },
+    {
+      "epoch": 1.6697241473026225,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.956422407935861e-05,
+      "loss": 1.1616,
+      "step": 24575
+    },
+    {
+      "epoch": 1.6700638673732844,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.956210082891698e-05,
+      "loss": 1.031,
+      "step": 24580
+    },
+    {
+      "epoch": 1.6704035874439462,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.9559977578475335e-05,
+      "loss": 0.9943,
+      "step": 24585
+    },
+    {
+      "epoch": 1.6707433075146079,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.95578543280337e-05,
+      "loss": 1.0252,
+      "step": 24590
+    },
+    {
+      "epoch": 1.6710830275852697,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.955573107759207e-05,
+      "loss": 1.0849,
+      "step": 24595
+    },
+    {
+      "epoch": 1.6714227476559316,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.955360782715043e-05,
+      "loss": 1.0757,
+      "step": 24600
+    },
+    {
+      "epoch": 1.6717624677265932,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.955148457670879e-05,
+      "loss": 0.9992,
+      "step": 24605
+    },
+    {
+      "epoch": 1.672102187797255,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.954936132626716e-05,
+      "loss": 1.1253,
+      "step": 24610
+    },
+    {
+      "epoch": 1.672441907867917,
+      "grad_norm": 6.125,
+      "learning_rate": 3.954723807582552e-05,
+      "loss": 1.0805,
+      "step": 24615
+    },
+    {
+      "epoch": 1.6727816279385785,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.954511482538388e-05,
+      "loss": 0.941,
+      "step": 24620
+    },
+    {
+      "epoch": 1.6731213480092404,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.9542991574942254e-05,
+      "loss": 1.063,
+      "step": 24625
+    },
+    {
+      "epoch": 1.6734610680799022,
+      "grad_norm": 5.875,
+      "learning_rate": 3.954086832450061e-05,
+      "loss": 1.1247,
+      "step": 24630
+    },
+    {
+      "epoch": 1.6738007881505639,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.9538745074058975e-05,
+      "loss": 1.1316,
+      "step": 24635
+    },
+    {
+      "epoch": 1.6741405082212257,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.953662182361734e-05,
+      "loss": 1.0481,
+      "step": 24640
+    },
+    {
+      "epoch": 1.6744802282918876,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.95344985731757e-05,
+      "loss": 1.0547,
+      "step": 24645
+    },
+    {
+      "epoch": 1.6748199483625492,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.953237532273407e-05,
+      "loss": 1.0043,
+      "step": 24650
+    },
+    {
+      "epoch": 1.675159668433211,
+      "grad_norm": 5.75,
+      "learning_rate": 3.953025207229243e-05,
+      "loss": 0.9704,
+      "step": 24655
+    },
+    {
+      "epoch": 1.675499388503873,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.9528128821850795e-05,
+      "loss": 1.0356,
+      "step": 24660
+    },
+    {
+      "epoch": 1.6758391085745346,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.952600557140916e-05,
+      "loss": 1.1461,
+      "step": 24665
+    },
+    {
+      "epoch": 1.6761788286451964,
+      "grad_norm": 5.625,
+      "learning_rate": 3.952388232096752e-05,
+      "loss": 0.98,
+      "step": 24670
+    },
+    {
+      "epoch": 1.6765185487158583,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.952175907052589e-05,
+      "loss": 1.0159,
+      "step": 24675
+    },
+    {
+      "epoch": 1.67685826878652,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.951963582008425e-05,
+      "loss": 1.0091,
+      "step": 24680
+    },
+    {
+      "epoch": 1.6771979888571815,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.9517512569642615e-05,
+      "loss": 1.0515,
+      "step": 24685
+    },
+    {
+      "epoch": 1.6775377089278436,
+      "grad_norm": 7.25,
+      "learning_rate": 3.951538931920098e-05,
+      "loss": 1.1049,
+      "step": 24690
+    },
+    {
+      "epoch": 1.6778774289985052,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.951326606875934e-05,
+      "loss": 1.0205,
+      "step": 24695
+    },
+    {
+      "epoch": 1.6782171490691669,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.951114281831771e-05,
+      "loss": 1.002,
+      "step": 24700
+    },
+    {
+      "epoch": 1.678556869139829,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.950901956787607e-05,
+      "loss": 1.0647,
+      "step": 24705
+    },
+    {
+      "epoch": 1.6788965892104906,
+      "grad_norm": 6.0,
+      "learning_rate": 3.950689631743444e-05,
+      "loss": 1.0979,
+      "step": 24710
+    },
+    {
+      "epoch": 1.6792363092811522,
+      "grad_norm": 5.25,
+      "learning_rate": 3.95047730669928e-05,
+      "loss": 1.1506,
+      "step": 24715
+    },
+    {
+      "epoch": 1.6795760293518143,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.950264981655116e-05,
+      "loss": 1.1153,
+      "step": 24720
+    },
+    {
+      "epoch": 1.679915749422476,
+      "grad_norm": 4.90625,
+      "learning_rate": 3.950052656610953e-05,
+      "loss": 1.0449,
+      "step": 24725
+    },
+    {
+      "epoch": 1.6802554694931375,
+      "grad_norm": 6.375,
+      "learning_rate": 3.949840331566789e-05,
+      "loss": 1.0824,
+      "step": 24730
+    },
+    {
+      "epoch": 1.6805951895637994,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.9496280065226255e-05,
+      "loss": 1.0999,
+      "step": 24735
+    },
+    {
+      "epoch": 1.6809349096344612,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.949415681478462e-05,
+      "loss": 1.0941,
+      "step": 24740
+    },
+    {
+      "epoch": 1.6812746297051229,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.949203356434298e-05,
+      "loss": 1.1664,
+      "step": 24745
+    },
+    {
+      "epoch": 1.6816143497757847,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.948991031390135e-05,
+      "loss": 0.9816,
+      "step": 24750
+    },
+    {
+      "epoch": 1.6819540698464466,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.948778706345971e-05,
+      "loss": 0.994,
+      "step": 24755
+    },
+    {
+      "epoch": 1.6822937899171082,
+      "grad_norm": 6.5,
+      "learning_rate": 3.9485663813018075e-05,
+      "loss": 1.0519,
+      "step": 24760
+    },
+    {
+      "epoch": 1.68263350998777,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.948354056257644e-05,
+      "loss": 1.0819,
+      "step": 24765
+    },
+    {
+      "epoch": 1.682973230058432,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.94814173121348e-05,
+      "loss": 1.0543,
+      "step": 24770
+    },
+    {
+      "epoch": 1.6833129501290935,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.947929406169317e-05,
+      "loss": 1.0005,
+      "step": 24775
+    },
+    {
+      "epoch": 1.6836526701997554,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.947717081125153e-05,
+      "loss": 1.0642,
+      "step": 24780
+    },
+    {
+      "epoch": 1.6839923902704172,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.9475047560809895e-05,
+      "loss": 1.0663,
+      "step": 24785
+    },
+    {
+      "epoch": 1.6843321103410789,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.947292431036826e-05,
+      "loss": 1.0983,
+      "step": 24790
+    },
+    {
+      "epoch": 1.6846718304117407,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.947080105992662e-05,
+      "loss": 1.0106,
+      "step": 24795
+    },
+    {
+      "epoch": 1.6850115504824026,
+      "grad_norm": 5.875,
+      "learning_rate": 3.946867780948499e-05,
+      "loss": 1.072,
+      "step": 24800
+    },
+    {
+      "epoch": 1.6853512705530642,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.946655455904335e-05,
+      "loss": 1.0813,
+      "step": 24805
+    },
+    {
+      "epoch": 1.685690990623726,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.9464431308601715e-05,
+      "loss": 0.9654,
+      "step": 24810
+    },
+    {
+      "epoch": 1.686030710694388,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.946230805816008e-05,
+      "loss": 1.0804,
+      "step": 24815
+    },
+    {
+      "epoch": 1.6863704307650496,
+      "grad_norm": 5.875,
+      "learning_rate": 3.946018480771844e-05,
+      "loss": 1.1225,
+      "step": 24820
+    },
+    {
+      "epoch": 1.6867101508357114,
+      "grad_norm": 5.75,
+      "learning_rate": 3.945806155727681e-05,
+      "loss": 0.9419,
+      "step": 24825
+    },
+    {
+      "epoch": 1.6870498709063733,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.945593830683517e-05,
+      "loss": 1.1021,
+      "step": 24830
+    },
+    {
+      "epoch": 1.687389590977035,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.9453815056393535e-05,
+      "loss": 1.0688,
+      "step": 24835
+    },
+    {
+      "epoch": 1.6877293110476967,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.945169180595189e-05,
+      "loss": 1.1212,
+      "step": 24840
+    },
+    {
+      "epoch": 1.6880690311183586,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.944956855551026e-05,
+      "loss": 1.0524,
+      "step": 24845
+    },
+    {
+      "epoch": 1.6884087511890202,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.944744530506863e-05,
+      "loss": 1.0833,
+      "step": 24850
+    },
+    {
+      "epoch": 1.6887484712596819,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.9445322054626984e-05,
+      "loss": 1.046,
+      "step": 24855
+    },
+    {
+      "epoch": 1.689088191330344,
+      "grad_norm": 6.125,
+      "learning_rate": 3.9443198804185355e-05,
+      "loss": 1.0731,
+      "step": 24860
+    },
+    {
+      "epoch": 1.6894279114010056,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.944107555374372e-05,
+      "loss": 1.0544,
+      "step": 24865
+    },
+    {
+      "epoch": 1.6897676314716672,
+      "grad_norm": 6.375,
+      "learning_rate": 3.9438952303302076e-05,
+      "loss": 1.0828,
+      "step": 24870
+    },
+    {
+      "epoch": 1.6901073515423293,
+      "grad_norm": 5.625,
+      "learning_rate": 3.943682905286045e-05,
+      "loss": 1.1193,
+      "step": 24875
+    },
+    {
+      "epoch": 1.690447071612991,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.943470580241881e-05,
+      "loss": 1.2482,
+      "step": 24880
+    },
+    {
+      "epoch": 1.6907867916836525,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.943258255197717e-05,
+      "loss": 1.0766,
+      "step": 24885
+    },
+    {
+      "epoch": 1.6911265117543146,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.943045930153554e-05,
+      "loss": 1.063,
+      "step": 24890
+    },
+    {
+      "epoch": 1.6914662318249762,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.94283360510939e-05,
+      "loss": 1.0886,
+      "step": 24895
+    },
+    {
+      "epoch": 1.6918059518956379,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.942621280065226e-05,
+      "loss": 1.0396,
+      "step": 24900
+    },
+    {
+      "epoch": 1.6921456719662997,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.942408955021063e-05,
+      "loss": 0.9664,
+      "step": 24905
+    },
+    {
+      "epoch": 1.6924853920369616,
+      "grad_norm": 6.375,
+      "learning_rate": 3.9421966299768995e-05,
+      "loss": 1.1111,
+      "step": 24910
+    },
+    {
+      "epoch": 1.6928251121076232,
+      "grad_norm": 5.75,
+      "learning_rate": 3.941984304932735e-05,
+      "loss": 1.0258,
+      "step": 24915
+    },
+    {
+      "epoch": 1.693164832178285,
+      "grad_norm": 6.0,
+      "learning_rate": 3.941771979888572e-05,
+      "loss": 1.0465,
+      "step": 24920
+    },
+    {
+      "epoch": 1.693504552248947,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.941559654844408e-05,
+      "loss": 1.0098,
+      "step": 24925
+    },
+    {
+      "epoch": 1.6938442723196085,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.9413473298002445e-05,
+      "loss": 1.0488,
+      "step": 24930
+    },
+    {
+      "epoch": 1.6941839923902704,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.9411350047560815e-05,
+      "loss": 1.2128,
+      "step": 24935
+    },
+    {
+      "epoch": 1.6945237124609323,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.940922679711917e-05,
+      "loss": 1.0552,
+      "step": 24940
+    },
+    {
+      "epoch": 1.6948634325315939,
+      "grad_norm": 6.625,
+      "learning_rate": 3.9407103546677537e-05,
+      "loss": 1.1309,
+      "step": 24945
+    },
+    {
+      "epoch": 1.6952031526022557,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.940498029623591e-05,
+      "loss": 1.0138,
+      "step": 24950
+    },
+    {
+      "epoch": 1.6955428726729176,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.9402857045794265e-05,
+      "loss": 1.0561,
+      "step": 24955
+    },
+    {
+      "epoch": 1.6958825927435792,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.940073379535263e-05,
+      "loss": 1.0381,
+      "step": 24960
+    },
+    {
+      "epoch": 1.696222312814241,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.9398610544911e-05,
+      "loss": 0.9565,
+      "step": 24965
+    },
+    {
+      "epoch": 1.696562032884903,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.9396487294469357e-05,
+      "loss": 1.0255,
+      "step": 24970
+    },
+    {
+      "epoch": 1.6969017529555646,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.939436404402772e-05,
+      "loss": 1.0386,
+      "step": 24975
+    },
+    {
+      "epoch": 1.6972414730262264,
+      "grad_norm": 6.25,
+      "learning_rate": 3.939224079358609e-05,
+      "loss": 1.0844,
+      "step": 24980
+    },
+    {
+      "epoch": 1.6975811930968883,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.939011754314445e-05,
+      "loss": 0.9445,
+      "step": 24985
+    },
+    {
+      "epoch": 1.69792091316755,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.938799429270281e-05,
+      "loss": 1.0565,
+      "step": 24990
+    },
+    {
+      "epoch": 1.6982606332382117,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.9385871042261177e-05,
+      "loss": 1.089,
+      "step": 24995
+    },
+    {
+      "epoch": 1.6986003533088736,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.938374779181954e-05,
+      "loss": 1.115,
+      "step": 25000
+    },
+    {
+      "epoch": 1.6989400733795352,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.9381624541377905e-05,
+      "loss": 1.0594,
+      "step": 25005
+    },
+    {
+      "epoch": 1.699279793450197,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.937950129093627e-05,
+      "loss": 1.0382,
+      "step": 25010
+    },
+    {
+      "epoch": 1.699619513520859,
+      "grad_norm": 5.625,
+      "learning_rate": 3.937737804049463e-05,
+      "loss": 1.0554,
+      "step": 25015
+    },
+    {
+      "epoch": 1.6999592335915206,
+      "grad_norm": 6.5,
+      "learning_rate": 3.9375254790052997e-05,
+      "loss": 1.1424,
+      "step": 25020
+    },
+    {
+      "epoch": 1.7002989536621822,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.937313153961136e-05,
+      "loss": 1.0556,
+      "step": 25025
+    },
+    {
+      "epoch": 1.7006386737328443,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.9371008289169725e-05,
+      "loss": 1.0445,
+      "step": 25030
+    },
+    {
+      "epoch": 1.700978393803506,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.936888503872809e-05,
+      "loss": 0.9411,
+      "step": 25035
+    },
+    {
+      "epoch": 1.7013181138741675,
+      "grad_norm": 6.0,
+      "learning_rate": 3.936676178828645e-05,
+      "loss": 1.1064,
+      "step": 25040
+    },
+    {
+      "epoch": 1.7016578339448296,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.936463853784482e-05,
+      "loss": 1.1015,
+      "step": 25045
+    },
+    {
+      "epoch": 1.7019975540154912,
+      "grad_norm": 6.0,
+      "learning_rate": 3.936251528740319e-05,
+      "loss": 1.0849,
+      "step": 25050
+    },
+    {
+      "epoch": 1.7023372740861529,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.9360392036961545e-05,
+      "loss": 1.105,
+      "step": 25055
+    },
+    {
+      "epoch": 1.702676994156815,
+      "grad_norm": 6.0,
+      "learning_rate": 3.935826878651991e-05,
+      "loss": 1.0204,
+      "step": 25060
+    },
+    {
+      "epoch": 1.7030167142274766,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.935614553607828e-05,
+      "loss": 0.9781,
+      "step": 25065
+    },
+    {
+      "epoch": 1.7033564342981382,
+      "grad_norm": 5.75,
+      "learning_rate": 3.935402228563664e-05,
+      "loss": 1.0768,
+      "step": 25070
+    },
+    {
+      "epoch": 1.7036961543688,
+      "grad_norm": 6.25,
+      "learning_rate": 3.9351899035195e-05,
+      "loss": 1.0368,
+      "step": 25075
+    },
+    {
+      "epoch": 1.704035874439462,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.9349775784753365e-05,
+      "loss": 1.0636,
+      "step": 25080
+    },
+    {
+      "epoch": 1.7043755945101235,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.934765253431173e-05,
+      "loss": 1.0041,
+      "step": 25085
+    },
+    {
+      "epoch": 1.7047153145807854,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.934552928387009e-05,
+      "loss": 1.0625,
+      "step": 25090
+    },
+    {
+      "epoch": 1.7050550346514473,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.934340603342846e-05,
+      "loss": 1.0687,
+      "step": 25095
+    },
+    {
+      "epoch": 1.7053947547221089,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.934128278298682e-05,
+      "loss": 1.1425,
+      "step": 25100
+    },
+    {
+      "epoch": 1.7057344747927707,
+      "grad_norm": 6.375,
+      "learning_rate": 3.9339159532545185e-05,
+      "loss": 1.0993,
+      "step": 25105
+    },
+    {
+      "epoch": 1.7060741948634326,
+      "grad_norm": 6.0,
+      "learning_rate": 3.933703628210355e-05,
+      "loss": 1.1042,
+      "step": 25110
+    },
+    {
+      "epoch": 1.7064139149340942,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.933491303166191e-05,
+      "loss": 1.0784,
+      "step": 25115
+    },
+    {
+      "epoch": 1.706753635004756,
+      "grad_norm": 7.0,
+      "learning_rate": 3.933278978122028e-05,
+      "loss": 1.0033,
+      "step": 25120
+    },
+    {
+      "epoch": 1.707093355075418,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.933066653077864e-05,
+      "loss": 1.0139,
+      "step": 25125
+    },
+    {
+      "epoch": 1.7074330751460796,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.9328543280337005e-05,
+      "loss": 1.116,
+      "step": 25130
+    },
+    {
+      "epoch": 1.7077727952167414,
+      "grad_norm": 5.1875,
+      "learning_rate": 3.932642002989537e-05,
+      "loss": 1.1549,
+      "step": 25135
+    },
+    {
+      "epoch": 1.7081125152874033,
+      "grad_norm": 6.0,
+      "learning_rate": 3.932429677945373e-05,
+      "loss": 1.0977,
+      "step": 25140
+    },
+    {
+      "epoch": 1.708452235358065,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.93221735290121e-05,
+      "loss": 1.0174,
+      "step": 25145
+    },
+    {
+      "epoch": 1.7087919554287267,
+      "grad_norm": 5.875,
+      "learning_rate": 3.932005027857046e-05,
+      "loss": 1.0087,
+      "step": 25150
+    },
+    {
+      "epoch": 1.7091316754993886,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.9317927028128825e-05,
+      "loss": 1.1421,
+      "step": 25155
+    },
+    {
+      "epoch": 1.7094713955700502,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.931580377768719e-05,
+      "loss": 1.1902,
+      "step": 25160
+    },
+    {
+      "epoch": 1.709811115640712,
+      "grad_norm": 6.125,
+      "learning_rate": 3.931368052724555e-05,
+      "loss": 1.0257,
+      "step": 25165
+    },
+    {
+      "epoch": 1.710150835711374,
+      "grad_norm": 6.375,
+      "learning_rate": 3.931155727680392e-05,
+      "loss": 1.0569,
+      "step": 25170
+    },
+    {
+      "epoch": 1.7104905557820356,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.930943402636228e-05,
+      "loss": 1.128,
+      "step": 25175
+    },
+    {
+      "epoch": 1.7108302758526974,
+      "grad_norm": 6.0,
+      "learning_rate": 3.9307310775920645e-05,
+      "loss": 1.1004,
+      "step": 25180
+    },
+    {
+      "epoch": 1.7111699959233593,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.930518752547901e-05,
+      "loss": 1.002,
+      "step": 25185
+    },
+    {
+      "epoch": 1.711509715994021,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.930306427503737e-05,
+      "loss": 1.0108,
+      "step": 25190
+    },
+    {
+      "epoch": 1.7118494360646825,
+      "grad_norm": 6.5,
+      "learning_rate": 3.930094102459573e-05,
+      "loss": 1.113,
+      "step": 25195
+    },
+    {
+      "epoch": 1.7121891561353446,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.92988177741541e-05,
+      "loss": 1.0813,
+      "step": 25200
+    },
+    {
+      "epoch": 1.7125288762060062,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.9296694523712465e-05,
+      "loss": 1.0367,
+      "step": 25205
+    },
+    {
+      "epoch": 1.7128685962766679,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.929457127327082e-05,
+      "loss": 1.101,
+      "step": 25210
+    },
+    {
+      "epoch": 1.71320831634733,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.929244802282919e-05,
+      "loss": 1.0614,
+      "step": 25215
+    },
+    {
+      "epoch": 1.7135480364179916,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.929032477238756e-05,
+      "loss": 1.1031,
+      "step": 25220
+    },
+    {
+      "epoch": 1.7138877564886532,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.9288201521945914e-05,
+      "loss": 1.0802,
+      "step": 25225
+    },
+    {
+      "epoch": 1.7142274765593153,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.9286078271504285e-05,
+      "loss": 1.1111,
+      "step": 25230
+    },
+    {
+      "epoch": 1.714567196629977,
+      "grad_norm": 6.75,
+      "learning_rate": 3.928395502106265e-05,
+      "loss": 1.0318,
+      "step": 25235
+    },
+    {
+      "epoch": 1.7149069167006386,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.9281831770621006e-05,
+      "loss": 1.1874,
+      "step": 25240
+    },
+    {
+      "epoch": 1.7152466367713004,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.927970852017938e-05,
+      "loss": 0.9844,
+      "step": 25245
+    },
+    {
+      "epoch": 1.7155863568419623,
+      "grad_norm": 6.25,
+      "learning_rate": 3.927758526973774e-05,
+      "loss": 1.0772,
+      "step": 25250
+    },
+    {
+      "epoch": 1.7159260769126239,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.92754620192961e-05,
+      "loss": 1.0789,
+      "step": 25255
+    },
+    {
+      "epoch": 1.7162657969832857,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.927333876885447e-05,
+      "loss": 1.0336,
+      "step": 25260
+    },
+    {
+      "epoch": 1.7166055170539476,
+      "grad_norm": 5.3125,
+      "learning_rate": 3.927121551841283e-05,
+      "loss": 1.1135,
+      "step": 25265
+    },
+    {
+      "epoch": 1.7169452371246092,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.926909226797119e-05,
+      "loss": 1.1043,
+      "step": 25270
+    },
+    {
+      "epoch": 1.717284957195271,
+      "grad_norm": 5.875,
+      "learning_rate": 3.926696901752956e-05,
+      "loss": 1.1132,
+      "step": 25275
+    },
+    {
+      "epoch": 1.717624677265933,
+      "grad_norm": 6.5,
+      "learning_rate": 3.926484576708792e-05,
+      "loss": 1.0655,
+      "step": 25280
+    },
+    {
+      "epoch": 1.7179643973365946,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.926272251664628e-05,
+      "loss": 1.0695,
+      "step": 25285
+    },
+    {
+      "epoch": 1.7183041174072564,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.926059926620465e-05,
+      "loss": 1.0813,
+      "step": 25290
+    },
+    {
+      "epoch": 1.7186438374779183,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.925847601576301e-05,
+      "loss": 1.0211,
+      "step": 25295
+    },
+    {
+      "epoch": 1.71898355754858,
+      "grad_norm": 6.125,
+      "learning_rate": 3.9256352765321374e-05,
+      "loss": 1.0556,
+      "step": 25300
+    },
+    {
+      "epoch": 1.7193232776192418,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.9254229514879745e-05,
+      "loss": 1.089,
+      "step": 25305
+    },
+    {
+      "epoch": 1.7196629976899036,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.92521062644381e-05,
+      "loss": 1.1583,
+      "step": 25310
+    },
+    {
+      "epoch": 1.7200027177605652,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.9249983013996466e-05,
+      "loss": 1.0704,
+      "step": 25315
+    },
+    {
+      "epoch": 1.720342437831227,
+      "grad_norm": 5.5,
+      "learning_rate": 3.924785976355484e-05,
+      "loss": 1.0262,
+      "step": 25320
+    },
+    {
+      "epoch": 1.720682157901889,
+      "grad_norm": 6.0,
+      "learning_rate": 3.9245736513113194e-05,
+      "loss": 1.0959,
+      "step": 25325
+    },
+    {
+      "epoch": 1.7210218779725506,
+      "grad_norm": 6.125,
+      "learning_rate": 3.924361326267156e-05,
+      "loss": 1.1743,
+      "step": 25330
+    },
+    {
+      "epoch": 1.7213615980432124,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.924149001222993e-05,
+      "loss": 0.9895,
+      "step": 25335
+    },
+    {
+      "epoch": 1.7217013181138743,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9239366761788286e-05,
+      "loss": 1.051,
+      "step": 25340
+    },
+    {
+      "epoch": 1.722041038184536,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.923724351134665e-05,
+      "loss": 1.026,
+      "step": 25345
+    },
+    {
+      "epoch": 1.7223807582551978,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.9235120260905014e-05,
+      "loss": 1.0481,
+      "step": 25350
+    },
+    {
+      "epoch": 1.7227204783258596,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.923299701046338e-05,
+      "loss": 1.0641,
+      "step": 25355
+    },
+    {
+      "epoch": 1.7230601983965212,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.923087376002174e-05,
+      "loss": 1.0773,
+      "step": 25360
+    },
+    {
+      "epoch": 1.7233999184671829,
+      "grad_norm": 6.375,
+      "learning_rate": 3.9228750509580106e-05,
+      "loss": 1.0355,
+      "step": 25365
+    },
+    {
+      "epoch": 1.723739638537845,
+      "grad_norm": 6.125,
+      "learning_rate": 3.922662725913847e-05,
+      "loss": 1.1266,
+      "step": 25370
+    },
+    {
+      "epoch": 1.7240793586085066,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.9224504008696834e-05,
+      "loss": 1.067,
+      "step": 25375
+    },
+    {
+      "epoch": 1.7244190786791682,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.92223807582552e-05,
+      "loss": 1.0481,
+      "step": 25380
+    },
+    {
+      "epoch": 1.7247587987498303,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.922025750781356e-05,
+      "loss": 1.0662,
+      "step": 25385
+    },
+    {
+      "epoch": 1.725098518820492,
+      "grad_norm": 6.25,
+      "learning_rate": 3.921813425737193e-05,
+      "loss": 1.1056,
+      "step": 25390
+    },
+    {
+      "epoch": 1.7254382388911536,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.921601100693029e-05,
+      "loss": 1.1206,
+      "step": 25395
+    },
+    {
+      "epoch": 1.7257779589618156,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.9213887756488654e-05,
+      "loss": 1.0564,
+      "step": 25400
+    },
+    {
+      "epoch": 1.7261176790324773,
+      "grad_norm": 5.5,
+      "learning_rate": 3.9211764506047025e-05,
+      "loss": 1.053,
+      "step": 25405
+    },
+    {
+      "epoch": 1.726457399103139,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.920964125560538e-05,
+      "loss": 1.0685,
+      "step": 25410
+    },
+    {
+      "epoch": 1.7267971191738007,
+      "grad_norm": 6.5,
+      "learning_rate": 3.9207518005163746e-05,
+      "loss": 0.9973,
+      "step": 25415
+    },
+    {
+      "epoch": 1.7271368392444626,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.920539475472211e-05,
+      "loss": 1.1249,
+      "step": 25420
+    },
+    {
+      "epoch": 1.7274765593151242,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.9203271504280474e-05,
+      "loss": 1.0494,
+      "step": 25425
+    },
+    {
+      "epoch": 1.727816279385786,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.920114825383884e-05,
+      "loss": 0.966,
+      "step": 25430
+    },
+    {
+      "epoch": 1.728155999456448,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.91990250033972e-05,
+      "loss": 0.9511,
+      "step": 25435
+    },
+    {
+      "epoch": 1.7284957195271096,
+      "grad_norm": 6.0,
+      "learning_rate": 3.9196901752955566e-05,
+      "loss": 1.0216,
+      "step": 25440
+    },
+    {
+      "epoch": 1.7288354395977714,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.919477850251393e-05,
+      "loss": 1.015,
+      "step": 25445
+    },
+    {
+      "epoch": 1.7291751596684333,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9192655252072294e-05,
+      "loss": 1.085,
+      "step": 25450
+    },
+    {
+      "epoch": 1.729514879739095,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.919053200163066e-05,
+      "loss": 1.1158,
+      "step": 25455
+    },
+    {
+      "epoch": 1.7298545998097568,
+      "grad_norm": 6.125,
+      "learning_rate": 3.918840875118902e-05,
+      "loss": 1.0796,
+      "step": 25460
+    },
+    {
+      "epoch": 1.7301943198804186,
+      "grad_norm": 9.3125,
+      "learning_rate": 3.9186285500747386e-05,
+      "loss": 0.9426,
+      "step": 25465
+    },
+    {
+      "epoch": 1.7305340399510802,
+      "grad_norm": 5.25,
+      "learning_rate": 3.918416225030575e-05,
+      "loss": 0.9782,
+      "step": 25470
+    },
+    {
+      "epoch": 1.730873760021742,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.9182038999864114e-05,
+      "loss": 1.0536,
+      "step": 25475
+    },
+    {
+      "epoch": 1.731213480092404,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.917991574942248e-05,
+      "loss": 1.0821,
+      "step": 25480
+    },
+    {
+      "epoch": 1.7315532001630656,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.917779249898084e-05,
+      "loss": 1.0955,
+      "step": 25485
+    },
+    {
+      "epoch": 1.7318929202337274,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.9175669248539206e-05,
+      "loss": 1.0346,
+      "step": 25490
+    },
+    {
+      "epoch": 1.7322326403043893,
+      "grad_norm": 6.125,
+      "learning_rate": 3.917354599809757e-05,
+      "loss": 1.116,
+      "step": 25495
+    },
+    {
+      "epoch": 1.732572360375051,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.9171422747655934e-05,
+      "loss": 1.1262,
+      "step": 25500
+    },
+    {
+      "epoch": 1.7329120804457128,
+      "grad_norm": 5.25,
+      "learning_rate": 3.91692994972143e-05,
+      "loss": 1.0533,
+      "step": 25505
+    },
+    {
+      "epoch": 1.7332518005163746,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.916717624677266e-05,
+      "loss": 1.1864,
+      "step": 25510
+    },
+    {
+      "epoch": 1.7335915205870362,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.9165052996331026e-05,
+      "loss": 1.0348,
+      "step": 25515
+    },
+    {
+      "epoch": 1.733931240657698,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.916292974588939e-05,
+      "loss": 1.0451,
+      "step": 25520
+    },
+    {
+      "epoch": 1.73427096072836,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.9160806495447754e-05,
+      "loss": 1.2817,
+      "step": 25525
+    },
+    {
+      "epoch": 1.7346106807990216,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.915868324500612e-05,
+      "loss": 1.1186,
+      "step": 25530
+    },
+    {
+      "epoch": 1.7349504008696832,
+      "grad_norm": 5.625,
+      "learning_rate": 3.915655999456448e-05,
+      "loss": 1.0786,
+      "step": 25535
+    },
+    {
+      "epoch": 1.7352901209403453,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.9154436744122846e-05,
+      "loss": 1.0628,
+      "step": 25540
+    },
+    {
+      "epoch": 1.735629841011007,
+      "grad_norm": 6.0,
+      "learning_rate": 3.915231349368121e-05,
+      "loss": 1.0811,
+      "step": 25545
+    },
+    {
+      "epoch": 1.7359695610816686,
+      "grad_norm": 5.5,
+      "learning_rate": 3.915019024323957e-05,
+      "loss": 1.0443,
+      "step": 25550
+    },
+    {
+      "epoch": 1.7363092811523306,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.914806699279794e-05,
+      "loss": 1.0587,
+      "step": 25555
+    },
+    {
+      "epoch": 1.7366490012229923,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.91459437423563e-05,
+      "loss": 1.1399,
+      "step": 25560
+    },
+    {
+      "epoch": 1.736988721293654,
+      "grad_norm": 6.25,
+      "learning_rate": 3.914382049191466e-05,
+      "loss": 1.0611,
+      "step": 25565
+    },
+    {
+      "epoch": 1.737328441364316,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.914169724147303e-05,
+      "loss": 1.0442,
+      "step": 25570
+    },
+    {
+      "epoch": 1.7376681614349776,
+      "grad_norm": 6.875,
+      "learning_rate": 3.9139573991031394e-05,
+      "loss": 1.102,
+      "step": 25575
+    },
+    {
+      "epoch": 1.7380078815056392,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.913745074058975e-05,
+      "loss": 1.0295,
+      "step": 25580
+    },
+    {
+      "epoch": 1.738347601576301,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.913532749014812e-05,
+      "loss": 1.0234,
+      "step": 25585
+    },
+    {
+      "epoch": 1.738687321646963,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.9133204239706486e-05,
+      "loss": 1.1409,
+      "step": 25590
+    },
+    {
+      "epoch": 1.7390270417176246,
+      "grad_norm": 6.0,
+      "learning_rate": 3.9131080989264843e-05,
+      "loss": 1.1512,
+      "step": 25595
+    },
+    {
+      "epoch": 1.7393667617882864,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.9128957738823214e-05,
+      "loss": 1.0421,
+      "step": 25600
+    },
+    {
+      "epoch": 1.7397064818589483,
+      "grad_norm": 5.3125,
+      "learning_rate": 3.912683448838158e-05,
+      "loss": 1.0769,
+      "step": 25605
+    },
+    {
+      "epoch": 1.74004620192961,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.9124711237939936e-05,
+      "loss": 1.0525,
+      "step": 25610
+    },
+    {
+      "epoch": 1.7403859220002718,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.9122587987498306e-05,
+      "loss": 1.0961,
+      "step": 25615
+    },
+    {
+      "epoch": 1.7407256420709336,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.9120464737056664e-05,
+      "loss": 1.0873,
+      "step": 25620
+    },
+    {
+      "epoch": 1.7410653621415952,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.911834148661503e-05,
+      "loss": 1.0565,
+      "step": 25625
+    },
+    {
+      "epoch": 1.741405082212257,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.91162182361734e-05,
+      "loss": 1.1104,
+      "step": 25630
+    },
+    {
+      "epoch": 1.741744802282919,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.9114094985731756e-05,
+      "loss": 1.0251,
+      "step": 25635
+    },
+    {
+      "epoch": 1.7420845223535806,
+      "grad_norm": 4.90625,
+      "learning_rate": 3.911197173529012e-05,
+      "loss": 1.0977,
+      "step": 25640
+    },
+    {
+      "epoch": 1.7424242424242424,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.910984848484849e-05,
+      "loss": 1.0788,
+      "step": 25645
+    },
+    {
+      "epoch": 1.7427639624949043,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.910772523440685e-05,
+      "loss": 1.0598,
+      "step": 25650
+    },
+    {
+      "epoch": 1.743103682565566,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.910560198396521e-05,
+      "loss": 1.0782,
+      "step": 25655
+    },
+    {
+      "epoch": 1.7434434026362278,
+      "grad_norm": 5.5,
+      "learning_rate": 3.910347873352358e-05,
+      "loss": 1.0631,
+      "step": 25660
+    },
+    {
+      "epoch": 1.7437831227068896,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.910135548308194e-05,
+      "loss": 1.0311,
+      "step": 25665
+    },
+    {
+      "epoch": 1.7441228427775513,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9099232232640304e-05,
+      "loss": 1.189,
+      "step": 25670
+    },
+    {
+      "epoch": 1.744462562848213,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.9097108982198674e-05,
+      "loss": 1.0502,
+      "step": 25675
+    },
+    {
+      "epoch": 1.744802282918875,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.909498573175703e-05,
+      "loss": 1.1632,
+      "step": 25680
+    },
+    {
+      "epoch": 1.7451420029895366,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9092862481315396e-05,
+      "loss": 0.9584,
+      "step": 25685
+    },
+    {
+      "epoch": 1.7454817230601984,
+      "grad_norm": 5.75,
+      "learning_rate": 3.9090739230873766e-05,
+      "loss": 1.0848,
+      "step": 25690
+    },
+    {
+      "epoch": 1.7458214431308603,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.9088615980432124e-05,
+      "loss": 1.0238,
+      "step": 25695
+    },
+    {
+      "epoch": 1.746161163201522,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.908649272999049e-05,
+      "loss": 1.0953,
+      "step": 25700
+    },
+    {
+      "epoch": 1.7465008832721836,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.908436947954885e-05,
+      "loss": 1.0627,
+      "step": 25705
+    },
+    {
+      "epoch": 1.7468406033428456,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.9082246229107216e-05,
+      "loss": 1.0225,
+      "step": 25710
+    },
+    {
+      "epoch": 1.7471803234135073,
+      "grad_norm": 7.25,
+      "learning_rate": 3.908012297866558e-05,
+      "loss": 1.0199,
+      "step": 25715
+    },
+    {
+      "epoch": 1.747520043484169,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.9077999728223944e-05,
+      "loss": 1.1718,
+      "step": 25720
+    },
+    {
+      "epoch": 1.747859763554831,
+      "grad_norm": 6.125,
+      "learning_rate": 3.907587647778231e-05,
+      "loss": 0.9788,
+      "step": 25725
+    },
+    {
+      "epoch": 1.7481994836254926,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.907375322734068e-05,
+      "loss": 1.0649,
+      "step": 25730
+    },
+    {
+      "epoch": 1.7485392036961542,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.9071629976899036e-05,
+      "loss": 0.9807,
+      "step": 25735
+    },
+    {
+      "epoch": 1.7488789237668163,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.90695067264574e-05,
+      "loss": 1.042,
+      "step": 25740
+    },
+    {
+      "epoch": 1.749218643837478,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.906738347601577e-05,
+      "loss": 1.0543,
+      "step": 25745
+    },
+    {
+      "epoch": 1.7495583639081396,
+      "grad_norm": 6.0,
+      "learning_rate": 3.906526022557413e-05,
+      "loss": 1.0023,
+      "step": 25750
+    },
+    {
+      "epoch": 1.7498980839788014,
+      "grad_norm": 7.0,
+      "learning_rate": 3.906313697513249e-05,
+      "loss": 1.0889,
+      "step": 25755
+    },
+    {
+      "epoch": 1.7502378040494633,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.906101372469086e-05,
+      "loss": 1.1318,
+      "step": 25760
+    },
+    {
+      "epoch": 1.750577524120125,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.905889047424922e-05,
+      "loss": 1.0416,
+      "step": 25765
+    },
+    {
+      "epoch": 1.7509172441907868,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.9056767223807584e-05,
+      "loss": 1.1102,
+      "step": 25770
+    },
+    {
+      "epoch": 1.7512569642614486,
+      "grad_norm": 6.0,
+      "learning_rate": 3.905464397336595e-05,
+      "loss": 1.0692,
+      "step": 25775
+    },
+    {
+      "epoch": 1.7515966843321102,
+      "grad_norm": 6.25,
+      "learning_rate": 3.905252072292431e-05,
+      "loss": 1.097,
+      "step": 25780
+    },
+    {
+      "epoch": 1.751936404402772,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.9050397472482676e-05,
+      "loss": 1.0237,
+      "step": 25785
+    },
+    {
+      "epoch": 1.752276124473434,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.904827422204104e-05,
+      "loss": 1.0754,
+      "step": 25790
+    },
+    {
+      "epoch": 1.7526158445440956,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.9046150971599404e-05,
+      "loss": 1.1522,
+      "step": 25795
+    },
+    {
+      "epoch": 1.7529555646147574,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.904402772115777e-05,
+      "loss": 1.3762,
+      "step": 25800
+    },
+    {
+      "epoch": 1.7532952846854193,
+      "grad_norm": 6.75,
+      "learning_rate": 3.904190447071613e-05,
+      "loss": 1.1159,
+      "step": 25805
+    },
+    {
+      "epoch": 1.753635004756081,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.9039781220274496e-05,
+      "loss": 1.0489,
+      "step": 25810
+    },
+    {
+      "epoch": 1.7539747248267428,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.903765796983286e-05,
+      "loss": 1.0581,
+      "step": 25815
+    },
+    {
+      "epoch": 1.7543144448974046,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.9035534719391224e-05,
+      "loss": 0.9873,
+      "step": 25820
+    },
+    {
+      "epoch": 1.7546541649680663,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.903341146894959e-05,
+      "loss": 1.028,
+      "step": 25825
+    },
+    {
+      "epoch": 1.754993885038728,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.903128821850795e-05,
+      "loss": 0.9848,
+      "step": 25830
+    },
+    {
+      "epoch": 1.75533360510939,
+      "grad_norm": 6.5,
+      "learning_rate": 3.9029164968066316e-05,
+      "loss": 1.0164,
+      "step": 25835
+    },
+    {
+      "epoch": 1.7556733251800516,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.902704171762468e-05,
+      "loss": 1.1356,
+      "step": 25840
+    },
+    {
+      "epoch": 1.7560130452507134,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.9024918467183044e-05,
+      "loss": 1.1112,
+      "step": 25845
+    },
+    {
+      "epoch": 1.7563527653213753,
+      "grad_norm": 5.875,
+      "learning_rate": 3.902279521674141e-05,
+      "loss": 1.0855,
+      "step": 25850
+    },
+    {
+      "epoch": 1.756692485392037,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.902067196629977e-05,
+      "loss": 1.0474,
+      "step": 25855
+    },
+    {
+      "epoch": 1.7570322054626988,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.9018548715858136e-05,
+      "loss": 0.9952,
+      "step": 25860
+    },
+    {
+      "epoch": 1.7573719255333606,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.90164254654165e-05,
+      "loss": 1.087,
+      "step": 25865
+    },
+    {
+      "epoch": 1.7577116456040223,
+      "grad_norm": 6.0,
+      "learning_rate": 3.9014302214974864e-05,
+      "loss": 1.0242,
+      "step": 25870
+    },
+    {
+      "epoch": 1.758051365674684,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.901217896453323e-05,
+      "loss": 1.0647,
+      "step": 25875
+    },
+    {
+      "epoch": 1.758391085745346,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.901005571409159e-05,
+      "loss": 1.0909,
+      "step": 25880
+    },
+    {
+      "epoch": 1.7587308058160076,
+      "grad_norm": 5.875,
+      "learning_rate": 3.9007932463649956e-05,
+      "loss": 1.0758,
+      "step": 25885
+    },
+    {
+      "epoch": 1.7590705258866692,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.900580921320832e-05,
+      "loss": 1.0844,
+      "step": 25890
+    },
+    {
+      "epoch": 1.7594102459573313,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.9003685962766684e-05,
+      "loss": 0.9992,
+      "step": 25895
+    },
+    {
+      "epoch": 1.759749966027993,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.900156271232505e-05,
+      "loss": 1.0536,
+      "step": 25900
+    },
+    {
+      "epoch": 1.7600896860986546,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.8999439461883405e-05,
+      "loss": 1.0451,
+      "step": 25905
+    },
+    {
+      "epoch": 1.7604294061693166,
+      "grad_norm": 5.75,
+      "learning_rate": 3.8997316211441776e-05,
+      "loss": 1.0164,
+      "step": 25910
+    },
+    {
+      "epoch": 1.7607691262399783,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.899519296100014e-05,
+      "loss": 1.0596,
+      "step": 25915
+    },
+    {
+      "epoch": 1.76110884631064,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.89930697105585e-05,
+      "loss": 1.0724,
+      "step": 25920
+    },
+    {
+      "epoch": 1.7614485663813018,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.899094646011687e-05,
+      "loss": 1.0681,
+      "step": 25925
+    },
+    {
+      "epoch": 1.7617882864519636,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.898882320967523e-05,
+      "loss": 1.084,
+      "step": 25930
+    },
+    {
+      "epoch": 1.7621280065226252,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.898669995923359e-05,
+      "loss": 1.0638,
+      "step": 25935
+    },
+    {
+      "epoch": 1.762467726593287,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.898457670879196e-05,
+      "loss": 1.0539,
+      "step": 25940
+    },
+    {
+      "epoch": 1.762807446663949,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.8982453458350324e-05,
+      "loss": 1.1317,
+      "step": 25945
+    },
+    {
+      "epoch": 1.7631471667346106,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.898033020790868e-05,
+      "loss": 1.0382,
+      "step": 25950
+    },
+    {
+      "epoch": 1.7634868868052724,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.897820695746705e-05,
+      "loss": 1.1182,
+      "step": 25955
+    },
+    {
+      "epoch": 1.7638266068759343,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.8976083707025416e-05,
+      "loss": 1.0212,
+      "step": 25960
+    },
+    {
+      "epoch": 1.764166326946596,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.897396045658377e-05,
+      "loss": 1.081,
+      "step": 25965
+    },
+    {
+      "epoch": 1.7645060470172578,
+      "grad_norm": 6.25,
+      "learning_rate": 3.8971837206142144e-05,
+      "loss": 1.0895,
+      "step": 25970
+    },
+    {
+      "epoch": 1.7648457670879196,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.89697139557005e-05,
+      "loss": 1.0742,
+      "step": 25975
+    },
+    {
+      "epoch": 1.7651854871585813,
+      "grad_norm": 4.78125,
+      "learning_rate": 3.8967590705258865e-05,
+      "loss": 0.945,
+      "step": 25980
+    },
+    {
+      "epoch": 1.765525207229243,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.8965467454817236e-05,
+      "loss": 1.0273,
+      "step": 25985
+    },
+    {
+      "epoch": 1.765864927299905,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.896334420437559e-05,
+      "loss": 1.0821,
+      "step": 25990
+    },
+    {
+      "epoch": 1.7662046473705666,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.896122095393396e-05,
+      "loss": 1.0632,
+      "step": 25995
+    },
+    {
+      "epoch": 1.7665443674412284,
+      "grad_norm": 6.375,
+      "learning_rate": 3.895909770349233e-05,
+      "loss": 1.0457,
+      "step": 26000
+    },
+    {
+      "epoch": 1.7668840875118903,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.8956974453050685e-05,
+      "loss": 1.078,
+      "step": 26005
+    },
+    {
+      "epoch": 1.767223807582552,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.895485120260905e-05,
+      "loss": 1.1201,
+      "step": 26010
+    },
+    {
+      "epoch": 1.7675635276532138,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.895272795216742e-05,
+      "loss": 1.0828,
+      "step": 26015
+    },
+    {
+      "epoch": 1.7679032477238756,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.895060470172578e-05,
+      "loss": 1.0476,
+      "step": 26020
+    },
+    {
+      "epoch": 1.7682429677945373,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.894848145128414e-05,
+      "loss": 0.8924,
+      "step": 26025
+    },
+    {
+      "epoch": 1.7685826878651991,
+      "grad_norm": 6.0,
+      "learning_rate": 3.894635820084251e-05,
+      "loss": 0.9658,
+      "step": 26030
+    },
+    {
+      "epoch": 1.768922407935861,
+      "grad_norm": 5.75,
+      "learning_rate": 3.894423495040087e-05,
+      "loss": 1.0589,
+      "step": 26035
+    },
+    {
+      "epoch": 1.7692621280065226,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.894211169995923e-05,
+      "loss": 1.0915,
+      "step": 26040
+    },
+    {
+      "epoch": 1.7696018480771842,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.89399884495176e-05,
+      "loss": 1.0598,
+      "step": 26045
+    },
+    {
+      "epoch": 1.7699415681478463,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.893786519907596e-05,
+      "loss": 1.0323,
+      "step": 26050
+    },
+    {
+      "epoch": 1.770281288218508,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.8935741948634325e-05,
+      "loss": 1.061,
+      "step": 26055
+    },
+    {
+      "epoch": 1.7706210082891696,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.893361869819269e-05,
+      "loss": 1.0783,
+      "step": 26060
+    },
+    {
+      "epoch": 1.7709607283598316,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.893149544775105e-05,
+      "loss": 1.0387,
+      "step": 26065
+    },
+    {
+      "epoch": 1.7713004484304933,
+      "grad_norm": 5.875,
+      "learning_rate": 3.8929372197309424e-05,
+      "loss": 1.0357,
+      "step": 26070
+    },
+    {
+      "epoch": 1.771640168501155,
+      "grad_norm": 6.0,
+      "learning_rate": 3.892724894686778e-05,
+      "loss": 0.9822,
+      "step": 26075
+    },
+    {
+      "epoch": 1.771979888571817,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.8925125696426145e-05,
+      "loss": 1.1033,
+      "step": 26080
+    },
+    {
+      "epoch": 1.7723196086424786,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.8923002445984516e-05,
+      "loss": 0.9833,
+      "step": 26085
+    },
+    {
+      "epoch": 1.7726593287131402,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.892087919554287e-05,
+      "loss": 1.0556,
+      "step": 26090
+    },
+    {
+      "epoch": 1.772999048783802,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.891875594510124e-05,
+      "loss": 0.97,
+      "step": 26095
+    },
+    {
+      "epoch": 1.773338768854464,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.891663269465961e-05,
+      "loss": 1.0372,
+      "step": 26100
+    },
+    {
+      "epoch": 1.7736784889251256,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.8914509444217965e-05,
+      "loss": 1.0999,
+      "step": 26105
+    },
+    {
+      "epoch": 1.7740182089957874,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.891238619377633e-05,
+      "loss": 1.0369,
+      "step": 26110
+    },
+    {
+      "epoch": 1.7743579290664493,
+      "grad_norm": 6.75,
+      "learning_rate": 3.89102629433347e-05,
+      "loss": 1.0826,
+      "step": 26115
+    },
+    {
+      "epoch": 1.774697649137111,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.890813969289306e-05,
+      "loss": 1.0409,
+      "step": 26120
+    },
+    {
+      "epoch": 1.7750373692077728,
+      "grad_norm": 5.75,
+      "learning_rate": 3.890601644245142e-05,
+      "loss": 1.0575,
+      "step": 26125
+    },
+    {
+      "epoch": 1.7753770892784346,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.8903893192009785e-05,
+      "loss": 1.0909,
+      "step": 26130
+    },
+    {
+      "epoch": 1.7757168093490963,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.890176994156815e-05,
+      "loss": 0.976,
+      "step": 26135
+    },
+    {
+      "epoch": 1.7760565294197581,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.889964669112651e-05,
+      "loss": 1.0212,
+      "step": 26140
+    },
+    {
+      "epoch": 1.77639624949042,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.889752344068488e-05,
+      "loss": 0.974,
+      "step": 26145
+    },
+    {
+      "epoch": 1.7767359695610816,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.889540019024324e-05,
+      "loss": 1.0282,
+      "step": 26150
+    },
+    {
+      "epoch": 1.7770756896317434,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8893276939801605e-05,
+      "loss": 1.0401,
+      "step": 26155
+    },
+    {
+      "epoch": 1.7774154097024053,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.889115368935997e-05,
+      "loss": 0.9768,
+      "step": 26160
+    },
+    {
+      "epoch": 1.777755129773067,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.888903043891833e-05,
+      "loss": 1.0773,
+      "step": 26165
+    },
+    {
+      "epoch": 1.7780948498437288,
+      "grad_norm": 6.625,
+      "learning_rate": 3.88869071884767e-05,
+      "loss": 1.1183,
+      "step": 26170
+    },
+    {
+      "epoch": 1.7784345699143906,
+      "grad_norm": 5.875,
+      "learning_rate": 3.888478393803506e-05,
+      "loss": 1.0175,
+      "step": 26175
+    },
+    {
+      "epoch": 1.7787742899850523,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.8882660687593425e-05,
+      "loss": 1.1098,
+      "step": 26180
+    },
+    {
+      "epoch": 1.7791140100557141,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.888053743715179e-05,
+      "loss": 1.1185,
+      "step": 26185
+    },
+    {
+      "epoch": 1.779453730126376,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.887841418671015e-05,
+      "loss": 1.0479,
+      "step": 26190
+    },
+    {
+      "epoch": 1.7797934501970376,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.887629093626852e-05,
+      "loss": 1.0177,
+      "step": 26195
+    },
+    {
+      "epoch": 1.7801331702676995,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.887416768582688e-05,
+      "loss": 0.9493,
+      "step": 26200
+    },
+    {
+      "epoch": 1.7804728903383613,
+      "grad_norm": 6.125,
+      "learning_rate": 3.8872044435385245e-05,
+      "loss": 0.9689,
+      "step": 26205
+    },
+    {
+      "epoch": 1.780812610409023,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.886992118494361e-05,
+      "loss": 1.1185,
+      "step": 26210
+    },
+    {
+      "epoch": 1.7811523304796846,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.886779793450197e-05,
+      "loss": 1.0452,
+      "step": 26215
+    },
+    {
+      "epoch": 1.7814920505503467,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.886567468406034e-05,
+      "loss": 1.032,
+      "step": 26220
+    },
+    {
+      "epoch": 1.7818317706210083,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.88635514336187e-05,
+      "loss": 1.0891,
+      "step": 26225
+    },
+    {
+      "epoch": 1.78217149069167,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.8861428183177065e-05,
+      "loss": 1.1462,
+      "step": 26230
+    },
+    {
+      "epoch": 1.782511210762332,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.885930493273543e-05,
+      "loss": 1.0576,
+      "step": 26235
+    },
+    {
+      "epoch": 1.7828509308329936,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.885718168229379e-05,
+      "loss": 1.0264,
+      "step": 26240
+    },
+    {
+      "epoch": 1.7831906509036552,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.885505843185215e-05,
+      "loss": 1.1166,
+      "step": 26245
+    },
+    {
+      "epoch": 1.7835303709743173,
+      "grad_norm": 6.625,
+      "learning_rate": 3.885293518141052e-05,
+      "loss": 1.1191,
+      "step": 26250
+    },
+    {
+      "epoch": 1.783870091044979,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.8850811930968885e-05,
+      "loss": 1.0372,
+      "step": 26255
+    },
+    {
+      "epoch": 1.7842098111156406,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.884868868052724e-05,
+      "loss": 1.0621,
+      "step": 26260
+    },
+    {
+      "epoch": 1.7845495311863024,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.884656543008561e-05,
+      "loss": 1.1006,
+      "step": 26265
+    },
+    {
+      "epoch": 1.7848892512569643,
+      "grad_norm": 5.875,
+      "learning_rate": 3.884444217964398e-05,
+      "loss": 1.1109,
+      "step": 26270
+    },
+    {
+      "epoch": 1.785228971327626,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.8842318929202334e-05,
+      "loss": 1.0864,
+      "step": 26275
+    },
+    {
+      "epoch": 1.7855686913982878,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.8840195678760705e-05,
+      "loss": 1.0518,
+      "step": 26280
+    },
+    {
+      "epoch": 1.7859084114689496,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.883807242831907e-05,
+      "loss": 1.1886,
+      "step": 26285
+    },
+    {
+      "epoch": 1.7862481315396113,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.8835949177877426e-05,
+      "loss": 1.0512,
+      "step": 26290
+    },
+    {
+      "epoch": 1.7865878516102731,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.88338259274358e-05,
+      "loss": 1.0986,
+      "step": 26295
+    },
+    {
+      "epoch": 1.786927571680935,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.883170267699416e-05,
+      "loss": 1.0993,
+      "step": 26300
+    },
+    {
+      "epoch": 1.7872672917515966,
+      "grad_norm": 6.0,
+      "learning_rate": 3.882957942655252e-05,
+      "loss": 1.02,
+      "step": 26305
+    },
+    {
+      "epoch": 1.7876070118222585,
+      "grad_norm": 6.375,
+      "learning_rate": 3.882745617611089e-05,
+      "loss": 1.0746,
+      "step": 26310
+    },
+    {
+      "epoch": 1.7879467318929203,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.882533292566925e-05,
+      "loss": 1.1261,
+      "step": 26315
+    },
+    {
+      "epoch": 1.788286451963582,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.882320967522761e-05,
+      "loss": 1.112,
+      "step": 26320
+    },
+    {
+      "epoch": 1.7886261720342438,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.882108642478598e-05,
+      "loss": 1.0155,
+      "step": 26325
+    },
+    {
+      "epoch": 1.7889658921049056,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.881896317434434e-05,
+      "loss": 1.0361,
+      "step": 26330
+    },
+    {
+      "epoch": 1.7893056121755673,
+      "grad_norm": 5.75,
+      "learning_rate": 3.88168399239027e-05,
+      "loss": 0.9814,
+      "step": 26335
+    },
+    {
+      "epoch": 1.7896453322462291,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.881471667346107e-05,
+      "loss": 1.0178,
+      "step": 26340
+    },
+    {
+      "epoch": 1.789985052316891,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.881259342301943e-05,
+      "loss": 1.0338,
+      "step": 26345
+    },
+    {
+      "epoch": 1.7903247723875526,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.8810470172577795e-05,
+      "loss": 1.1131,
+      "step": 26350
+    },
+    {
+      "epoch": 1.7906644924582145,
+      "grad_norm": 6.5,
+      "learning_rate": 3.8808346922136165e-05,
+      "loss": 1.0799,
+      "step": 26355
+    },
+    {
+      "epoch": 1.7910042125288763,
+      "grad_norm": 5.875,
+      "learning_rate": 3.880622367169452e-05,
+      "loss": 1.0191,
+      "step": 26360
+    },
+    {
+      "epoch": 1.791343932599538,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.8804100421252887e-05,
+      "loss": 1.0927,
+      "step": 26365
+    },
+    {
+      "epoch": 1.7916836526701998,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.880197717081126e-05,
+      "loss": 1.0011,
+      "step": 26370
+    },
+    {
+      "epoch": 1.7920233727408617,
+      "grad_norm": 5.5,
+      "learning_rate": 3.8799853920369615e-05,
+      "loss": 1.1225,
+      "step": 26375
+    },
+    {
+      "epoch": 1.7923630928115233,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.879773066992798e-05,
+      "loss": 1.1067,
+      "step": 26380
+    },
+    {
+      "epoch": 1.792702812882185,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.879560741948635e-05,
+      "loss": 1.0693,
+      "step": 26385
+    },
+    {
+      "epoch": 1.793042532952847,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.8793484169044707e-05,
+      "loss": 1.096,
+      "step": 26390
+    },
+    {
+      "epoch": 1.7933822530235086,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.879136091860307e-05,
+      "loss": 1.1074,
+      "step": 26395
+    },
+    {
+      "epoch": 1.7937219730941703,
+      "grad_norm": 7.125,
+      "learning_rate": 3.8789237668161435e-05,
+      "loss": 1.0309,
+      "step": 26400
+    },
+    {
+      "epoch": 1.7940616931648323,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.87871144177198e-05,
+      "loss": 1.0324,
+      "step": 26405
+    },
+    {
+      "epoch": 1.794401413235494,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.878499116727817e-05,
+      "loss": 1.1187,
+      "step": 26410
+    },
+    {
+      "epoch": 1.7947411333061556,
+      "grad_norm": 4.9375,
+      "learning_rate": 3.8782867916836527e-05,
+      "loss": 1.0694,
+      "step": 26415
+    },
+    {
+      "epoch": 1.7950808533768177,
+      "grad_norm": 5.5,
+      "learning_rate": 3.878074466639489e-05,
+      "loss": 1.0632,
+      "step": 26420
+    },
+    {
+      "epoch": 1.7954205734474793,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.877862141595326e-05,
+      "loss": 1.0749,
+      "step": 26425
+    },
+    {
+      "epoch": 1.795760293518141,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.877649816551162e-05,
+      "loss": 1.0112,
+      "step": 26430
+    },
+    {
+      "epoch": 1.7961000135888028,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.877437491506998e-05,
+      "loss": 1.098,
+      "step": 26435
+    },
+    {
+      "epoch": 1.7964397336594646,
+      "grad_norm": 5.375,
+      "learning_rate": 3.877225166462835e-05,
+      "loss": 1.1697,
+      "step": 26440
+    },
+    {
+      "epoch": 1.7967794537301263,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.877012841418671e-05,
+      "loss": 1.082,
+      "step": 26445
+    },
+    {
+      "epoch": 1.7971191738007881,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.8768005163745075e-05,
+      "loss": 0.9362,
+      "step": 26450
+    },
+    {
+      "epoch": 1.79745889387145,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.8765881913303445e-05,
+      "loss": 1.0556,
+      "step": 26455
+    },
+    {
+      "epoch": 1.7977986139421116,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.87637586628618e-05,
+      "loss": 1.0528,
+      "step": 26460
+    },
+    {
+      "epoch": 1.7981383340127735,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8761635412420167e-05,
+      "loss": 1.0737,
+      "step": 26465
+    },
+    {
+      "epoch": 1.7984780540834353,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.875951216197853e-05,
+      "loss": 1.0754,
+      "step": 26470
+    },
+    {
+      "epoch": 1.798817774154097,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.8757388911536895e-05,
+      "loss": 1.0685,
+      "step": 26475
+    },
+    {
+      "epoch": 1.7991574942247588,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.875526566109526e-05,
+      "loss": 1.0988,
+      "step": 26480
+    },
+    {
+      "epoch": 1.7994972142954206,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.875314241065362e-05,
+      "loss": 1.0547,
+      "step": 26485
+    },
+    {
+      "epoch": 1.7998369343660823,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.875101916021199e-05,
+      "loss": 1.0498,
+      "step": 26490
+    },
+    {
+      "epoch": 1.8001766544367441,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.874889590977035e-05,
+      "loss": 1.0255,
+      "step": 26495
+    },
+    {
+      "epoch": 1.800516374507406,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.8746772659328715e-05,
+      "loss": 1.0314,
+      "step": 26500
+    },
+    {
+      "epoch": 1.8008560945780676,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.874464940888708e-05,
+      "loss": 0.9921,
+      "step": 26505
+    },
+    {
+      "epoch": 1.8011958146487295,
+      "grad_norm": 5.5,
+      "learning_rate": 3.874252615844544e-05,
+      "loss": 1.0617,
+      "step": 26510
+    },
+    {
+      "epoch": 1.8015355347193913,
+      "grad_norm": 6.125,
+      "learning_rate": 3.874040290800381e-05,
+      "loss": 1.1011,
+      "step": 26515
+    },
+    {
+      "epoch": 1.801875254790053,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.873827965756217e-05,
+      "loss": 0.994,
+      "step": 26520
+    },
+    {
+      "epoch": 1.8022149748607148,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.8736156407120535e-05,
+      "loss": 1.1383,
+      "step": 26525
+    },
+    {
+      "epoch": 1.8025546949313767,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.87340331566789e-05,
+      "loss": 1.0258,
+      "step": 26530
+    },
+    {
+      "epoch": 1.8028944150020383,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.873190990623726e-05,
+      "loss": 1.1122,
+      "step": 26535
+    },
+    {
+      "epoch": 1.8032341350727001,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.872978665579563e-05,
+      "loss": 1.0103,
+      "step": 26540
+    },
+    {
+      "epoch": 1.803573855143362,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.872766340535399e-05,
+      "loss": 1.0813,
+      "step": 26545
+    },
+    {
+      "epoch": 1.8039135752140236,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8725540154912355e-05,
+      "loss": 1.0552,
+      "step": 26550
+    },
+    {
+      "epoch": 1.8042532952846853,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.872341690447072e-05,
+      "loss": 1.0988,
+      "step": 26555
+    },
+    {
+      "epoch": 1.8045930153553473,
+      "grad_norm": 6.875,
+      "learning_rate": 3.872129365402908e-05,
+      "loss": 1.109,
+      "step": 26560
+    },
+    {
+      "epoch": 1.804932735426009,
+      "grad_norm": 6.25,
+      "learning_rate": 3.871917040358745e-05,
+      "loss": 1.0576,
+      "step": 26565
+    },
+    {
+      "epoch": 1.8052724554966706,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.871704715314581e-05,
+      "loss": 1.1429,
+      "step": 26570
+    },
+    {
+      "epoch": 1.8056121755673327,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.8714923902704175e-05,
+      "loss": 1.0289,
+      "step": 26575
+    },
+    {
+      "epoch": 1.8059518956379943,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.871280065226254e-05,
+      "loss": 1.0087,
+      "step": 26580
+    },
+    {
+      "epoch": 1.806291615708656,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.87106774018209e-05,
+      "loss": 1.1353,
+      "step": 26585
+    },
+    {
+      "epoch": 1.806631335779318,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.870855415137927e-05,
+      "loss": 1.0755,
+      "step": 26590
+    },
+    {
+      "epoch": 1.8069710558499796,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.870643090093763e-05,
+      "loss": 0.9352,
+      "step": 26595
+    },
+    {
+      "epoch": 1.8073107759206413,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.870430765049599e-05,
+      "loss": 1.0452,
+      "step": 26600
+    },
+    {
+      "epoch": 1.8076504959913031,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.870218440005436e-05,
+      "loss": 1.0856,
+      "step": 26605
+    },
+    {
+      "epoch": 1.807990216061965,
+      "grad_norm": 5.375,
+      "learning_rate": 3.870006114961272e-05,
+      "loss": 1.1087,
+      "step": 26610
+    },
+    {
+      "epoch": 1.8083299361326266,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.869793789917108e-05,
+      "loss": 1.0271,
+      "step": 26615
+    },
+    {
+      "epoch": 1.8086696562032885,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.869581464872945e-05,
+      "loss": 1.0776,
+      "step": 26620
+    },
+    {
+      "epoch": 1.8090093762739503,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8693691398287815e-05,
+      "loss": 1.0518,
+      "step": 26625
+    },
+    {
+      "epoch": 1.809349096344612,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.869156814784617e-05,
+      "loss": 1.0795,
+      "step": 26630
+    },
+    {
+      "epoch": 1.8096888164152738,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.868944489740454e-05,
+      "loss": 1.1575,
+      "step": 26635
+    },
+    {
+      "epoch": 1.8100285364859356,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.868732164696291e-05,
+      "loss": 1.0648,
+      "step": 26640
+    },
+    {
+      "epoch": 1.8103682565565973,
+      "grad_norm": 5.75,
+      "learning_rate": 3.8685198396521264e-05,
+      "loss": 1.0216,
+      "step": 26645
+    },
+    {
+      "epoch": 1.8107079766272591,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.8683075146079635e-05,
+      "loss": 1.1712,
+      "step": 26650
+    },
+    {
+      "epoch": 1.811047696697921,
+      "grad_norm": 5.625,
+      "learning_rate": 3.8680951895638e-05,
+      "loss": 1.0396,
+      "step": 26655
+    },
+    {
+      "epoch": 1.8113874167685826,
+      "grad_norm": 5.375,
+      "learning_rate": 3.8678828645196356e-05,
+      "loss": 1.1355,
+      "step": 26660
+    },
+    {
+      "epoch": 1.8117271368392445,
+      "grad_norm": 6.25,
+      "learning_rate": 3.867670539475473e-05,
+      "loss": 1.0376,
+      "step": 26665
+    },
+    {
+      "epoch": 1.8120668569099063,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.8674582144313084e-05,
+      "loss": 1.0915,
+      "step": 26670
+    },
+    {
+      "epoch": 1.812406576980568,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.867245889387145e-05,
+      "loss": 1.0719,
+      "step": 26675
+    },
+    {
+      "epoch": 1.8127462970512298,
+      "grad_norm": 6.25,
+      "learning_rate": 3.867033564342982e-05,
+      "loss": 0.9826,
+      "step": 26680
+    },
+    {
+      "epoch": 1.8130860171218917,
+      "grad_norm": 4.9375,
+      "learning_rate": 3.8668212392988176e-05,
+      "loss": 1.0573,
+      "step": 26685
+    },
+    {
+      "epoch": 1.8134257371925533,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.866608914254654e-05,
+      "loss": 1.1417,
+      "step": 26690
+    },
+    {
+      "epoch": 1.8137654572632151,
+      "grad_norm": 6.125,
+      "learning_rate": 3.866396589210491e-05,
+      "loss": 1.0926,
+      "step": 26695
+    },
+    {
+      "epoch": 1.814105177333877,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.866184264166327e-05,
+      "loss": 1.1088,
+      "step": 26700
+    },
+    {
+      "epoch": 1.8144448974045386,
+      "grad_norm": 6.375,
+      "learning_rate": 3.865971939122163e-05,
+      "loss": 1.1862,
+      "step": 26705
+    },
+    {
+      "epoch": 1.8147846174752005,
+      "grad_norm": 6.25,
+      "learning_rate": 3.865759614078e-05,
+      "loss": 1.108,
+      "step": 26710
+    },
+    {
+      "epoch": 1.8151243375458623,
+      "grad_norm": 5.5,
+      "learning_rate": 3.865547289033836e-05,
+      "loss": 1.177,
+      "step": 26715
+    },
+    {
+      "epoch": 1.815464057616524,
+      "grad_norm": 5.875,
+      "learning_rate": 3.8653349639896724e-05,
+      "loss": 0.9878,
+      "step": 26720
+    },
+    {
+      "epoch": 1.8158037776871856,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.8651226389455095e-05,
+      "loss": 1.0107,
+      "step": 26725
+    },
+    {
+      "epoch": 1.8161434977578477,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.864910313901345e-05,
+      "loss": 0.9896,
+      "step": 26730
+    },
+    {
+      "epoch": 1.8164832178285093,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.8646979888571816e-05,
+      "loss": 1.1212,
+      "step": 26735
+    },
+    {
+      "epoch": 1.816822937899171,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.864485663813019e-05,
+      "loss": 1.0123,
+      "step": 26740
+    },
+    {
+      "epoch": 1.817162657969833,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.8642733387688544e-05,
+      "loss": 1.1352,
+      "step": 26745
+    },
+    {
+      "epoch": 1.8175023780404946,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.8640610137246915e-05,
+      "loss": 0.9639,
+      "step": 26750
+    },
+    {
+      "epoch": 1.8178420981111563,
+      "grad_norm": 6.0,
+      "learning_rate": 3.863848688680527e-05,
+      "loss": 1.1516,
+      "step": 26755
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.8636363636363636e-05,
+      "loss": 1.1615,
+      "step": 26760
+    },
+    {
+      "epoch": 1.81852153825248,
+      "grad_norm": 6.25,
+      "learning_rate": 3.863424038592201e-05,
+      "loss": 0.9932,
+      "step": 26765
+    },
+    {
+      "epoch": 1.8188612583231416,
+      "grad_norm": 5.875,
+      "learning_rate": 3.8632117135480364e-05,
+      "loss": 1.0318,
+      "step": 26770
+    },
+    {
+      "epoch": 1.8192009783938035,
+      "grad_norm": 6.0,
+      "learning_rate": 3.862999388503873e-05,
+      "loss": 1.0583,
+      "step": 26775
+    },
+    {
+      "epoch": 1.8195406984644653,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.86278706345971e-05,
+      "loss": 1.009,
+      "step": 26780
+    },
+    {
+      "epoch": 1.819880418535127,
+      "grad_norm": 6.25,
+      "learning_rate": 3.8625747384155456e-05,
+      "loss": 1.1017,
+      "step": 26785
+    },
+    {
+      "epoch": 1.8202201386057888,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.862362413371382e-05,
+      "loss": 1.0162,
+      "step": 26790
+    },
+    {
+      "epoch": 1.8205598586764506,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.862150088327219e-05,
+      "loss": 1.0309,
+      "step": 26795
+    },
+    {
+      "epoch": 1.8208995787471123,
+      "grad_norm": 5.875,
+      "learning_rate": 3.861937763283055e-05,
+      "loss": 1.1231,
+      "step": 26800
+    },
+    {
+      "epoch": 1.8212392988177741,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.861725438238891e-05,
+      "loss": 1.1082,
+      "step": 26805
+    },
+    {
+      "epoch": 1.821579018888436,
+      "grad_norm": 6.25,
+      "learning_rate": 3.861513113194728e-05,
+      "loss": 1.0279,
+      "step": 26810
+    },
+    {
+      "epoch": 1.8219187389590976,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.861300788150564e-05,
+      "loss": 1.0531,
+      "step": 26815
+    },
+    {
+      "epoch": 1.8222584590297595,
+      "grad_norm": 6.625,
+      "learning_rate": 3.8610884631064004e-05,
+      "loss": 0.9543,
+      "step": 26820
+    },
+    {
+      "epoch": 1.8225981791004213,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.860876138062237e-05,
+      "loss": 1.114,
+      "step": 26825
+    },
+    {
+      "epoch": 1.822937899171083,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.860663813018073e-05,
+      "loss": 1.0358,
+      "step": 26830
+    },
+    {
+      "epoch": 1.8232776192417448,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.8604514879739096e-05,
+      "loss": 1.0529,
+      "step": 26835
+    },
+    {
+      "epoch": 1.8236173393124067,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.860239162929746e-05,
+      "loss": 1.0959,
+      "step": 26840
+    },
+    {
+      "epoch": 1.8239570593830683,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.8600268378855824e-05,
+      "loss": 1.0646,
+      "step": 26845
+    },
+    {
+      "epoch": 1.8242967794537301,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.859814512841419e-05,
+      "loss": 1.0804,
+      "step": 26850
+    },
+    {
+      "epoch": 1.824636499524392,
+      "grad_norm": 5.375,
+      "learning_rate": 3.859602187797255e-05,
+      "loss": 1.0628,
+      "step": 26855
+    },
+    {
+      "epoch": 1.8249762195950536,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8593898627530916e-05,
+      "loss": 1.1237,
+      "step": 26860
+    },
+    {
+      "epoch": 1.8253159396657155,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.859177537708928e-05,
+      "loss": 1.1573,
+      "step": 26865
+    },
+    {
+      "epoch": 1.8256556597363773,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.8589652126647644e-05,
+      "loss": 1.0116,
+      "step": 26870
+    },
+    {
+      "epoch": 1.825995379807039,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.858752887620601e-05,
+      "loss": 1.0903,
+      "step": 26875
+    },
+    {
+      "epoch": 1.8263350998777008,
+      "grad_norm": 6.25,
+      "learning_rate": 3.858540562576437e-05,
+      "loss": 1.0419,
+      "step": 26880
+    },
+    {
+      "epoch": 1.8266748199483627,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8583282375322736e-05,
+      "loss": 1.0554,
+      "step": 26885
+    },
+    {
+      "epoch": 1.8270145400190243,
+      "grad_norm": 6.0,
+      "learning_rate": 3.85811591248811e-05,
+      "loss": 0.9869,
+      "step": 26890
+    },
+    {
+      "epoch": 1.827354260089686,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.8579035874439464e-05,
+      "loss": 1.0429,
+      "step": 26895
+    },
+    {
+      "epoch": 1.827693980160348,
+      "grad_norm": 6.625,
+      "learning_rate": 3.857691262399783e-05,
+      "loss": 1.0798,
+      "step": 26900
+    },
+    {
+      "epoch": 1.8280337002310096,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.857478937355619e-05,
+      "loss": 1.0549,
+      "step": 26905
+    },
+    {
+      "epoch": 1.8283734203016713,
+      "grad_norm": 6.625,
+      "learning_rate": 3.8572666123114556e-05,
+      "loss": 1.0396,
+      "step": 26910
+    },
+    {
+      "epoch": 1.8287131403723333,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.857054287267292e-05,
+      "loss": 1.1136,
+      "step": 26915
+    },
+    {
+      "epoch": 1.829052860442995,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.8568419622231284e-05,
+      "loss": 0.9789,
+      "step": 26920
+    },
+    {
+      "epoch": 1.8293925805136566,
+      "grad_norm": 6.25,
+      "learning_rate": 3.856629637178965e-05,
+      "loss": 1.0752,
+      "step": 26925
+    },
+    {
+      "epoch": 1.8297323005843187,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.856417312134801e-05,
+      "loss": 1.0135,
+      "step": 26930
+    },
+    {
+      "epoch": 1.8300720206549803,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8562049870906376e-05,
+      "loss": 1.1284,
+      "step": 26935
+    },
+    {
+      "epoch": 1.830411740725642,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.855992662046474e-05,
+      "loss": 1.0958,
+      "step": 26940
+    },
+    {
+      "epoch": 1.8307514607963038,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8557803370023104e-05,
+      "loss": 1.0411,
+      "step": 26945
+    },
+    {
+      "epoch": 1.8310911808669657,
+      "grad_norm": 6.125,
+      "learning_rate": 3.855568011958147e-05,
+      "loss": 0.9773,
+      "step": 26950
+    },
+    {
+      "epoch": 1.8314309009376273,
+      "grad_norm": 6.625,
+      "learning_rate": 3.8553556869139825e-05,
+      "loss": 1.0223,
+      "step": 26955
+    },
+    {
+      "epoch": 1.8317706210082891,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.8551433618698196e-05,
+      "loss": 1.1152,
+      "step": 26960
+    },
+    {
+      "epoch": 1.832110341078951,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.854931036825656e-05,
+      "loss": 0.9577,
+      "step": 26965
+    },
+    {
+      "epoch": 1.8324500611496126,
+      "grad_norm": 6.625,
+      "learning_rate": 3.854718711781492e-05,
+      "loss": 1.1035,
+      "step": 26970
+    },
+    {
+      "epoch": 1.8327897812202745,
+      "grad_norm": 6.375,
+      "learning_rate": 3.854506386737329e-05,
+      "loss": 0.9953,
+      "step": 26975
+    },
+    {
+      "epoch": 1.8331295012909363,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.854294061693165e-05,
+      "loss": 1.0591,
+      "step": 26980
+    },
+    {
+      "epoch": 1.833469221361598,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.854081736649001e-05,
+      "loss": 1.0871,
+      "step": 26985
+    },
+    {
+      "epoch": 1.8338089414322598,
+      "grad_norm": 6.125,
+      "learning_rate": 3.853869411604838e-05,
+      "loss": 1.1237,
+      "step": 26990
+    },
+    {
+      "epoch": 1.8341486615029217,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.8536570865606744e-05,
+      "loss": 1.044,
+      "step": 26995
+    },
+    {
+      "epoch": 1.8344883815735833,
+      "grad_norm": 6.5,
+      "learning_rate": 3.85344476151651e-05,
+      "loss": 1.0481,
+      "step": 27000
+    },
+    {
+      "epoch": 1.8348281016442451,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.853232436472347e-05,
+      "loss": 1.1349,
+      "step": 27005
+    },
+    {
+      "epoch": 1.835167821714907,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8530201114281836e-05,
+      "loss": 1.006,
+      "step": 27010
+    },
+    {
+      "epoch": 1.8355075417855686,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.8528077863840193e-05,
+      "loss": 1.063,
+      "step": 27015
+    },
+    {
+      "epoch": 1.8358472618562305,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8525954613398564e-05,
+      "loss": 1.0108,
+      "step": 27020
+    },
+    {
+      "epoch": 1.8361869819268923,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.852383136295692e-05,
+      "loss": 1.0491,
+      "step": 27025
+    },
+    {
+      "epoch": 1.836526701997554,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.8521708112515285e-05,
+      "loss": 1.096,
+      "step": 27030
+    },
+    {
+      "epoch": 1.8368664220682158,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.8519584862073656e-05,
+      "loss": 1.2082,
+      "step": 27035
+    },
+    {
+      "epoch": 1.8372061421388777,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.8517461611632013e-05,
+      "loss": 1.0405,
+      "step": 27040
+    },
+    {
+      "epoch": 1.8375458622095393,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.851533836119038e-05,
+      "loss": 1.0365,
+      "step": 27045
+    },
+    {
+      "epoch": 1.8378855822802012,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.851321511074875e-05,
+      "loss": 0.9773,
+      "step": 27050
+    },
+    {
+      "epoch": 1.838225302350863,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.8511091860307106e-05,
+      "loss": 1.0581,
+      "step": 27055
+    },
+    {
+      "epoch": 1.8385650224215246,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.850896860986547e-05,
+      "loss": 1.1366,
+      "step": 27060
+    },
+    {
+      "epoch": 1.8389047424921863,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.850684535942384e-05,
+      "loss": 1.0153,
+      "step": 27065
+    },
+    {
+      "epoch": 1.8392444625628483,
+      "grad_norm": 5.5,
+      "learning_rate": 3.85047221089822e-05,
+      "loss": 1.1038,
+      "step": 27070
+    },
+    {
+      "epoch": 1.83958418263351,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.850259885854056e-05,
+      "loss": 1.1511,
+      "step": 27075
+    },
+    {
+      "epoch": 1.8399239027041716,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.850047560809893e-05,
+      "loss": 1.1588,
+      "step": 27080
+    },
+    {
+      "epoch": 1.8402636227748337,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.849835235765729e-05,
+      "loss": 0.9435,
+      "step": 27085
+    },
+    {
+      "epoch": 1.8406033428454953,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.849622910721566e-05,
+      "loss": 1.0341,
+      "step": 27090
+    },
+    {
+      "epoch": 1.840943062916157,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.849410585677402e-05,
+      "loss": 1.0359,
+      "step": 27095
+    },
+    {
+      "epoch": 1.841282782986819,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.849198260633238e-05,
+      "loss": 1.087,
+      "step": 27100
+    },
+    {
+      "epoch": 1.8416225030574807,
+      "grad_norm": 6.125,
+      "learning_rate": 3.848985935589075e-05,
+      "loss": 0.9756,
+      "step": 27105
+    },
+    {
+      "epoch": 1.8419622231281423,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.848773610544911e-05,
+      "loss": 1.0854,
+      "step": 27110
+    },
+    {
+      "epoch": 1.8423019431988041,
+      "grad_norm": 6.125,
+      "learning_rate": 3.8485612855007474e-05,
+      "loss": 1.0758,
+      "step": 27115
+    },
+    {
+      "epoch": 1.842641663269466,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.8483489604565844e-05,
+      "loss": 1.0286,
+      "step": 27120
+    },
+    {
+      "epoch": 1.8429813833401276,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.84813663541242e-05,
+      "loss": 1.1158,
+      "step": 27125
+    },
+    {
+      "epoch": 1.8433211034107895,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.8479243103682566e-05,
+      "loss": 1.1087,
+      "step": 27130
+    },
+    {
+      "epoch": 1.8436608234814513,
+      "grad_norm": 5.375,
+      "learning_rate": 3.8477119853240936e-05,
+      "loss": 1.0817,
+      "step": 27135
+    },
+    {
+      "epoch": 1.844000543552113,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8474996602799294e-05,
+      "loss": 1.0486,
+      "step": 27140
+    },
+    {
+      "epoch": 1.8443402636227748,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.847287335235766e-05,
+      "loss": 1.0569,
+      "step": 27145
+    },
+    {
+      "epoch": 1.8446799836934367,
+      "grad_norm": 5.75,
+      "learning_rate": 3.847075010191603e-05,
+      "loss": 1.0332,
+      "step": 27150
+    },
+    {
+      "epoch": 1.8450197037640983,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.8468626851474386e-05,
+      "loss": 1.0803,
+      "step": 27155
+    },
+    {
+      "epoch": 1.8453594238347601,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.846650360103275e-05,
+      "loss": 1.1536,
+      "step": 27160
+    },
+    {
+      "epoch": 1.845699143905422,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.846438035059112e-05,
+      "loss": 0.9908,
+      "step": 27165
+    },
+    {
+      "epoch": 1.8460388639760836,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.846225710014948e-05,
+      "loss": 1.1913,
+      "step": 27170
+    },
+    {
+      "epoch": 1.8463785840467455,
+      "grad_norm": 4.875,
+      "learning_rate": 3.846013384970784e-05,
+      "loss": 1.0481,
+      "step": 27175
+    },
+    {
+      "epoch": 1.8467183041174073,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.8458010599266206e-05,
+      "loss": 1.0105,
+      "step": 27180
+    },
+    {
+      "epoch": 1.847058024188069,
+      "grad_norm": 6.0,
+      "learning_rate": 3.845588734882457e-05,
+      "loss": 1.1194,
+      "step": 27185
+    },
+    {
+      "epoch": 1.8473977442587308,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.8453764098382934e-05,
+      "loss": 1.1799,
+      "step": 27190
+    },
+    {
+      "epoch": 1.8477374643293927,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.84516408479413e-05,
+      "loss": 1.144,
+      "step": 27195
+    },
+    {
+      "epoch": 1.8480771844000543,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.844951759749966e-05,
+      "loss": 1.0045,
+      "step": 27200
+    },
+    {
+      "epoch": 1.8484169044707162,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.8447394347058026e-05,
+      "loss": 1.1315,
+      "step": 27205
+    },
+    {
+      "epoch": 1.848756624541378,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.844527109661639e-05,
+      "loss": 0.9654,
+      "step": 27210
+    },
+    {
+      "epoch": 1.8490963446120396,
+      "grad_norm": 6.0,
+      "learning_rate": 3.8443147846174754e-05,
+      "loss": 1.085,
+      "step": 27215
+    },
+    {
+      "epoch": 1.8494360646827015,
+      "grad_norm": 5.625,
+      "learning_rate": 3.844102459573312e-05,
+      "loss": 1.0627,
+      "step": 27220
+    },
+    {
+      "epoch": 1.8497757847533634,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.843890134529148e-05,
+      "loss": 0.9121,
+      "step": 27225
+    },
+    {
+      "epoch": 1.850115504824025,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.8436778094849846e-05,
+      "loss": 0.928,
+      "step": 27230
+    },
+    {
+      "epoch": 1.8504552248946866,
+      "grad_norm": 5.75,
+      "learning_rate": 3.843465484440821e-05,
+      "loss": 1.066,
+      "step": 27235
+    },
+    {
+      "epoch": 1.8507949449653487,
+      "grad_norm": 6.0,
+      "learning_rate": 3.8432531593966574e-05,
+      "loss": 1.098,
+      "step": 27240
+    },
+    {
+      "epoch": 1.8511346650360103,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.843040834352494e-05,
+      "loss": 1.0545,
+      "step": 27245
+    },
+    {
+      "epoch": 1.851474385106672,
+      "grad_norm": 5.625,
+      "learning_rate": 3.84282850930833e-05,
+      "loss": 1.0565,
+      "step": 27250
+    },
+    {
+      "epoch": 1.851814105177334,
+      "grad_norm": 5.125,
+      "learning_rate": 3.8426161842641666e-05,
+      "loss": 0.9355,
+      "step": 27255
+    },
+    {
+      "epoch": 1.8521538252479957,
+      "grad_norm": 6.125,
+      "learning_rate": 3.842403859220003e-05,
+      "loss": 1.057,
+      "step": 27260
+    },
+    {
+      "epoch": 1.8524935453186573,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.8421915341758394e-05,
+      "loss": 0.9911,
+      "step": 27265
+    },
+    {
+      "epoch": 1.8528332653893194,
+      "grad_norm": 6.125,
+      "learning_rate": 3.841979209131676e-05,
+      "loss": 1.0411,
+      "step": 27270
+    },
+    {
+      "epoch": 1.853172985459981,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.841766884087512e-05,
+      "loss": 1.0702,
+      "step": 27275
+    },
+    {
+      "epoch": 1.8535127055306426,
+      "grad_norm": 6.5,
+      "learning_rate": 3.8415545590433486e-05,
+      "loss": 0.9981,
+      "step": 27280
+    },
+    {
+      "epoch": 1.8538524256013045,
+      "grad_norm": 6.5,
+      "learning_rate": 3.841342233999185e-05,
+      "loss": 1.0206,
+      "step": 27285
+    },
+    {
+      "epoch": 1.8541921456719663,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.8411299089550214e-05,
+      "loss": 1.05,
+      "step": 27290
+    },
+    {
+      "epoch": 1.854531865742628,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.840917583910858e-05,
+      "loss": 1.0529,
+      "step": 27295
+    },
+    {
+      "epoch": 1.8548715858132898,
+      "grad_norm": 6.875,
+      "learning_rate": 3.840705258866694e-05,
+      "loss": 1.1005,
+      "step": 27300
+    },
+    {
+      "epoch": 1.8552113058839517,
+      "grad_norm": 5.875,
+      "learning_rate": 3.8404929338225306e-05,
+      "loss": 1.1276,
+      "step": 27305
+    },
+    {
+      "epoch": 1.8555510259546133,
+      "grad_norm": 6.375,
+      "learning_rate": 3.840280608778366e-05,
+      "loss": 0.9859,
+      "step": 27310
+    },
+    {
+      "epoch": 1.8558907460252752,
+      "grad_norm": 5.75,
+      "learning_rate": 3.8400682837342034e-05,
+      "loss": 1.1201,
+      "step": 27315
+    },
+    {
+      "epoch": 1.856230466095937,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.83985595869004e-05,
+      "loss": 1.0896,
+      "step": 27320
+    },
+    {
+      "epoch": 1.8565701861665986,
+      "grad_norm": 6.0,
+      "learning_rate": 3.8396436336458755e-05,
+      "loss": 1.0193,
+      "step": 27325
+    },
+    {
+      "epoch": 1.8569099062372605,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8394313086017126e-05,
+      "loss": 0.9997,
+      "step": 27330
+    },
+    {
+      "epoch": 1.8572496263079223,
+      "grad_norm": 6.5,
+      "learning_rate": 3.839218983557549e-05,
+      "loss": 1.0422,
+      "step": 27335
+    },
+    {
+      "epoch": 1.857589346378584,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.839006658513385e-05,
+      "loss": 1.0877,
+      "step": 27340
+    },
+    {
+      "epoch": 1.8579290664492458,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.838794333469222e-05,
+      "loss": 1.0541,
+      "step": 27345
+    },
+    {
+      "epoch": 1.8582687865199077,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.838582008425058e-05,
+      "loss": 1.0556,
+      "step": 27350
+    },
+    {
+      "epoch": 1.8586085065905693,
+      "grad_norm": 6.25,
+      "learning_rate": 3.838369683380894e-05,
+      "loss": 1.1201,
+      "step": 27355
+    },
+    {
+      "epoch": 1.8589482266612312,
+      "grad_norm": 5.875,
+      "learning_rate": 3.838157358336731e-05,
+      "loss": 0.9513,
+      "step": 27360
+    },
+    {
+      "epoch": 1.859287946731893,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.8379450332925674e-05,
+      "loss": 0.9947,
+      "step": 27365
+    },
+    {
+      "epoch": 1.8596276668025546,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.837732708248403e-05,
+      "loss": 1.08,
+      "step": 27370
+    },
+    {
+      "epoch": 1.8599673868732165,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.83752038320424e-05,
+      "loss": 1.0072,
+      "step": 27375
+    },
+    {
+      "epoch": 1.8603071069438784,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.837308058160076e-05,
+      "loss": 1.0912,
+      "step": 27380
+    },
+    {
+      "epoch": 1.86064682701454,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.837095733115912e-05,
+      "loss": 1.0901,
+      "step": 27385
+    },
+    {
+      "epoch": 1.8609865470852018,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.8368834080717494e-05,
+      "loss": 1.0295,
+      "step": 27390
+    },
+    {
+      "epoch": 1.8613262671558637,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.836671083027585e-05,
+      "loss": 1.0793,
+      "step": 27395
+    },
+    {
+      "epoch": 1.8616659872265253,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.8364587579834215e-05,
+      "loss": 1.1409,
+      "step": 27400
+    },
+    {
+      "epoch": 1.862005707297187,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.8362464329392586e-05,
+      "loss": 1.0045,
+      "step": 27405
+    },
+    {
+      "epoch": 1.862345427367849,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.836034107895094e-05,
+      "loss": 1.1535,
+      "step": 27410
+    },
+    {
+      "epoch": 1.8626851474385107,
+      "grad_norm": 6.375,
+      "learning_rate": 3.835821782850931e-05,
+      "loss": 1.1101,
+      "step": 27415
+    },
+    {
+      "epoch": 1.8630248675091723,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.835609457806768e-05,
+      "loss": 1.0642,
+      "step": 27420
+    },
+    {
+      "epoch": 1.8633645875798344,
+      "grad_norm": 6.25,
+      "learning_rate": 3.8353971327626035e-05,
+      "loss": 1.0459,
+      "step": 27425
+    },
+    {
+      "epoch": 1.863704307650496,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.8351848077184406e-05,
+      "loss": 1.0387,
+      "step": 27430
+    },
+    {
+      "epoch": 1.8640440277211576,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.834972482674277e-05,
+      "loss": 1.1088,
+      "step": 27435
+    },
+    {
+      "epoch": 1.8643837477918197,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.834760157630113e-05,
+      "loss": 1.0607,
+      "step": 27440
+    },
+    {
+      "epoch": 1.8647234678624813,
+      "grad_norm": 5.625,
+      "learning_rate": 3.83454783258595e-05,
+      "loss": 1.1035,
+      "step": 27445
+    },
+    {
+      "epoch": 1.865063187933143,
+      "grad_norm": 5.75,
+      "learning_rate": 3.8343355075417855e-05,
+      "loss": 1.0531,
+      "step": 27450
+    },
+    {
+      "epoch": 1.8654029080038048,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.834123182497622e-05,
+      "loss": 0.9998,
+      "step": 27455
+    },
+    {
+      "epoch": 1.8657426280744667,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.833910857453459e-05,
+      "loss": 1.0759,
+      "step": 27460
+    },
+    {
+      "epoch": 1.8660823481451283,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.833698532409295e-05,
+      "loss": 1.1405,
+      "step": 27465
+    },
+    {
+      "epoch": 1.8664220682157902,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.833486207365131e-05,
+      "loss": 0.9589,
+      "step": 27470
+    },
+    {
+      "epoch": 1.866761788286452,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.833273882320968e-05,
+      "loss": 1.0236,
+      "step": 27475
+    },
+    {
+      "epoch": 1.8671015083571136,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.833061557276804e-05,
+      "loss": 1.1147,
+      "step": 27480
+    },
+    {
+      "epoch": 1.8674412284277755,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.83284923223264e-05,
+      "loss": 1.0492,
+      "step": 27485
+    },
+    {
+      "epoch": 1.8677809484984373,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.8326369071884774e-05,
+      "loss": 1.1093,
+      "step": 27490
+    },
+    {
+      "epoch": 1.868120668569099,
+      "grad_norm": 6.0,
+      "learning_rate": 3.832424582144313e-05,
+      "loss": 1.1146,
+      "step": 27495
+    },
+    {
+      "epoch": 1.8684603886397608,
+      "grad_norm": 5.75,
+      "learning_rate": 3.8322122571001495e-05,
+      "loss": 1.1163,
+      "step": 27500
+    },
+    {
+      "epoch": 1.8688001087104227,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.8319999320559866e-05,
+      "loss": 1.0855,
+      "step": 27505
+    },
+    {
+      "epoch": 1.8691398287810843,
+      "grad_norm": 6.25,
+      "learning_rate": 3.831787607011822e-05,
+      "loss": 1.0718,
+      "step": 27510
+    },
+    {
+      "epoch": 1.8694795488517462,
+      "grad_norm": 6.25,
+      "learning_rate": 3.831575281967659e-05,
+      "loss": 1.0793,
+      "step": 27515
+    },
+    {
+      "epoch": 1.869819268922408,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.831362956923496e-05,
+      "loss": 1.1227,
+      "step": 27520
+    },
+    {
+      "epoch": 1.8701589889930696,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.8311506318793315e-05,
+      "loss": 1.1991,
+      "step": 27525
+    },
+    {
+      "epoch": 1.8704987090637315,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.830938306835168e-05,
+      "loss": 1.0559,
+      "step": 27530
+    },
+    {
+      "epoch": 1.8708384291343934,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.830725981791004e-05,
+      "loss": 1.0364,
+      "step": 27535
+    },
+    {
+      "epoch": 1.871178149205055,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.830513656746841e-05,
+      "loss": 1.1004,
+      "step": 27540
+    },
+    {
+      "epoch": 1.8715178692757168,
+      "grad_norm": 4.65625,
+      "learning_rate": 3.830301331702677e-05,
+      "loss": 0.9608,
+      "step": 27545
+    },
+    {
+      "epoch": 1.8718575893463787,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.8300890066585135e-05,
+      "loss": 1.0968,
+      "step": 27550
+    },
+    {
+      "epoch": 1.8721973094170403,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.82987668161435e-05,
+      "loss": 1.1464,
+      "step": 27555
+    },
+    {
+      "epoch": 1.8725370294877022,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.829664356570186e-05,
+      "loss": 1.0038,
+      "step": 27560
+    },
+    {
+      "epoch": 1.872876749558364,
+      "grad_norm": 6.125,
+      "learning_rate": 3.829452031526023e-05,
+      "loss": 1.0525,
+      "step": 27565
+    },
+    {
+      "epoch": 1.8732164696290257,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.829239706481859e-05,
+      "loss": 1.0484,
+      "step": 27570
+    },
+    {
+      "epoch": 1.8735561896996873,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.8290273814376955e-05,
+      "loss": 1.002,
+      "step": 27575
+    },
+    {
+      "epoch": 1.8738959097703494,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.828815056393532e-05,
+      "loss": 1.0419,
+      "step": 27580
+    },
+    {
+      "epoch": 1.874235629841011,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.828602731349368e-05,
+      "loss": 1.1324,
+      "step": 27585
+    },
+    {
+      "epoch": 1.8745753499116726,
+      "grad_norm": 6.375,
+      "learning_rate": 3.828390406305205e-05,
+      "loss": 1.0988,
+      "step": 27590
+    },
+    {
+      "epoch": 1.8749150699823347,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.828178081261041e-05,
+      "loss": 1.0246,
+      "step": 27595
+    },
+    {
+      "epoch": 1.8752547900529963,
+      "grad_norm": 5.625,
+      "learning_rate": 3.8279657562168775e-05,
+      "loss": 1.0611,
+      "step": 27600
+    },
+    {
+      "epoch": 1.875594510123658,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.827753431172714e-05,
+      "loss": 1.1317,
+      "step": 27605
+    },
+    {
+      "epoch": 1.87593423019432,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.82754110612855e-05,
+      "loss": 1.1084,
+      "step": 27610
+    },
+    {
+      "epoch": 1.8762739502649817,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.827328781084387e-05,
+      "loss": 1.0488,
+      "step": 27615
+    },
+    {
+      "epoch": 1.8766136703356433,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.827116456040223e-05,
+      "loss": 1.0522,
+      "step": 27620
+    },
+    {
+      "epoch": 1.8769533904063052,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8269041309960595e-05,
+      "loss": 1.0471,
+      "step": 27625
+    },
+    {
+      "epoch": 1.877293110476967,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.826691805951896e-05,
+      "loss": 1.0891,
+      "step": 27630
+    },
+    {
+      "epoch": 1.8776328305476286,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.826479480907732e-05,
+      "loss": 0.9584,
+      "step": 27635
+    },
+    {
+      "epoch": 1.8779725506182905,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.826267155863569e-05,
+      "loss": 1.0711,
+      "step": 27640
+    },
+    {
+      "epoch": 1.8783122706889523,
+      "grad_norm": 6.0,
+      "learning_rate": 3.826054830819405e-05,
+      "loss": 0.9643,
+      "step": 27645
+    },
+    {
+      "epoch": 1.878651990759614,
+      "grad_norm": 6.25,
+      "learning_rate": 3.825842505775241e-05,
+      "loss": 1.0339,
+      "step": 27650
+    },
+    {
+      "epoch": 1.8789917108302758,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.825630180731078e-05,
+      "loss": 1.0863,
+      "step": 27655
+    },
+    {
+      "epoch": 1.8793314309009377,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.825417855686914e-05,
+      "loss": 1.1143,
+      "step": 27660
+    },
+    {
+      "epoch": 1.8796711509715993,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.82520553064275e-05,
+      "loss": 1.1711,
+      "step": 27665
+    },
+    {
+      "epoch": 1.8800108710422612,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.824993205598587e-05,
+      "loss": 1.0994,
+      "step": 27670
+    },
+    {
+      "epoch": 1.880350591112923,
+      "grad_norm": 6.375,
+      "learning_rate": 3.8247808805544235e-05,
+      "loss": 1.0167,
+      "step": 27675
+    },
+    {
+      "epoch": 1.8806903111835847,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.824568555510259e-05,
+      "loss": 1.0114,
+      "step": 27680
+    },
+    {
+      "epoch": 1.8810300312542465,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.824356230466096e-05,
+      "loss": 0.99,
+      "step": 27685
+    },
+    {
+      "epoch": 1.8813697513249084,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.824143905421933e-05,
+      "loss": 1.0405,
+      "step": 27690
+    },
+    {
+      "epoch": 1.88170947139557,
+      "grad_norm": 4.71875,
+      "learning_rate": 3.8239315803777684e-05,
+      "loss": 1.1178,
+      "step": 27695
+    },
+    {
+      "epoch": 1.8820491914662318,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.8237192553336055e-05,
+      "loss": 0.9757,
+      "step": 27700
+    },
+    {
+      "epoch": 1.8823889115368937,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.823506930289442e-05,
+      "loss": 1.1207,
+      "step": 27705
+    },
+    {
+      "epoch": 1.8827286316075553,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.8232946052452776e-05,
+      "loss": 1.0935,
+      "step": 27710
+    },
+    {
+      "epoch": 1.8830683516782172,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.823082280201115e-05,
+      "loss": 1.1168,
+      "step": 27715
+    },
+    {
+      "epoch": 1.883408071748879,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.822869955156951e-05,
+      "loss": 1.0826,
+      "step": 27720
+    },
+    {
+      "epoch": 1.8837477918195407,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.822657630112787e-05,
+      "loss": 1.0807,
+      "step": 27725
+    },
+    {
+      "epoch": 1.8840875118902025,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.822445305068624e-05,
+      "loss": 1.0022,
+      "step": 27730
+    },
+    {
+      "epoch": 1.8844272319608644,
+      "grad_norm": 5.875,
+      "learning_rate": 3.8222329800244596e-05,
+      "loss": 1.1272,
+      "step": 27735
+    },
+    {
+      "epoch": 1.884766952031526,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.822020654980296e-05,
+      "loss": 1.0869,
+      "step": 27740
+    },
+    {
+      "epoch": 1.8851066721021879,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.821808329936133e-05,
+      "loss": 1.0869,
+      "step": 27745
+    },
+    {
+      "epoch": 1.8854463921728497,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.821596004891969e-05,
+      "loss": 1.1323,
+      "step": 27750
+    },
+    {
+      "epoch": 1.8857861122435113,
+      "grad_norm": 6.125,
+      "learning_rate": 3.821383679847805e-05,
+      "loss": 0.9865,
+      "step": 27755
+    },
+    {
+      "epoch": 1.886125832314173,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.821171354803642e-05,
+      "loss": 1.0566,
+      "step": 27760
+    },
+    {
+      "epoch": 1.886465552384835,
+      "grad_norm": 6.5,
+      "learning_rate": 3.820959029759478e-05,
+      "loss": 1.1321,
+      "step": 27765
+    },
+    {
+      "epoch": 1.8868052724554967,
+      "grad_norm": 6.625,
+      "learning_rate": 3.820746704715315e-05,
+      "loss": 0.9662,
+      "step": 27770
+    },
+    {
+      "epoch": 1.8871449925261583,
+      "grad_norm": 5.09375,
+      "learning_rate": 3.8205343796711515e-05,
+      "loss": 1.1259,
+      "step": 27775
+    },
+    {
+      "epoch": 1.8874847125968204,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.820322054626987e-05,
+      "loss": 1.1673,
+      "step": 27780
+    },
+    {
+      "epoch": 1.887824432667482,
+      "grad_norm": 6.125,
+      "learning_rate": 3.820109729582824e-05,
+      "loss": 0.9822,
+      "step": 27785
+    },
+    {
+      "epoch": 1.8881641527381436,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.819897404538661e-05,
+      "loss": 1.1189,
+      "step": 27790
+    },
+    {
+      "epoch": 1.8885038728088055,
+      "grad_norm": 6.5,
+      "learning_rate": 3.8196850794944965e-05,
+      "loss": 1.1111,
+      "step": 27795
+    },
+    {
+      "epoch": 1.8888435928794673,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.8194727544503335e-05,
+      "loss": 1.1027,
+      "step": 27800
+    },
+    {
+      "epoch": 1.889183312950129,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.819260429406169e-05,
+      "loss": 0.9907,
+      "step": 27805
+    },
+    {
+      "epoch": 1.8895230330207908,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8190481043620057e-05,
+      "loss": 1.0501,
+      "step": 27810
+    },
+    {
+      "epoch": 1.8898627530914527,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.818835779317843e-05,
+      "loss": 1.0106,
+      "step": 27815
+    },
+    {
+      "epoch": 1.8902024731621143,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.8186234542736785e-05,
+      "loss": 1.045,
+      "step": 27820
+    },
+    {
+      "epoch": 1.8905421932327762,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.818411129229515e-05,
+      "loss": 1.1092,
+      "step": 27825
+    },
+    {
+      "epoch": 1.890881913303438,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.818198804185352e-05,
+      "loss": 1.072,
+      "step": 27830
+    },
+    {
+      "epoch": 1.8912216333740997,
+      "grad_norm": 5.625,
+      "learning_rate": 3.8179864791411877e-05,
+      "loss": 1.0187,
+      "step": 27835
+    },
+    {
+      "epoch": 1.8915613534447615,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.817774154097024e-05,
+      "loss": 1.11,
+      "step": 27840
+    },
+    {
+      "epoch": 1.8919010735154234,
+      "grad_norm": 6.5,
+      "learning_rate": 3.817561829052861e-05,
+      "loss": 1.0616,
+      "step": 27845
+    },
+    {
+      "epoch": 1.892240793586085,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.817349504008697e-05,
+      "loss": 1.0247,
+      "step": 27850
+    },
+    {
+      "epoch": 1.8925805136567468,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.817137178964533e-05,
+      "loss": 1.1173,
+      "step": 27855
+    },
+    {
+      "epoch": 1.8929202337274087,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.81692485392037e-05,
+      "loss": 1.0879,
+      "step": 27860
+    },
+    {
+      "epoch": 1.8932599537980703,
+      "grad_norm": 6.125,
+      "learning_rate": 3.816712528876206e-05,
+      "loss": 1.0339,
+      "step": 27865
+    },
+    {
+      "epoch": 1.8935996738687322,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.8165002038320425e-05,
+      "loss": 1.1212,
+      "step": 27870
+    },
+    {
+      "epoch": 1.893939393939394,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.816287878787879e-05,
+      "loss": 1.1548,
+      "step": 27875
+    },
+    {
+      "epoch": 1.8942791140100557,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.816075553743715e-05,
+      "loss": 1.0543,
+      "step": 27880
+    },
+    {
+      "epoch": 1.8946188340807175,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.8158632286995517e-05,
+      "loss": 1.1134,
+      "step": 27885
+    },
+    {
+      "epoch": 1.8949585541513794,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.815650903655388e-05,
+      "loss": 1.0565,
+      "step": 27890
+    },
+    {
+      "epoch": 1.895298274222041,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.8154385786112245e-05,
+      "loss": 1.0795,
+      "step": 27895
+    },
+    {
+      "epoch": 1.8956379942927029,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.815226253567061e-05,
+      "loss": 1.1153,
+      "step": 27900
+    },
+    {
+      "epoch": 1.8959777143633647,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.815013928522897e-05,
+      "loss": 1.0679,
+      "step": 27905
+    },
+    {
+      "epoch": 1.8963174344340263,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.814801603478734e-05,
+      "loss": 1.0293,
+      "step": 27910
+    },
+    {
+      "epoch": 1.8966571545046882,
+      "grad_norm": 5.625,
+      "learning_rate": 3.81458927843457e-05,
+      "loss": 1.1169,
+      "step": 27915
+    },
+    {
+      "epoch": 1.89699687457535,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.8143769533904065e-05,
+      "loss": 1.1041,
+      "step": 27920
+    },
+    {
+      "epoch": 1.8973365946460117,
+      "grad_norm": 5.625,
+      "learning_rate": 3.814164628346243e-05,
+      "loss": 0.9933,
+      "step": 27925
+    },
+    {
+      "epoch": 1.8976763147166733,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.813952303302079e-05,
+      "loss": 1.1306,
+      "step": 27930
+    },
+    {
+      "epoch": 1.8980160347873354,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.813739978257916e-05,
+      "loss": 1.0769,
+      "step": 27935
+    },
+    {
+      "epoch": 1.898355754857997,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.813527653213752e-05,
+      "loss": 1.0416,
+      "step": 27940
+    },
+    {
+      "epoch": 1.8986954749286586,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.8133153281695885e-05,
+      "loss": 1.0538,
+      "step": 27945
+    },
+    {
+      "epoch": 1.8990351949993207,
+      "grad_norm": 6.625,
+      "learning_rate": 3.813103003125425e-05,
+      "loss": 1.0383,
+      "step": 27950
+    },
+    {
+      "epoch": 1.8993749150699824,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.812890678081261e-05,
+      "loss": 1.0999,
+      "step": 27955
+    },
+    {
+      "epoch": 1.899714635140644,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.812678353037098e-05,
+      "loss": 1.0762,
+      "step": 27960
+    },
+    {
+      "epoch": 1.9000543552113058,
+      "grad_norm": 6.25,
+      "learning_rate": 3.812466027992934e-05,
+      "loss": 1.1044,
+      "step": 27965
+    },
+    {
+      "epoch": 1.9003940752819677,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.8122537029487705e-05,
+      "loss": 1.1451,
+      "step": 27970
+    },
+    {
+      "epoch": 1.9007337953526293,
+      "grad_norm": 6.375,
+      "learning_rate": 3.812041377904607e-05,
+      "loss": 0.8764,
+      "step": 27975
+    },
+    {
+      "epoch": 1.9010735154232912,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.811829052860443e-05,
+      "loss": 1.0316,
+      "step": 27980
+    },
+    {
+      "epoch": 1.901413235493953,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.81161672781628e-05,
+      "loss": 1.0765,
+      "step": 27985
+    },
+    {
+      "epoch": 1.9017529555646147,
+      "grad_norm": 6.125,
+      "learning_rate": 3.811404402772116e-05,
+      "loss": 1.0854,
+      "step": 27990
+    },
+    {
+      "epoch": 1.9020926756352765,
+      "grad_norm": 7.375,
+      "learning_rate": 3.8111920777279525e-05,
+      "loss": 1.08,
+      "step": 27995
+    },
+    {
+      "epoch": 1.9024323957059384,
+      "grad_norm": 5.5,
+      "learning_rate": 3.810979752683789e-05,
+      "loss": 1.0761,
+      "step": 28000
+    },
+    {
+      "epoch": 1.9027721157766,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.8107674276396246e-05,
+      "loss": 1.0776,
+      "step": 28005
+    },
+    {
+      "epoch": 1.9031118358472618,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.810555102595462e-05,
+      "loss": 1.1015,
+      "step": 28010
+    },
+    {
+      "epoch": 1.9034515559179237,
+      "grad_norm": 6.5,
+      "learning_rate": 3.810342777551298e-05,
+      "loss": 1.0631,
+      "step": 28015
+    },
+    {
+      "epoch": 1.9037912759885853,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.810130452507134e-05,
+      "loss": 1.1238,
+      "step": 28020
+    },
+    {
+      "epoch": 1.9041309960592472,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.809918127462971e-05,
+      "loss": 1.0815,
+      "step": 28025
+    },
+    {
+      "epoch": 1.904470716129909,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.809705802418807e-05,
+      "loss": 1.0786,
+      "step": 28030
+    },
+    {
+      "epoch": 1.9048104362005707,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.809493477374643e-05,
+      "loss": 1.1975,
+      "step": 28035
+    },
+    {
+      "epoch": 1.9051501562712325,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.80928115233048e-05,
+      "loss": 1.0287,
+      "step": 28040
+    },
+    {
+      "epoch": 1.9054898763418944,
+      "grad_norm": 6.25,
+      "learning_rate": 3.8090688272863165e-05,
+      "loss": 1.0993,
+      "step": 28045
+    },
+    {
+      "epoch": 1.905829596412556,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.808856502242152e-05,
+      "loss": 1.1742,
+      "step": 28050
+    },
+    {
+      "epoch": 1.9061693164832179,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.808644177197989e-05,
+      "loss": 1.0324,
+      "step": 28055
+    },
+    {
+      "epoch": 1.9065090365538797,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.808431852153826e-05,
+      "loss": 1.0845,
+      "step": 28060
+    },
+    {
+      "epoch": 1.9068487566245413,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.8082195271096614e-05,
+      "loss": 1.132,
+      "step": 28065
+    },
+    {
+      "epoch": 1.9071884766952032,
+      "grad_norm": 6.625,
+      "learning_rate": 3.8080072020654985e-05,
+      "loss": 1.0462,
+      "step": 28070
+    },
+    {
+      "epoch": 1.907528196765865,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.807794877021334e-05,
+      "loss": 1.0436,
+      "step": 28075
+    },
+    {
+      "epoch": 1.9078679168365267,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.8075825519771706e-05,
+      "loss": 1.0407,
+      "step": 28080
+    },
+    {
+      "epoch": 1.9082076369071885,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.807370226933008e-05,
+      "loss": 1.046,
+      "step": 28085
+    },
+    {
+      "epoch": 1.9085473569778504,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.8071579018888434e-05,
+      "loss": 1.0725,
+      "step": 28090
+    },
+    {
+      "epoch": 1.908887077048512,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.80694557684468e-05,
+      "loss": 1.0684,
+      "step": 28095
+    },
+    {
+      "epoch": 1.9092267971191736,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.806733251800517e-05,
+      "loss": 0.9776,
+      "step": 28100
+    },
+    {
+      "epoch": 1.9095665171898357,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.8065209267563526e-05,
+      "loss": 1.1064,
+      "step": 28105
+    },
+    {
+      "epoch": 1.9099062372604974,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.80630860171219e-05,
+      "loss": 1.0763,
+      "step": 28110
+    },
+    {
+      "epoch": 1.910245957331159,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.806096276668026e-05,
+      "loss": 1.0943,
+      "step": 28115
+    },
+    {
+      "epoch": 1.910585677401821,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.805883951623862e-05,
+      "loss": 1.0832,
+      "step": 28120
+    },
+    {
+      "epoch": 1.9109253974724827,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.805671626579699e-05,
+      "loss": 1.0322,
+      "step": 28125
+    },
+    {
+      "epoch": 1.9112651175431443,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.805459301535535e-05,
+      "loss": 1.0276,
+      "step": 28130
+    },
+    {
+      "epoch": 1.9116048376138062,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.805246976491371e-05,
+      "loss": 1.0075,
+      "step": 28135
+    },
+    {
+      "epoch": 1.911944557684468,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.805034651447208e-05,
+      "loss": 1.0284,
+      "step": 28140
+    },
+    {
+      "epoch": 1.9122842777551297,
+      "grad_norm": 6.5,
+      "learning_rate": 3.8048223264030445e-05,
+      "loss": 1.0317,
+      "step": 28145
+    },
+    {
+      "epoch": 1.9126239978257915,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.80461000135888e-05,
+      "loss": 1.0294,
+      "step": 28150
+    },
+    {
+      "epoch": 1.9129637178964534,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.804397676314717e-05,
+      "loss": 1.1722,
+      "step": 28155
+    },
+    {
+      "epoch": 1.913303437967115,
+      "grad_norm": 6.125,
+      "learning_rate": 3.804185351270553e-05,
+      "loss": 1.0553,
+      "step": 28160
+    },
+    {
+      "epoch": 1.9136431580377768,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.8039730262263894e-05,
+      "loss": 1.1097,
+      "step": 28165
+    },
+    {
+      "epoch": 1.9139828781084387,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.8037607011822265e-05,
+      "loss": 1.0492,
+      "step": 28170
+    },
+    {
+      "epoch": 1.9143225981791003,
+      "grad_norm": 6.0,
+      "learning_rate": 3.803548376138062e-05,
+      "loss": 1.0328,
+      "step": 28175
+    },
+    {
+      "epoch": 1.9146623182497622,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.8033360510938986e-05,
+      "loss": 1.0647,
+      "step": 28180
+    },
+    {
+      "epoch": 1.915002038320424,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.803123726049736e-05,
+      "loss": 1.0107,
+      "step": 28185
+    },
+    {
+      "epoch": 1.9153417583910857,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.8029114010055714e-05,
+      "loss": 1.0117,
+      "step": 28190
+    },
+    {
+      "epoch": 1.9156814784617475,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.802699075961408e-05,
+      "loss": 1.0255,
+      "step": 28195
+    },
+    {
+      "epoch": 1.9160211985324094,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.802486750917245e-05,
+      "loss": 1.045,
+      "step": 28200
+    },
+    {
+      "epoch": 1.916360918603071,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.8022744258730806e-05,
+      "loss": 1.0335,
+      "step": 28205
+    },
+    {
+      "epoch": 1.9167006386737329,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.802062100828917e-05,
+      "loss": 1.0121,
+      "step": 28210
+    },
+    {
+      "epoch": 1.9170403587443947,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.801849775784754e-05,
+      "loss": 1.0624,
+      "step": 28215
+    },
+    {
+      "epoch": 1.9173800788150563,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.80163745074059e-05,
+      "loss": 1.0593,
+      "step": 28220
+    },
+    {
+      "epoch": 1.9177197988857182,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.801425125696426e-05,
+      "loss": 0.9945,
+      "step": 28225
+    },
+    {
+      "epoch": 1.91805951895638,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.8012128006522626e-05,
+      "loss": 1.0002,
+      "step": 28230
+    },
+    {
+      "epoch": 1.9183992390270417,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.801000475608099e-05,
+      "loss": 1.0561,
+      "step": 28235
+    },
+    {
+      "epoch": 1.9187389590977035,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.8007881505639354e-05,
+      "loss": 1.1017,
+      "step": 28240
+    },
+    {
+      "epoch": 1.9190786791683654,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.800575825519772e-05,
+      "loss": 1.0788,
+      "step": 28245
+    },
+    {
+      "epoch": 1.919418399239027,
+      "grad_norm": 6.375,
+      "learning_rate": 3.800363500475608e-05,
+      "loss": 1.1105,
+      "step": 28250
+    },
+    {
+      "epoch": 1.9197581193096889,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.8001511754314446e-05,
+      "loss": 1.0453,
+      "step": 28255
+    },
+    {
+      "epoch": 1.9200978393803507,
+      "grad_norm": 6.125,
+      "learning_rate": 3.799938850387281e-05,
+      "loss": 1.0202,
+      "step": 28260
+    },
+    {
+      "epoch": 1.9204375594510124,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.7997265253431174e-05,
+      "loss": 1.1338,
+      "step": 28265
+    },
+    {
+      "epoch": 1.920777279521674,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.799514200298954e-05,
+      "loss": 0.9697,
+      "step": 28270
+    },
+    {
+      "epoch": 1.921116999592336,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.79930187525479e-05,
+      "loss": 1.0853,
+      "step": 28275
+    },
+    {
+      "epoch": 1.9214567196629977,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7990895502106266e-05,
+      "loss": 0.9959,
+      "step": 28280
+    },
+    {
+      "epoch": 1.9217964397336593,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.798877225166463e-05,
+      "loss": 1.1086,
+      "step": 28285
+    },
+    {
+      "epoch": 1.9221361598043214,
+      "grad_norm": 6.0,
+      "learning_rate": 3.7986649001222994e-05,
+      "loss": 1.1642,
+      "step": 28290
+    },
+    {
+      "epoch": 1.922475879874983,
+      "grad_norm": 6.125,
+      "learning_rate": 3.798452575078136e-05,
+      "loss": 1.0674,
+      "step": 28295
+    },
+    {
+      "epoch": 1.9228155999456447,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.798240250033972e-05,
+      "loss": 1.1072,
+      "step": 28300
+    },
+    {
+      "epoch": 1.9231553200163065,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.7980279249898086e-05,
+      "loss": 1.0449,
+      "step": 28305
+    },
+    {
+      "epoch": 1.9234950400869684,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.797815599945645e-05,
+      "loss": 1.0676,
+      "step": 28310
+    },
+    {
+      "epoch": 1.92383476015763,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.7976032749014814e-05,
+      "loss": 1.0679,
+      "step": 28315
+    },
+    {
+      "epoch": 1.9241744802282919,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.797390949857318e-05,
+      "loss": 1.1166,
+      "step": 28320
+    },
+    {
+      "epoch": 1.9245142002989537,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.797178624813154e-05,
+      "loss": 1.0125,
+      "step": 28325
+    },
+    {
+      "epoch": 1.9248539203696153,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.7969662997689906e-05,
+      "loss": 0.9919,
+      "step": 28330
+    },
+    {
+      "epoch": 1.9251936404402772,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.796753974724827e-05,
+      "loss": 1.1125,
+      "step": 28335
+    },
+    {
+      "epoch": 1.925533360510939,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.7965416496806634e-05,
+      "loss": 1.1221,
+      "step": 28340
+    },
+    {
+      "epoch": 1.9258730805816007,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.7963293246365e-05,
+      "loss": 1.0302,
+      "step": 28345
+    },
+    {
+      "epoch": 1.9262128006522625,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.796116999592336e-05,
+      "loss": 1.0897,
+      "step": 28350
+    },
+    {
+      "epoch": 1.9265525207229244,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.7959046745481726e-05,
+      "loss": 1.0549,
+      "step": 28355
+    },
+    {
+      "epoch": 1.926892240793586,
+      "grad_norm": 6.625,
+      "learning_rate": 3.7956923495040083e-05,
+      "loss": 1.0539,
+      "step": 28360
+    },
+    {
+      "epoch": 1.9272319608642479,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.7954800244598454e-05,
+      "loss": 1.1116,
+      "step": 28365
+    },
+    {
+      "epoch": 1.9275716809349097,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.795267699415682e-05,
+      "loss": 1.1183,
+      "step": 28370
+    },
+    {
+      "epoch": 1.9279114010055713,
+      "grad_norm": 6.5,
+      "learning_rate": 3.7950553743715175e-05,
+      "loss": 1.113,
+      "step": 28375
+    },
+    {
+      "epoch": 1.9282511210762332,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.7948430493273546e-05,
+      "loss": 1.0187,
+      "step": 28380
+    },
+    {
+      "epoch": 1.928590841146895,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.794630724283191e-05,
+      "loss": 1.0432,
+      "step": 28385
+    },
+    {
+      "epoch": 1.9289305612175567,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.794418399239027e-05,
+      "loss": 1.0322,
+      "step": 28390
+    },
+    {
+      "epoch": 1.9292702812882185,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.794206074194864e-05,
+      "loss": 1.1082,
+      "step": 28395
+    },
+    {
+      "epoch": 1.9296100013588804,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.7939937491507e-05,
+      "loss": 1.0582,
+      "step": 28400
+    },
+    {
+      "epoch": 1.929949721429542,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.793781424106536e-05,
+      "loss": 1.1307,
+      "step": 28405
+    },
+    {
+      "epoch": 1.9302894415002039,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.793569099062373e-05,
+      "loss": 1.0772,
+      "step": 28410
+    },
+    {
+      "epoch": 1.9306291615708657,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7933567740182094e-05,
+      "loss": 1.0155,
+      "step": 28415
+    },
+    {
+      "epoch": 1.9309688816415274,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.793144448974045e-05,
+      "loss": 1.0745,
+      "step": 28420
+    },
+    {
+      "epoch": 1.9313086017121892,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.792932123929882e-05,
+      "loss": 1.0183,
+      "step": 28425
+    },
+    {
+      "epoch": 1.931648321782851,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.792719798885718e-05,
+      "loss": 1.0994,
+      "step": 28430
+    },
+    {
+      "epoch": 1.9319880418535127,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.7925074738415543e-05,
+      "loss": 1.1013,
+      "step": 28435
+    },
+    {
+      "epoch": 1.9323277619241743,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.7922951487973914e-05,
+      "loss": 1.0542,
+      "step": 28440
+    },
+    {
+      "epoch": 1.9326674819948364,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.792082823753227e-05,
+      "loss": 1.0405,
+      "step": 28445
+    },
+    {
+      "epoch": 1.933007202065498,
+      "grad_norm": 6.125,
+      "learning_rate": 3.791870498709064e-05,
+      "loss": 0.9917,
+      "step": 28450
+    },
+    {
+      "epoch": 1.9333469221361597,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.7916581736649006e-05,
+      "loss": 1.1334,
+      "step": 28455
+    },
+    {
+      "epoch": 1.9336866422068217,
+      "grad_norm": 5.875,
+      "learning_rate": 3.7914458486207363e-05,
+      "loss": 1.0619,
+      "step": 28460
+    },
+    {
+      "epoch": 1.9340263622774834,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.7912335235765734e-05,
+      "loss": 0.992,
+      "step": 28465
+    },
+    {
+      "epoch": 1.934366082348145,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.79102119853241e-05,
+      "loss": 1.0688,
+      "step": 28470
+    },
+    {
+      "epoch": 1.9347058024188069,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.7908088734882456e-05,
+      "loss": 1.0555,
+      "step": 28475
+    },
+    {
+      "epoch": 1.9350455224894687,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.7905965484440826e-05,
+      "loss": 1.1496,
+      "step": 28480
+    },
+    {
+      "epoch": 1.9353852425601303,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.790384223399919e-05,
+      "loss": 1.1341,
+      "step": 28485
+    },
+    {
+      "epoch": 1.9357249626307922,
+      "grad_norm": 5.03125,
+      "learning_rate": 3.790171898355755e-05,
+      "loss": 0.9713,
+      "step": 28490
+    },
+    {
+      "epoch": 1.936064682701454,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.789959573311592e-05,
+      "loss": 1.0938,
+      "step": 28495
+    },
+    {
+      "epoch": 1.9364044027721157,
+      "grad_norm": 6.75,
+      "learning_rate": 3.7897472482674276e-05,
+      "loss": 1.1063,
+      "step": 28500
+    },
+    {
+      "epoch": 1.9367441228427775,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.789534923223264e-05,
+      "loss": 1.0777,
+      "step": 28505
+    },
+    {
+      "epoch": 1.9370838429134394,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.789322598179101e-05,
+      "loss": 1.0817,
+      "step": 28510
+    },
+    {
+      "epoch": 1.937423562984101,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.789110273134937e-05,
+      "loss": 1.1236,
+      "step": 28515
+    },
+    {
+      "epoch": 1.9377632830547629,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.788897948090773e-05,
+      "loss": 1.0479,
+      "step": 28520
+    },
+    {
+      "epoch": 1.9381030031254247,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.78868562304661e-05,
+      "loss": 1.063,
+      "step": 28525
+    },
+    {
+      "epoch": 1.9384427231960863,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.788473298002446e-05,
+      "loss": 1.0574,
+      "step": 28530
+    },
+    {
+      "epoch": 1.9387824432667482,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.7882609729582824e-05,
+      "loss": 1.1019,
+      "step": 28535
+    },
+    {
+      "epoch": 1.93912216333741,
+      "grad_norm": 5.875,
+      "learning_rate": 3.7880486479141194e-05,
+      "loss": 1.0595,
+      "step": 28540
+    },
+    {
+      "epoch": 1.9394618834080717,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.787836322869955e-05,
+      "loss": 1.0791,
+      "step": 28545
+    },
+    {
+      "epoch": 1.9398016034787335,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.7876239978257916e-05,
+      "loss": 0.9898,
+      "step": 28550
+    },
+    {
+      "epoch": 1.9401413235493954,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.7874116727816286e-05,
+      "loss": 1.0858,
+      "step": 28555
+    },
+    {
+      "epoch": 1.940481043620057,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7871993477374644e-05,
+      "loss": 1.0386,
+      "step": 28560
+    },
+    {
+      "epoch": 1.9408207636907189,
+      "grad_norm": 6.125,
+      "learning_rate": 3.786987022693301e-05,
+      "loss": 1.0784,
+      "step": 28565
+    },
+    {
+      "epoch": 1.9411604837613807,
+      "grad_norm": 6.5,
+      "learning_rate": 3.786774697649138e-05,
+      "loss": 1.1035,
+      "step": 28570
+    },
+    {
+      "epoch": 1.9415002038320424,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.7865623726049736e-05,
+      "loss": 0.9907,
+      "step": 28575
+    },
+    {
+      "epoch": 1.9418399239027042,
+      "grad_norm": 6.125,
+      "learning_rate": 3.78635004756081e-05,
+      "loss": 1.03,
+      "step": 28580
+    },
+    {
+      "epoch": 1.942179643973366,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.7861377225166464e-05,
+      "loss": 1.0124,
+      "step": 28585
+    },
+    {
+      "epoch": 1.9425193640440277,
+      "grad_norm": 6.0,
+      "learning_rate": 3.785925397472483e-05,
+      "loss": 1.0431,
+      "step": 28590
+    },
+    {
+      "epoch": 1.9428590841146895,
+      "grad_norm": 5.875,
+      "learning_rate": 3.785713072428319e-05,
+      "loss": 1.0545,
+      "step": 28595
+    },
+    {
+      "epoch": 1.9431988041853514,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.7855007473841556e-05,
+      "loss": 1.0667,
+      "step": 28600
+    },
+    {
+      "epoch": 1.943538524256013,
+      "grad_norm": 6.0,
+      "learning_rate": 3.785288422339992e-05,
+      "loss": 1.0544,
+      "step": 28605
+    },
+    {
+      "epoch": 1.9438782443266747,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.7850760972958284e-05,
+      "loss": 1.0047,
+      "step": 28610
+    },
+    {
+      "epoch": 1.9442179643973367,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.784863772251665e-05,
+      "loss": 1.096,
+      "step": 28615
+    },
+    {
+      "epoch": 1.9445576844679984,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.784651447207501e-05,
+      "loss": 0.9755,
+      "step": 28620
+    },
+    {
+      "epoch": 1.94489740453866,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.7844391221633376e-05,
+      "loss": 1.1036,
+      "step": 28625
+    },
+    {
+      "epoch": 1.945237124609322,
+      "grad_norm": 6.125,
+      "learning_rate": 3.784226797119174e-05,
+      "loss": 1.0011,
+      "step": 28630
+    },
+    {
+      "epoch": 1.9455768446799837,
+      "grad_norm": 6.125,
+      "learning_rate": 3.7840144720750104e-05,
+      "loss": 0.9715,
+      "step": 28635
+    },
+    {
+      "epoch": 1.9459165647506453,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.783802147030847e-05,
+      "loss": 1.0558,
+      "step": 28640
+    },
+    {
+      "epoch": 1.9462562848213072,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.783589821986683e-05,
+      "loss": 1.0331,
+      "step": 28645
+    },
+    {
+      "epoch": 1.946596004891969,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7833774969425196e-05,
+      "loss": 0.9888,
+      "step": 28650
+    },
+    {
+      "epoch": 1.9469357249626307,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.783165171898356e-05,
+      "loss": 1.0393,
+      "step": 28655
+    },
+    {
+      "epoch": 1.9472754450332925,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.7829528468541924e-05,
+      "loss": 1.1032,
+      "step": 28660
+    },
+    {
+      "epoch": 1.9476151651039544,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.782740521810029e-05,
+      "loss": 0.9994,
+      "step": 28665
+    },
+    {
+      "epoch": 1.947954885174616,
+      "grad_norm": 5.875,
+      "learning_rate": 3.782528196765865e-05,
+      "loss": 1.0193,
+      "step": 28670
+    },
+    {
+      "epoch": 1.9482946052452779,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.7823158717217016e-05,
+      "loss": 0.9802,
+      "step": 28675
+    },
+    {
+      "epoch": 1.9486343253159397,
+      "grad_norm": 5.375,
+      "learning_rate": 3.782103546677538e-05,
+      "loss": 1.0782,
+      "step": 28680
+    },
+    {
+      "epoch": 1.9489740453866014,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.7818912216333744e-05,
+      "loss": 1.0662,
+      "step": 28685
+    },
+    {
+      "epoch": 1.9493137654572632,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.781678896589211e-05,
+      "loss": 1.0986,
+      "step": 28690
+    },
+    {
+      "epoch": 1.949653485527925,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.781466571545047e-05,
+      "loss": 1.0684,
+      "step": 28695
+    },
+    {
+      "epoch": 1.9499932055985867,
+      "grad_norm": 5.75,
+      "learning_rate": 3.781254246500883e-05,
+      "loss": 1.1021,
+      "step": 28700
+    },
+    {
+      "epoch": 1.9503329256692485,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.78104192145672e-05,
+      "loss": 0.9908,
+      "step": 28705
+    },
+    {
+      "epoch": 1.9506726457399104,
+      "grad_norm": 5.875,
+      "learning_rate": 3.7808295964125564e-05,
+      "loss": 1.0481,
+      "step": 28710
+    },
+    {
+      "epoch": 1.951012365810572,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.780617271368392e-05,
+      "loss": 1.1489,
+      "step": 28715
+    },
+    {
+      "epoch": 1.9513520858812339,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.780404946324229e-05,
+      "loss": 1.0569,
+      "step": 28720
+    },
+    {
+      "epoch": 1.9516918059518957,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.7801926212800656e-05,
+      "loss": 1.1218,
+      "step": 28725
+    },
+    {
+      "epoch": 1.9520315260225574,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.779980296235901e-05,
+      "loss": 1.0097,
+      "step": 28730
+    },
+    {
+      "epoch": 1.9523712460932192,
+      "grad_norm": 5.75,
+      "learning_rate": 3.7797679711917384e-05,
+      "loss": 1.0777,
+      "step": 28735
+    },
+    {
+      "epoch": 1.952710966163881,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.779555646147575e-05,
+      "loss": 1.0319,
+      "step": 28740
+    },
+    {
+      "epoch": 1.9530506862345427,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.7793433211034105e-05,
+      "loss": 1.0718,
+      "step": 28745
+    },
+    {
+      "epoch": 1.9533904063052046,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.7791309960592476e-05,
+      "loss": 1.0923,
+      "step": 28750
+    },
+    {
+      "epoch": 1.9537301263758664,
+      "grad_norm": 6.125,
+      "learning_rate": 3.778918671015084e-05,
+      "loss": 1.0568,
+      "step": 28755
+    },
+    {
+      "epoch": 1.954069846446528,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.77870634597092e-05,
+      "loss": 1.0601,
+      "step": 28760
+    },
+    {
+      "epoch": 1.95440956651719,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.778494020926757e-05,
+      "loss": 1.1185,
+      "step": 28765
+    },
+    {
+      "epoch": 1.9547492865878517,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.778281695882593e-05,
+      "loss": 1.0512,
+      "step": 28770
+    },
+    {
+      "epoch": 1.9550890066585134,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.778069370838429e-05,
+      "loss": 1.082,
+      "step": 28775
+    },
+    {
+      "epoch": 1.955428726729175,
+      "grad_norm": 6.875,
+      "learning_rate": 3.777857045794266e-05,
+      "loss": 1.1167,
+      "step": 28780
+    },
+    {
+      "epoch": 1.955768446799837,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.777644720750102e-05,
+      "loss": 1.0697,
+      "step": 28785
+    },
+    {
+      "epoch": 1.9561081668704987,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.777432395705939e-05,
+      "loss": 1.0333,
+      "step": 28790
+    },
+    {
+      "epoch": 1.9564478869411603,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.777220070661775e-05,
+      "loss": 1.153,
+      "step": 28795
+    },
+    {
+      "epoch": 1.9567876070118224,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.777007745617611e-05,
+      "loss": 1.0979,
+      "step": 28800
+    },
+    {
+      "epoch": 1.957127327082484,
+      "grad_norm": 5.75,
+      "learning_rate": 3.776795420573448e-05,
+      "loss": 0.9834,
+      "step": 28805
+    },
+    {
+      "epoch": 1.9574670471531457,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.7765830955292844e-05,
+      "loss": 1.0958,
+      "step": 28810
+    },
+    {
+      "epoch": 1.9578067672238075,
+      "grad_norm": 6.125,
+      "learning_rate": 3.77637077048512e-05,
+      "loss": 1.001,
+      "step": 28815
+    },
+    {
+      "epoch": 1.9581464872944694,
+      "grad_norm": 6.0,
+      "learning_rate": 3.776158445440957e-05,
+      "loss": 1.0807,
+      "step": 28820
+    },
+    {
+      "epoch": 1.958486207365131,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.7759461203967936e-05,
+      "loss": 1.0346,
+      "step": 28825
+    },
+    {
+      "epoch": 1.9588259274357929,
+      "grad_norm": 5.875,
+      "learning_rate": 3.775733795352629e-05,
+      "loss": 1.0169,
+      "step": 28830
+    },
+    {
+      "epoch": 1.9591656475064547,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.7755214703084664e-05,
+      "loss": 0.9488,
+      "step": 28835
+    },
+    {
+      "epoch": 1.9595053675771164,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.775309145264303e-05,
+      "loss": 1.0635,
+      "step": 28840
+    },
+    {
+      "epoch": 1.9598450876477782,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.7750968202201385e-05,
+      "loss": 1.0493,
+      "step": 28845
+    },
+    {
+      "epoch": 1.96018480771844,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.7748844951759756e-05,
+      "loss": 1.0796,
+      "step": 28850
+    },
+    {
+      "epoch": 1.9605245277891017,
+      "grad_norm": 5.75,
+      "learning_rate": 3.774672170131811e-05,
+      "loss": 1.1298,
+      "step": 28855
+    },
+    {
+      "epoch": 1.9608642478597635,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.774459845087648e-05,
+      "loss": 1.0802,
+      "step": 28860
+    },
+    {
+      "epoch": 1.9612039679304254,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.774247520043485e-05,
+      "loss": 1.0597,
+      "step": 28865
+    },
+    {
+      "epoch": 1.961543688001087,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.7740351949993205e-05,
+      "loss": 1.078,
+      "step": 28870
+    },
+    {
+      "epoch": 1.9618834080717489,
+      "grad_norm": 6.875,
+      "learning_rate": 3.773822869955157e-05,
+      "loss": 1.0204,
+      "step": 28875
+    },
+    {
+      "epoch": 1.9622231281424107,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.773610544910994e-05,
+      "loss": 1.0136,
+      "step": 28880
+    },
+    {
+      "epoch": 1.9625628482130724,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.77339821986683e-05,
+      "loss": 0.9707,
+      "step": 28885
+    },
+    {
+      "epoch": 1.9629025682837342,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.773185894822666e-05,
+      "loss": 1.0906,
+      "step": 28890
+    },
+    {
+      "epoch": 1.963242288354396,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.772973569778503e-05,
+      "loss": 0.9752,
+      "step": 28895
+    },
+    {
+      "epoch": 1.9635820084250577,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.772761244734339e-05,
+      "loss": 1.0679,
+      "step": 28900
+    },
+    {
+      "epoch": 1.9639217284957196,
+      "grad_norm": 6.875,
+      "learning_rate": 3.772548919690175e-05,
+      "loss": 1.0545,
+      "step": 28905
+    },
+    {
+      "epoch": 1.9642614485663814,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7723365946460124e-05,
+      "loss": 1.0531,
+      "step": 28910
+    },
+    {
+      "epoch": 1.964601168637043,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.772124269601848e-05,
+      "loss": 1.1355,
+      "step": 28915
+    },
+    {
+      "epoch": 1.964940888707705,
+      "grad_norm": 6.625,
+      "learning_rate": 3.7719119445576845e-05,
+      "loss": 1.2154,
+      "step": 28920
+    },
+    {
+      "epoch": 1.9652806087783667,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.771699619513521e-05,
+      "loss": 1.1375,
+      "step": 28925
+    },
+    {
+      "epoch": 1.9656203288490284,
+      "grad_norm": 6.0,
+      "learning_rate": 3.771487294469357e-05,
+      "loss": 1.0159,
+      "step": 28930
+    },
+    {
+      "epoch": 1.9659600489196902,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.771274969425194e-05,
+      "loss": 0.9619,
+      "step": 28935
+    },
+    {
+      "epoch": 1.966299768990352,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.77106264438103e-05,
+      "loss": 1.0777,
+      "step": 28940
+    },
+    {
+      "epoch": 1.9666394890610137,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.7708503193368665e-05,
+      "loss": 1.103,
+      "step": 28945
+    },
+    {
+      "epoch": 1.9669792091316753,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.770637994292703e-05,
+      "loss": 1.033,
+      "step": 28950
+    },
+    {
+      "epoch": 1.9673189292023374,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.770425669248539e-05,
+      "loss": 0.9997,
+      "step": 28955
+    },
+    {
+      "epoch": 1.967658649272999,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.770213344204376e-05,
+      "loss": 1.077,
+      "step": 28960
+    },
+    {
+      "epoch": 1.9679983693436607,
+      "grad_norm": 6.25,
+      "learning_rate": 3.770001019160212e-05,
+      "loss": 0.9356,
+      "step": 28965
+    },
+    {
+      "epoch": 1.9683380894143228,
+      "grad_norm": 6.25,
+      "learning_rate": 3.7697886941160485e-05,
+      "loss": 1.0946,
+      "step": 28970
+    },
+    {
+      "epoch": 1.9686778094849844,
+      "grad_norm": 5.625,
+      "learning_rate": 3.769576369071885e-05,
+      "loss": 1.0048,
+      "step": 28975
+    },
+    {
+      "epoch": 1.969017529555646,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.769364044027721e-05,
+      "loss": 1.0591,
+      "step": 28980
+    },
+    {
+      "epoch": 1.9693572496263079,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.769151718983558e-05,
+      "loss": 1.0127,
+      "step": 28985
+    },
+    {
+      "epoch": 1.9696969696969697,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.768939393939394e-05,
+      "loss": 1.1948,
+      "step": 28990
+    },
+    {
+      "epoch": 1.9700366897676314,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.7687270688952305e-05,
+      "loss": 1.0718,
+      "step": 28995
+    },
+    {
+      "epoch": 1.9703764098382932,
+      "grad_norm": 6.25,
+      "learning_rate": 3.768514743851067e-05,
+      "loss": 1.0382,
+      "step": 29000
+    },
+    {
+      "epoch": 1.970716129908955,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.768302418806903e-05,
+      "loss": 1.1116,
+      "step": 29005
+    },
+    {
+      "epoch": 1.9710558499796167,
+      "grad_norm": 6.0,
+      "learning_rate": 3.76809009376274e-05,
+      "loss": 1.0399,
+      "step": 29010
+    },
+    {
+      "epoch": 1.9713955700502785,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.767877768718576e-05,
+      "loss": 1.0319,
+      "step": 29015
+    },
+    {
+      "epoch": 1.9717352901209404,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.7676654436744125e-05,
+      "loss": 1.0574,
+      "step": 29020
+    },
+    {
+      "epoch": 1.972075010191602,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.767453118630249e-05,
+      "loss": 1.0878,
+      "step": 29025
+    },
+    {
+      "epoch": 1.9724147302622639,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.767240793586085e-05,
+      "loss": 1.0693,
+      "step": 29030
+    },
+    {
+      "epoch": 1.9727544503329257,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.767028468541922e-05,
+      "loss": 1.078,
+      "step": 29035
+    },
+    {
+      "epoch": 1.9730941704035874,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.766816143497758e-05,
+      "loss": 1.0176,
+      "step": 29040
+    },
+    {
+      "epoch": 1.9734338904742492,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.7666038184535945e-05,
+      "loss": 1.0246,
+      "step": 29045
+    },
+    {
+      "epoch": 1.973773610544911,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.766391493409431e-05,
+      "loss": 0.9968,
+      "step": 29050
+    },
+    {
+      "epoch": 1.9741133306155727,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.7661791683652666e-05,
+      "loss": 1.0908,
+      "step": 29055
+    },
+    {
+      "epoch": 1.9744530506862346,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.765966843321104e-05,
+      "loss": 1.0732,
+      "step": 29060
+    },
+    {
+      "epoch": 1.9747927707568964,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.76575451827694e-05,
+      "loss": 1.0791,
+      "step": 29065
+    },
+    {
+      "epoch": 1.975132490827558,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.765542193232776e-05,
+      "loss": 0.9804,
+      "step": 29070
+    },
+    {
+      "epoch": 1.97547221089822,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.765329868188613e-05,
+      "loss": 1.0872,
+      "step": 29075
+    },
+    {
+      "epoch": 1.9758119309688817,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.765117543144449e-05,
+      "loss": 1.0843,
+      "step": 29080
+    },
+    {
+      "epoch": 1.9761516510395434,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.764905218100285e-05,
+      "loss": 1.1091,
+      "step": 29085
+    },
+    {
+      "epoch": 1.9764913711102052,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.764692893056122e-05,
+      "loss": 1.064,
+      "step": 29090
+    },
+    {
+      "epoch": 1.976831091180867,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.7644805680119585e-05,
+      "loss": 1.0222,
+      "step": 29095
+    },
+    {
+      "epoch": 1.9771708112515287,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.764268242967794e-05,
+      "loss": 1.0178,
+      "step": 29100
+    },
+    {
+      "epoch": 1.9775105313221906,
+      "grad_norm": 4.5,
+      "learning_rate": 3.764055917923631e-05,
+      "loss": 1.0303,
+      "step": 29105
+    },
+    {
+      "epoch": 1.9778502513928524,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.763843592879468e-05,
+      "loss": 0.9707,
+      "step": 29110
+    },
+    {
+      "epoch": 1.978189971463514,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7636312678353034e-05,
+      "loss": 1.0751,
+      "step": 29115
+    },
+    {
+      "epoch": 1.9785296915341757,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.7634189427911405e-05,
+      "loss": 1.0035,
+      "step": 29120
+    },
+    {
+      "epoch": 1.9788694116048378,
+      "grad_norm": 5.625,
+      "learning_rate": 3.763206617746976e-05,
+      "loss": 1.1142,
+      "step": 29125
+    },
+    {
+      "epoch": 1.9792091316754994,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.762994292702813e-05,
+      "loss": 1.0675,
+      "step": 29130
+    },
+    {
+      "epoch": 1.979548851746161,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.76278196765865e-05,
+      "loss": 1.0142,
+      "step": 29135
+    },
+    {
+      "epoch": 1.979888571816823,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.7625696426144854e-05,
+      "loss": 1.0368,
+      "step": 29140
+    },
+    {
+      "epoch": 1.9802282918874847,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.7623573175703225e-05,
+      "loss": 1.042,
+      "step": 29145
+    },
+    {
+      "epoch": 1.9805680119581464,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.762144992526159e-05,
+      "loss": 1.0747,
+      "step": 29150
+    },
+    {
+      "epoch": 1.9809077320288082,
+      "grad_norm": 5.625,
+      "learning_rate": 3.7619326674819946e-05,
+      "loss": 1.0884,
+      "step": 29155
+    },
+    {
+      "epoch": 1.98124745209947,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.761720342437832e-05,
+      "loss": 1.0131,
+      "step": 29160
+    },
+    {
+      "epoch": 1.9815871721701317,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.761508017393668e-05,
+      "loss": 1.0012,
+      "step": 29165
+    },
+    {
+      "epoch": 1.9819268922407935,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.761295692349504e-05,
+      "loss": 1.056,
+      "step": 29170
+    },
+    {
+      "epoch": 1.9822666123114554,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.761083367305341e-05,
+      "loss": 1.1538,
+      "step": 29175
+    },
+    {
+      "epoch": 1.982606332382117,
+      "grad_norm": 5.875,
+      "learning_rate": 3.760871042261177e-05,
+      "loss": 1.0548,
+      "step": 29180
+    },
+    {
+      "epoch": 1.9829460524527789,
+      "grad_norm": 6.125,
+      "learning_rate": 3.760658717217013e-05,
+      "loss": 1.0157,
+      "step": 29185
+    },
+    {
+      "epoch": 1.9832857725234407,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.76044639217285e-05,
+      "loss": 1.0939,
+      "step": 29190
+    },
+    {
+      "epoch": 1.9836254925941024,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.7602340671286865e-05,
+      "loss": 1.0553,
+      "step": 29195
+    },
+    {
+      "epoch": 1.9839652126647642,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.760021742084522e-05,
+      "loss": 1.0443,
+      "step": 29200
+    },
+    {
+      "epoch": 1.984304932735426,
+      "grad_norm": 6.625,
+      "learning_rate": 3.759809417040359e-05,
+      "loss": 1.0947,
+      "step": 29205
+    },
+    {
+      "epoch": 1.9846446528060877,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.759597091996195e-05,
+      "loss": 1.0,
+      "step": 29210
+    },
+    {
+      "epoch": 1.9849843728767496,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.7593847669520315e-05,
+      "loss": 1.0899,
+      "step": 29215
+    },
+    {
+      "epoch": 1.9853240929474114,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.7591724419078685e-05,
+      "loss": 1.0298,
+      "step": 29220
+    },
+    {
+      "epoch": 1.985663813018073,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.758960116863704e-05,
+      "loss": 1.0123,
+      "step": 29225
+    },
+    {
+      "epoch": 1.986003533088735,
+      "grad_norm": 6.25,
+      "learning_rate": 3.7587477918195407e-05,
+      "loss": 1.0379,
+      "step": 29230
+    },
+    {
+      "epoch": 1.9863432531593967,
+      "grad_norm": 5.75,
+      "learning_rate": 3.758535466775378e-05,
+      "loss": 1.0163,
+      "step": 29235
+    },
+    {
+      "epoch": 1.9866829732300584,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.7583231417312135e-05,
+      "loss": 1.1045,
+      "step": 29240
+    },
+    {
+      "epoch": 1.9870226933007202,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.75811081668705e-05,
+      "loss": 1.1234,
+      "step": 29245
+    },
+    {
+      "epoch": 1.987362413371382,
+      "grad_norm": 6.375,
+      "learning_rate": 3.757898491642887e-05,
+      "loss": 1.0235,
+      "step": 29250
+    },
+    {
+      "epoch": 1.9877021334420437,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.7576861665987227e-05,
+      "loss": 1.0587,
+      "step": 29255
+    },
+    {
+      "epoch": 1.9880418535127056,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.757473841554559e-05,
+      "loss": 1.117,
+      "step": 29260
+    },
+    {
+      "epoch": 1.9883815735833674,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.757261516510396e-05,
+      "loss": 1.0479,
+      "step": 29265
+    },
+    {
+      "epoch": 1.988721293654029,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.757049191466232e-05,
+      "loss": 1.0866,
+      "step": 29270
+    },
+    {
+      "epoch": 1.989061013724691,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.756836866422068e-05,
+      "loss": 0.9852,
+      "step": 29275
+    },
+    {
+      "epoch": 1.9894007337953528,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.7566245413779047e-05,
+      "loss": 1.1646,
+      "step": 29280
+    },
+    {
+      "epoch": 1.9897404538660144,
+      "grad_norm": 5.09375,
+      "learning_rate": 3.756412216333741e-05,
+      "loss": 1.1633,
+      "step": 29285
+    },
+    {
+      "epoch": 1.990080173936676,
+      "grad_norm": 5.75,
+      "learning_rate": 3.7561998912895775e-05,
+      "loss": 1.0718,
+      "step": 29290
+    },
+    {
+      "epoch": 1.990419894007338,
+      "grad_norm": 5.75,
+      "learning_rate": 3.755987566245414e-05,
+      "loss": 1.1213,
+      "step": 29295
+    },
+    {
+      "epoch": 1.9907596140779997,
+      "grad_norm": 5.125,
+      "learning_rate": 3.75577524120125e-05,
+      "loss": 1.081,
+      "step": 29300
+    },
+    {
+      "epoch": 1.9910993341486614,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.7555629161570867e-05,
+      "loss": 1.0519,
+      "step": 29305
+    },
+    {
+      "epoch": 1.9914390542193234,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.755350591112923e-05,
+      "loss": 1.0578,
+      "step": 29310
+    },
+    {
+      "epoch": 1.991778774289985,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7551382660687595e-05,
+      "loss": 1.1129,
+      "step": 29315
+    },
+    {
+      "epoch": 1.9921184943606467,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.754925941024596e-05,
+      "loss": 1.1076,
+      "step": 29320
+    },
+    {
+      "epoch": 1.9924582144313085,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.754713615980432e-05,
+      "loss": 1.0437,
+      "step": 29325
+    },
+    {
+      "epoch": 1.9927979345019704,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.7545012909362687e-05,
+      "loss": 1.0458,
+      "step": 29330
+    },
+    {
+      "epoch": 1.993137654572632,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.754288965892105e-05,
+      "loss": 1.0773,
+      "step": 29335
+    },
+    {
+      "epoch": 1.9934773746432939,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.7540766408479415e-05,
+      "loss": 1.0605,
+      "step": 29340
+    },
+    {
+      "epoch": 1.9938170947139557,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.753864315803778e-05,
+      "loss": 0.9143,
+      "step": 29345
+    },
+    {
+      "epoch": 1.9941568147846174,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.753651990759614e-05,
+      "loss": 1.0154,
+      "step": 29350
+    },
+    {
+      "epoch": 1.9944965348552792,
+      "grad_norm": 6.625,
+      "learning_rate": 3.753439665715451e-05,
+      "loss": 0.9778,
+      "step": 29355
+    },
+    {
+      "epoch": 1.994836254925941,
+      "grad_norm": 6.0,
+      "learning_rate": 3.753227340671287e-05,
+      "loss": 1.059,
+      "step": 29360
+    },
+    {
+      "epoch": 1.9951759749966027,
+      "grad_norm": 6.125,
+      "learning_rate": 3.7530150156271235e-05,
+      "loss": 1.1135,
+      "step": 29365
+    },
+    {
+      "epoch": 1.9955156950672646,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.75280269058296e-05,
+      "loss": 1.075,
+      "step": 29370
+    },
+    {
+      "epoch": 1.9958554151379264,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.752590365538796e-05,
+      "loss": 1.0047,
+      "step": 29375
+    },
+    {
+      "epoch": 1.996195135208588,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.752378040494633e-05,
+      "loss": 1.0741,
+      "step": 29380
+    },
+    {
+      "epoch": 1.99653485527925,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.752165715450469e-05,
+      "loss": 0.9647,
+      "step": 29385
+    },
+    {
+      "epoch": 1.9968745753499118,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.7519533904063055e-05,
+      "loss": 0.9627,
+      "step": 29390
+    },
+    {
+      "epoch": 1.9972142954205734,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.751741065362142e-05,
+      "loss": 1.0899,
+      "step": 29395
+    },
+    {
+      "epoch": 1.9975540154912352,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.751528740317978e-05,
+      "loss": 1.1823,
+      "step": 29400
+    },
+    {
+      "epoch": 1.997893735561897,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.751316415273815e-05,
+      "loss": 1.0237,
+      "step": 29405
+    },
+    {
+      "epoch": 1.9982334556325587,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7511040902296504e-05,
+      "loss": 1.0119,
+      "step": 29410
+    },
+    {
+      "epoch": 1.9985731757032206,
+      "grad_norm": 5.625,
+      "learning_rate": 3.7508917651854875e-05,
+      "loss": 1.0226,
+      "step": 29415
+    },
+    {
+      "epoch": 1.9989128957738824,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.750679440141324e-05,
+      "loss": 1.0931,
+      "step": 29420
+    },
+    {
+      "epoch": 1.999252615844544,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.7504671150971596e-05,
+      "loss": 1.0418,
+      "step": 29425
+    },
+    {
+      "epoch": 1.999592335915206,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.750254790052997e-05,
+      "loss": 1.0239,
+      "step": 29430
+    },
+    {
+      "epoch": 1.9999320559858678,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.750042465008833e-05,
+      "loss": 1.1435,
+      "step": 29435
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 1.6046582460403442,
+      "eval_runtime": 146.6008,
+      "eval_samples_per_second": 70.388,
+      "eval_steps_per_second": 8.799,
+      "step": 29436
+    },
+    {
+      "epoch": 2.0002717760565294,
+      "grad_norm": 4.1875,
+      "learning_rate": 3.749830139964669e-05,
+      "loss": 0.9432,
+      "step": 29440
+    },
+    {
+      "epoch": 2.000611496127191,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.749617814920506e-05,
+      "loss": 1.0118,
+      "step": 29445
+    },
+    {
+      "epoch": 2.000951216197853,
+      "grad_norm": 6.375,
+      "learning_rate": 3.749405489876342e-05,
+      "loss": 0.9981,
+      "step": 29450
+    },
+    {
+      "epoch": 2.0012909362685147,
+      "grad_norm": 6.0,
+      "learning_rate": 3.749193164832178e-05,
+      "loss": 0.8998,
+      "step": 29455
+    },
+    {
+      "epoch": 2.0016306563391764,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.748980839788015e-05,
+      "loss": 0.9131,
+      "step": 29460
+    },
+    {
+      "epoch": 2.0019703764098384,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.7487685147438515e-05,
+      "loss": 0.8879,
+      "step": 29465
+    },
+    {
+      "epoch": 2.0023100964805,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.748556189699687e-05,
+      "loss": 0.8777,
+      "step": 29470
+    },
+    {
+      "epoch": 2.0026498165511617,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.748343864655524e-05,
+      "loss": 0.904,
+      "step": 29475
+    },
+    {
+      "epoch": 2.0029895366218238,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.74813153961136e-05,
+      "loss": 0.9461,
+      "step": 29480
+    },
+    {
+      "epoch": 2.0033292566924854,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.747919214567197e-05,
+      "loss": 0.9313,
+      "step": 29485
+    },
+    {
+      "epoch": 2.003668976763147,
+      "grad_norm": 6.125,
+      "learning_rate": 3.7477068895230335e-05,
+      "loss": 0.9351,
+      "step": 29490
+    },
+    {
+      "epoch": 2.004008696833809,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.747494564478869e-05,
+      "loss": 1.1042,
+      "step": 29495
+    },
+    {
+      "epoch": 2.0043484169044707,
+      "grad_norm": 6.25,
+      "learning_rate": 3.747282239434706e-05,
+      "loss": 0.9716,
+      "step": 29500
+    },
+    {
+      "epoch": 2.0046881369751324,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.747069914390543e-05,
+      "loss": 1.0188,
+      "step": 29505
+    },
+    {
+      "epoch": 2.0050278570457944,
+      "grad_norm": 6.0,
+      "learning_rate": 3.7468575893463784e-05,
+      "loss": 1.0083,
+      "step": 29510
+    },
+    {
+      "epoch": 2.005367577116456,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.7466452643022155e-05,
+      "loss": 0.9694,
+      "step": 29515
+    },
+    {
+      "epoch": 2.0057072971871177,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.746432939258052e-05,
+      "loss": 0.9634,
+      "step": 29520
+    },
+    {
+      "epoch": 2.00604701725778,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7462206142138876e-05,
+      "loss": 0.9561,
+      "step": 29525
+    },
+    {
+      "epoch": 2.0063867373284414,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.746008289169725e-05,
+      "loss": 0.9871,
+      "step": 29530
+    },
+    {
+      "epoch": 2.006726457399103,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.745795964125561e-05,
+      "loss": 1.0025,
+      "step": 29535
+    },
+    {
+      "epoch": 2.007066177469765,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.745583639081397e-05,
+      "loss": 1.0369,
+      "step": 29540
+    },
+    {
+      "epoch": 2.0074058975404268,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.745371314037234e-05,
+      "loss": 1.0803,
+      "step": 29545
+    },
+    {
+      "epoch": 2.0077456176110884,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.74515898899307e-05,
+      "loss": 0.9667,
+      "step": 29550
+    },
+    {
+      "epoch": 2.0080853376817505,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.744946663948906e-05,
+      "loss": 0.962,
+      "step": 29555
+    },
+    {
+      "epoch": 2.008425057752412,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.744734338904743e-05,
+      "loss": 0.9612,
+      "step": 29560
+    },
+    {
+      "epoch": 2.0087647778230737,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.744522013860579e-05,
+      "loss": 0.9204,
+      "step": 29565
+    },
+    {
+      "epoch": 2.0091044978937354,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.744309688816415e-05,
+      "loss": 0.991,
+      "step": 29570
+    },
+    {
+      "epoch": 2.0094442179643974,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.744097363772252e-05,
+      "loss": 0.932,
+      "step": 29575
+    },
+    {
+      "epoch": 2.009783938035059,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.743885038728088e-05,
+      "loss": 0.99,
+      "step": 29580
+    },
+    {
+      "epoch": 2.0101236581057207,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.7436727136839244e-05,
+      "loss": 0.9927,
+      "step": 29585
+    },
+    {
+      "epoch": 2.0104633781763828,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.7434603886397615e-05,
+      "loss": 0.9261,
+      "step": 29590
+    },
+    {
+      "epoch": 2.0108030982470444,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.743248063595597e-05,
+      "loss": 0.9886,
+      "step": 29595
+    },
+    {
+      "epoch": 2.011142818317706,
+      "grad_norm": 6.25,
+      "learning_rate": 3.7430357385514336e-05,
+      "loss": 0.9915,
+      "step": 29600
+    },
+    {
+      "epoch": 2.011482538388368,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.742823413507271e-05,
+      "loss": 0.9749,
+      "step": 29605
+    },
+    {
+      "epoch": 2.0118222584590297,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.7426110884631064e-05,
+      "loss": 0.8911,
+      "step": 29610
+    },
+    {
+      "epoch": 2.0121619785296914,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.742398763418943e-05,
+      "loss": 0.9228,
+      "step": 29615
+    },
+    {
+      "epoch": 2.0125016986003534,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.74218643837478e-05,
+      "loss": 0.9817,
+      "step": 29620
+    },
+    {
+      "epoch": 2.012841418671015,
+      "grad_norm": 6.25,
+      "learning_rate": 3.7419741133306156e-05,
+      "loss": 0.8835,
+      "step": 29625
+    },
+    {
+      "epoch": 2.0131811387416767,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.741761788286452e-05,
+      "loss": 0.989,
+      "step": 29630
+    },
+    {
+      "epoch": 2.0135208588123388,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.7415494632422884e-05,
+      "loss": 0.9729,
+      "step": 29635
+    },
+    {
+      "epoch": 2.0138605788830004,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.741337138198125e-05,
+      "loss": 0.8724,
+      "step": 29640
+    },
+    {
+      "epoch": 2.014200298953662,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.741124813153961e-05,
+      "loss": 1.0403,
+      "step": 29645
+    },
+    {
+      "epoch": 2.014540019024324,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.7409124881097976e-05,
+      "loss": 0.9947,
+      "step": 29650
+    },
+    {
+      "epoch": 2.0148797390949857,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.740700163065634e-05,
+      "loss": 1.0045,
+      "step": 29655
+    },
+    {
+      "epoch": 2.0152194591656474,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.7404878380214704e-05,
+      "loss": 0.914,
+      "step": 29660
+    },
+    {
+      "epoch": 2.0155591792363095,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.740275512977307e-05,
+      "loss": 0.9777,
+      "step": 29665
+    },
+    {
+      "epoch": 2.015898899306971,
+      "grad_norm": 6.375,
+      "learning_rate": 3.740063187933143e-05,
+      "loss": 0.9791,
+      "step": 29670
+    },
+    {
+      "epoch": 2.0162386193776327,
+      "grad_norm": 6.625,
+      "learning_rate": 3.7398508628889796e-05,
+      "loss": 0.8987,
+      "step": 29675
+    },
+    {
+      "epoch": 2.016578339448295,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.739638537844816e-05,
+      "loss": 0.9772,
+      "step": 29680
+    },
+    {
+      "epoch": 2.0169180595189564,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.7394262128006524e-05,
+      "loss": 0.9427,
+      "step": 29685
+    },
+    {
+      "epoch": 2.017257779589618,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.739213887756489e-05,
+      "loss": 1.0366,
+      "step": 29690
+    },
+    {
+      "epoch": 2.01759749966028,
+      "grad_norm": 7.125,
+      "learning_rate": 3.739001562712325e-05,
+      "loss": 1.0343,
+      "step": 29695
+    },
+    {
+      "epoch": 2.0179372197309418,
+      "grad_norm": 6.75,
+      "learning_rate": 3.7387892376681616e-05,
+      "loss": 0.9349,
+      "step": 29700
+    },
+    {
+      "epoch": 2.0182769398016034,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.738576912623998e-05,
+      "loss": 0.916,
+      "step": 29705
+    },
+    {
+      "epoch": 2.0186166598722655,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.7383645875798344e-05,
+      "loss": 0.9466,
+      "step": 29710
+    },
+    {
+      "epoch": 2.018956379942927,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.738152262535671e-05,
+      "loss": 0.9946,
+      "step": 29715
+    },
+    {
+      "epoch": 2.0192961000135887,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.737939937491507e-05,
+      "loss": 0.9679,
+      "step": 29720
+    },
+    {
+      "epoch": 2.0196358200842504,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7377276124473436e-05,
+      "loss": 0.9859,
+      "step": 29725
+    },
+    {
+      "epoch": 2.0199755401549124,
+      "grad_norm": 6.375,
+      "learning_rate": 3.73751528740318e-05,
+      "loss": 0.7889,
+      "step": 29730
+    },
+    {
+      "epoch": 2.020315260225574,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.7373029623590164e-05,
+      "loss": 0.9746,
+      "step": 29735
+    },
+    {
+      "epoch": 2.0206549802962357,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.737090637314853e-05,
+      "loss": 0.9083,
+      "step": 29740
+    },
+    {
+      "epoch": 2.0209947003668978,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.736878312270689e-05,
+      "loss": 0.9921,
+      "step": 29745
+    },
+    {
+      "epoch": 2.0213344204375594,
+      "grad_norm": 7.25,
+      "learning_rate": 3.7366659872265256e-05,
+      "loss": 0.8367,
+      "step": 29750
+    },
+    {
+      "epoch": 2.021674140508221,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.736453662182362e-05,
+      "loss": 1.0716,
+      "step": 29755
+    },
+    {
+      "epoch": 2.022013860578883,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.7362413371381984e-05,
+      "loss": 0.9867,
+      "step": 29760
+    },
+    {
+      "epoch": 2.0223535806495447,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.736029012094034e-05,
+      "loss": 0.9912,
+      "step": 29765
+    },
+    {
+      "epoch": 2.0226933007202064,
+      "grad_norm": 7.125,
+      "learning_rate": 3.735816687049871e-05,
+      "loss": 0.9168,
+      "step": 29770
+    },
+    {
+      "epoch": 2.0230330207908684,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.7356043620057076e-05,
+      "loss": 0.8524,
+      "step": 29775
+    },
+    {
+      "epoch": 2.02337274086153,
+      "grad_norm": 6.75,
+      "learning_rate": 3.7353920369615433e-05,
+      "loss": 0.9447,
+      "step": 29780
+    },
+    {
+      "epoch": 2.0237124609321917,
+      "grad_norm": 6.75,
+      "learning_rate": 3.7351797119173804e-05,
+      "loss": 0.9232,
+      "step": 29785
+    },
+    {
+      "epoch": 2.024052181002854,
+      "grad_norm": 5.875,
+      "learning_rate": 3.734967386873217e-05,
+      "loss": 0.9342,
+      "step": 29790
+    },
+    {
+      "epoch": 2.0243919010735154,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7347550618290525e-05,
+      "loss": 0.9322,
+      "step": 29795
+    },
+    {
+      "epoch": 2.024731621144177,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.7345427367848896e-05,
+      "loss": 1.008,
+      "step": 29800
+    },
+    {
+      "epoch": 2.025071341214839,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.734330411740726e-05,
+      "loss": 0.9774,
+      "step": 29805
+    },
+    {
+      "epoch": 2.0254110612855007,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.734118086696562e-05,
+      "loss": 0.9924,
+      "step": 29810
+    },
+    {
+      "epoch": 2.0257507813561624,
+      "grad_norm": 6.625,
+      "learning_rate": 3.733905761652399e-05,
+      "loss": 0.8929,
+      "step": 29815
+    },
+    {
+      "epoch": 2.0260905014268245,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.733693436608235e-05,
+      "loss": 1.0443,
+      "step": 29820
+    },
+    {
+      "epoch": 2.026430221497486,
+      "grad_norm": 5.40625,
+      "learning_rate": 3.7334811115640716e-05,
+      "loss": 0.9138,
+      "step": 29825
+    },
+    {
+      "epoch": 2.0267699415681477,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.733268786519908e-05,
+      "loss": 0.9815,
+      "step": 29830
+    },
+    {
+      "epoch": 2.02710966163881,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.733056461475744e-05,
+      "loss": 0.8649,
+      "step": 29835
+    },
+    {
+      "epoch": 2.0274493817094714,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.732844136431581e-05,
+      "loss": 1.0128,
+      "step": 29840
+    },
+    {
+      "epoch": 2.027789101780133,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.732631811387417e-05,
+      "loss": 0.8784,
+      "step": 29845
+    },
+    {
+      "epoch": 2.028128821850795,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.732419486343253e-05,
+      "loss": 0.9872,
+      "step": 29850
+    },
+    {
+      "epoch": 2.0284685419214568,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.73220716129909e-05,
+      "loss": 1.0246,
+      "step": 29855
+    },
+    {
+      "epoch": 2.0288082619921184,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7319948362549264e-05,
+      "loss": 0.9694,
+      "step": 29860
+    },
+    {
+      "epoch": 2.0291479820627805,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.731782511210762e-05,
+      "loss": 0.9075,
+      "step": 29865
+    },
+    {
+      "epoch": 2.029487702133442,
+      "grad_norm": 5.75,
+      "learning_rate": 3.731570186166599e-05,
+      "loss": 0.9634,
+      "step": 29870
+    },
+    {
+      "epoch": 2.0298274222041037,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.7313578611224356e-05,
+      "loss": 0.9034,
+      "step": 29875
+    },
+    {
+      "epoch": 2.030167142274766,
+      "grad_norm": 7.75,
+      "learning_rate": 3.7311455360782713e-05,
+      "loss": 0.9749,
+      "step": 29880
+    },
+    {
+      "epoch": 2.0305068623454274,
+      "grad_norm": 6.0,
+      "learning_rate": 3.7309332110341084e-05,
+      "loss": 0.9935,
+      "step": 29885
+    },
+    {
+      "epoch": 2.030846582416089,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.730720885989945e-05,
+      "loss": 0.9177,
+      "step": 29890
+    },
+    {
+      "epoch": 2.031186302486751,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.7305085609457805e-05,
+      "loss": 0.9849,
+      "step": 29895
+    },
+    {
+      "epoch": 2.0315260225574128,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7302962359016176e-05,
+      "loss": 0.8638,
+      "step": 29900
+    },
+    {
+      "epoch": 2.0318657426280744,
+      "grad_norm": 7.125,
+      "learning_rate": 3.7300839108574534e-05,
+      "loss": 0.9269,
+      "step": 29905
+    },
+    {
+      "epoch": 2.032205462698736,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.72987158581329e-05,
+      "loss": 1.0243,
+      "step": 29910
+    },
+    {
+      "epoch": 2.032545182769398,
+      "grad_norm": 6.125,
+      "learning_rate": 3.729659260769127e-05,
+      "loss": 0.8868,
+      "step": 29915
+    },
+    {
+      "epoch": 2.0328849028400597,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.7294469357249626e-05,
+      "loss": 0.9016,
+      "step": 29920
+    },
+    {
+      "epoch": 2.0332246229107214,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.729234610680799e-05,
+      "loss": 0.8533,
+      "step": 29925
+    },
+    {
+      "epoch": 2.0335643429813834,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.729022285636636e-05,
+      "loss": 1.0204,
+      "step": 29930
+    },
+    {
+      "epoch": 2.033904063052045,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.728809960592472e-05,
+      "loss": 1.0339,
+      "step": 29935
+    },
+    {
+      "epoch": 2.0342437831227067,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.728597635548308e-05,
+      "loss": 0.9971,
+      "step": 29940
+    },
+    {
+      "epoch": 2.034583503193369,
+      "grad_norm": 6.75,
+      "learning_rate": 3.728385310504145e-05,
+      "loss": 0.9925,
+      "step": 29945
+    },
+    {
+      "epoch": 2.0349232232640304,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.728172985459981e-05,
+      "loss": 1.0049,
+      "step": 29950
+    },
+    {
+      "epoch": 2.035262943334692,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.7279606604158174e-05,
+      "loss": 1.005,
+      "step": 29955
+    },
+    {
+      "epoch": 2.035602663405354,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.7277483353716544e-05,
+      "loss": 0.9918,
+      "step": 29960
+    },
+    {
+      "epoch": 2.0359423834760157,
+      "grad_norm": 6.375,
+      "learning_rate": 3.72753601032749e-05,
+      "loss": 1.0277,
+      "step": 29965
+    },
+    {
+      "epoch": 2.0362821035466774,
+      "grad_norm": 6.5,
+      "learning_rate": 3.7273236852833266e-05,
+      "loss": 0.912,
+      "step": 29970
+    },
+    {
+      "epoch": 2.0366218236173395,
+      "grad_norm": 6.125,
+      "learning_rate": 3.7271113602391636e-05,
+      "loss": 1.03,
+      "step": 29975
+    },
+    {
+      "epoch": 2.036961543688001,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.7268990351949994e-05,
+      "loss": 0.9417,
+      "step": 29980
+    },
+    {
+      "epoch": 2.0373012637586627,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.726686710150836e-05,
+      "loss": 0.928,
+      "step": 29985
+    },
+    {
+      "epoch": 2.037640983829325,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.726474385106672e-05,
+      "loss": 0.9734,
+      "step": 29990
+    },
+    {
+      "epoch": 2.0379807038999864,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.7262620600625086e-05,
+      "loss": 1.0392,
+      "step": 29995
+    },
+    {
+      "epoch": 2.038320423970648,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.726049735018345e-05,
+      "loss": 1.0381,
+      "step": 30000
+    },
+    {
+      "epoch": 2.03866014404131,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.7258374099741814e-05,
+      "loss": 0.9145,
+      "step": 30005
+    },
+    {
+      "epoch": 2.0389998641119718,
+      "grad_norm": 6.75,
+      "learning_rate": 3.725625084930018e-05,
+      "loss": 1.0287,
+      "step": 30010
+    },
+    {
+      "epoch": 2.0393395841826334,
+      "grad_norm": 5.75,
+      "learning_rate": 3.725412759885854e-05,
+      "loss": 0.9706,
+      "step": 30015
+    },
+    {
+      "epoch": 2.0396793042532955,
+      "grad_norm": 7.0,
+      "learning_rate": 3.7252004348416906e-05,
+      "loss": 1.0394,
+      "step": 30020
+    },
+    {
+      "epoch": 2.040019024323957,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.724988109797527e-05,
+      "loss": 0.945,
+      "step": 30025
+    },
+    {
+      "epoch": 2.0403587443946187,
+      "grad_norm": 6.625,
+      "learning_rate": 3.7247757847533634e-05,
+      "loss": 0.9162,
+      "step": 30030
+    },
+    {
+      "epoch": 2.040698464465281,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.7245634597092e-05,
+      "loss": 0.8787,
+      "step": 30035
+    },
+    {
+      "epoch": 2.0410381845359424,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.724351134665036e-05,
+      "loss": 0.9168,
+      "step": 30040
+    },
+    {
+      "epoch": 2.041377904606604,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7241388096208726e-05,
+      "loss": 0.9699,
+      "step": 30045
+    },
+    {
+      "epoch": 2.041717624677266,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.723926484576709e-05,
+      "loss": 1.0472,
+      "step": 30050
+    },
+    {
+      "epoch": 2.0420573447479278,
+      "grad_norm": 6.25,
+      "learning_rate": 3.7237141595325454e-05,
+      "loss": 0.8951,
+      "step": 30055
+    },
+    {
+      "epoch": 2.0423970648185894,
+      "grad_norm": 7.5,
+      "learning_rate": 3.723501834488382e-05,
+      "loss": 0.9448,
+      "step": 30060
+    },
+    {
+      "epoch": 2.042736784889251,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.723289509444218e-05,
+      "loss": 0.939,
+      "step": 30065
+    },
+    {
+      "epoch": 2.043076504959913,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.7230771844000546e-05,
+      "loss": 0.9302,
+      "step": 30070
+    },
+    {
+      "epoch": 2.0434162250305747,
+      "grad_norm": 6.75,
+      "learning_rate": 3.722864859355891e-05,
+      "loss": 0.9758,
+      "step": 30075
+    },
+    {
+      "epoch": 2.0437559451012364,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.7226525343117274e-05,
+      "loss": 0.9667,
+      "step": 30080
+    },
+    {
+      "epoch": 2.0440956651718984,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.722440209267564e-05,
+      "loss": 0.9471,
+      "step": 30085
+    },
+    {
+      "epoch": 2.04443538524256,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.7222278842234e-05,
+      "loss": 0.9803,
+      "step": 30090
+    },
+    {
+      "epoch": 2.0447751053132217,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7220155591792366e-05,
+      "loss": 0.9154,
+      "step": 30095
+    },
+    {
+      "epoch": 2.045114825383884,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.721803234135073e-05,
+      "loss": 0.9837,
+      "step": 30100
+    },
+    {
+      "epoch": 2.0454545454545454,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.721590909090909e-05,
+      "loss": 0.9733,
+      "step": 30105
+    },
+    {
+      "epoch": 2.045794265525207,
+      "grad_norm": 6.125,
+      "learning_rate": 3.721378584046746e-05,
+      "loss": 0.8899,
+      "step": 30110
+    },
+    {
+      "epoch": 2.046133985595869,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.721166259002582e-05,
+      "loss": 1.0458,
+      "step": 30115
+    },
+    {
+      "epoch": 2.0464737056665308,
+      "grad_norm": 6.25,
+      "learning_rate": 3.720953933958418e-05,
+      "loss": 0.9921,
+      "step": 30120
+    },
+    {
+      "epoch": 2.0468134257371924,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.720741608914255e-05,
+      "loss": 0.8688,
+      "step": 30125
+    },
+    {
+      "epoch": 2.0471531458078545,
+      "grad_norm": 6.0,
+      "learning_rate": 3.7205292838700914e-05,
+      "loss": 0.9223,
+      "step": 30130
+    },
+    {
+      "epoch": 2.047492865878516,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.720316958825927e-05,
+      "loss": 0.9804,
+      "step": 30135
+    },
+    {
+      "epoch": 2.0478325859491777,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.720104633781764e-05,
+      "loss": 0.9957,
+      "step": 30140
+    },
+    {
+      "epoch": 2.04817230601984,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7198923087376006e-05,
+      "loss": 1.0852,
+      "step": 30145
+    },
+    {
+      "epoch": 2.0485120260905014,
+      "grad_norm": 6.125,
+      "learning_rate": 3.719679983693436e-05,
+      "loss": 0.9003,
+      "step": 30150
+    },
+    {
+      "epoch": 2.048851746161163,
+      "grad_norm": 7.125,
+      "learning_rate": 3.7194676586492734e-05,
+      "loss": 0.9663,
+      "step": 30155
+    },
+    {
+      "epoch": 2.049191466231825,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.71925533360511e-05,
+      "loss": 0.9613,
+      "step": 30160
+    },
+    {
+      "epoch": 2.0495311863024868,
+      "grad_norm": 6.5,
+      "learning_rate": 3.719043008560946e-05,
+      "loss": 0.9752,
+      "step": 30165
+    },
+    {
+      "epoch": 2.0498709063731484,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.7188306835167826e-05,
+      "loss": 1.0045,
+      "step": 30170
+    },
+    {
+      "epoch": 2.0502106264438105,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.718618358472619e-05,
+      "loss": 1.0198,
+      "step": 30175
+    },
+    {
+      "epoch": 2.050550346514472,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.7184060334284554e-05,
+      "loss": 1.0689,
+      "step": 30180
+    },
+    {
+      "epoch": 2.0508900665851337,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.718193708384292e-05,
+      "loss": 0.9797,
+      "step": 30185
+    },
+    {
+      "epoch": 2.051229786655796,
+      "grad_norm": 6.625,
+      "learning_rate": 3.7179813833401275e-05,
+      "loss": 0.8996,
+      "step": 30190
+    },
+    {
+      "epoch": 2.0515695067264574,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.7177690582959646e-05,
+      "loss": 0.9527,
+      "step": 30195
+    },
+    {
+      "epoch": 2.051909226797119,
+      "grad_norm": 6.625,
+      "learning_rate": 3.717556733251801e-05,
+      "loss": 0.9739,
+      "step": 30200
+    },
+    {
+      "epoch": 2.052248946867781,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.717344408207637e-05,
+      "loss": 0.9693,
+      "step": 30205
+    },
+    {
+      "epoch": 2.0525886669384428,
+      "grad_norm": 5.34375,
+      "learning_rate": 3.717132083163474e-05,
+      "loss": 0.9587,
+      "step": 30210
+    },
+    {
+      "epoch": 2.0529283870091044,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.71691975811931e-05,
+      "loss": 0.9544,
+      "step": 30215
+    },
+    {
+      "epoch": 2.0532681070797665,
+      "grad_norm": 6.875,
+      "learning_rate": 3.716707433075146e-05,
+      "loss": 1.0315,
+      "step": 30220
+    },
+    {
+      "epoch": 2.053607827150428,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.716495108030983e-05,
+      "loss": 0.945,
+      "step": 30225
+    },
+    {
+      "epoch": 2.0539475472210897,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.7162827829868194e-05,
+      "loss": 0.9798,
+      "step": 30230
+    },
+    {
+      "epoch": 2.054287267291752,
+      "grad_norm": 6.5,
+      "learning_rate": 3.716070457942655e-05,
+      "loss": 1.0016,
+      "step": 30235
+    },
+    {
+      "epoch": 2.0546269873624134,
+      "grad_norm": 7.125,
+      "learning_rate": 3.715858132898492e-05,
+      "loss": 1.0405,
+      "step": 30240
+    },
+    {
+      "epoch": 2.054966707433075,
+      "grad_norm": 6.375,
+      "learning_rate": 3.7156458078543286e-05,
+      "loss": 0.9602,
+      "step": 30245
+    },
+    {
+      "epoch": 2.0553064275037367,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.715433482810164e-05,
+      "loss": 0.9747,
+      "step": 30250
+    },
+    {
+      "epoch": 2.055646147574399,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.7152211577660014e-05,
+      "loss": 1.0112,
+      "step": 30255
+    },
+    {
+      "epoch": 2.0559858676450604,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.715008832721837e-05,
+      "loss": 0.9381,
+      "step": 30260
+    },
+    {
+      "epoch": 2.056325587715722,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.7147965076776735e-05,
+      "loss": 0.9183,
+      "step": 30265
+    },
+    {
+      "epoch": 2.056665307786384,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.7145841826335106e-05,
+      "loss": 0.8893,
+      "step": 30270
+    },
+    {
+      "epoch": 2.0570050278570458,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.714371857589346e-05,
+      "loss": 0.9896,
+      "step": 30275
+    },
+    {
+      "epoch": 2.0573447479277074,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.714159532545183e-05,
+      "loss": 0.8895,
+      "step": 30280
+    },
+    {
+      "epoch": 2.0576844679983695,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.71394720750102e-05,
+      "loss": 0.9047,
+      "step": 30285
+    },
+    {
+      "epoch": 2.058024188069031,
+      "grad_norm": 6.0,
+      "learning_rate": 3.7137348824568555e-05,
+      "loss": 0.9712,
+      "step": 30290
+    },
+    {
+      "epoch": 2.0583639081396927,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.713522557412692e-05,
+      "loss": 0.9734,
+      "step": 30295
+    },
+    {
+      "epoch": 2.058703628210355,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.713310232368529e-05,
+      "loss": 1.0435,
+      "step": 30300
+    },
+    {
+      "epoch": 2.0590433482810164,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.713097907324365e-05,
+      "loss": 0.9884,
+      "step": 30305
+    },
+    {
+      "epoch": 2.059383068351678,
+      "grad_norm": 7.25,
+      "learning_rate": 3.712885582280201e-05,
+      "loss": 0.9201,
+      "step": 30310
+    },
+    {
+      "epoch": 2.05972278842234,
+      "grad_norm": 6.75,
+      "learning_rate": 3.712673257236038e-05,
+      "loss": 1.0395,
+      "step": 30315
+    },
+    {
+      "epoch": 2.0600625084930018,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.712460932191874e-05,
+      "loss": 1.0133,
+      "step": 30320
+    },
+    {
+      "epoch": 2.0604022285636634,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.71224860714771e-05,
+      "loss": 0.8873,
+      "step": 30325
+    },
+    {
+      "epoch": 2.0607419486343255,
+      "grad_norm": 6.0,
+      "learning_rate": 3.712036282103547e-05,
+      "loss": 0.9285,
+      "step": 30330
+    },
+    {
+      "epoch": 2.061081668704987,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.711823957059383e-05,
+      "loss": 0.8956,
+      "step": 30335
+    },
+    {
+      "epoch": 2.0614213887756487,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.7116116320152195e-05,
+      "loss": 0.9993,
+      "step": 30340
+    },
+    {
+      "epoch": 2.061761108846311,
+      "grad_norm": 6.25,
+      "learning_rate": 3.711399306971056e-05,
+      "loss": 0.9742,
+      "step": 30345
+    },
+    {
+      "epoch": 2.0621008289169724,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.711186981926892e-05,
+      "loss": 1.0474,
+      "step": 30350
+    },
+    {
+      "epoch": 2.062440548987634,
+      "grad_norm": 6.875,
+      "learning_rate": 3.710974656882729e-05,
+      "loss": 0.9853,
+      "step": 30355
+    },
+    {
+      "epoch": 2.062780269058296,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.710762331838565e-05,
+      "loss": 0.9866,
+      "step": 30360
+    },
+    {
+      "epoch": 2.0631199891289578,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.7105500067944015e-05,
+      "loss": 0.9837,
+      "step": 30365
+    },
+    {
+      "epoch": 2.0634597091996194,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.710337681750238e-05,
+      "loss": 0.9811,
+      "step": 30370
+    },
+    {
+      "epoch": 2.0637994292702815,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.710125356706074e-05,
+      "loss": 0.9158,
+      "step": 30375
+    },
+    {
+      "epoch": 2.064139149340943,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.709913031661911e-05,
+      "loss": 0.9872,
+      "step": 30380
+    },
+    {
+      "epoch": 2.0644788694116047,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.709700706617747e-05,
+      "loss": 0.9759,
+      "step": 30385
+    },
+    {
+      "epoch": 2.064818589482267,
+      "grad_norm": 6.5,
+      "learning_rate": 3.7094883815735835e-05,
+      "loss": 0.9993,
+      "step": 30390
+    },
+    {
+      "epoch": 2.0651583095529285,
+      "grad_norm": 5.875,
+      "learning_rate": 3.70927605652942e-05,
+      "loss": 0.9031,
+      "step": 30395
+    },
+    {
+      "epoch": 2.06549802962359,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.709063731485256e-05,
+      "loss": 0.9682,
+      "step": 30400
+    },
+    {
+      "epoch": 2.0658377496942517,
+      "grad_norm": 5.125,
+      "learning_rate": 3.708851406441093e-05,
+      "loss": 0.8509,
+      "step": 30405
+    },
+    {
+      "epoch": 2.066177469764914,
+      "grad_norm": 7.125,
+      "learning_rate": 3.708639081396929e-05,
+      "loss": 1.017,
+      "step": 30410
+    },
+    {
+      "epoch": 2.0665171898355754,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.7084267563527655e-05,
+      "loss": 0.9149,
+      "step": 30415
+    },
+    {
+      "epoch": 2.066856909906237,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.708214431308602e-05,
+      "loss": 0.9242,
+      "step": 30420
+    },
+    {
+      "epoch": 2.067196629976899,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.708002106264438e-05,
+      "loss": 1.0294,
+      "step": 30425
+    },
+    {
+      "epoch": 2.0675363500475608,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.707789781220275e-05,
+      "loss": 0.9246,
+      "step": 30430
+    },
+    {
+      "epoch": 2.0678760701182224,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.707577456176111e-05,
+      "loss": 0.9505,
+      "step": 30435
+    },
+    {
+      "epoch": 2.0682157901888845,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.7073651311319475e-05,
+      "loss": 0.9894,
+      "step": 30440
+    },
+    {
+      "epoch": 2.068555510259546,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.707152806087784e-05,
+      "loss": 0.9017,
+      "step": 30445
+    },
+    {
+      "epoch": 2.0688952303302077,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.70694048104362e-05,
+      "loss": 0.998,
+      "step": 30450
+    },
+    {
+      "epoch": 2.06923495040087,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.706728155999457e-05,
+      "loss": 0.8635,
+      "step": 30455
+    },
+    {
+      "epoch": 2.0695746704715314,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.7065158309552924e-05,
+      "loss": 0.9852,
+      "step": 30460
+    },
+    {
+      "epoch": 2.069914390542193,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.7063035059111295e-05,
+      "loss": 0.8783,
+      "step": 30465
+    },
+    {
+      "epoch": 2.070254110612855,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.706091180866966e-05,
+      "loss": 0.9797,
+      "step": 30470
+    },
+    {
+      "epoch": 2.0705938306835168,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.7058788558228016e-05,
+      "loss": 0.9363,
+      "step": 30475
+    },
+    {
+      "epoch": 2.0709335507541784,
+      "grad_norm": 7.0,
+      "learning_rate": 3.705666530778639e-05,
+      "loss": 0.9532,
+      "step": 30480
+    },
+    {
+      "epoch": 2.0712732708248405,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.705454205734475e-05,
+      "loss": 0.9291,
+      "step": 30485
+    },
+    {
+      "epoch": 2.071612990895502,
+      "grad_norm": 6.75,
+      "learning_rate": 3.705241880690311e-05,
+      "loss": 0.8813,
+      "step": 30490
+    },
+    {
+      "epoch": 2.0719527109661637,
+      "grad_norm": 6.375,
+      "learning_rate": 3.705029555646148e-05,
+      "loss": 0.9141,
+      "step": 30495
+    },
+    {
+      "epoch": 2.072292431036826,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.704817230601984e-05,
+      "loss": 0.9516,
+      "step": 30500
+    },
+    {
+      "epoch": 2.0726321511074874,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.704604905557821e-05,
+      "loss": 0.9165,
+      "step": 30505
+    },
+    {
+      "epoch": 2.072971871178149,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.704392580513657e-05,
+      "loss": 0.9718,
+      "step": 30510
+    },
+    {
+      "epoch": 2.073311591248811,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.7041802554694935e-05,
+      "loss": 0.9242,
+      "step": 30515
+    },
+    {
+      "epoch": 2.073651311319473,
+      "grad_norm": 7.125,
+      "learning_rate": 3.70396793042533e-05,
+      "loss": 0.968,
+      "step": 30520
+    },
+    {
+      "epoch": 2.0739910313901344,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.703755605381166e-05,
+      "loss": 0.9661,
+      "step": 30525
+    },
+    {
+      "epoch": 2.0743307514607965,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.703543280337002e-05,
+      "loss": 0.9497,
+      "step": 30530
+    },
+    {
+      "epoch": 2.074670471531458,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.703330955292839e-05,
+      "loss": 0.9377,
+      "step": 30535
+    },
+    {
+      "epoch": 2.0750101916021197,
+      "grad_norm": 5.875,
+      "learning_rate": 3.7031186302486755e-05,
+      "loss": 0.9417,
+      "step": 30540
+    },
+    {
+      "epoch": 2.075349911672782,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.702906305204511e-05,
+      "loss": 1.074,
+      "step": 30545
+    },
+    {
+      "epoch": 2.0756896317434435,
+      "grad_norm": 6.125,
+      "learning_rate": 3.702693980160348e-05,
+      "loss": 0.9446,
+      "step": 30550
+    },
+    {
+      "epoch": 2.076029351814105,
+      "grad_norm": 6.5,
+      "learning_rate": 3.702481655116185e-05,
+      "loss": 1.0039,
+      "step": 30555
+    },
+    {
+      "epoch": 2.076369071884767,
+      "grad_norm": 6.125,
+      "learning_rate": 3.7022693300720204e-05,
+      "loss": 0.9446,
+      "step": 30560
+    },
+    {
+      "epoch": 2.076708791955429,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.7020570050278575e-05,
+      "loss": 0.9642,
+      "step": 30565
+    },
+    {
+      "epoch": 2.0770485120260904,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.701844679983694e-05,
+      "loss": 0.9221,
+      "step": 30570
+    },
+    {
+      "epoch": 2.0773882320967525,
+      "grad_norm": 7.25,
+      "learning_rate": 3.7016323549395296e-05,
+      "loss": 0.9231,
+      "step": 30575
+    },
+    {
+      "epoch": 2.077727952167414,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.701420029895367e-05,
+      "loss": 0.9008,
+      "step": 30580
+    },
+    {
+      "epoch": 2.0780676722380758,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.701207704851203e-05,
+      "loss": 0.9767,
+      "step": 30585
+    },
+    {
+      "epoch": 2.0784073923087374,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.700995379807039e-05,
+      "loss": 0.9733,
+      "step": 30590
+    },
+    {
+      "epoch": 2.0787471123793995,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.700783054762876e-05,
+      "loss": 0.8523,
+      "step": 30595
+    },
+    {
+      "epoch": 2.079086832450061,
+      "grad_norm": 6.25,
+      "learning_rate": 3.700570729718712e-05,
+      "loss": 1.0234,
+      "step": 30600
+    },
+    {
+      "epoch": 2.0794265525207227,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.700358404674548e-05,
+      "loss": 1.0046,
+      "step": 30605
+    },
+    {
+      "epoch": 2.079766272591385,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.700146079630385e-05,
+      "loss": 0.967,
+      "step": 30610
+    },
+    {
+      "epoch": 2.0801059926620464,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.699933754586221e-05,
+      "loss": 0.9957,
+      "step": 30615
+    },
+    {
+      "epoch": 2.080445712732708,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.699721429542057e-05,
+      "loss": 0.8845,
+      "step": 30620
+    },
+    {
+      "epoch": 2.08078543280337,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.699509104497894e-05,
+      "loss": 0.9552,
+      "step": 30625
+    },
+    {
+      "epoch": 2.0811251528740318,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.69929677945373e-05,
+      "loss": 1.0184,
+      "step": 30630
+    },
+    {
+      "epoch": 2.0814648729446934,
+      "grad_norm": 5.875,
+      "learning_rate": 3.6990844544095665e-05,
+      "loss": 0.9379,
+      "step": 30635
+    },
+    {
+      "epoch": 2.0818045930153555,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.6988721293654035e-05,
+      "loss": 0.9792,
+      "step": 30640
+    },
+    {
+      "epoch": 2.082144313086017,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.698659804321239e-05,
+      "loss": 0.9836,
+      "step": 30645
+    },
+    {
+      "epoch": 2.0824840331566787,
+      "grad_norm": 6.125,
+      "learning_rate": 3.6984474792770757e-05,
+      "loss": 1.0596,
+      "step": 30650
+    },
+    {
+      "epoch": 2.082823753227341,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.698235154232913e-05,
+      "loss": 1.0256,
+      "step": 30655
+    },
+    {
+      "epoch": 2.0831634732980024,
+      "grad_norm": 6.0,
+      "learning_rate": 3.6980228291887485e-05,
+      "loss": 0.9464,
+      "step": 30660
+    },
+    {
+      "epoch": 2.083503193368664,
+      "grad_norm": 6.5,
+      "learning_rate": 3.697810504144585e-05,
+      "loss": 0.951,
+      "step": 30665
+    },
+    {
+      "epoch": 2.083842913439326,
+      "grad_norm": 6.625,
+      "learning_rate": 3.697598179100422e-05,
+      "loss": 0.922,
+      "step": 30670
+    },
+    {
+      "epoch": 2.084182633509988,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.6973858540562577e-05,
+      "loss": 0.9198,
+      "step": 30675
+    },
+    {
+      "epoch": 2.0845223535806494,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.697173529012094e-05,
+      "loss": 0.9917,
+      "step": 30680
+    },
+    {
+      "epoch": 2.0848620736513115,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6969612039679305e-05,
+      "loss": 1.0085,
+      "step": 30685
+    },
+    {
+      "epoch": 2.085201793721973,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.696748878923767e-05,
+      "loss": 0.9461,
+      "step": 30690
+    },
+    {
+      "epoch": 2.0855415137926347,
+      "grad_norm": 6.75,
+      "learning_rate": 3.696536553879603e-05,
+      "loss": 0.9368,
+      "step": 30695
+    },
+    {
+      "epoch": 2.085881233863297,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.6963242288354397e-05,
+      "loss": 0.9413,
+      "step": 30700
+    },
+    {
+      "epoch": 2.0862209539339585,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.696111903791276e-05,
+      "loss": 0.9962,
+      "step": 30705
+    },
+    {
+      "epoch": 2.08656067400462,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.6958995787471125e-05,
+      "loss": 1.0062,
+      "step": 30710
+    },
+    {
+      "epoch": 2.086900394075282,
+      "grad_norm": 6.125,
+      "learning_rate": 3.695687253702949e-05,
+      "loss": 0.8722,
+      "step": 30715
+    },
+    {
+      "epoch": 2.087240114145944,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.695474928658785e-05,
+      "loss": 0.8839,
+      "step": 30720
+    },
+    {
+      "epoch": 2.0875798342166054,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.6952626036146217e-05,
+      "loss": 0.9583,
+      "step": 30725
+    },
+    {
+      "epoch": 2.0879195542872675,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.695050278570458e-05,
+      "loss": 0.8959,
+      "step": 30730
+    },
+    {
+      "epoch": 2.088259274357929,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.6948379535262945e-05,
+      "loss": 0.8715,
+      "step": 30735
+    },
+    {
+      "epoch": 2.0885989944285908,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.694625628482131e-05,
+      "loss": 0.9032,
+      "step": 30740
+    },
+    {
+      "epoch": 2.0889387144992524,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.694413303437967e-05,
+      "loss": 0.8908,
+      "step": 30745
+    },
+    {
+      "epoch": 2.0892784345699145,
+      "grad_norm": 6.5,
+      "learning_rate": 3.6942009783938037e-05,
+      "loss": 0.91,
+      "step": 30750
+    },
+    {
+      "epoch": 2.089618154640576,
+      "grad_norm": 5.875,
+      "learning_rate": 3.69398865334964e-05,
+      "loss": 0.9331,
+      "step": 30755
+    },
+    {
+      "epoch": 2.0899578747112377,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.6937763283054765e-05,
+      "loss": 0.9212,
+      "step": 30760
+    },
+    {
+      "epoch": 2.0902975947819,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.693564003261313e-05,
+      "loss": 1.038,
+      "step": 30765
+    },
+    {
+      "epoch": 2.0906373148525614,
+      "grad_norm": 7.125,
+      "learning_rate": 3.693351678217149e-05,
+      "loss": 0.9467,
+      "step": 30770
+    },
+    {
+      "epoch": 2.090977034923223,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.693139353172986e-05,
+      "loss": 0.8688,
+      "step": 30775
+    },
+    {
+      "epoch": 2.091316754993885,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.692927028128822e-05,
+      "loss": 1.0028,
+      "step": 30780
+    },
+    {
+      "epoch": 2.0916564750645468,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.6927147030846585e-05,
+      "loss": 0.9824,
+      "step": 30785
+    },
+    {
+      "epoch": 2.0919961951352084,
+      "grad_norm": 6.25,
+      "learning_rate": 3.692502378040495e-05,
+      "loss": 0.9918,
+      "step": 30790
+    },
+    {
+      "epoch": 2.0923359152058705,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.692290052996331e-05,
+      "loss": 0.8661,
+      "step": 30795
+    },
+    {
+      "epoch": 2.092675635276532,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.692077727952168e-05,
+      "loss": 1.0151,
+      "step": 30800
+    },
+    {
+      "epoch": 2.0930153553471937,
+      "grad_norm": 6.625,
+      "learning_rate": 3.691865402908004e-05,
+      "loss": 1.0407,
+      "step": 30805
+    },
+    {
+      "epoch": 2.093355075417856,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.6916530778638405e-05,
+      "loss": 0.9771,
+      "step": 30810
+    },
+    {
+      "epoch": 2.0936947954885174,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.691440752819676e-05,
+      "loss": 1.0623,
+      "step": 30815
+    },
+    {
+      "epoch": 2.094034515559179,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.691228427775513e-05,
+      "loss": 1.0399,
+      "step": 30820
+    },
+    {
+      "epoch": 2.094374235629841,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.69101610273135e-05,
+      "loss": 0.9391,
+      "step": 30825
+    },
+    {
+      "epoch": 2.094713955700503,
+      "grad_norm": 6.0,
+      "learning_rate": 3.6908037776871854e-05,
+      "loss": 0.9754,
+      "step": 30830
+    },
+    {
+      "epoch": 2.0950536757711644,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.6905914526430225e-05,
+      "loss": 0.8564,
+      "step": 30835
+    },
+    {
+      "epoch": 2.0953933958418265,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.690379127598859e-05,
+      "loss": 0.9403,
+      "step": 30840
+    },
+    {
+      "epoch": 2.095733115912488,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.690166802554695e-05,
+      "loss": 1.0783,
+      "step": 30845
+    },
+    {
+      "epoch": 2.0960728359831498,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.689954477510532e-05,
+      "loss": 1.1814,
+      "step": 30850
+    },
+    {
+      "epoch": 2.096412556053812,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.689742152466368e-05,
+      "loss": 0.9929,
+      "step": 30855
+    },
+    {
+      "epoch": 2.0967522761244735,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.6895298274222045e-05,
+      "loss": 1.0386,
+      "step": 30860
+    },
+    {
+      "epoch": 2.097091996195135,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.689317502378041e-05,
+      "loss": 0.8904,
+      "step": 30865
+    },
+    {
+      "epoch": 2.097431716265797,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.689105177333877e-05,
+      "loss": 1.0415,
+      "step": 30870
+    },
+    {
+      "epoch": 2.097771436336459,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.688892852289714e-05,
+      "loss": 0.9162,
+      "step": 30875
+    },
+    {
+      "epoch": 2.0981111564071204,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.68868052724555e-05,
+      "loss": 0.9056,
+      "step": 30880
+    },
+    {
+      "epoch": 2.0984508764777825,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.688468202201386e-05,
+      "loss": 0.928,
+      "step": 30885
+    },
+    {
+      "epoch": 2.098790596548444,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.688255877157223e-05,
+      "loss": 0.9372,
+      "step": 30890
+    },
+    {
+      "epoch": 2.0991303166191058,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.688043552113059e-05,
+      "loss": 0.9109,
+      "step": 30895
+    },
+    {
+      "epoch": 2.099470036689768,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.687831227068895e-05,
+      "loss": 0.9074,
+      "step": 30900
+    },
+    {
+      "epoch": 2.0998097567604295,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.687618902024732e-05,
+      "loss": 0.9977,
+      "step": 30905
+    },
+    {
+      "epoch": 2.100149476831091,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.6874065769805685e-05,
+      "loss": 0.9191,
+      "step": 30910
+    },
+    {
+      "epoch": 2.100489196901753,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.687194251936404e-05,
+      "loss": 0.974,
+      "step": 30915
+    },
+    {
+      "epoch": 2.100828916972415,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.686981926892241e-05,
+      "loss": 0.9222,
+      "step": 30920
+    },
+    {
+      "epoch": 2.1011686370430764,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.686769601848078e-05,
+      "loss": 1.0037,
+      "step": 30925
+    },
+    {
+      "epoch": 2.101508357113738,
+      "grad_norm": 7.125,
+      "learning_rate": 3.6865572768039134e-05,
+      "loss": 1.0041,
+      "step": 30930
+    },
+    {
+      "epoch": 2.1018480771844,
+      "grad_norm": 7.125,
+      "learning_rate": 3.6863449517597505e-05,
+      "loss": 1.0923,
+      "step": 30935
+    },
+    {
+      "epoch": 2.1021877972550618,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.686132626715587e-05,
+      "loss": 0.9384,
+      "step": 30940
+    },
+    {
+      "epoch": 2.1025275173257234,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.6859203016714226e-05,
+      "loss": 0.9191,
+      "step": 30945
+    },
+    {
+      "epoch": 2.1028672373963855,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.68570797662726e-05,
+      "loss": 0.9597,
+      "step": 30950
+    },
+    {
+      "epoch": 2.103206957467047,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.6854956515830954e-05,
+      "loss": 0.9234,
+      "step": 30955
+    },
+    {
+      "epoch": 2.1035466775377087,
+      "grad_norm": 6.625,
+      "learning_rate": 3.685283326538932e-05,
+      "loss": 0.9507,
+      "step": 30960
+    },
+    {
+      "epoch": 2.103886397608371,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.685071001494769e-05,
+      "loss": 1.0434,
+      "step": 30965
+    },
+    {
+      "epoch": 2.1042261176790324,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.6848586764506046e-05,
+      "loss": 0.8654,
+      "step": 30970
+    },
+    {
+      "epoch": 2.104565837749694,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.684646351406441e-05,
+      "loss": 0.9364,
+      "step": 30975
+    },
+    {
+      "epoch": 2.104905557820356,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.684434026362278e-05,
+      "loss": 0.9413,
+      "step": 30980
+    },
+    {
+      "epoch": 2.105245277891018,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.684221701318114e-05,
+      "loss": 0.951,
+      "step": 30985
+    },
+    {
+      "epoch": 2.1055849979616794,
+      "grad_norm": 4.625,
+      "learning_rate": 3.68400937627395e-05,
+      "loss": 0.9642,
+      "step": 30990
+    },
+    {
+      "epoch": 2.1059247180323415,
+      "grad_norm": 6.5,
+      "learning_rate": 3.683797051229787e-05,
+      "loss": 1.0939,
+      "step": 30995
+    },
+    {
+      "epoch": 2.106264438103003,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.683584726185623e-05,
+      "loss": 0.9511,
+      "step": 31000
+    },
+    {
+      "epoch": 2.1066041581736648,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.6833724011414594e-05,
+      "loss": 1.0463,
+      "step": 31005
+    },
+    {
+      "epoch": 2.106943878244327,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.6831600760972965e-05,
+      "loss": 0.9623,
+      "step": 31010
+    },
+    {
+      "epoch": 2.1072835983149885,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.682947751053132e-05,
+      "loss": 0.9823,
+      "step": 31015
+    },
+    {
+      "epoch": 2.10762331838565,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.6827354260089686e-05,
+      "loss": 0.9995,
+      "step": 31020
+    },
+    {
+      "epoch": 2.107963038456312,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.682523100964806e-05,
+      "loss": 0.914,
+      "step": 31025
+    },
+    {
+      "epoch": 2.108302758526974,
+      "grad_norm": 6.5,
+      "learning_rate": 3.6823107759206414e-05,
+      "loss": 1.022,
+      "step": 31030
+    },
+    {
+      "epoch": 2.1086424785976354,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.682098450876478e-05,
+      "loss": 0.947,
+      "step": 31035
+    },
+    {
+      "epoch": 2.1089821986682975,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.681886125832314e-05,
+      "loss": 0.9653,
+      "step": 31040
+    },
+    {
+      "epoch": 2.109321918738959,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.6816738007881506e-05,
+      "loss": 0.9466,
+      "step": 31045
+    },
+    {
+      "epoch": 2.1096616388096208,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.681461475743987e-05,
+      "loss": 0.9832,
+      "step": 31050
+    },
+    {
+      "epoch": 2.110001358880283,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.6812491506998234e-05,
+      "loss": 0.9747,
+      "step": 31055
+    },
+    {
+      "epoch": 2.1103410789509445,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.68103682565566e-05,
+      "loss": 1.0098,
+      "step": 31060
+    },
+    {
+      "epoch": 2.110680799021606,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.680824500611496e-05,
+      "loss": 1.0436,
+      "step": 31065
+    },
+    {
+      "epoch": 2.111020519092268,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.6806121755673326e-05,
+      "loss": 0.914,
+      "step": 31070
+    },
+    {
+      "epoch": 2.11136023916293,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.680399850523169e-05,
+      "loss": 1.1059,
+      "step": 31075
+    },
+    {
+      "epoch": 2.1116999592335914,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.6801875254790054e-05,
+      "loss": 0.9979,
+      "step": 31080
+    },
+    {
+      "epoch": 2.112039679304253,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.679975200434842e-05,
+      "loss": 0.98,
+      "step": 31085
+    },
+    {
+      "epoch": 2.112379399374915,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.679762875390678e-05,
+      "loss": 0.9913,
+      "step": 31090
+    },
+    {
+      "epoch": 2.1127191194455768,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.6795505503465146e-05,
+      "loss": 0.9413,
+      "step": 31095
+    },
+    {
+      "epoch": 2.1130588395162384,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.679338225302351e-05,
+      "loss": 0.9171,
+      "step": 31100
+    },
+    {
+      "epoch": 2.1133985595869005,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.6791259002581874e-05,
+      "loss": 0.923,
+      "step": 31105
+    },
+    {
+      "epoch": 2.113738279657562,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.678913575214024e-05,
+      "loss": 1.03,
+      "step": 31110
+    },
+    {
+      "epoch": 2.1140779997282237,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.67870125016986e-05,
+      "loss": 0.9387,
+      "step": 31115
+    },
+    {
+      "epoch": 2.114417719798886,
+      "grad_norm": 6.5,
+      "learning_rate": 3.6784889251256966e-05,
+      "loss": 1.0177,
+      "step": 31120
+    },
+    {
+      "epoch": 2.1147574398695475,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.678276600081533e-05,
+      "loss": 1.0029,
+      "step": 31125
+    },
+    {
+      "epoch": 2.115097159940209,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.6780642750373694e-05,
+      "loss": 0.9924,
+      "step": 31130
+    },
+    {
+      "epoch": 2.115436880010871,
+      "grad_norm": 6.375,
+      "learning_rate": 3.677851949993206e-05,
+      "loss": 0.9152,
+      "step": 31135
+    },
+    {
+      "epoch": 2.115776600081533,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.677639624949042e-05,
+      "loss": 0.9146,
+      "step": 31140
+    },
+    {
+      "epoch": 2.1161163201521944,
+      "grad_norm": 6.0,
+      "learning_rate": 3.6774272999048786e-05,
+      "loss": 0.9721,
+      "step": 31145
+    },
+    {
+      "epoch": 2.1164560402228565,
+      "grad_norm": 6.5,
+      "learning_rate": 3.677214974860715e-05,
+      "loss": 1.044,
+      "step": 31150
+    },
+    {
+      "epoch": 2.116795760293518,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.677002649816551e-05,
+      "loss": 0.92,
+      "step": 31155
+    },
+    {
+      "epoch": 2.1171354803641798,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.676790324772388e-05,
+      "loss": 0.9618,
+      "step": 31160
+    },
+    {
+      "epoch": 2.117475200434842,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.676577999728224e-05,
+      "loss": 0.9193,
+      "step": 31165
+    },
+    {
+      "epoch": 2.1178149205055035,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.67636567468406e-05,
+      "loss": 0.9607,
+      "step": 31170
+    },
+    {
+      "epoch": 2.118154640576165,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.676153349639897e-05,
+      "loss": 0.9291,
+      "step": 31175
+    },
+    {
+      "epoch": 2.118494360646827,
+      "grad_norm": 6.375,
+      "learning_rate": 3.6759410245957334e-05,
+      "loss": 0.9097,
+      "step": 31180
+    },
+    {
+      "epoch": 2.118834080717489,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.67572869955157e-05,
+      "loss": 0.995,
+      "step": 31185
+    },
+    {
+      "epoch": 2.1191738007881504,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.675516374507406e-05,
+      "loss": 0.944,
+      "step": 31190
+    },
+    {
+      "epoch": 2.1195135208588125,
+      "grad_norm": 6.875,
+      "learning_rate": 3.6753040494632426e-05,
+      "loss": 0.9221,
+      "step": 31195
+    },
+    {
+      "epoch": 2.119853240929474,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.675091724419079e-05,
+      "loss": 1.0341,
+      "step": 31200
+    },
+    {
+      "epoch": 2.1201929610001358,
+      "grad_norm": 6.625,
+      "learning_rate": 3.6748793993749154e-05,
+      "loss": 0.9344,
+      "step": 31205
+    },
+    {
+      "epoch": 2.120532681070798,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.674667074330752e-05,
+      "loss": 1.0664,
+      "step": 31210
+    },
+    {
+      "epoch": 2.1208724011414595,
+      "grad_norm": 6.75,
+      "learning_rate": 3.674454749286588e-05,
+      "loss": 0.8934,
+      "step": 31215
+    },
+    {
+      "epoch": 2.121212121212121,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.6742424242424246e-05,
+      "loss": 0.9267,
+      "step": 31220
+    },
+    {
+      "epoch": 2.121551841282783,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.674030099198261e-05,
+      "loss": 0.9393,
+      "step": 31225
+    },
+    {
+      "epoch": 2.121891561353445,
+      "grad_norm": 7.0,
+      "learning_rate": 3.6738177741540974e-05,
+      "loss": 0.8841,
+      "step": 31230
+    },
+    {
+      "epoch": 2.1222312814241064,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.673605449109934e-05,
+      "loss": 0.8731,
+      "step": 31235
+    },
+    {
+      "epoch": 2.1225710014947685,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6733931240657695e-05,
+      "loss": 0.9665,
+      "step": 31240
+    },
+    {
+      "epoch": 2.12291072156543,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.6731807990216066e-05,
+      "loss": 1.0079,
+      "step": 31245
+    },
+    {
+      "epoch": 2.123250441636092,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.672968473977443e-05,
+      "loss": 0.9443,
+      "step": 31250
+    },
+    {
+      "epoch": 2.123590161706754,
+      "grad_norm": 6.75,
+      "learning_rate": 3.672756148933279e-05,
+      "loss": 0.9298,
+      "step": 31255
+    },
+    {
+      "epoch": 2.1239298817774155,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.672543823889116e-05,
+      "loss": 0.9718,
+      "step": 31260
+    },
+    {
+      "epoch": 2.124269601848077,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.672331498844952e-05,
+      "loss": 0.9378,
+      "step": 31265
+    },
+    {
+      "epoch": 2.1246093219187387,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.672119173800788e-05,
+      "loss": 0.9116,
+      "step": 31270
+    },
+    {
+      "epoch": 2.124949041989401,
+      "grad_norm": 6.125,
+      "learning_rate": 3.671906848756625e-05,
+      "loss": 0.8169,
+      "step": 31275
+    },
+    {
+      "epoch": 2.1252887620600625,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.6716945237124614e-05,
+      "loss": 0.9209,
+      "step": 31280
+    },
+    {
+      "epoch": 2.125628482130724,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.671482198668297e-05,
+      "loss": 0.9674,
+      "step": 31285
+    },
+    {
+      "epoch": 2.125968202201386,
+      "grad_norm": 6.375,
+      "learning_rate": 3.671269873624134e-05,
+      "loss": 0.9147,
+      "step": 31290
+    },
+    {
+      "epoch": 2.126307922272048,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6710575485799706e-05,
+      "loss": 0.9234,
+      "step": 31295
+    },
+    {
+      "epoch": 2.1266476423427094,
+      "grad_norm": 6.875,
+      "learning_rate": 3.6708452235358063e-05,
+      "loss": 0.8727,
+      "step": 31300
+    },
+    {
+      "epoch": 2.1269873624133715,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.6706328984916434e-05,
+      "loss": 0.9463,
+      "step": 31305
+    },
+    {
+      "epoch": 2.127327082484033,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.670420573447479e-05,
+      "loss": 0.9711,
+      "step": 31310
+    },
+    {
+      "epoch": 2.1276668025546948,
+      "grad_norm": 6.75,
+      "learning_rate": 3.6702082484033155e-05,
+      "loss": 0.9054,
+      "step": 31315
+    },
+    {
+      "epoch": 2.128006522625357,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.6699959233591526e-05,
+      "loss": 0.9102,
+      "step": 31320
+    },
+    {
+      "epoch": 2.1283462426960185,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6697835983149883e-05,
+      "loss": 1.0084,
+      "step": 31325
+    },
+    {
+      "epoch": 2.12868596276668,
+      "grad_norm": 6.75,
+      "learning_rate": 3.669571273270825e-05,
+      "loss": 0.9088,
+      "step": 31330
+    },
+    {
+      "epoch": 2.129025682837342,
+      "grad_norm": 6.375,
+      "learning_rate": 3.669358948226662e-05,
+      "loss": 0.929,
+      "step": 31335
+    },
+    {
+      "epoch": 2.129365402908004,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.6691466231824976e-05,
+      "loss": 1.0558,
+      "step": 31340
+    },
+    {
+      "epoch": 2.1297051229786654,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.668934298138334e-05,
+      "loss": 0.969,
+      "step": 31345
+    },
+    {
+      "epoch": 2.1300448430493275,
+      "grad_norm": 6.75,
+      "learning_rate": 3.668721973094171e-05,
+      "loss": 1.0276,
+      "step": 31350
+    },
+    {
+      "epoch": 2.130384563119989,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.668509648050007e-05,
+      "loss": 0.8881,
+      "step": 31355
+    },
+    {
+      "epoch": 2.1307242831906508,
+      "grad_norm": 5.15625,
+      "learning_rate": 3.668297323005843e-05,
+      "loss": 0.9021,
+      "step": 31360
+    },
+    {
+      "epoch": 2.131064003261313,
+      "grad_norm": 6.625,
+      "learning_rate": 3.66808499796168e-05,
+      "loss": 1.0339,
+      "step": 31365
+    },
+    {
+      "epoch": 2.1314037233319745,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.667872672917516e-05,
+      "loss": 0.9062,
+      "step": 31370
+    },
+    {
+      "epoch": 2.131743443402636,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6676603478733524e-05,
+      "loss": 1.0005,
+      "step": 31375
+    },
+    {
+      "epoch": 2.132083163473298,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.667448022829189e-05,
+      "loss": 0.8606,
+      "step": 31380
+    },
+    {
+      "epoch": 2.13242288354396,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.667235697785025e-05,
+      "loss": 1.0191,
+      "step": 31385
+    },
+    {
+      "epoch": 2.1327626036146214,
+      "grad_norm": 6.875,
+      "learning_rate": 3.6670233727408616e-05,
+      "loss": 0.9351,
+      "step": 31390
+    },
+    {
+      "epoch": 2.1331023236852835,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.666811047696698e-05,
+      "loss": 0.9074,
+      "step": 31395
+    },
+    {
+      "epoch": 2.133442043755945,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.6665987226525344e-05,
+      "loss": 0.882,
+      "step": 31400
+    },
+    {
+      "epoch": 2.133781763826607,
+      "grad_norm": 6.125,
+      "learning_rate": 3.666386397608371e-05,
+      "loss": 0.9408,
+      "step": 31405
+    },
+    {
+      "epoch": 2.134121483897269,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.666174072564207e-05,
+      "loss": 1.016,
+      "step": 31410
+    },
+    {
+      "epoch": 2.1344612039679305,
+      "grad_norm": 6.875,
+      "learning_rate": 3.6659617475200436e-05,
+      "loss": 0.9771,
+      "step": 31415
+    },
+    {
+      "epoch": 2.134800924038592,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.66574942247588e-05,
+      "loss": 1.1096,
+      "step": 31420
+    },
+    {
+      "epoch": 2.1351406441092537,
+      "grad_norm": 6.75,
+      "learning_rate": 3.6655370974317164e-05,
+      "loss": 0.929,
+      "step": 31425
+    },
+    {
+      "epoch": 2.135480364179916,
+      "grad_norm": 7.0,
+      "learning_rate": 3.665324772387553e-05,
+      "loss": 1.2188,
+      "step": 31430
+    },
+    {
+      "epoch": 2.1358200842505775,
+      "grad_norm": 6.625,
+      "learning_rate": 3.665112447343389e-05,
+      "loss": 0.9652,
+      "step": 31435
+    },
+    {
+      "epoch": 2.136159804321239,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.6649001222992256e-05,
+      "loss": 0.876,
+      "step": 31440
+    },
+    {
+      "epoch": 2.136499524391901,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.664687797255062e-05,
+      "loss": 0.9148,
+      "step": 31445
+    },
+    {
+      "epoch": 2.136839244462563,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.6644754722108984e-05,
+      "loss": 0.8998,
+      "step": 31450
+    },
+    {
+      "epoch": 2.1371789645332244,
+      "grad_norm": 6.5,
+      "learning_rate": 3.664263147166735e-05,
+      "loss": 0.9235,
+      "step": 31455
+    },
+    {
+      "epoch": 2.1375186846038865,
+      "grad_norm": 7.875,
+      "learning_rate": 3.664050822122571e-05,
+      "loss": 0.9226,
+      "step": 31460
+    },
+    {
+      "epoch": 2.137858404674548,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.6638384970784076e-05,
+      "loss": 1.0089,
+      "step": 31465
+    },
+    {
+      "epoch": 2.1381981247452098,
+      "grad_norm": 7.0,
+      "learning_rate": 3.663626172034244e-05,
+      "loss": 0.8938,
+      "step": 31470
+    },
+    {
+      "epoch": 2.138537844815872,
+      "grad_norm": 6.375,
+      "learning_rate": 3.6634138469900804e-05,
+      "loss": 0.9092,
+      "step": 31475
+    },
+    {
+      "epoch": 2.1388775648865335,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.663201521945917e-05,
+      "loss": 0.9619,
+      "step": 31480
+    },
+    {
+      "epoch": 2.139217284957195,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.662989196901753e-05,
+      "loss": 0.9947,
+      "step": 31485
+    },
+    {
+      "epoch": 2.139557005027857,
+      "grad_norm": 7.25,
+      "learning_rate": 3.6627768718575896e-05,
+      "loss": 0.9724,
+      "step": 31490
+    },
+    {
+      "epoch": 2.139896725098519,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.662564546813426e-05,
+      "loss": 0.843,
+      "step": 31495
+    },
+    {
+      "epoch": 2.1402364451691804,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6623522217692624e-05,
+      "loss": 0.9663,
+      "step": 31500
+    },
+    {
+      "epoch": 2.1405761652398425,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.662139896725099e-05,
+      "loss": 0.9367,
+      "step": 31505
+    },
+    {
+      "epoch": 2.140915885310504,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.6619275716809345e-05,
+      "loss": 0.9858,
+      "step": 31510
+    },
+    {
+      "epoch": 2.1412556053811658,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.6617152466367716e-05,
+      "loss": 0.9527,
+      "step": 31515
+    },
+    {
+      "epoch": 2.141595325451828,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.661502921592608e-05,
+      "loss": 0.9524,
+      "step": 31520
+    },
+    {
+      "epoch": 2.1419350455224895,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.6612905965484444e-05,
+      "loss": 0.9248,
+      "step": 31525
+    },
+    {
+      "epoch": 2.142274765593151,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.661078271504281e-05,
+      "loss": 0.8525,
+      "step": 31530
+    },
+    {
+      "epoch": 2.142614485663813,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.660865946460117e-05,
+      "loss": 0.9564,
+      "step": 31535
+    },
+    {
+      "epoch": 2.142954205734475,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.6606536214159536e-05,
+      "loss": 0.9721,
+      "step": 31540
+    },
+    {
+      "epoch": 2.1432939258051364,
+      "grad_norm": 5.875,
+      "learning_rate": 3.66044129637179e-05,
+      "loss": 1.0007,
+      "step": 31545
+    },
+    {
+      "epoch": 2.1436336458757985,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.6602289713276264e-05,
+      "loss": 1.0242,
+      "step": 31550
+    },
+    {
+      "epoch": 2.14397336594646,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.660016646283463e-05,
+      "loss": 0.9776,
+      "step": 31555
+    },
+    {
+      "epoch": 2.144313086017122,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.659804321239299e-05,
+      "loss": 0.9564,
+      "step": 31560
+    },
+    {
+      "epoch": 2.144652806087784,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.6595919961951356e-05,
+      "loss": 0.9223,
+      "step": 31565
+    },
+    {
+      "epoch": 2.1449925261584455,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.659379671150972e-05,
+      "loss": 0.9789,
+      "step": 31570
+    },
+    {
+      "epoch": 2.145332246229107,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.6591673461068084e-05,
+      "loss": 0.9682,
+      "step": 31575
+    },
+    {
+      "epoch": 2.145671966299769,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.658955021062644e-05,
+      "loss": 0.9611,
+      "step": 31580
+    },
+    {
+      "epoch": 2.146011686370431,
+      "grad_norm": 7.0,
+      "learning_rate": 3.658742696018481e-05,
+      "loss": 0.9477,
+      "step": 31585
+    },
+    {
+      "epoch": 2.1463514064410925,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.6585303709743176e-05,
+      "loss": 0.9659,
+      "step": 31590
+    },
+    {
+      "epoch": 2.1466911265117545,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.658318045930153e-05,
+      "loss": 0.9188,
+      "step": 31595
+    },
+    {
+      "epoch": 2.147030846582416,
+      "grad_norm": 6.625,
+      "learning_rate": 3.6581057208859904e-05,
+      "loss": 0.9752,
+      "step": 31600
+    },
+    {
+      "epoch": 2.147370566653078,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.657893395841827e-05,
+      "loss": 0.9539,
+      "step": 31605
+    },
+    {
+      "epoch": 2.14771028672374,
+      "grad_norm": 7.0,
+      "learning_rate": 3.6576810707976625e-05,
+      "loss": 0.9364,
+      "step": 31610
+    },
+    {
+      "epoch": 2.1480500067944015,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6574687457534996e-05,
+      "loss": 1.022,
+      "step": 31615
+    },
+    {
+      "epoch": 2.148389726865063,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.657256420709336e-05,
+      "loss": 0.9183,
+      "step": 31620
+    },
+    {
+      "epoch": 2.1487294469357248,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.657044095665172e-05,
+      "loss": 1.0013,
+      "step": 31625
+    },
+    {
+      "epoch": 2.149069167006387,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.656831770621009e-05,
+      "loss": 0.9859,
+      "step": 31630
+    },
+    {
+      "epoch": 2.1494088870770485,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.656619445576845e-05,
+      "loss": 0.8368,
+      "step": 31635
+    },
+    {
+      "epoch": 2.14974860714771,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.656407120532681e-05,
+      "loss": 0.9642,
+      "step": 31640
+    },
+    {
+      "epoch": 2.150088327218372,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.656194795488518e-05,
+      "loss": 0.9151,
+      "step": 31645
+    },
+    {
+      "epoch": 2.150428047289034,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.6559824704443544e-05,
+      "loss": 0.9933,
+      "step": 31650
+    },
+    {
+      "epoch": 2.1507677673596954,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.65577014540019e-05,
+      "loss": 0.9628,
+      "step": 31655
+    },
+    {
+      "epoch": 2.1511074874303575,
+      "grad_norm": 6.625,
+      "learning_rate": 3.655557820356027e-05,
+      "loss": 0.955,
+      "step": 31660
+    },
+    {
+      "epoch": 2.151447207501019,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.655345495311863e-05,
+      "loss": 0.9529,
+      "step": 31665
+    },
+    {
+      "epoch": 2.1517869275716808,
+      "grad_norm": 7.0,
+      "learning_rate": 3.655133170267699e-05,
+      "loss": 0.9714,
+      "step": 31670
+    },
+    {
+      "epoch": 2.152126647642343,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.6549208452235364e-05,
+      "loss": 0.9136,
+      "step": 31675
+    },
+    {
+      "epoch": 2.1524663677130045,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.654708520179372e-05,
+      "loss": 0.9339,
+      "step": 31680
+    },
+    {
+      "epoch": 2.152806087783666,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.6544961951352085e-05,
+      "loss": 0.9159,
+      "step": 31685
+    },
+    {
+      "epoch": 2.153145807854328,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.6542838700910456e-05,
+      "loss": 0.9582,
+      "step": 31690
+    },
+    {
+      "epoch": 2.15348552792499,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.654071545046881e-05,
+      "loss": 0.9567,
+      "step": 31695
+    },
+    {
+      "epoch": 2.1538252479956514,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.653859220002718e-05,
+      "loss": 0.9598,
+      "step": 31700
+    },
+    {
+      "epoch": 2.1541649680663135,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.653646894958555e-05,
+      "loss": 0.9075,
+      "step": 31705
+    },
+    {
+      "epoch": 2.154504688136975,
+      "grad_norm": 7.125,
+      "learning_rate": 3.6534345699143905e-05,
+      "loss": 0.9397,
+      "step": 31710
+    },
+    {
+      "epoch": 2.154844408207637,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.653222244870227e-05,
+      "loss": 0.9751,
+      "step": 31715
+    },
+    {
+      "epoch": 2.155184128278299,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.653009919826064e-05,
+      "loss": 0.916,
+      "step": 31720
+    },
+    {
+      "epoch": 2.1555238483489605,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.6527975947819e-05,
+      "loss": 0.9811,
+      "step": 31725
+    },
+    {
+      "epoch": 2.155863568419622,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.652585269737736e-05,
+      "loss": 1.0097,
+      "step": 31730
+    },
+    {
+      "epoch": 2.156203288490284,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.6523729446935725e-05,
+      "loss": 0.945,
+      "step": 31735
+    },
+    {
+      "epoch": 2.156543008560946,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.652160619649409e-05,
+      "loss": 0.9729,
+      "step": 31740
+    },
+    {
+      "epoch": 2.1568827286316075,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.651948294605245e-05,
+      "loss": 0.9554,
+      "step": 31745
+    },
+    {
+      "epoch": 2.1572224487022695,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.651735969561082e-05,
+      "loss": 1.0064,
+      "step": 31750
+    },
+    {
+      "epoch": 2.157562168772931,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.651523644516918e-05,
+      "loss": 0.8627,
+      "step": 31755
+    },
+    {
+      "epoch": 2.157901888843593,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.6513113194727545e-05,
+      "loss": 0.8622,
+      "step": 31760
+    },
+    {
+      "epoch": 2.1582416089142544,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.651098994428591e-05,
+      "loss": 0.9084,
+      "step": 31765
+    },
+    {
+      "epoch": 2.1585813289849165,
+      "grad_norm": 6.75,
+      "learning_rate": 3.650886669384427e-05,
+      "loss": 0.9053,
+      "step": 31770
+    },
+    {
+      "epoch": 2.158921049055578,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.650674344340264e-05,
+      "loss": 0.9399,
+      "step": 31775
+    },
+    {
+      "epoch": 2.1592607691262398,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.6504620192961e-05,
+      "loss": 0.9474,
+      "step": 31780
+    },
+    {
+      "epoch": 2.159600489196902,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.6502496942519365e-05,
+      "loss": 0.8723,
+      "step": 31785
+    },
+    {
+      "epoch": 2.1599402092675635,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.650037369207773e-05,
+      "loss": 0.9753,
+      "step": 31790
+    },
+    {
+      "epoch": 2.160279929338225,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.649825044163609e-05,
+      "loss": 0.9927,
+      "step": 31795
+    },
+    {
+      "epoch": 2.160619649408887,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.649612719119446e-05,
+      "loss": 0.9395,
+      "step": 31800
+    },
+    {
+      "epoch": 2.160959369479549,
+      "grad_norm": 8.875,
+      "learning_rate": 3.649400394075282e-05,
+      "loss": 1.0236,
+      "step": 31805
+    },
+    {
+      "epoch": 2.1612990895502104,
+      "grad_norm": 6.625,
+      "learning_rate": 3.6491880690311185e-05,
+      "loss": 0.9405,
+      "step": 31810
+    },
+    {
+      "epoch": 2.1616388096208725,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.648975743986955e-05,
+      "loss": 0.9734,
+      "step": 31815
+    },
+    {
+      "epoch": 2.161978529691534,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.648763418942791e-05,
+      "loss": 0.9869,
+      "step": 31820
+    },
+    {
+      "epoch": 2.1623182497621958,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.648551093898628e-05,
+      "loss": 0.976,
+      "step": 31825
+    },
+    {
+      "epoch": 2.162657969832858,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.648338768854464e-05,
+      "loss": 0.9204,
+      "step": 31830
+    },
+    {
+      "epoch": 2.1629976899035195,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.6481264438103005e-05,
+      "loss": 0.9675,
+      "step": 31835
+    },
+    {
+      "epoch": 2.163337409974181,
+      "grad_norm": 6.25,
+      "learning_rate": 3.647914118766137e-05,
+      "loss": 0.9212,
+      "step": 31840
+    },
+    {
+      "epoch": 2.163677130044843,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.647701793721973e-05,
+      "loss": 0.9393,
+      "step": 31845
+    },
+    {
+      "epoch": 2.164016850115505,
+      "grad_norm": 6.625,
+      "learning_rate": 3.64748946867781e-05,
+      "loss": 0.9885,
+      "step": 31850
+    },
+    {
+      "epoch": 2.1643565701861665,
+      "grad_norm": 7.0,
+      "learning_rate": 3.647277143633646e-05,
+      "loss": 0.9915,
+      "step": 31855
+    },
+    {
+      "epoch": 2.1646962902568285,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.6470648185894825e-05,
+      "loss": 0.9388,
+      "step": 31860
+    },
+    {
+      "epoch": 2.16503601032749,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.646852493545319e-05,
+      "loss": 0.9324,
+      "step": 31865
+    },
+    {
+      "epoch": 2.165375730398152,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.646640168501155e-05,
+      "loss": 0.9737,
+      "step": 31870
+    },
+    {
+      "epoch": 2.165715450468814,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.646427843456992e-05,
+      "loss": 0.9604,
+      "step": 31875
+    },
+    {
+      "epoch": 2.1660551705394755,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.646215518412828e-05,
+      "loss": 0.9007,
+      "step": 31880
+    },
+    {
+      "epoch": 2.166394890610137,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.6460031933686645e-05,
+      "loss": 0.9064,
+      "step": 31885
+    },
+    {
+      "epoch": 2.166734610680799,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.645790868324501e-05,
+      "loss": 0.9298,
+      "step": 31890
+    },
+    {
+      "epoch": 2.167074330751461,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.645578543280337e-05,
+      "loss": 0.9266,
+      "step": 31895
+    },
+    {
+      "epoch": 2.1674140508221225,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.645366218236174e-05,
+      "loss": 1.0075,
+      "step": 31900
+    },
+    {
+      "epoch": 2.1677537708927845,
+      "grad_norm": 5.875,
+      "learning_rate": 3.64515389319201e-05,
+      "loss": 1.0169,
+      "step": 31905
+    },
+    {
+      "epoch": 2.168093490963446,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.6449415681478465e-05,
+      "loss": 0.9308,
+      "step": 31910
+    },
+    {
+      "epoch": 2.168433211034108,
+      "grad_norm": 7.25,
+      "learning_rate": 3.644729243103683e-05,
+      "loss": 0.9135,
+      "step": 31915
+    },
+    {
+      "epoch": 2.16877293110477,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.644516918059519e-05,
+      "loss": 0.9216,
+      "step": 31920
+    },
+    {
+      "epoch": 2.1691126511754315,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.644304593015356e-05,
+      "loss": 0.9567,
+      "step": 31925
+    },
+    {
+      "epoch": 2.169452371246093,
+      "grad_norm": 7.125,
+      "learning_rate": 3.644092267971192e-05,
+      "loss": 0.8925,
+      "step": 31930
+    },
+    {
+      "epoch": 2.169792091316755,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.643879942927028e-05,
+      "loss": 0.9477,
+      "step": 31935
+    },
+    {
+      "epoch": 2.170131811387417,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.643667617882865e-05,
+      "loss": 1.1238,
+      "step": 31940
+    },
+    {
+      "epoch": 2.1704715314580785,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.643455292838701e-05,
+      "loss": 1.0134,
+      "step": 31945
+    },
+    {
+      "epoch": 2.1708112515287405,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.643242967794537e-05,
+      "loss": 0.9839,
+      "step": 31950
+    },
+    {
+      "epoch": 2.171150971599402,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.643030642750374e-05,
+      "loss": 0.9952,
+      "step": 31955
+    },
+    {
+      "epoch": 2.171490691670064,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.6428183177062105e-05,
+      "loss": 0.9448,
+      "step": 31960
+    },
+    {
+      "epoch": 2.1718304117407254,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.642605992662046e-05,
+      "loss": 0.9737,
+      "step": 31965
+    },
+    {
+      "epoch": 2.1721701318113875,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.642393667617883e-05,
+      "loss": 0.9335,
+      "step": 31970
+    },
+    {
+      "epoch": 2.172509851882049,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.64218134257372e-05,
+      "loss": 0.8427,
+      "step": 31975
+    },
+    {
+      "epoch": 2.172849571952711,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.6419690175295554e-05,
+      "loss": 0.9836,
+      "step": 31980
+    },
+    {
+      "epoch": 2.173189292023373,
+      "grad_norm": 8.0,
+      "learning_rate": 3.6417566924853925e-05,
+      "loss": 0.9945,
+      "step": 31985
+    },
+    {
+      "epoch": 2.1735290120940345,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.641544367441229e-05,
+      "loss": 0.9373,
+      "step": 31990
+    },
+    {
+      "epoch": 2.173868732164696,
+      "grad_norm": 6.75,
+      "learning_rate": 3.6413320423970646e-05,
+      "loss": 0.8786,
+      "step": 31995
+    },
+    {
+      "epoch": 2.174208452235358,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.641119717352902e-05,
+      "loss": 0.96,
+      "step": 32000
+    },
+    {
+      "epoch": 2.17454817230602,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.640907392308738e-05,
+      "loss": 1.0583,
+      "step": 32005
+    },
+    {
+      "epoch": 2.1748878923766815,
+      "grad_norm": 6.125,
+      "learning_rate": 3.640695067264574e-05,
+      "loss": 0.9082,
+      "step": 32010
+    },
+    {
+      "epoch": 2.1752276124473435,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.640482742220411e-05,
+      "loss": 1.034,
+      "step": 32015
+    },
+    {
+      "epoch": 2.175567332518005,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.6402704171762466e-05,
+      "loss": 0.8889,
+      "step": 32020
+    },
+    {
+      "epoch": 2.175907052588667,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.640058092132083e-05,
+      "loss": 1.0226,
+      "step": 32025
+    },
+    {
+      "epoch": 2.176246772659329,
+      "grad_norm": 6.875,
+      "learning_rate": 3.63984576708792e-05,
+      "loss": 0.9245,
+      "step": 32030
+    },
+    {
+      "epoch": 2.1765864927299905,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.639633442043756e-05,
+      "loss": 0.9411,
+      "step": 32035
+    },
+    {
+      "epoch": 2.176926212800652,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.639421116999592e-05,
+      "loss": 0.9071,
+      "step": 32040
+    },
+    {
+      "epoch": 2.177265932871314,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.639208791955429e-05,
+      "loss": 0.9132,
+      "step": 32045
+    },
+    {
+      "epoch": 2.177605652941976,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.638996466911265e-05,
+      "loss": 0.9031,
+      "step": 32050
+    },
+    {
+      "epoch": 2.1779453730126375,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.6387841418671015e-05,
+      "loss": 1.0051,
+      "step": 32055
+    },
+    {
+      "epoch": 2.1782850930832995,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.6385718168229385e-05,
+      "loss": 0.9772,
+      "step": 32060
+    },
+    {
+      "epoch": 2.178624813153961,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.638359491778774e-05,
+      "loss": 0.9813,
+      "step": 32065
+    },
+    {
+      "epoch": 2.178964533224623,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.6381471667346107e-05,
+      "loss": 1.0401,
+      "step": 32070
+    },
+    {
+      "epoch": 2.179304253295285,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.637934841690448e-05,
+      "loss": 0.9907,
+      "step": 32075
+    },
+    {
+      "epoch": 2.1796439733659465,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.6377225166462835e-05,
+      "loss": 0.9951,
+      "step": 32080
+    },
+    {
+      "epoch": 2.179983693436608,
+      "grad_norm": 6.875,
+      "learning_rate": 3.63751019160212e-05,
+      "loss": 1.0572,
+      "step": 32085
+    },
+    {
+      "epoch": 2.18032341350727,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.637297866557956e-05,
+      "loss": 0.9043,
+      "step": 32090
+    },
+    {
+      "epoch": 2.180663133577932,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6370855415137927e-05,
+      "loss": 0.9954,
+      "step": 32095
+    },
+    {
+      "epoch": 2.1810028536485935,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.636873216469629e-05,
+      "loss": 0.9116,
+      "step": 32100
+    },
+    {
+      "epoch": 2.181342573719255,
+      "grad_norm": 6.0,
+      "learning_rate": 3.6366608914254655e-05,
+      "loss": 0.966,
+      "step": 32105
+    },
+    {
+      "epoch": 2.181682293789917,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.636448566381302e-05,
+      "loss": 0.9344,
+      "step": 32110
+    },
+    {
+      "epoch": 2.182022013860579,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.636236241337138e-05,
+      "loss": 0.957,
+      "step": 32115
+    },
+    {
+      "epoch": 2.1823617339312404,
+      "grad_norm": 6.5,
+      "learning_rate": 3.6360239162929747e-05,
+      "loss": 0.9781,
+      "step": 32120
+    },
+    {
+      "epoch": 2.1827014540019025,
+      "grad_norm": 6.625,
+      "learning_rate": 3.635811591248811e-05,
+      "loss": 0.9463,
+      "step": 32125
+    },
+    {
+      "epoch": 2.183041174072564,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.6355992662046475e-05,
+      "loss": 0.9386,
+      "step": 32130
+    },
+    {
+      "epoch": 2.183380894143226,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.635386941160484e-05,
+      "loss": 1.0632,
+      "step": 32135
+    },
+    {
+      "epoch": 2.183720614213888,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.63517461611632e-05,
+      "loss": 0.9334,
+      "step": 32140
+    },
+    {
+      "epoch": 2.1840603342845495,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.6349622910721567e-05,
+      "loss": 0.9774,
+      "step": 32145
+    },
+    {
+      "epoch": 2.184400054355211,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.634749966027993e-05,
+      "loss": 1.0211,
+      "step": 32150
+    },
+    {
+      "epoch": 2.184739774425873,
+      "grad_norm": 6.5,
+      "learning_rate": 3.6345376409838295e-05,
+      "loss": 0.9177,
+      "step": 32155
+    },
+    {
+      "epoch": 2.185079494496535,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.634325315939666e-05,
+      "loss": 1.0052,
+      "step": 32160
+    },
+    {
+      "epoch": 2.1854192145671965,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.634112990895502e-05,
+      "loss": 0.9853,
+      "step": 32165
+    },
+    {
+      "epoch": 2.1857589346378585,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.6339006658513387e-05,
+      "loss": 0.9987,
+      "step": 32170
+    },
+    {
+      "epoch": 2.18609865470852,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.633688340807175e-05,
+      "loss": 0.9353,
+      "step": 32175
+    },
+    {
+      "epoch": 2.186438374779182,
+      "grad_norm": 4.96875,
+      "learning_rate": 3.6334760157630115e-05,
+      "loss": 0.9259,
+      "step": 32180
+    },
+    {
+      "epoch": 2.186778094849844,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.633263690718848e-05,
+      "loss": 0.9147,
+      "step": 32185
+    },
+    {
+      "epoch": 2.1871178149205055,
+      "grad_norm": 6.625,
+      "learning_rate": 3.633051365674684e-05,
+      "loss": 1.049,
+      "step": 32190
+    },
+    {
+      "epoch": 2.187457534991167,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.6328390406305207e-05,
+      "loss": 1.0235,
+      "step": 32195
+    },
+    {
+      "epoch": 2.187797255061829,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.632626715586357e-05,
+      "loss": 0.9482,
+      "step": 32200
+    },
+    {
+      "epoch": 2.188136975132491,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.6324143905421935e-05,
+      "loss": 0.877,
+      "step": 32205
+    },
+    {
+      "epoch": 2.1884766952031525,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.63220206549803e-05,
+      "loss": 0.9445,
+      "step": 32210
+    },
+    {
+      "epoch": 2.1888164152738145,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.631989740453866e-05,
+      "loss": 0.9234,
+      "step": 32215
+    },
+    {
+      "epoch": 2.189156135344476,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.631777415409703e-05,
+      "loss": 0.9593,
+      "step": 32220
+    },
+    {
+      "epoch": 2.189495855415138,
+      "grad_norm": 6.25,
+      "learning_rate": 3.631565090365539e-05,
+      "loss": 0.8223,
+      "step": 32225
+    },
+    {
+      "epoch": 2.1898355754858,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.6313527653213755e-05,
+      "loss": 0.9764,
+      "step": 32230
+    },
+    {
+      "epoch": 2.1901752955564615,
+      "grad_norm": 7.125,
+      "learning_rate": 3.631140440277212e-05,
+      "loss": 0.9815,
+      "step": 32235
+    },
+    {
+      "epoch": 2.190515015627123,
+      "grad_norm": 6.5,
+      "learning_rate": 3.630928115233048e-05,
+      "loss": 0.9643,
+      "step": 32240
+    },
+    {
+      "epoch": 2.190854735697785,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.630715790188885e-05,
+      "loss": 1.0025,
+      "step": 32245
+    },
+    {
+      "epoch": 2.191194455768447,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.630503465144721e-05,
+      "loss": 0.9416,
+      "step": 32250
+    },
+    {
+      "epoch": 2.1915341758391085,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.6302911401005575e-05,
+      "loss": 0.9253,
+      "step": 32255
+    },
+    {
+      "epoch": 2.1918738959097706,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.630078815056394e-05,
+      "loss": 0.9629,
+      "step": 32260
+    },
+    {
+      "epoch": 2.192213615980432,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.62986649001223e-05,
+      "loss": 0.9577,
+      "step": 32265
+    },
+    {
+      "epoch": 2.192553336051094,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.629654164968067e-05,
+      "loss": 0.9311,
+      "step": 32270
+    },
+    {
+      "epoch": 2.192893056121756,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.629441839923903e-05,
+      "loss": 1.0104,
+      "step": 32275
+    },
+    {
+      "epoch": 2.1932327761924175,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.6292295148797395e-05,
+      "loss": 0.9607,
+      "step": 32280
+    },
+    {
+      "epoch": 2.193572496263079,
+      "grad_norm": 6.125,
+      "learning_rate": 3.629017189835576e-05,
+      "loss": 0.9203,
+      "step": 32285
+    },
+    {
+      "epoch": 2.1939122163337412,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.6288048647914116e-05,
+      "loss": 0.9571,
+      "step": 32290
+    },
+    {
+      "epoch": 2.194251936404403,
+      "grad_norm": 6.375,
+      "learning_rate": 3.628592539747249e-05,
+      "loss": 0.9341,
+      "step": 32295
+    },
+    {
+      "epoch": 2.1945916564750645,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.628380214703085e-05,
+      "loss": 0.9568,
+      "step": 32300
+    },
+    {
+      "epoch": 2.194931376545726,
+      "grad_norm": 6.75,
+      "learning_rate": 3.628167889658921e-05,
+      "loss": 0.911,
+      "step": 32305
+    },
+    {
+      "epoch": 2.195271096616388,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.627955564614758e-05,
+      "loss": 1.0077,
+      "step": 32310
+    },
+    {
+      "epoch": 2.19561081668705,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.627743239570594e-05,
+      "loss": 0.9686,
+      "step": 32315
+    },
+    {
+      "epoch": 2.1959505367577115,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.62753091452643e-05,
+      "loss": 0.9231,
+      "step": 32320
+    },
+    {
+      "epoch": 2.1962902568283735,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.627318589482267e-05,
+      "loss": 0.9915,
+      "step": 32325
+    },
+    {
+      "epoch": 2.196629976899035,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.6271062644381035e-05,
+      "loss": 0.9523,
+      "step": 32330
+    },
+    {
+      "epoch": 2.196969696969697,
+      "grad_norm": 7.125,
+      "learning_rate": 3.626893939393939e-05,
+      "loss": 0.9432,
+      "step": 32335
+    },
+    {
+      "epoch": 2.197309417040359,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.626681614349776e-05,
+      "loss": 1.0655,
+      "step": 32340
+    },
+    {
+      "epoch": 2.1976491371110205,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.626469289305613e-05,
+      "loss": 1.0436,
+      "step": 32345
+    },
+    {
+      "epoch": 2.197988857181682,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6262569642614484e-05,
+      "loss": 0.9894,
+      "step": 32350
+    },
+    {
+      "epoch": 2.198328577252344,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.6260446392172855e-05,
+      "loss": 0.8665,
+      "step": 32355
+    },
+    {
+      "epoch": 2.198668297323006,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.625832314173121e-05,
+      "loss": 1.0154,
+      "step": 32360
+    },
+    {
+      "epoch": 2.1990080173936675,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.6256199891289576e-05,
+      "loss": 0.9311,
+      "step": 32365
+    },
+    {
+      "epoch": 2.1993477374643295,
+      "grad_norm": 6.5,
+      "learning_rate": 3.625407664084795e-05,
+      "loss": 0.8828,
+      "step": 32370
+    },
+    {
+      "epoch": 2.199687457534991,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.6251953390406304e-05,
+      "loss": 0.9036,
+      "step": 32375
+    },
+    {
+      "epoch": 2.200027177605653,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.624983013996467e-05,
+      "loss": 1.0081,
+      "step": 32380
+    },
+    {
+      "epoch": 2.200366897676315,
+      "grad_norm": 6.875,
+      "learning_rate": 3.624770688952304e-05,
+      "loss": 0.9053,
+      "step": 32385
+    },
+    {
+      "epoch": 2.2007066177469765,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.6245583639081396e-05,
+      "loss": 0.9591,
+      "step": 32390
+    },
+    {
+      "epoch": 2.201046337817638,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.624346038863976e-05,
+      "loss": 0.9583,
+      "step": 32395
+    },
+    {
+      "epoch": 2.2013860578883,
+      "grad_norm": 6.875,
+      "learning_rate": 3.624133713819813e-05,
+      "loss": 1.0772,
+      "step": 32400
+    },
+    {
+      "epoch": 2.201725777958962,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.623921388775649e-05,
+      "loss": 0.9641,
+      "step": 32405
+    },
+    {
+      "epoch": 2.2020654980296235,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.623709063731485e-05,
+      "loss": 0.9713,
+      "step": 32410
+    },
+    {
+      "epoch": 2.2024052181002856,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.623496738687322e-05,
+      "loss": 1.0376,
+      "step": 32415
+    },
+    {
+      "epoch": 2.202744938170947,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.623284413643158e-05,
+      "loss": 0.9894,
+      "step": 32420
+    },
+    {
+      "epoch": 2.203084658241609,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.6230720885989944e-05,
+      "loss": 1.0424,
+      "step": 32425
+    },
+    {
+      "epoch": 2.203424378312271,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6228597635548315e-05,
+      "loss": 0.9311,
+      "step": 32430
+    },
+    {
+      "epoch": 2.2037640983829325,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.622647438510667e-05,
+      "loss": 0.9893,
+      "step": 32435
+    },
+    {
+      "epoch": 2.204103818453594,
+      "grad_norm": 7.125,
+      "learning_rate": 3.6224351134665036e-05,
+      "loss": 1.0261,
+      "step": 32440
+    },
+    {
+      "epoch": 2.204443538524256,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.62222278842234e-05,
+      "loss": 1.0187,
+      "step": 32445
+    },
+    {
+      "epoch": 2.204783258594918,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.6220104633781764e-05,
+      "loss": 0.9104,
+      "step": 32450
+    },
+    {
+      "epoch": 2.2051229786655795,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.621798138334013e-05,
+      "loss": 0.9407,
+      "step": 32455
+    },
+    {
+      "epoch": 2.205462698736241,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.621585813289849e-05,
+      "loss": 1.0118,
+      "step": 32460
+    },
+    {
+      "epoch": 2.205802418806903,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.6213734882456856e-05,
+      "loss": 0.9563,
+      "step": 32465
+    },
+    {
+      "epoch": 2.206142138877565,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.621161163201522e-05,
+      "loss": 0.9514,
+      "step": 32470
+    },
+    {
+      "epoch": 2.2064818589482265,
+      "grad_norm": 6.375,
+      "learning_rate": 3.6209488381573584e-05,
+      "loss": 1.029,
+      "step": 32475
+    },
+    {
+      "epoch": 2.2068215790188885,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.620736513113195e-05,
+      "loss": 0.9651,
+      "step": 32480
+    },
+    {
+      "epoch": 2.20716129908955,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.620524188069031e-05,
+      "loss": 0.9252,
+      "step": 32485
+    },
+    {
+      "epoch": 2.207501019160212,
+      "grad_norm": 6.75,
+      "learning_rate": 3.6203118630248676e-05,
+      "loss": 1.0383,
+      "step": 32490
+    },
+    {
+      "epoch": 2.207840739230874,
+      "grad_norm": 6.625,
+      "learning_rate": 3.620099537980704e-05,
+      "loss": 1.0222,
+      "step": 32495
+    },
+    {
+      "epoch": 2.2081804593015355,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.6198872129365404e-05,
+      "loss": 1.0075,
+      "step": 32500
+    },
+    {
+      "epoch": 2.208520179372197,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.619674887892377e-05,
+      "loss": 0.8881,
+      "step": 32505
+    },
+    {
+      "epoch": 2.208859899442859,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.619462562848213e-05,
+      "loss": 1.0405,
+      "step": 32510
+    },
+    {
+      "epoch": 2.209199619513521,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.6192502378040496e-05,
+      "loss": 0.974,
+      "step": 32515
+    },
+    {
+      "epoch": 2.2095393395841825,
+      "grad_norm": 7.125,
+      "learning_rate": 3.619037912759886e-05,
+      "loss": 0.9582,
+      "step": 32520
+    },
+    {
+      "epoch": 2.2098790596548445,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.6188255877157224e-05,
+      "loss": 0.9538,
+      "step": 32525
+    },
+    {
+      "epoch": 2.210218779725506,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.618613262671559e-05,
+      "loss": 0.9663,
+      "step": 32530
+    },
+    {
+      "epoch": 2.210558499796168,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.618400937627395e-05,
+      "loss": 0.9533,
+      "step": 32535
+    },
+    {
+      "epoch": 2.21089821986683,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.6181886125832316e-05,
+      "loss": 1.0638,
+      "step": 32540
+    },
+    {
+      "epoch": 2.2112379399374915,
+      "grad_norm": 6.375,
+      "learning_rate": 3.617976287539068e-05,
+      "loss": 1.0464,
+      "step": 32545
+    },
+    {
+      "epoch": 2.211577660008153,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.6177639624949044e-05,
+      "loss": 0.9456,
+      "step": 32550
+    },
+    {
+      "epoch": 2.211917380078815,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.617551637450741e-05,
+      "loss": 0.9029,
+      "step": 32555
+    },
+    {
+      "epoch": 2.212257100149477,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.617339312406577e-05,
+      "loss": 1.004,
+      "step": 32560
+    },
+    {
+      "epoch": 2.2125968202201385,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.6171269873624136e-05,
+      "loss": 0.9608,
+      "step": 32565
+    },
+    {
+      "epoch": 2.2129365402908006,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.61691466231825e-05,
+      "loss": 0.93,
+      "step": 32570
+    },
+    {
+      "epoch": 2.213276260361462,
+      "grad_norm": 6.625,
+      "learning_rate": 3.6167023372740864e-05,
+      "loss": 1.0428,
+      "step": 32575
+    },
+    {
+      "epoch": 2.213615980432124,
+      "grad_norm": 7.375,
+      "learning_rate": 3.616490012229923e-05,
+      "loss": 0.9784,
+      "step": 32580
+    },
+    {
+      "epoch": 2.213955700502786,
+      "grad_norm": 6.375,
+      "learning_rate": 3.616277687185759e-05,
+      "loss": 0.9113,
+      "step": 32585
+    },
+    {
+      "epoch": 2.2142954205734475,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6160653621415956e-05,
+      "loss": 1.0465,
+      "step": 32590
+    },
+    {
+      "epoch": 2.214635140644109,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.615853037097432e-05,
+      "loss": 0.869,
+      "step": 32595
+    },
+    {
+      "epoch": 2.2149748607147712,
+      "grad_norm": 6.625,
+      "learning_rate": 3.6156407120532684e-05,
+      "loss": 1.0114,
+      "step": 32600
+    },
+    {
+      "epoch": 2.215314580785433,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.615428387009105e-05,
+      "loss": 0.9136,
+      "step": 32605
+    },
+    {
+      "epoch": 2.2156543008560945,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.615216061964941e-05,
+      "loss": 0.9347,
+      "step": 32610
+    },
+    {
+      "epoch": 2.2159940209267566,
+      "grad_norm": 6.875,
+      "learning_rate": 3.6150037369207776e-05,
+      "loss": 0.9431,
+      "step": 32615
+    },
+    {
+      "epoch": 2.216333740997418,
+      "grad_norm": 6.875,
+      "learning_rate": 3.614791411876614e-05,
+      "loss": 0.9415,
+      "step": 32620
+    },
+    {
+      "epoch": 2.21667346106808,
+      "grad_norm": 7.125,
+      "learning_rate": 3.6145790868324504e-05,
+      "loss": 0.8619,
+      "step": 32625
+    },
+    {
+      "epoch": 2.217013181138742,
+      "grad_norm": 7.5,
+      "learning_rate": 3.614366761788287e-05,
+      "loss": 0.9461,
+      "step": 32630
+    },
+    {
+      "epoch": 2.2173529012094035,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.614154436744123e-05,
+      "loss": 0.9763,
+      "step": 32635
+    },
+    {
+      "epoch": 2.217692621280065,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.6139421116999596e-05,
+      "loss": 0.9152,
+      "step": 32640
+    },
+    {
+      "epoch": 2.218032341350727,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.6137297866557953e-05,
+      "loss": 0.9755,
+      "step": 32645
+    },
+    {
+      "epoch": 2.218372061421389,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.6135174616116324e-05,
+      "loss": 0.9655,
+      "step": 32650
+    },
+    {
+      "epoch": 2.2187117814920505,
+      "grad_norm": 4.875,
+      "learning_rate": 3.613305136567469e-05,
+      "loss": 0.9782,
+      "step": 32655
+    },
+    {
+      "epoch": 2.219051501562712,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.6130928115233045e-05,
+      "loss": 0.9824,
+      "step": 32660
+    },
+    {
+      "epoch": 2.219391221633374,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.6128804864791416e-05,
+      "loss": 0.9888,
+      "step": 32665
+    },
+    {
+      "epoch": 2.219730941704036,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.612668161434978e-05,
+      "loss": 0.9636,
+      "step": 32670
+    },
+    {
+      "epoch": 2.2200706617746975,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.612455836390814e-05,
+      "loss": 0.9935,
+      "step": 32675
+    },
+    {
+      "epoch": 2.2204103818453595,
+      "grad_norm": 6.625,
+      "learning_rate": 3.612243511346651e-05,
+      "loss": 0.9455,
+      "step": 32680
+    },
+    {
+      "epoch": 2.220750101916021,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.612031186302487e-05,
+      "loss": 0.9304,
+      "step": 32685
+    },
+    {
+      "epoch": 2.221089821986683,
+      "grad_norm": 6.75,
+      "learning_rate": 3.611818861258323e-05,
+      "loss": 0.9246,
+      "step": 32690
+    },
+    {
+      "epoch": 2.221429542057345,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.61160653621416e-05,
+      "loss": 0.9561,
+      "step": 32695
+    },
+    {
+      "epoch": 2.2217692621280065,
+      "grad_norm": 7.0,
+      "learning_rate": 3.6113942111699964e-05,
+      "loss": 1.0186,
+      "step": 32700
+    },
+    {
+      "epoch": 2.222108982198668,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.611181886125832e-05,
+      "loss": 1.049,
+      "step": 32705
+    },
+    {
+      "epoch": 2.2224487022693302,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.610969561081669e-05,
+      "loss": 0.9757,
+      "step": 32710
+    },
+    {
+      "epoch": 2.222788422339992,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.610757236037505e-05,
+      "loss": 0.9835,
+      "step": 32715
+    },
+    {
+      "epoch": 2.2231281424106535,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.6105449109933413e-05,
+      "loss": 0.9288,
+      "step": 32720
+    },
+    {
+      "epoch": 2.2234678624813156,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.6103325859491784e-05,
+      "loss": 0.9966,
+      "step": 32725
+    },
+    {
+      "epoch": 2.223807582551977,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.610120260905014e-05,
+      "loss": 0.982,
+      "step": 32730
+    },
+    {
+      "epoch": 2.224147302622639,
+      "grad_norm": 6.0,
+      "learning_rate": 3.6099079358608505e-05,
+      "loss": 0.9231,
+      "step": 32735
+    },
+    {
+      "epoch": 2.224487022693301,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.6096956108166876e-05,
+      "loss": 0.9907,
+      "step": 32740
+    },
+    {
+      "epoch": 2.2248267427639625,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.6094832857725233e-05,
+      "loss": 0.9594,
+      "step": 32745
+    },
+    {
+      "epoch": 2.225166462834624,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.60927096072836e-05,
+      "loss": 0.862,
+      "step": 32750
+    },
+    {
+      "epoch": 2.2255061829052862,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.609058635684197e-05,
+      "loss": 0.8847,
+      "step": 32755
+    },
+    {
+      "epoch": 2.225845902975948,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6088463106400325e-05,
+      "loss": 0.9644,
+      "step": 32760
+    },
+    {
+      "epoch": 2.2261856230466095,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.608633985595869e-05,
+      "loss": 0.9692,
+      "step": 32765
+    },
+    {
+      "epoch": 2.2265253431172716,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.608421660551706e-05,
+      "loss": 0.9909,
+      "step": 32770
+    },
+    {
+      "epoch": 2.226865063187933,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.608209335507542e-05,
+      "loss": 0.969,
+      "step": 32775
+    },
+    {
+      "epoch": 2.227204783258595,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.607997010463378e-05,
+      "loss": 1.0001,
+      "step": 32780
+    },
+    {
+      "epoch": 2.2275445033292565,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.6077846854192146e-05,
+      "loss": 1.0043,
+      "step": 32785
+    },
+    {
+      "epoch": 2.2278842233999185,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.607572360375051e-05,
+      "loss": 1.0095,
+      "step": 32790
+    },
+    {
+      "epoch": 2.22822394347058,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.6073600353308874e-05,
+      "loss": 0.8589,
+      "step": 32795
+    },
+    {
+      "epoch": 2.228563663541242,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.607147710286724e-05,
+      "loss": 0.9577,
+      "step": 32800
+    },
+    {
+      "epoch": 2.228903383611904,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.60693538524256e-05,
+      "loss": 0.9747,
+      "step": 32805
+    },
+    {
+      "epoch": 2.2292431036825655,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.6067230601983966e-05,
+      "loss": 0.942,
+      "step": 32810
+    },
+    {
+      "epoch": 2.229582823753227,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.606510735154233e-05,
+      "loss": 0.9618,
+      "step": 32815
+    },
+    {
+      "epoch": 2.229922543823889,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.6062984101100694e-05,
+      "loss": 0.9689,
+      "step": 32820
+    },
+    {
+      "epoch": 2.230262263894551,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.606086085065906e-05,
+      "loss": 0.9775,
+      "step": 32825
+    },
+    {
+      "epoch": 2.2306019839652125,
+      "grad_norm": 6.625,
+      "learning_rate": 3.605873760021742e-05,
+      "loss": 1.0184,
+      "step": 32830
+    },
+    {
+      "epoch": 2.2309417040358746,
+      "grad_norm": 7.375,
+      "learning_rate": 3.6056614349775786e-05,
+      "loss": 0.9413,
+      "step": 32835
+    },
+    {
+      "epoch": 2.231281424106536,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.605449109933415e-05,
+      "loss": 0.9792,
+      "step": 32840
+    },
+    {
+      "epoch": 2.231621144177198,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.6052367848892514e-05,
+      "loss": 0.9718,
+      "step": 32845
+    },
+    {
+      "epoch": 2.23196086424786,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.605024459845088e-05,
+      "loss": 0.961,
+      "step": 32850
+    },
+    {
+      "epoch": 2.2323005843185215,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.604812134800924e-05,
+      "loss": 1.0433,
+      "step": 32855
+    },
+    {
+      "epoch": 2.232640304389183,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.6045998097567606e-05,
+      "loss": 0.9012,
+      "step": 32860
+    },
+    {
+      "epoch": 2.2329800244598452,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.604387484712597e-05,
+      "loss": 0.9569,
+      "step": 32865
+    },
+    {
+      "epoch": 2.233319744530507,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.6041751596684334e-05,
+      "loss": 0.9691,
+      "step": 32870
+    },
+    {
+      "epoch": 2.2336594646011685,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.60396283462427e-05,
+      "loss": 0.9936,
+      "step": 32875
+    },
+    {
+      "epoch": 2.2339991846718306,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.603750509580106e-05,
+      "loss": 0.932,
+      "step": 32880
+    },
+    {
+      "epoch": 2.234338904742492,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.6035381845359426e-05,
+      "loss": 0.9622,
+      "step": 32885
+    },
+    {
+      "epoch": 2.234678624813154,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.603325859491779e-05,
+      "loss": 0.9591,
+      "step": 32890
+    },
+    {
+      "epoch": 2.235018344883816,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.6031135344476154e-05,
+      "loss": 0.9898,
+      "step": 32895
+    },
+    {
+      "epoch": 2.2353580649544775,
+      "grad_norm": 6.25,
+      "learning_rate": 3.602901209403452e-05,
+      "loss": 0.9901,
+      "step": 32900
+    },
+    {
+      "epoch": 2.235697785025139,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.602688884359288e-05,
+      "loss": 0.9075,
+      "step": 32905
+    },
+    {
+      "epoch": 2.2360375050958012,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.6024765593151246e-05,
+      "loss": 0.9921,
+      "step": 32910
+    },
+    {
+      "epoch": 2.236377225166463,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.602264234270961e-05,
+      "loss": 0.9802,
+      "step": 32915
+    },
+    {
+      "epoch": 2.2367169452371245,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.6020519092267974e-05,
+      "loss": 1.0624,
+      "step": 32920
+    },
+    {
+      "epoch": 2.2370566653077866,
+      "grad_norm": 6.0,
+      "learning_rate": 3.601839584182634e-05,
+      "loss": 1.0279,
+      "step": 32925
+    },
+    {
+      "epoch": 2.237396385378448,
+      "grad_norm": 7.0,
+      "learning_rate": 3.60162725913847e-05,
+      "loss": 1.006,
+      "step": 32930
+    },
+    {
+      "epoch": 2.23773610544911,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.6014149340943066e-05,
+      "loss": 0.8937,
+      "step": 32935
+    },
+    {
+      "epoch": 2.238075825519772,
+      "grad_norm": 6.375,
+      "learning_rate": 3.601202609050143e-05,
+      "loss": 1.0103,
+      "step": 32940
+    },
+    {
+      "epoch": 2.2384155455904335,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.6009902840059794e-05,
+      "loss": 0.8995,
+      "step": 32945
+    },
+    {
+      "epoch": 2.238755265661095,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.600777958961816e-05,
+      "loss": 1.0269,
+      "step": 32950
+    },
+    {
+      "epoch": 2.2390949857317572,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.600565633917652e-05,
+      "loss": 0.9412,
+      "step": 32955
+    },
+    {
+      "epoch": 2.239434705802419,
+      "grad_norm": 6.0,
+      "learning_rate": 3.6003533088734886e-05,
+      "loss": 0.9449,
+      "step": 32960
+    },
+    {
+      "epoch": 2.2397744258730805,
+      "grad_norm": 6.25,
+      "learning_rate": 3.600140983829325e-05,
+      "loss": 0.939,
+      "step": 32965
+    },
+    {
+      "epoch": 2.2401141459437426,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.5999286587851614e-05,
+      "loss": 0.9904,
+      "step": 32970
+    },
+    {
+      "epoch": 2.240453866014404,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.599716333740998e-05,
+      "loss": 0.997,
+      "step": 32975
+    },
+    {
+      "epoch": 2.240793586085066,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.599504008696834e-05,
+      "loss": 1.053,
+      "step": 32980
+    },
+    {
+      "epoch": 2.2411333061557275,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.59929168365267e-05,
+      "loss": 0.9309,
+      "step": 32985
+    },
+    {
+      "epoch": 2.2414730262263896,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.599079358608507e-05,
+      "loss": 0.9195,
+      "step": 32990
+    },
+    {
+      "epoch": 2.241812746297051,
+      "grad_norm": 7.125,
+      "learning_rate": 3.5988670335643434e-05,
+      "loss": 0.9356,
+      "step": 32995
+    },
+    {
+      "epoch": 2.242152466367713,
+      "grad_norm": 6.5,
+      "learning_rate": 3.598654708520179e-05,
+      "loss": 0.9705,
+      "step": 33000
+    },
+    {
+      "epoch": 2.242492186438375,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.598442383476016e-05,
+      "loss": 1.0023,
+      "step": 33005
+    },
+    {
+      "epoch": 2.2428319065090365,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.5982300584318526e-05,
+      "loss": 0.961,
+      "step": 33010
+    },
+    {
+      "epoch": 2.243171626579698,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.598017733387688e-05,
+      "loss": 1.0406,
+      "step": 33015
+    },
+    {
+      "epoch": 2.2435113466503602,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.5978054083435254e-05,
+      "loss": 0.9308,
+      "step": 33020
+    },
+    {
+      "epoch": 2.243851066721022,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.597593083299362e-05,
+      "loss": 0.945,
+      "step": 33025
+    },
+    {
+      "epoch": 2.2441907867916835,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5973807582551975e-05,
+      "loss": 0.8763,
+      "step": 33030
+    },
+    {
+      "epoch": 2.2445305068623456,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.5971684332110346e-05,
+      "loss": 0.9093,
+      "step": 33035
+    },
+    {
+      "epoch": 2.244870226933007,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.596956108166871e-05,
+      "loss": 0.9676,
+      "step": 33040
+    },
+    {
+      "epoch": 2.245209947003669,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.596743783122707e-05,
+      "loss": 1.0168,
+      "step": 33045
+    },
+    {
+      "epoch": 2.245549667074331,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.596531458078544e-05,
+      "loss": 1.0536,
+      "step": 33050
+    },
+    {
+      "epoch": 2.2458893871449925,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.59631913303438e-05,
+      "loss": 0.9868,
+      "step": 33055
+    },
+    {
+      "epoch": 2.246229107215654,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.596106807990216e-05,
+      "loss": 0.8604,
+      "step": 33060
+    },
+    {
+      "epoch": 2.2465688272863162,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.595894482946053e-05,
+      "loss": 0.932,
+      "step": 33065
+    },
+    {
+      "epoch": 2.246908547356978,
+      "grad_norm": 7.25,
+      "learning_rate": 3.595682157901889e-05,
+      "loss": 1.0851,
+      "step": 33070
+    },
+    {
+      "epoch": 2.2472482674276395,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.595469832857725e-05,
+      "loss": 0.983,
+      "step": 33075
+    },
+    {
+      "epoch": 2.2475879874983016,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.595257507813562e-05,
+      "loss": 1.0063,
+      "step": 33080
+    },
+    {
+      "epoch": 2.247927707568963,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.595045182769398e-05,
+      "loss": 0.8983,
+      "step": 33085
+    },
+    {
+      "epoch": 2.248267427639625,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.594832857725234e-05,
+      "loss": 1.0131,
+      "step": 33090
+    },
+    {
+      "epoch": 2.248607147710287,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.5946205326810714e-05,
+      "loss": 0.9652,
+      "step": 33095
+    },
+    {
+      "epoch": 2.2489468677809485,
+      "grad_norm": 6.5,
+      "learning_rate": 3.594408207636907e-05,
+      "loss": 0.957,
+      "step": 33100
+    },
+    {
+      "epoch": 2.24928658785161,
+      "grad_norm": 6.25,
+      "learning_rate": 3.5941958825927435e-05,
+      "loss": 0.963,
+      "step": 33105
+    },
+    {
+      "epoch": 2.2496263079222723,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.5939835575485806e-05,
+      "loss": 0.9379,
+      "step": 33110
+    },
+    {
+      "epoch": 2.249966027992934,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.593771232504416e-05,
+      "loss": 1.017,
+      "step": 33115
+    },
+    {
+      "epoch": 2.2503057480635955,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.593558907460253e-05,
+      "loss": 0.9494,
+      "step": 33120
+    },
+    {
+      "epoch": 2.250645468134257,
+      "grad_norm": 6.25,
+      "learning_rate": 3.59334658241609e-05,
+      "loss": 0.959,
+      "step": 33125
+    },
+    {
+      "epoch": 2.250985188204919,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5931342573719255e-05,
+      "loss": 0.9541,
+      "step": 33130
+    },
+    {
+      "epoch": 2.251324908275581,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.592921932327762e-05,
+      "loss": 0.9102,
+      "step": 33135
+    },
+    {
+      "epoch": 2.2516646283462425,
+      "grad_norm": 7.375,
+      "learning_rate": 3.592709607283598e-05,
+      "loss": 1.0386,
+      "step": 33140
+    },
+    {
+      "epoch": 2.2520043484169046,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.592497282239435e-05,
+      "loss": 1.0746,
+      "step": 33145
+    },
+    {
+      "epoch": 2.252344068487566,
+      "grad_norm": 8.0,
+      "learning_rate": 3.592284957195271e-05,
+      "loss": 0.8862,
+      "step": 33150
+    },
+    {
+      "epoch": 2.252683788558228,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.5920726321511075e-05,
+      "loss": 1.0749,
+      "step": 33155
+    },
+    {
+      "epoch": 2.25302350862889,
+      "grad_norm": 6.875,
+      "learning_rate": 3.591860307106944e-05,
+      "loss": 1.0066,
+      "step": 33160
+    },
+    {
+      "epoch": 2.2533632286995515,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.59164798206278e-05,
+      "loss": 0.8696,
+      "step": 33165
+    },
+    {
+      "epoch": 2.253702948770213,
+      "grad_norm": 6.125,
+      "learning_rate": 3.591435657018617e-05,
+      "loss": 0.9403,
+      "step": 33170
+    },
+    {
+      "epoch": 2.2540426688408752,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.591223331974453e-05,
+      "loss": 0.9609,
+      "step": 33175
+    },
+    {
+      "epoch": 2.254382388911537,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5910110069302895e-05,
+      "loss": 1.0285,
+      "step": 33180
+    },
+    {
+      "epoch": 2.2547221089821985,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.590798681886126e-05,
+      "loss": 1.0109,
+      "step": 33185
+    },
+    {
+      "epoch": 2.2550618290528606,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.590586356841962e-05,
+      "loss": 1.0201,
+      "step": 33190
+    },
+    {
+      "epoch": 2.255401549123522,
+      "grad_norm": 6.625,
+      "learning_rate": 3.590374031797799e-05,
+      "loss": 1.0889,
+      "step": 33195
+    },
+    {
+      "epoch": 2.255741269194184,
+      "grad_norm": 6.875,
+      "learning_rate": 3.590161706753635e-05,
+      "loss": 0.9255,
+      "step": 33200
+    },
+    {
+      "epoch": 2.256080989264846,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.5899493817094715e-05,
+      "loss": 0.9625,
+      "step": 33205
+    },
+    {
+      "epoch": 2.2564207093355075,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.589737056665308e-05,
+      "loss": 0.9489,
+      "step": 33210
+    },
+    {
+      "epoch": 2.256760429406169,
+      "grad_norm": 6.625,
+      "learning_rate": 3.589524731621144e-05,
+      "loss": 0.9838,
+      "step": 33215
+    },
+    {
+      "epoch": 2.2571001494768312,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.589312406576981e-05,
+      "loss": 1.0346,
+      "step": 33220
+    },
+    {
+      "epoch": 2.257439869547493,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.589100081532817e-05,
+      "loss": 1.0572,
+      "step": 33225
+    },
+    {
+      "epoch": 2.2577795896181545,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.5888877564886535e-05,
+      "loss": 1.0184,
+      "step": 33230
+    },
+    {
+      "epoch": 2.2581193096888166,
+      "grad_norm": 6.25,
+      "learning_rate": 3.58867543144449e-05,
+      "loss": 0.9906,
+      "step": 33235
+    },
+    {
+      "epoch": 2.258459029759478,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.588463106400326e-05,
+      "loss": 0.933,
+      "step": 33240
+    },
+    {
+      "epoch": 2.25879874983014,
+      "grad_norm": 7.0,
+      "learning_rate": 3.588250781356163e-05,
+      "loss": 0.9853,
+      "step": 33245
+    },
+    {
+      "epoch": 2.259138469900802,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.588038456311999e-05,
+      "loss": 1.0491,
+      "step": 33250
+    },
+    {
+      "epoch": 2.2594781899714635,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.5878261312678355e-05,
+      "loss": 0.9168,
+      "step": 33255
+    },
+    {
+      "epoch": 2.259817910042125,
+      "grad_norm": 6.625,
+      "learning_rate": 3.587613806223672e-05,
+      "loss": 1.0668,
+      "step": 33260
+    },
+    {
+      "epoch": 2.2601576301127873,
+      "grad_norm": 7.875,
+      "learning_rate": 3.587401481179508e-05,
+      "loss": 1.0067,
+      "step": 33265
+    },
+    {
+      "epoch": 2.260497350183449,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.587189156135345e-05,
+      "loss": 0.9467,
+      "step": 33270
+    },
+    {
+      "epoch": 2.2608370702541105,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.586976831091181e-05,
+      "loss": 0.9892,
+      "step": 33275
+    },
+    {
+      "epoch": 2.2611767903247726,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.5867645060470175e-05,
+      "loss": 0.9364,
+      "step": 33280
+    },
+    {
+      "epoch": 2.261516510395434,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.586552181002854e-05,
+      "loss": 1.0057,
+      "step": 33285
+    },
+    {
+      "epoch": 2.261856230466096,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.58633985595869e-05,
+      "loss": 0.904,
+      "step": 33290
+    },
+    {
+      "epoch": 2.262195950536758,
+      "grad_norm": 7.25,
+      "learning_rate": 3.586127530914527e-05,
+      "loss": 0.951,
+      "step": 33295
+    },
+    {
+      "epoch": 2.2625356706074196,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.585915205870363e-05,
+      "loss": 0.922,
+      "step": 33300
+    },
+    {
+      "epoch": 2.262875390678081,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.5857028808261995e-05,
+      "loss": 0.9676,
+      "step": 33305
+    },
+    {
+      "epoch": 2.2632151107487433,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.585490555782036e-05,
+      "loss": 0.8892,
+      "step": 33310
+    },
+    {
+      "epoch": 2.263554830819405,
+      "grad_norm": 8.25,
+      "learning_rate": 3.585278230737872e-05,
+      "loss": 0.9629,
+      "step": 33315
+    },
+    {
+      "epoch": 2.2638945508900665,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.585065905693709e-05,
+      "loss": 0.9542,
+      "step": 33320
+    },
+    {
+      "epoch": 2.2642342709607286,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.584853580649545e-05,
+      "loss": 0.93,
+      "step": 33325
+    },
+    {
+      "epoch": 2.2645739910313902,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.5846412556053815e-05,
+      "loss": 0.9309,
+      "step": 33330
+    },
+    {
+      "epoch": 2.264913711102052,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.584428930561218e-05,
+      "loss": 0.9165,
+      "step": 33335
+    },
+    {
+      "epoch": 2.2652534311727135,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.5842166055170536e-05,
+      "loss": 0.9388,
+      "step": 33340
+    },
+    {
+      "epoch": 2.2655931512433756,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.584004280472891e-05,
+      "loss": 0.9207,
+      "step": 33345
+    },
+    {
+      "epoch": 2.265932871314037,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.583791955428727e-05,
+      "loss": 0.9496,
+      "step": 33350
+    },
+    {
+      "epoch": 2.266272591384699,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.583579630384563e-05,
+      "loss": 1.0063,
+      "step": 33355
+    },
+    {
+      "epoch": 2.266612311455361,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.5833673053404e-05,
+      "loss": 1.0936,
+      "step": 33360
+    },
+    {
+      "epoch": 2.2669520315260225,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.583154980296236e-05,
+      "loss": 0.9853,
+      "step": 33365
+    },
+    {
+      "epoch": 2.267291751596684,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.582942655252072e-05,
+      "loss": 0.9297,
+      "step": 33370
+    },
+    {
+      "epoch": 2.2676314716673462,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.582730330207909e-05,
+      "loss": 1.0302,
+      "step": 33375
+    },
+    {
+      "epoch": 2.267971191738008,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5825180051637455e-05,
+      "loss": 0.939,
+      "step": 33380
+    },
+    {
+      "epoch": 2.2683109118086695,
+      "grad_norm": 6.625,
+      "learning_rate": 3.582305680119581e-05,
+      "loss": 0.9298,
+      "step": 33385
+    },
+    {
+      "epoch": 2.2686506318793316,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.582093355075418e-05,
+      "loss": 0.9884,
+      "step": 33390
+    },
+    {
+      "epoch": 2.268990351949993,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.581881030031255e-05,
+      "loss": 1.295,
+      "step": 33395
+    },
+    {
+      "epoch": 2.269330072020655,
+      "grad_norm": 6.875,
+      "learning_rate": 3.5816687049870904e-05,
+      "loss": 1.0054,
+      "step": 33400
+    },
+    {
+      "epoch": 2.269669792091317,
+      "grad_norm": 7.375,
+      "learning_rate": 3.5814563799429275e-05,
+      "loss": 0.9449,
+      "step": 33405
+    },
+    {
+      "epoch": 2.2700095121619785,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.581244054898763e-05,
+      "loss": 0.814,
+      "step": 33410
+    },
+    {
+      "epoch": 2.27034923223264,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.5810317298545996e-05,
+      "loss": 0.8992,
+      "step": 33415
+    },
+    {
+      "epoch": 2.2706889523033023,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.580819404810437e-05,
+      "loss": 0.9626,
+      "step": 33420
+    },
+    {
+      "epoch": 2.271028672373964,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.5806070797662724e-05,
+      "loss": 0.9956,
+      "step": 33425
+    },
+    {
+      "epoch": 2.2713683924446255,
+      "grad_norm": 5.75,
+      "learning_rate": 3.580394754722109e-05,
+      "loss": 1.0649,
+      "step": 33430
+    },
+    {
+      "epoch": 2.2717081125152876,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.580182429677946e-05,
+      "loss": 0.906,
+      "step": 33435
+    },
+    {
+      "epoch": 2.2720478325859492,
+      "grad_norm": 6.25,
+      "learning_rate": 3.5799701046337816e-05,
+      "loss": 0.9537,
+      "step": 33440
+    },
+    {
+      "epoch": 2.272387552656611,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.579757779589618e-05,
+      "loss": 0.9859,
+      "step": 33445
+    },
+    {
+      "epoch": 2.2727272727272725,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.579545454545455e-05,
+      "loss": 0.9218,
+      "step": 33450
+    },
+    {
+      "epoch": 2.2730669927979346,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.579333129501291e-05,
+      "loss": 0.9584,
+      "step": 33455
+    },
+    {
+      "epoch": 2.273406712868596,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.579120804457127e-05,
+      "loss": 0.968,
+      "step": 33460
+    },
+    {
+      "epoch": 2.273746432939258,
+      "grad_norm": 6.75,
+      "learning_rate": 3.578908479412964e-05,
+      "loss": 0.9792,
+      "step": 33465
+    },
+    {
+      "epoch": 2.27408615300992,
+      "grad_norm": 7.25,
+      "learning_rate": 3.5786961543688e-05,
+      "loss": 1.0024,
+      "step": 33470
+    },
+    {
+      "epoch": 2.2744258730805815,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.5784838293246364e-05,
+      "loss": 0.9476,
+      "step": 33475
+    },
+    {
+      "epoch": 2.274765593151243,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.5782715042804735e-05,
+      "loss": 0.9734,
+      "step": 33480
+    },
+    {
+      "epoch": 2.2751053132219052,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.578059179236309e-05,
+      "loss": 0.9416,
+      "step": 33485
+    },
+    {
+      "epoch": 2.275445033292567,
+      "grad_norm": 6.25,
+      "learning_rate": 3.5778468541921457e-05,
+      "loss": 0.9703,
+      "step": 33490
+    },
+    {
+      "epoch": 2.2757847533632285,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.577634529147982e-05,
+      "loss": 1.0834,
+      "step": 33495
+    },
+    {
+      "epoch": 2.2761244734338906,
+      "grad_norm": 7.0,
+      "learning_rate": 3.5774222041038185e-05,
+      "loss": 0.9802,
+      "step": 33500
+    },
+    {
+      "epoch": 2.276464193504552,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.577209879059655e-05,
+      "loss": 1.024,
+      "step": 33505
+    },
+    {
+      "epoch": 2.276803913575214,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.576997554015491e-05,
+      "loss": 0.9321,
+      "step": 33510
+    },
+    {
+      "epoch": 2.277143633645876,
+      "grad_norm": 6.625,
+      "learning_rate": 3.5767852289713277e-05,
+      "loss": 1.0036,
+      "step": 33515
+    },
+    {
+      "epoch": 2.2774833537165375,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.576572903927164e-05,
+      "loss": 1.044,
+      "step": 33520
+    },
+    {
+      "epoch": 2.277823073787199,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.5763605788830005e-05,
+      "loss": 0.9644,
+      "step": 33525
+    },
+    {
+      "epoch": 2.2781627938578612,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.576148253838837e-05,
+      "loss": 1.0274,
+      "step": 33530
+    },
+    {
+      "epoch": 2.278502513928523,
+      "grad_norm": 6.625,
+      "learning_rate": 3.575935928794673e-05,
+      "loss": 0.92,
+      "step": 33535
+    },
+    {
+      "epoch": 2.2788422339991845,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.5757236037505097e-05,
+      "loss": 0.9576,
+      "step": 33540
+    },
+    {
+      "epoch": 2.2791819540698466,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.575511278706346e-05,
+      "loss": 0.8834,
+      "step": 33545
+    },
+    {
+      "epoch": 2.279521674140508,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.5752989536621825e-05,
+      "loss": 0.8255,
+      "step": 33550
+    },
+    {
+      "epoch": 2.27986139421117,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.575086628618019e-05,
+      "loss": 0.9847,
+      "step": 33555
+    },
+    {
+      "epoch": 2.280201114281832,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.574874303573855e-05,
+      "loss": 0.9422,
+      "step": 33560
+    },
+    {
+      "epoch": 2.2805408343524936,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5746619785296917e-05,
+      "loss": 0.9745,
+      "step": 33565
+    },
+    {
+      "epoch": 2.280880554423155,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.574449653485528e-05,
+      "loss": 1.0044,
+      "step": 33570
+    },
+    {
+      "epoch": 2.2812202744938173,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.5742373284413645e-05,
+      "loss": 0.8977,
+      "step": 33575
+    },
+    {
+      "epoch": 2.281559994564479,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.574025003397201e-05,
+      "loss": 0.9419,
+      "step": 33580
+    },
+    {
+      "epoch": 2.2818997146351405,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.573812678353037e-05,
+      "loss": 0.9527,
+      "step": 33585
+    },
+    {
+      "epoch": 2.2822394347058026,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5736003533088737e-05,
+      "loss": 0.9384,
+      "step": 33590
+    },
+    {
+      "epoch": 2.2825791547764642,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.57338802826471e-05,
+      "loss": 1.0118,
+      "step": 33595
+    },
+    {
+      "epoch": 2.282918874847126,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.5731757032205465e-05,
+      "loss": 1.0094,
+      "step": 33600
+    },
+    {
+      "epoch": 2.283258594917788,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.572963378176383e-05,
+      "loss": 0.8649,
+      "step": 33605
+    },
+    {
+      "epoch": 2.2835983149884496,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.572751053132219e-05,
+      "loss": 0.9405,
+      "step": 33610
+    },
+    {
+      "epoch": 2.283938035059111,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.5725387280880557e-05,
+      "loss": 0.9913,
+      "step": 33615
+    },
+    {
+      "epoch": 2.2842777551297733,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.572326403043892e-05,
+      "loss": 1.0199,
+      "step": 33620
+    },
+    {
+      "epoch": 2.284617475200435,
+      "grad_norm": 6.75,
+      "learning_rate": 3.5721140779997285e-05,
+      "loss": 1.1089,
+      "step": 33625
+    },
+    {
+      "epoch": 2.2849571952710965,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.571901752955565e-05,
+      "loss": 0.9485,
+      "step": 33630
+    },
+    {
+      "epoch": 2.2852969153417586,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.571689427911401e-05,
+      "loss": 0.9779,
+      "step": 33635
+    },
+    {
+      "epoch": 2.2856366354124202,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.571477102867238e-05,
+      "loss": 1.0612,
+      "step": 33640
+    },
+    {
+      "epoch": 2.285976355483082,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.571264777823074e-05,
+      "loss": 1.007,
+      "step": 33645
+    },
+    {
+      "epoch": 2.286316075553744,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.5710524527789105e-05,
+      "loss": 0.93,
+      "step": 33650
+    },
+    {
+      "epoch": 2.2866557956244056,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.570840127734747e-05,
+      "loss": 0.9292,
+      "step": 33655
+    },
+    {
+      "epoch": 2.286995515695067,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.570627802690583e-05,
+      "loss": 1.0022,
+      "step": 33660
+    },
+    {
+      "epoch": 2.2873352357657293,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.57041547764642e-05,
+      "loss": 1.0248,
+      "step": 33665
+    },
+    {
+      "epoch": 2.287674955836391,
+      "grad_norm": 6.125,
+      "learning_rate": 3.570203152602256e-05,
+      "loss": 0.9547,
+      "step": 33670
+    },
+    {
+      "epoch": 2.2880146759070525,
+      "grad_norm": 6.75,
+      "learning_rate": 3.5699908275580925e-05,
+      "loss": 1.0051,
+      "step": 33675
+    },
+    {
+      "epoch": 2.288354395977714,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.569778502513929e-05,
+      "loss": 0.8958,
+      "step": 33680
+    },
+    {
+      "epoch": 2.2886941160483762,
+      "grad_norm": 6.125,
+      "learning_rate": 3.569566177469765e-05,
+      "loss": 1.0096,
+      "step": 33685
+    },
+    {
+      "epoch": 2.289033836119038,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.569353852425602e-05,
+      "loss": 0.9926,
+      "step": 33690
+    },
+    {
+      "epoch": 2.2893735561896995,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.5691415273814374e-05,
+      "loss": 1.0315,
+      "step": 33695
+    },
+    {
+      "epoch": 2.2897132762603616,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.5689292023372745e-05,
+      "loss": 0.9959,
+      "step": 33700
+    },
+    {
+      "epoch": 2.290052996331023,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.568716877293111e-05,
+      "loss": 1.0023,
+      "step": 33705
+    },
+    {
+      "epoch": 2.290392716401685,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.5685045522489466e-05,
+      "loss": 0.8968,
+      "step": 33710
+    },
+    {
+      "epoch": 2.290732436472347,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.568292227204784e-05,
+      "loss": 0.9287,
+      "step": 33715
+    },
+    {
+      "epoch": 2.2910721565430086,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.56807990216062e-05,
+      "loss": 0.984,
+      "step": 33720
+    },
+    {
+      "epoch": 2.29141187661367,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.567867577116456e-05,
+      "loss": 0.9803,
+      "step": 33725
+    },
+    {
+      "epoch": 2.2917515966843323,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.567655252072293e-05,
+      "loss": 1.0339,
+      "step": 33730
+    },
+    {
+      "epoch": 2.292091316754994,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.567442927028129e-05,
+      "loss": 1.0202,
+      "step": 33735
+    },
+    {
+      "epoch": 2.2924310368256555,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.567230601983965e-05,
+      "loss": 0.9851,
+      "step": 33740
+    },
+    {
+      "epoch": 2.2927707568963176,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.567018276939802e-05,
+      "loss": 0.9674,
+      "step": 33745
+    },
+    {
+      "epoch": 2.2931104769669792,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5668059518956385e-05,
+      "loss": 1.0063,
+      "step": 33750
+    },
+    {
+      "epoch": 2.293450197037641,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.566593626851474e-05,
+      "loss": 1.0202,
+      "step": 33755
+    },
+    {
+      "epoch": 2.293789917108303,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.566381301807311e-05,
+      "loss": 0.9507,
+      "step": 33760
+    },
+    {
+      "epoch": 2.2941296371789646,
+      "grad_norm": 7.0,
+      "learning_rate": 3.566168976763147e-05,
+      "loss": 0.8948,
+      "step": 33765
+    },
+    {
+      "epoch": 2.294469357249626,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.5659566517189834e-05,
+      "loss": 0.9392,
+      "step": 33770
+    },
+    {
+      "epoch": 2.2948090773202883,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.5657443266748205e-05,
+      "loss": 0.9396,
+      "step": 33775
+    },
+    {
+      "epoch": 2.29514879739095,
+      "grad_norm": 7.125,
+      "learning_rate": 3.565532001630656e-05,
+      "loss": 0.9953,
+      "step": 33780
+    },
+    {
+      "epoch": 2.2954885174616115,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.5653196765864926e-05,
+      "loss": 0.8966,
+      "step": 33785
+    },
+    {
+      "epoch": 2.295828237532273,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.56510735154233e-05,
+      "loss": 0.9339,
+      "step": 33790
+    },
+    {
+      "epoch": 2.2961679576029352,
+      "grad_norm": 6.875,
+      "learning_rate": 3.5648950264981654e-05,
+      "loss": 0.9018,
+      "step": 33795
+    },
+    {
+      "epoch": 2.296507677673597,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.564682701454002e-05,
+      "loss": 0.9264,
+      "step": 33800
+    },
+    {
+      "epoch": 2.2968473977442585,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.564470376409839e-05,
+      "loss": 0.9312,
+      "step": 33805
+    },
+    {
+      "epoch": 2.2971871178149206,
+      "grad_norm": 7.0,
+      "learning_rate": 3.5642580513656746e-05,
+      "loss": 0.9379,
+      "step": 33810
+    },
+    {
+      "epoch": 2.297526837885582,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.564045726321511e-05,
+      "loss": 1.0702,
+      "step": 33815
+    },
+    {
+      "epoch": 2.297866557956244,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.563833401277348e-05,
+      "loss": 1.0057,
+      "step": 33820
+    },
+    {
+      "epoch": 2.298206278026906,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.563621076233184e-05,
+      "loss": 0.8562,
+      "step": 33825
+    },
+    {
+      "epoch": 2.2985459980975675,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.56340875118902e-05,
+      "loss": 0.849,
+      "step": 33830
+    },
+    {
+      "epoch": 2.298885718168229,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.5631964261448566e-05,
+      "loss": 0.9192,
+      "step": 33835
+    },
+    {
+      "epoch": 2.2992254382388913,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.562984101100693e-05,
+      "loss": 1.0067,
+      "step": 33840
+    },
+    {
+      "epoch": 2.299565158309553,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.5627717760565294e-05,
+      "loss": 0.9687,
+      "step": 33845
+    },
+    {
+      "epoch": 2.2999048783802145,
+      "grad_norm": 6.875,
+      "learning_rate": 3.562559451012366e-05,
+      "loss": 0.9737,
+      "step": 33850
+    },
+    {
+      "epoch": 2.3002445984508766,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.562347125968202e-05,
+      "loss": 1.0005,
+      "step": 33855
+    },
+    {
+      "epoch": 2.300584318521538,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.5621348009240386e-05,
+      "loss": 0.8911,
+      "step": 33860
+    },
+    {
+      "epoch": 2.3009240385922,
+      "grad_norm": 6.875,
+      "learning_rate": 3.561922475879875e-05,
+      "loss": 1.0891,
+      "step": 33865
+    },
+    {
+      "epoch": 2.301263758662862,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.5617101508357114e-05,
+      "loss": 0.9093,
+      "step": 33870
+    },
+    {
+      "epoch": 2.3016034787335236,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.561497825791548e-05,
+      "loss": 0.8846,
+      "step": 33875
+    },
+    {
+      "epoch": 2.301943198804185,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.561285500747384e-05,
+      "loss": 0.9269,
+      "step": 33880
+    },
+    {
+      "epoch": 2.3022829188748473,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.5610731757032206e-05,
+      "loss": 1.0576,
+      "step": 33885
+    },
+    {
+      "epoch": 2.302622638945509,
+      "grad_norm": 6.625,
+      "learning_rate": 3.560860850659057e-05,
+      "loss": 1.023,
+      "step": 33890
+    },
+    {
+      "epoch": 2.3029623590161705,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.5606485256148934e-05,
+      "loss": 1.0523,
+      "step": 33895
+    },
+    {
+      "epoch": 2.3033020790868326,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.56043620057073e-05,
+      "loss": 0.9275,
+      "step": 33900
+    },
+    {
+      "epoch": 2.3036417991574942,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.560223875526567e-05,
+      "loss": 0.9467,
+      "step": 33905
+    },
+    {
+      "epoch": 2.303981519228156,
+      "grad_norm": 6.25,
+      "learning_rate": 3.5600115504824026e-05,
+      "loss": 0.9016,
+      "step": 33910
+    },
+    {
+      "epoch": 2.304321239298818,
+      "grad_norm": 5.875,
+      "learning_rate": 3.559799225438239e-05,
+      "loss": 0.9775,
+      "step": 33915
+    },
+    {
+      "epoch": 2.3046609593694796,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.5595869003940754e-05,
+      "loss": 0.9307,
+      "step": 33920
+    },
+    {
+      "epoch": 2.305000679440141,
+      "grad_norm": 6.625,
+      "learning_rate": 3.559374575349912e-05,
+      "loss": 0.9295,
+      "step": 33925
+    },
+    {
+      "epoch": 2.3053403995108033,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.559162250305748e-05,
+      "loss": 1.0019,
+      "step": 33930
+    },
+    {
+      "epoch": 2.305680119581465,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.5589499252615846e-05,
+      "loss": 0.964,
+      "step": 33935
+    },
+    {
+      "epoch": 2.3060198396521265,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.558737600217421e-05,
+      "loss": 0.9202,
+      "step": 33940
+    },
+    {
+      "epoch": 2.3063595597227886,
+      "grad_norm": 7.625,
+      "learning_rate": 3.5585252751732574e-05,
+      "loss": 0.999,
+      "step": 33945
+    },
+    {
+      "epoch": 2.3066992797934502,
+      "grad_norm": 6.5,
+      "learning_rate": 3.558312950129094e-05,
+      "loss": 0.9191,
+      "step": 33950
+    },
+    {
+      "epoch": 2.307038999864112,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.55810062508493e-05,
+      "loss": 0.9088,
+      "step": 33955
+    },
+    {
+      "epoch": 2.307378719934774,
+      "grad_norm": 9.125,
+      "learning_rate": 3.5578883000407666e-05,
+      "loss": 0.9871,
+      "step": 33960
+    },
+    {
+      "epoch": 2.3077184400054356,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.557675974996603e-05,
+      "loss": 0.9514,
+      "step": 33965
+    },
+    {
+      "epoch": 2.308058160076097,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.5574636499524394e-05,
+      "loss": 0.904,
+      "step": 33970
+    },
+    {
+      "epoch": 2.3083978801467593,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.557251324908276e-05,
+      "loss": 0.8887,
+      "step": 33975
+    },
+    {
+      "epoch": 2.308737600217421,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.557038999864112e-05,
+      "loss": 0.8742,
+      "step": 33980
+    },
+    {
+      "epoch": 2.3090773202880825,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.5568266748199486e-05,
+      "loss": 1.0295,
+      "step": 33985
+    },
+    {
+      "epoch": 2.3094170403587446,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.556614349775785e-05,
+      "loss": 0.9267,
+      "step": 33990
+    },
+    {
+      "epoch": 2.3097567604294063,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.5564020247316214e-05,
+      "loss": 0.9096,
+      "step": 33995
+    },
+    {
+      "epoch": 2.310096480500068,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.556189699687458e-05,
+      "loss": 0.8853,
+      "step": 34000
+    },
+    {
+      "epoch": 2.31043620057073,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.555977374643294e-05,
+      "loss": 1.0959,
+      "step": 34005
+    },
+    {
+      "epoch": 2.3107759206413916,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5557650495991306e-05,
+      "loss": 0.9945,
+      "step": 34010
+    },
+    {
+      "epoch": 2.311115640712053,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.555552724554967e-05,
+      "loss": 1.0031,
+      "step": 34015
+    },
+    {
+      "epoch": 2.311455360782715,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5553403995108034e-05,
+      "loss": 0.9584,
+      "step": 34020
+    },
+    {
+      "epoch": 2.311795080853377,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.55512807446664e-05,
+      "loss": 0.9744,
+      "step": 34025
+    },
+    {
+      "epoch": 2.3121348009240386,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.554915749422476e-05,
+      "loss": 0.8563,
+      "step": 34030
+    },
+    {
+      "epoch": 2.3124745209947,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.554703424378312e-05,
+      "loss": 1.0094,
+      "step": 34035
+    },
+    {
+      "epoch": 2.3128142410653623,
+      "grad_norm": 7.375,
+      "learning_rate": 3.554491099334149e-05,
+      "loss": 0.9727,
+      "step": 34040
+    },
+    {
+      "epoch": 2.313153961136024,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5542787742899854e-05,
+      "loss": 0.8923,
+      "step": 34045
+    },
+    {
+      "epoch": 2.3134936812066855,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.554066449245821e-05,
+      "loss": 0.9073,
+      "step": 34050
+    },
+    {
+      "epoch": 2.3138334012773476,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.553854124201658e-05,
+      "loss": 0.8975,
+      "step": 34055
+    },
+    {
+      "epoch": 2.3141731213480092,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.5536417991574946e-05,
+      "loss": 0.9705,
+      "step": 34060
+    },
+    {
+      "epoch": 2.314512841418671,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.5534294741133303e-05,
+      "loss": 1.0208,
+      "step": 34065
+    },
+    {
+      "epoch": 2.314852561489333,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.5532171490691674e-05,
+      "loss": 0.913,
+      "step": 34070
+    },
+    {
+      "epoch": 2.3151922815599946,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.553004824025004e-05,
+      "loss": 0.9365,
+      "step": 34075
+    },
+    {
+      "epoch": 2.315532001630656,
+      "grad_norm": 7.25,
+      "learning_rate": 3.5527924989808395e-05,
+      "loss": 0.8439,
+      "step": 34080
+    },
+    {
+      "epoch": 2.3158717217013183,
+      "grad_norm": 7.0,
+      "learning_rate": 3.5525801739366766e-05,
+      "loss": 0.9364,
+      "step": 34085
+    },
+    {
+      "epoch": 2.31621144177198,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.552367848892513e-05,
+      "loss": 0.9028,
+      "step": 34090
+    },
+    {
+      "epoch": 2.3165511618426415,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.552155523848349e-05,
+      "loss": 1.0524,
+      "step": 34095
+    },
+    {
+      "epoch": 2.3168908819133036,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.551943198804186e-05,
+      "loss": 0.9234,
+      "step": 34100
+    },
+    {
+      "epoch": 2.3172306019839652,
+      "grad_norm": 5.75,
+      "learning_rate": 3.551730873760022e-05,
+      "loss": 0.8853,
+      "step": 34105
+    },
+    {
+      "epoch": 2.317570322054627,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.551518548715858e-05,
+      "loss": 0.9759,
+      "step": 34110
+    },
+    {
+      "epoch": 2.317910042125289,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.551306223671695e-05,
+      "loss": 0.9916,
+      "step": 34115
+    },
+    {
+      "epoch": 2.3182497621959506,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.551093898627531e-05,
+      "loss": 1.0399,
+      "step": 34120
+    },
+    {
+      "epoch": 2.318589482266612,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.550881573583367e-05,
+      "loss": 1.012,
+      "step": 34125
+    },
+    {
+      "epoch": 2.318929202337274,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.550669248539204e-05,
+      "loss": 0.9592,
+      "step": 34130
+    },
+    {
+      "epoch": 2.319268922407936,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.55045692349504e-05,
+      "loss": 0.995,
+      "step": 34135
+    },
+    {
+      "epoch": 2.3196086424785975,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5502445984508763e-05,
+      "loss": 1.0507,
+      "step": 34140
+    },
+    {
+      "epoch": 2.319948362549259,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5500322734067134e-05,
+      "loss": 0.9034,
+      "step": 34145
+    },
+    {
+      "epoch": 2.3202880826199213,
+      "grad_norm": 6.5,
+      "learning_rate": 3.549819948362549e-05,
+      "loss": 0.99,
+      "step": 34150
+    },
+    {
+      "epoch": 2.320627802690583,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.5496076233183855e-05,
+      "loss": 0.9795,
+      "step": 34155
+    },
+    {
+      "epoch": 2.3209675227612445,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5493952982742226e-05,
+      "loss": 0.9733,
+      "step": 34160
+    },
+    {
+      "epoch": 2.3213072428319066,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.5491829732300583e-05,
+      "loss": 1.017,
+      "step": 34165
+    },
+    {
+      "epoch": 2.3216469629025682,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.548970648185895e-05,
+      "loss": 0.9357,
+      "step": 34170
+    },
+    {
+      "epoch": 2.32198668297323,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.548758323141732e-05,
+      "loss": 0.9902,
+      "step": 34175
+    },
+    {
+      "epoch": 2.322326403043892,
+      "grad_norm": 6.625,
+      "learning_rate": 3.5485459980975675e-05,
+      "loss": 0.8817,
+      "step": 34180
+    },
+    {
+      "epoch": 2.3226661231145536,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.548333673053404e-05,
+      "loss": 0.8688,
+      "step": 34185
+    },
+    {
+      "epoch": 2.323005843185215,
+      "grad_norm": 6.875,
+      "learning_rate": 3.5481213480092403e-05,
+      "loss": 0.9835,
+      "step": 34190
+    },
+    {
+      "epoch": 2.3233455632558773,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.547909022965077e-05,
+      "loss": 0.9339,
+      "step": 34195
+    },
+    {
+      "epoch": 2.323685283326539,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.547696697920913e-05,
+      "loss": 1.0587,
+      "step": 34200
+    },
+    {
+      "epoch": 2.3240250033972005,
+      "grad_norm": 7.125,
+      "learning_rate": 3.5474843728767496e-05,
+      "loss": 0.9283,
+      "step": 34205
+    },
+    {
+      "epoch": 2.3243647234678626,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.547272047832586e-05,
+      "loss": 1.1255,
+      "step": 34210
+    },
+    {
+      "epoch": 2.3247044435385242,
+      "grad_norm": 6.125,
+      "learning_rate": 3.5470597227884224e-05,
+      "loss": 0.9451,
+      "step": 34215
+    },
+    {
+      "epoch": 2.325044163609186,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.546847397744259e-05,
+      "loss": 1.0571,
+      "step": 34220
+    },
+    {
+      "epoch": 2.325383883679848,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.546635072700095e-05,
+      "loss": 0.9235,
+      "step": 34225
+    },
+    {
+      "epoch": 2.3257236037505096,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.5464227476559316e-05,
+      "loss": 0.9996,
+      "step": 34230
+    },
+    {
+      "epoch": 2.326063323821171,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.546210422611768e-05,
+      "loss": 0.9521,
+      "step": 34235
+    },
+    {
+      "epoch": 2.3264030438918333,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.5459980975676044e-05,
+      "loss": 0.9618,
+      "step": 34240
+    },
+    {
+      "epoch": 2.326742763962495,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5457857725234414e-05,
+      "loss": 1.0171,
+      "step": 34245
+    },
+    {
+      "epoch": 2.3270824840331565,
+      "grad_norm": 6.125,
+      "learning_rate": 3.545573447479277e-05,
+      "loss": 0.9922,
+      "step": 34250
+    },
+    {
+      "epoch": 2.3274222041038186,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.5453611224351136e-05,
+      "loss": 0.9914,
+      "step": 34255
+    },
+    {
+      "epoch": 2.3277619241744802,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5451487973909506e-05,
+      "loss": 0.9829,
+      "step": 34260
+    },
+    {
+      "epoch": 2.328101644245142,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.5449364723467864e-05,
+      "loss": 0.9276,
+      "step": 34265
+    },
+    {
+      "epoch": 2.328441364315804,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.544724147302623e-05,
+      "loss": 0.9298,
+      "step": 34270
+    },
+    {
+      "epoch": 2.3287810843864656,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.544511822258459e-05,
+      "loss": 0.9526,
+      "step": 34275
+    },
+    {
+      "epoch": 2.329120804457127,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.5442994972142956e-05,
+      "loss": 0.996,
+      "step": 34280
+    },
+    {
+      "epoch": 2.3294605245277893,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.544087172170132e-05,
+      "loss": 0.956,
+      "step": 34285
+    },
+    {
+      "epoch": 2.329800244598451,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.5438748471259684e-05,
+      "loss": 0.9455,
+      "step": 34290
+    },
+    {
+      "epoch": 2.3301399646691126,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.543662522081805e-05,
+      "loss": 0.9352,
+      "step": 34295
+    },
+    {
+      "epoch": 2.3304796847397746,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.543450197037641e-05,
+      "loss": 0.9358,
+      "step": 34300
+    },
+    {
+      "epoch": 2.3308194048104363,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5432378719934776e-05,
+      "loss": 0.979,
+      "step": 34305
+    },
+    {
+      "epoch": 2.331159124881098,
+      "grad_norm": 7.0,
+      "learning_rate": 3.543025546949314e-05,
+      "loss": 1.0144,
+      "step": 34310
+    },
+    {
+      "epoch": 2.33149884495176,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.5428132219051504e-05,
+      "loss": 0.9644,
+      "step": 34315
+    },
+    {
+      "epoch": 2.3318385650224216,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.542600896860987e-05,
+      "loss": 0.8841,
+      "step": 34320
+    },
+    {
+      "epoch": 2.3321782850930832,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.542388571816823e-05,
+      "loss": 0.9454,
+      "step": 34325
+    },
+    {
+      "epoch": 2.3325180051637453,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.5421762467726596e-05,
+      "loss": 0.9361,
+      "step": 34330
+    },
+    {
+      "epoch": 2.332857725234407,
+      "grad_norm": 7.0,
+      "learning_rate": 3.541963921728496e-05,
+      "loss": 0.9289,
+      "step": 34335
+    },
+    {
+      "epoch": 2.3331974453050686,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.5417515966843324e-05,
+      "loss": 0.9816,
+      "step": 34340
+    },
+    {
+      "epoch": 2.3335371653757306,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.541539271640169e-05,
+      "loss": 1.0013,
+      "step": 34345
+    },
+    {
+      "epoch": 2.3338768854463923,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.541326946596005e-05,
+      "loss": 0.9588,
+      "step": 34350
+    },
+    {
+      "epoch": 2.334216605517054,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.5411146215518416e-05,
+      "loss": 0.9983,
+      "step": 34355
+    },
+    {
+      "epoch": 2.3345563255877155,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.540902296507678e-05,
+      "loss": 1.004,
+      "step": 34360
+    },
+    {
+      "epoch": 2.3348960456583776,
+      "grad_norm": 7.5,
+      "learning_rate": 3.5406899714635144e-05,
+      "loss": 0.9338,
+      "step": 34365
+    },
+    {
+      "epoch": 2.3352357657290392,
+      "grad_norm": 6.0,
+      "learning_rate": 3.540477646419351e-05,
+      "loss": 0.9301,
+      "step": 34370
+    },
+    {
+      "epoch": 2.335575485799701,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.540265321375187e-05,
+      "loss": 1.0598,
+      "step": 34375
+    },
+    {
+      "epoch": 2.335915205870363,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.5400529963310236e-05,
+      "loss": 1.0187,
+      "step": 34380
+    },
+    {
+      "epoch": 2.3362549259410246,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.53984067128686e-05,
+      "loss": 0.9319,
+      "step": 34385
+    },
+    {
+      "epoch": 2.336594646011686,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.539628346242696e-05,
+      "loss": 0.9049,
+      "step": 34390
+    },
+    {
+      "epoch": 2.3369343660823483,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.539416021198533e-05,
+      "loss": 0.9326,
+      "step": 34395
+    },
+    {
+      "epoch": 2.33727408615301,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.539203696154369e-05,
+      "loss": 1.0218,
+      "step": 34400
+    },
+    {
+      "epoch": 2.3376138062236715,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.538991371110205e-05,
+      "loss": 0.9276,
+      "step": 34405
+    },
+    {
+      "epoch": 2.3379535262943336,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.538779046066042e-05,
+      "loss": 1.0282,
+      "step": 34410
+    },
+    {
+      "epoch": 2.3382932463649952,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.5385667210218784e-05,
+      "loss": 0.9723,
+      "step": 34415
+    },
+    {
+      "epoch": 2.338632966435657,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.538354395977714e-05,
+      "loss": 0.9544,
+      "step": 34420
+    },
+    {
+      "epoch": 2.338972686506319,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.538142070933551e-05,
+      "loss": 0.9726,
+      "step": 34425
+    },
+    {
+      "epoch": 2.3393124065769806,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.5379297458893876e-05,
+      "loss": 0.9703,
+      "step": 34430
+    },
+    {
+      "epoch": 2.339652126647642,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.537717420845223e-05,
+      "loss": 0.9981,
+      "step": 34435
+    },
+    {
+      "epoch": 2.3399918467183043,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5375050958010604e-05,
+      "loss": 0.9235,
+      "step": 34440
+    },
+    {
+      "epoch": 2.340331566788966,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.537292770756897e-05,
+      "loss": 0.9555,
+      "step": 34445
+    },
+    {
+      "epoch": 2.3406712868596276,
+      "grad_norm": 6.75,
+      "learning_rate": 3.5370804457127325e-05,
+      "loss": 1.0366,
+      "step": 34450
+    },
+    {
+      "epoch": 2.3410110069302896,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.5368681206685696e-05,
+      "loss": 1.0367,
+      "step": 34455
+    },
+    {
+      "epoch": 2.3413507270009513,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.536655795624406e-05,
+      "loss": 0.94,
+      "step": 34460
+    },
+    {
+      "epoch": 2.341690447071613,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.536443470580242e-05,
+      "loss": 0.9816,
+      "step": 34465
+    },
+    {
+      "epoch": 2.3420301671422745,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.536231145536079e-05,
+      "loss": 0.9715,
+      "step": 34470
+    },
+    {
+      "epoch": 2.3423698872129366,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.5360188204919145e-05,
+      "loss": 0.9162,
+      "step": 34475
+    },
+    {
+      "epoch": 2.3427096072835982,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.535806495447751e-05,
+      "loss": 1.0169,
+      "step": 34480
+    },
+    {
+      "epoch": 2.34304932735426,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.535594170403588e-05,
+      "loss": 0.974,
+      "step": 34485
+    },
+    {
+      "epoch": 2.343389047424922,
+      "grad_norm": 6.375,
+      "learning_rate": 3.535381845359424e-05,
+      "loss": 0.969,
+      "step": 34490
+    },
+    {
+      "epoch": 2.3437287674955836,
+      "grad_norm": 8.25,
+      "learning_rate": 3.53516952031526e-05,
+      "loss": 0.9591,
+      "step": 34495
+    },
+    {
+      "epoch": 2.344068487566245,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.534957195271097e-05,
+      "loss": 0.8939,
+      "step": 34500
+    },
+    {
+      "epoch": 2.3444082076369073,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.534744870226933e-05,
+      "loss": 0.9866,
+      "step": 34505
+    },
+    {
+      "epoch": 2.344747927707569,
+      "grad_norm": 6.75,
+      "learning_rate": 3.534532545182769e-05,
+      "loss": 1.0226,
+      "step": 34510
+    },
+    {
+      "epoch": 2.3450876477782305,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.5343202201386064e-05,
+      "loss": 0.9038,
+      "step": 34515
+    },
+    {
+      "epoch": 2.3454273678488926,
+      "grad_norm": 7.25,
+      "learning_rate": 3.534107895094442e-05,
+      "loss": 0.9055,
+      "step": 34520
+    },
+    {
+      "epoch": 2.3457670879195542,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.5338955700502785e-05,
+      "loss": 1.0341,
+      "step": 34525
+    },
+    {
+      "epoch": 2.346106807990216,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5336832450061156e-05,
+      "loss": 0.9971,
+      "step": 34530
+    },
+    {
+      "epoch": 2.346446528060878,
+      "grad_norm": 6.5,
+      "learning_rate": 3.533470919961951e-05,
+      "loss": 0.9673,
+      "step": 34535
+    },
+    {
+      "epoch": 2.3467862481315396,
+      "grad_norm": 7.0,
+      "learning_rate": 3.533258594917788e-05,
+      "loss": 0.9795,
+      "step": 34540
+    },
+    {
+      "epoch": 2.347125968202201,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.533046269873624e-05,
+      "loss": 0.9071,
+      "step": 34545
+    },
+    {
+      "epoch": 2.3474656882728633,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.5328339448294605e-05,
+      "loss": 0.9327,
+      "step": 34550
+    },
+    {
+      "epoch": 2.347805408343525,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.532621619785297e-05,
+      "loss": 0.9931,
+      "step": 34555
+    },
+    {
+      "epoch": 2.3481451284141865,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.532409294741133e-05,
+      "loss": 0.9467,
+      "step": 34560
+    },
+    {
+      "epoch": 2.3484848484848486,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.53219696969697e-05,
+      "loss": 0.9841,
+      "step": 34565
+    },
+    {
+      "epoch": 2.3488245685555103,
+      "grad_norm": 6.375,
+      "learning_rate": 3.531984644652806e-05,
+      "loss": 0.8933,
+      "step": 34570
+    },
+    {
+      "epoch": 2.349164288626172,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.5317723196086425e-05,
+      "loss": 0.9143,
+      "step": 34575
+    },
+    {
+      "epoch": 2.349504008696834,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.531559994564479e-05,
+      "loss": 0.967,
+      "step": 34580
+    },
+    {
+      "epoch": 2.3498437287674956,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.531347669520316e-05,
+      "loss": 1.0029,
+      "step": 34585
+    },
+    {
+      "epoch": 2.350183448838157,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.531135344476152e-05,
+      "loss": 0.9548,
+      "step": 34590
+    },
+    {
+      "epoch": 2.3505231689088193,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.530923019431988e-05,
+      "loss": 0.9187,
+      "step": 34595
+    },
+    {
+      "epoch": 2.350862888979481,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.530710694387825e-05,
+      "loss": 1.0201,
+      "step": 34600
+    },
+    {
+      "epoch": 2.3512026090501426,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.530498369343661e-05,
+      "loss": 0.9824,
+      "step": 34605
+    },
+    {
+      "epoch": 2.3515423291208046,
+      "grad_norm": 7.125,
+      "learning_rate": 3.530286044299497e-05,
+      "loss": 0.9528,
+      "step": 34610
+    },
+    {
+      "epoch": 2.3518820491914663,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.530073719255334e-05,
+      "loss": 1.0288,
+      "step": 34615
+    },
+    {
+      "epoch": 2.352221769262128,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.52986139421117e-05,
+      "loss": 1.0535,
+      "step": 34620
+    },
+    {
+      "epoch": 2.35256148933279,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.5296490691670065e-05,
+      "loss": 0.9456,
+      "step": 34625
+    },
+    {
+      "epoch": 2.3529012094034516,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.529436744122843e-05,
+      "loss": 0.9995,
+      "step": 34630
+    },
+    {
+      "epoch": 2.3532409294741132,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.529224419078679e-05,
+      "loss": 1.0109,
+      "step": 34635
+    },
+    {
+      "epoch": 2.3535806495447753,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.529012094034516e-05,
+      "loss": 0.9635,
+      "step": 34640
+    },
+    {
+      "epoch": 2.353920369615437,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.528799768990352e-05,
+      "loss": 0.9394,
+      "step": 34645
+    },
+    {
+      "epoch": 2.3542600896860986,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5285874439461885e-05,
+      "loss": 1.0085,
+      "step": 34650
+    },
+    {
+      "epoch": 2.3545998097567606,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.528375118902025e-05,
+      "loss": 0.9793,
+      "step": 34655
+    },
+    {
+      "epoch": 2.3549395298274223,
+      "grad_norm": 6.75,
+      "learning_rate": 3.528162793857861e-05,
+      "loss": 0.9474,
+      "step": 34660
+    },
+    {
+      "epoch": 2.355279249898084,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.527950468813698e-05,
+      "loss": 0.9714,
+      "step": 34665
+    },
+    {
+      "epoch": 2.355618969968746,
+      "grad_norm": 6.625,
+      "learning_rate": 3.527738143769534e-05,
+      "loss": 0.9897,
+      "step": 34670
+    },
+    {
+      "epoch": 2.3559586900394076,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.5275258187253705e-05,
+      "loss": 0.9682,
+      "step": 34675
+    },
+    {
+      "epoch": 2.3562984101100692,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.527313493681207e-05,
+      "loss": 0.887,
+      "step": 34680
+    },
+    {
+      "epoch": 2.3566381301807313,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.527101168637043e-05,
+      "loss": 1.0212,
+      "step": 34685
+    },
+    {
+      "epoch": 2.356977850251393,
+      "grad_norm": 7.5,
+      "learning_rate": 3.52688884359288e-05,
+      "loss": 0.9468,
+      "step": 34690
+    },
+    {
+      "epoch": 2.3573175703220546,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.526676518548716e-05,
+      "loss": 0.9356,
+      "step": 34695
+    },
+    {
+      "epoch": 2.357657290392716,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5264641935045525e-05,
+      "loss": 0.9593,
+      "step": 34700
+    },
+    {
+      "epoch": 2.3579970104633783,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.526251868460389e-05,
+      "loss": 0.9748,
+      "step": 34705
+    },
+    {
+      "epoch": 2.35833673053404,
+      "grad_norm": 7.75,
+      "learning_rate": 3.526039543416225e-05,
+      "loss": 0.9392,
+      "step": 34710
+    },
+    {
+      "epoch": 2.3586764506047015,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.525827218372062e-05,
+      "loss": 1.0164,
+      "step": 34715
+    },
+    {
+      "epoch": 2.3590161706753636,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.525614893327898e-05,
+      "loss": 0.9864,
+      "step": 34720
+    },
+    {
+      "epoch": 2.3593558907460253,
+      "grad_norm": 7.0,
+      "learning_rate": 3.5254025682837345e-05,
+      "loss": 1.0372,
+      "step": 34725
+    },
+    {
+      "epoch": 2.359695610816687,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.525190243239571e-05,
+      "loss": 0.9733,
+      "step": 34730
+    },
+    {
+      "epoch": 2.360035330887349,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.524977918195407e-05,
+      "loss": 1.0207,
+      "step": 34735
+    },
+    {
+      "epoch": 2.3603750509580106,
+      "grad_norm": 6.75,
+      "learning_rate": 3.524765593151244e-05,
+      "loss": 0.8788,
+      "step": 34740
+    },
+    {
+      "epoch": 2.360714771028672,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.5245532681070794e-05,
+      "loss": 0.9874,
+      "step": 34745
+    },
+    {
+      "epoch": 2.3610544910993343,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.5243409430629165e-05,
+      "loss": 1.0604,
+      "step": 34750
+    },
+    {
+      "epoch": 2.361394211169996,
+      "grad_norm": 5.75,
+      "learning_rate": 3.524128618018753e-05,
+      "loss": 0.9649,
+      "step": 34755
+    },
+    {
+      "epoch": 2.3617339312406576,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5239162929745886e-05,
+      "loss": 0.9081,
+      "step": 34760
+    },
+    {
+      "epoch": 2.3620736513113196,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.523703967930426e-05,
+      "loss": 0.9945,
+      "step": 34765
+    },
+    {
+      "epoch": 2.3624133713819813,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.523491642886262e-05,
+      "loss": 0.9475,
+      "step": 34770
+    },
+    {
+      "epoch": 2.362753091452643,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.523279317842098e-05,
+      "loss": 1.0235,
+      "step": 34775
+    },
+    {
+      "epoch": 2.363092811523305,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.523066992797935e-05,
+      "loss": 0.9468,
+      "step": 34780
+    },
+    {
+      "epoch": 2.3634325315939666,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.522854667753771e-05,
+      "loss": 0.9795,
+      "step": 34785
+    },
+    {
+      "epoch": 2.3637722516646282,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.522642342709607e-05,
+      "loss": 0.9599,
+      "step": 34790
+    },
+    {
+      "epoch": 2.3641119717352903,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.522430017665444e-05,
+      "loss": 0.8953,
+      "step": 34795
+    },
+    {
+      "epoch": 2.364451691805952,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.5222176926212805e-05,
+      "loss": 0.9392,
+      "step": 34800
+    },
+    {
+      "epoch": 2.3647914118766136,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.522005367577116e-05,
+      "loss": 0.9887,
+      "step": 34805
+    },
+    {
+      "epoch": 2.365131131947275,
+      "grad_norm": 6.375,
+      "learning_rate": 3.521793042532953e-05,
+      "loss": 0.9176,
+      "step": 34810
+    },
+    {
+      "epoch": 2.3654708520179373,
+      "grad_norm": 6.375,
+      "learning_rate": 3.521580717488789e-05,
+      "loss": 0.9775,
+      "step": 34815
+    },
+    {
+      "epoch": 2.365810572088599,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5213683924446254e-05,
+      "loss": 0.9413,
+      "step": 34820
+    },
+    {
+      "epoch": 2.3661502921592605,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.5211560674004625e-05,
+      "loss": 0.9182,
+      "step": 34825
+    },
+    {
+      "epoch": 2.3664900122299226,
+      "grad_norm": 6.375,
+      "learning_rate": 3.520943742356298e-05,
+      "loss": 0.9178,
+      "step": 34830
+    },
+    {
+      "epoch": 2.3668297323005842,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.5207314173121346e-05,
+      "loss": 1.0579,
+      "step": 34835
+    },
+    {
+      "epoch": 2.367169452371246,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.520519092267972e-05,
+      "loss": 1.053,
+      "step": 34840
+    },
+    {
+      "epoch": 2.367509172441908,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5203067672238074e-05,
+      "loss": 0.9622,
+      "step": 34845
+    },
+    {
+      "epoch": 2.3678488925125696,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.520094442179644e-05,
+      "loss": 1.0021,
+      "step": 34850
+    },
+    {
+      "epoch": 2.368188612583231,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.519882117135481e-05,
+      "loss": 0.9442,
+      "step": 34855
+    },
+    {
+      "epoch": 2.3685283326538933,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.5196697920913166e-05,
+      "loss": 0.9941,
+      "step": 34860
+    },
+    {
+      "epoch": 2.368868052724555,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.519457467047153e-05,
+      "loss": 0.9019,
+      "step": 34865
+    },
+    {
+      "epoch": 2.3692077727952165,
+      "grad_norm": 6.875,
+      "learning_rate": 3.51924514200299e-05,
+      "loss": 1.0346,
+      "step": 34870
+    },
+    {
+      "epoch": 2.3695474928658786,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.519032816958826e-05,
+      "loss": 1.0721,
+      "step": 34875
+    },
+    {
+      "epoch": 2.3698872129365403,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.518820491914662e-05,
+      "loss": 0.979,
+      "step": 34880
+    },
+    {
+      "epoch": 2.370226933007202,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.518608166870499e-05,
+      "loss": 1.0267,
+      "step": 34885
+    },
+    {
+      "epoch": 2.370566653077864,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.518395841826335e-05,
+      "loss": 1.0243,
+      "step": 34890
+    },
+    {
+      "epoch": 2.3709063731485256,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.5181835167821714e-05,
+      "loss": 0.897,
+      "step": 34895
+    },
+    {
+      "epoch": 2.3712460932191872,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.517971191738008e-05,
+      "loss": 0.9843,
+      "step": 34900
+    },
+    {
+      "epoch": 2.3715858132898493,
+      "grad_norm": 6.375,
+      "learning_rate": 3.517758866693844e-05,
+      "loss": 0.8989,
+      "step": 34905
+    },
+    {
+      "epoch": 2.371925533360511,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.5175465416496806e-05,
+      "loss": 0.9336,
+      "step": 34910
+    },
+    {
+      "epoch": 2.3722652534311726,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.517334216605517e-05,
+      "loss": 0.9962,
+      "step": 34915
+    },
+    {
+      "epoch": 2.3726049735018346,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.5171218915613535e-05,
+      "loss": 0.9666,
+      "step": 34920
+    },
+    {
+      "epoch": 2.3729446935724963,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.5169095665171905e-05,
+      "loss": 0.9134,
+      "step": 34925
+    },
+    {
+      "epoch": 2.373284413643158,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.516697241473026e-05,
+      "loss": 1.0302,
+      "step": 34930
+    },
+    {
+      "epoch": 2.37362413371382,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.5164849164288627e-05,
+      "loss": 0.9407,
+      "step": 34935
+    },
+    {
+      "epoch": 2.3739638537844816,
+      "grad_norm": 6.75,
+      "learning_rate": 3.5162725913847e-05,
+      "loss": 0.9922,
+      "step": 34940
+    },
+    {
+      "epoch": 2.3743035738551432,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.5160602663405355e-05,
+      "loss": 0.9132,
+      "step": 34945
+    },
+    {
+      "epoch": 2.3746432939258053,
+      "grad_norm": 6.25,
+      "learning_rate": 3.515847941296372e-05,
+      "loss": 0.991,
+      "step": 34950
+    },
+    {
+      "epoch": 2.374983013996467,
+      "grad_norm": 7.625,
+      "learning_rate": 3.515635616252209e-05,
+      "loss": 1.0265,
+      "step": 34955
+    },
+    {
+      "epoch": 2.3753227340671286,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.5154232912080447e-05,
+      "loss": 1.0008,
+      "step": 34960
+    },
+    {
+      "epoch": 2.3756624541377906,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.515210966163881e-05,
+      "loss": 1.0208,
+      "step": 34965
+    },
+    {
+      "epoch": 2.3760021742084523,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.5149986411197175e-05,
+      "loss": 0.9619,
+      "step": 34970
+    },
+    {
+      "epoch": 2.376341894279114,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.514786316075554e-05,
+      "loss": 1.0162,
+      "step": 34975
+    },
+    {
+      "epoch": 2.376681614349776,
+      "grad_norm": 6.625,
+      "learning_rate": 3.51457399103139e-05,
+      "loss": 0.9923,
+      "step": 34980
+    },
+    {
+      "epoch": 2.3770213344204376,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5143616659872267e-05,
+      "loss": 0.9098,
+      "step": 34985
+    },
+    {
+      "epoch": 2.3773610544910992,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.514149340943063e-05,
+      "loss": 1.03,
+      "step": 34990
+    },
+    {
+      "epoch": 2.3777007745617613,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.5139370158988995e-05,
+      "loss": 0.9107,
+      "step": 34995
+    },
+    {
+      "epoch": 2.378040494632423,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.513724690854736e-05,
+      "loss": 0.9578,
+      "step": 35000
+    },
+    {
+      "epoch": 2.3783802147030846,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.513512365810572e-05,
+      "loss": 0.8867,
+      "step": 35005
+    },
+    {
+      "epoch": 2.3787199347737467,
+      "grad_norm": 6.375,
+      "learning_rate": 3.5133000407664087e-05,
+      "loss": 0.9917,
+      "step": 35010
+    },
+    {
+      "epoch": 2.3790596548444083,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.513087715722245e-05,
+      "loss": 0.8962,
+      "step": 35015
+    },
+    {
+      "epoch": 2.37939937491507,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.5128753906780815e-05,
+      "loss": 0.9498,
+      "step": 35020
+    },
+    {
+      "epoch": 2.379739094985732,
+      "grad_norm": 6.625,
+      "learning_rate": 3.512663065633918e-05,
+      "loss": 0.9792,
+      "step": 35025
+    },
+    {
+      "epoch": 2.3800788150563936,
+      "grad_norm": 5.25,
+      "learning_rate": 3.512450740589754e-05,
+      "loss": 1.0919,
+      "step": 35030
+    },
+    {
+      "epoch": 2.3804185351270553,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.5122384155455907e-05,
+      "loss": 1.0214,
+      "step": 35035
+    },
+    {
+      "epoch": 2.380758255197717,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.512026090501427e-05,
+      "loss": 0.9296,
+      "step": 35040
+    },
+    {
+      "epoch": 2.381097975268379,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.5118137654572635e-05,
+      "loss": 0.8574,
+      "step": 35045
+    },
+    {
+      "epoch": 2.3814376953390406,
+      "grad_norm": 6.75,
+      "learning_rate": 3.5116014404131e-05,
+      "loss": 0.955,
+      "step": 35050
+    },
+    {
+      "epoch": 2.3817774154097022,
+      "grad_norm": 5.75,
+      "learning_rate": 3.511389115368936e-05,
+      "loss": 0.972,
+      "step": 35055
+    },
+    {
+      "epoch": 2.3821171354803643,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.511176790324773e-05,
+      "loss": 0.9647,
+      "step": 35060
+    },
+    {
+      "epoch": 2.382456855551026,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.510964465280609e-05,
+      "loss": 0.997,
+      "step": 35065
+    },
+    {
+      "epoch": 2.3827965756216876,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.5107521402364455e-05,
+      "loss": 0.9573,
+      "step": 35070
+    },
+    {
+      "epoch": 2.3831362956923496,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.510539815192282e-05,
+      "loss": 0.9245,
+      "step": 35075
+    },
+    {
+      "epoch": 2.3834760157630113,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.510327490148118e-05,
+      "loss": 0.9877,
+      "step": 35080
+    },
+    {
+      "epoch": 2.383815735833673,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.510115165103955e-05,
+      "loss": 0.9029,
+      "step": 35085
+    },
+    {
+      "epoch": 2.384155455904335,
+      "grad_norm": 7.5,
+      "learning_rate": 3.509902840059791e-05,
+      "loss": 0.9669,
+      "step": 35090
+    },
+    {
+      "epoch": 2.3844951759749966,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.5096905150156275e-05,
+      "loss": 0.9803,
+      "step": 35095
+    },
+    {
+      "epoch": 2.3848348960456582,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.509478189971463e-05,
+      "loss": 0.9558,
+      "step": 35100
+    },
+    {
+      "epoch": 2.3851746161163203,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.5092658649273e-05,
+      "loss": 1.0394,
+      "step": 35105
+    },
+    {
+      "epoch": 2.385514336186982,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.509053539883137e-05,
+      "loss": 0.9665,
+      "step": 35110
+    },
+    {
+      "epoch": 2.3858540562576436,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.5088412148389724e-05,
+      "loss": 1.0269,
+      "step": 35115
+    },
+    {
+      "epoch": 2.3861937763283056,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.5086288897948095e-05,
+      "loss": 0.8903,
+      "step": 35120
+    },
+    {
+      "epoch": 2.3865334963989673,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.508416564750646e-05,
+      "loss": 1.048,
+      "step": 35125
+    },
+    {
+      "epoch": 2.386873216469629,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.5082042397064816e-05,
+      "loss": 1.0685,
+      "step": 35130
+    },
+    {
+      "epoch": 2.387212936540291,
+      "grad_norm": 8.125,
+      "learning_rate": 3.507991914662319e-05,
+      "loss": 0.9026,
+      "step": 35135
+    },
+    {
+      "epoch": 2.3875526566109526,
+      "grad_norm": 6.625,
+      "learning_rate": 3.507779589618155e-05,
+      "loss": 0.9318,
+      "step": 35140
+    },
+    {
+      "epoch": 2.3878923766816142,
+      "grad_norm": 7.25,
+      "learning_rate": 3.507567264573991e-05,
+      "loss": 0.9107,
+      "step": 35145
+    },
+    {
+      "epoch": 2.388232096752276,
+      "grad_norm": 7.375,
+      "learning_rate": 3.507354939529828e-05,
+      "loss": 0.9671,
+      "step": 35150
+    },
+    {
+      "epoch": 2.388571816822938,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.507142614485664e-05,
+      "loss": 0.9909,
+      "step": 35155
+    },
+    {
+      "epoch": 2.3889115368935996,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.5069302894415e-05,
+      "loss": 1.0564,
+      "step": 35160
+    },
+    {
+      "epoch": 2.389251256964261,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.506717964397337e-05,
+      "loss": 0.9546,
+      "step": 35165
+    },
+    {
+      "epoch": 2.3895909770349233,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.506505639353173e-05,
+      "loss": 0.9592,
+      "step": 35170
+    },
+    {
+      "epoch": 2.389930697105585,
+      "grad_norm": 7.0,
+      "learning_rate": 3.506293314309009e-05,
+      "loss": 0.9812,
+      "step": 35175
+    },
+    {
+      "epoch": 2.3902704171762466,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.506080989264846e-05,
+      "loss": 1.032,
+      "step": 35180
+    },
+    {
+      "epoch": 2.3906101372469086,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.505868664220682e-05,
+      "loss": 1.0513,
+      "step": 35185
+    },
+    {
+      "epoch": 2.3909498573175703,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.5056563391765184e-05,
+      "loss": 0.9041,
+      "step": 35190
+    },
+    {
+      "epoch": 2.391289577388232,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.5054440141323555e-05,
+      "loss": 1.035,
+      "step": 35195
+    },
+    {
+      "epoch": 2.391629297458894,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.505231689088191e-05,
+      "loss": 1.0364,
+      "step": 35200
+    },
+    {
+      "epoch": 2.3919690175295556,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.5050193640440276e-05,
+      "loss": 0.9086,
+      "step": 35205
+    },
+    {
+      "epoch": 2.3923087376002172,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.504807038999865e-05,
+      "loss": 1.0378,
+      "step": 35210
+    },
+    {
+      "epoch": 2.3926484576708793,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.5045947139557004e-05,
+      "loss": 0.9219,
+      "step": 35215
+    },
+    {
+      "epoch": 2.392988177741541,
+      "grad_norm": 6.5,
+      "learning_rate": 3.504382388911537e-05,
+      "loss": 0.9333,
+      "step": 35220
+    },
+    {
+      "epoch": 2.3933278978122026,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.504170063867374e-05,
+      "loss": 1.0516,
+      "step": 35225
+    },
+    {
+      "epoch": 2.3936676178828646,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.5039577388232096e-05,
+      "loss": 0.9485,
+      "step": 35230
+    },
+    {
+      "epoch": 2.3940073379535263,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.503745413779046e-05,
+      "loss": 0.9179,
+      "step": 35235
+    },
+    {
+      "epoch": 2.394347058024188,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.5035330887348824e-05,
+      "loss": 0.9603,
+      "step": 35240
+    },
+    {
+      "epoch": 2.39468677809485,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.503320763690719e-05,
+      "loss": 0.9731,
+      "step": 35245
+    },
+    {
+      "epoch": 2.3950264981655116,
+      "grad_norm": 6.25,
+      "learning_rate": 3.503108438646555e-05,
+      "loss": 0.9109,
+      "step": 35250
+    },
+    {
+      "epoch": 2.3953662182361732,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.5028961136023916e-05,
+      "loss": 0.9236,
+      "step": 35255
+    },
+    {
+      "epoch": 2.3957059383068353,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.502683788558228e-05,
+      "loss": 0.9558,
+      "step": 35260
+    },
+    {
+      "epoch": 2.396045658377497,
+      "grad_norm": 6.375,
+      "learning_rate": 3.502471463514065e-05,
+      "loss": 0.9806,
+      "step": 35265
+    },
+    {
+      "epoch": 2.3963853784481586,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.502259138469901e-05,
+      "loss": 1.1041,
+      "step": 35270
+    },
+    {
+      "epoch": 2.3967250985188207,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.502046813425737e-05,
+      "loss": 0.9728,
+      "step": 35275
+    },
+    {
+      "epoch": 2.3970648185894823,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.501834488381574e-05,
+      "loss": 0.924,
+      "step": 35280
+    },
+    {
+      "epoch": 2.397404538660144,
+      "grad_norm": 6.625,
+      "learning_rate": 3.50162216333741e-05,
+      "loss": 0.9254,
+      "step": 35285
+    },
+    {
+      "epoch": 2.397744258730806,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.5014098382932464e-05,
+      "loss": 0.9274,
+      "step": 35290
+    },
+    {
+      "epoch": 2.3980839788014676,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.5011975132490835e-05,
+      "loss": 1.0551,
+      "step": 35295
+    },
+    {
+      "epoch": 2.3984236988721293,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.500985188204919e-05,
+      "loss": 1.0221,
+      "step": 35300
+    },
+    {
+      "epoch": 2.3987634189427913,
+      "grad_norm": 6.75,
+      "learning_rate": 3.5007728631607556e-05,
+      "loss": 1.0537,
+      "step": 35305
+    },
+    {
+      "epoch": 2.399103139013453,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.500560538116593e-05,
+      "loss": 0.9333,
+      "step": 35310
+    },
+    {
+      "epoch": 2.3994428590841146,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.5003482130724284e-05,
+      "loss": 0.8552,
+      "step": 35315
+    },
+    {
+      "epoch": 2.3997825791547767,
+      "grad_norm": 6.75,
+      "learning_rate": 3.500135888028265e-05,
+      "loss": 0.9122,
+      "step": 35320
+    },
+    {
+      "epoch": 2.4001222992254383,
+      "grad_norm": 7.125,
+      "learning_rate": 3.499923562984101e-05,
+      "loss": 0.94,
+      "step": 35325
+    },
+    {
+      "epoch": 2.4004620192961,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.4997112379399376e-05,
+      "loss": 0.9437,
+      "step": 35330
+    },
+    {
+      "epoch": 2.400801739366762,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.499498912895774e-05,
+      "loss": 0.9928,
+      "step": 35335
+    },
+    {
+      "epoch": 2.4011414594374236,
+      "grad_norm": 6.125,
+      "learning_rate": 3.4992865878516104e-05,
+      "loss": 0.9931,
+      "step": 35340
+    },
+    {
+      "epoch": 2.4014811795080853,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.499074262807447e-05,
+      "loss": 0.981,
+      "step": 35345
+    },
+    {
+      "epoch": 2.4018208995787473,
+      "grad_norm": 7.0,
+      "learning_rate": 3.498861937763283e-05,
+      "loss": 0.9198,
+      "step": 35350
+    },
+    {
+      "epoch": 2.402160619649409,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.4986496127191196e-05,
+      "loss": 0.984,
+      "step": 35355
+    },
+    {
+      "epoch": 2.4025003397200706,
+      "grad_norm": 6.375,
+      "learning_rate": 3.498437287674956e-05,
+      "loss": 1.0078,
+      "step": 35360
+    },
+    {
+      "epoch": 2.4028400597907327,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.4982249626307924e-05,
+      "loss": 0.9373,
+      "step": 35365
+    },
+    {
+      "epoch": 2.4031797798613943,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.498012637586629e-05,
+      "loss": 1.0256,
+      "step": 35370
+    },
+    {
+      "epoch": 2.403519499932056,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.497800312542465e-05,
+      "loss": 0.9614,
+      "step": 35375
+    },
+    {
+      "epoch": 2.4038592200027176,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.4975879874983016e-05,
+      "loss": 0.9919,
+      "step": 35380
+    },
+    {
+      "epoch": 2.4041989400733796,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.497375662454138e-05,
+      "loss": 0.9386,
+      "step": 35385
+    },
+    {
+      "epoch": 2.4045386601440413,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.4971633374099744e-05,
+      "loss": 1.0296,
+      "step": 35390
+    },
+    {
+      "epoch": 2.404878380214703,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.496951012365811e-05,
+      "loss": 0.9324,
+      "step": 35395
+    },
+    {
+      "epoch": 2.405218100285365,
+      "grad_norm": 6.375,
+      "learning_rate": 3.496738687321647e-05,
+      "loss": 1.0161,
+      "step": 35400
+    },
+    {
+      "epoch": 2.4055578203560266,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.4965263622774836e-05,
+      "loss": 0.9992,
+      "step": 35405
+    },
+    {
+      "epoch": 2.4058975404266882,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.49631403723332e-05,
+      "loss": 0.8976,
+      "step": 35410
+    },
+    {
+      "epoch": 2.4062372604973503,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.4961017121891564e-05,
+      "loss": 0.8912,
+      "step": 35415
+    },
+    {
+      "epoch": 2.406576980568012,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.495889387144993e-05,
+      "loss": 0.9398,
+      "step": 35420
+    },
+    {
+      "epoch": 2.4069167006386736,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.495677062100829e-05,
+      "loss": 1.0108,
+      "step": 35425
+    },
+    {
+      "epoch": 2.4072564207093357,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.4954647370566656e-05,
+      "loss": 0.9533,
+      "step": 35430
+    },
+    {
+      "epoch": 2.4075961407799973,
+      "grad_norm": 7.0,
+      "learning_rate": 3.495252412012502e-05,
+      "loss": 0.9523,
+      "step": 35435
+    },
+    {
+      "epoch": 2.407935860850659,
+      "grad_norm": 5.5,
+      "learning_rate": 3.495040086968338e-05,
+      "loss": 0.9824,
+      "step": 35440
+    },
+    {
+      "epoch": 2.408275580921321,
+      "grad_norm": 5.75,
+      "learning_rate": 3.494827761924175e-05,
+      "loss": 1.0508,
+      "step": 35445
+    },
+    {
+      "epoch": 2.4086153009919826,
+      "grad_norm": 7.125,
+      "learning_rate": 3.494615436880011e-05,
+      "loss": 1.0145,
+      "step": 35450
+    },
+    {
+      "epoch": 2.4089550210626443,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.494403111835847e-05,
+      "loss": 0.9381,
+      "step": 35455
+    },
+    {
+      "epoch": 2.4092947411333063,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.494190786791684e-05,
+      "loss": 0.9746,
+      "step": 35460
+    },
+    {
+      "epoch": 2.409634461203968,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.4939784617475204e-05,
+      "loss": 0.9014,
+      "step": 35465
+    },
+    {
+      "epoch": 2.4099741812746296,
+      "grad_norm": 7.125,
+      "learning_rate": 3.493766136703356e-05,
+      "loss": 1.0382,
+      "step": 35470
+    },
+    {
+      "epoch": 2.4103139013452917,
+      "grad_norm": 6.5,
+      "learning_rate": 3.493553811659193e-05,
+      "loss": 0.9077,
+      "step": 35475
+    },
+    {
+      "epoch": 2.4106536214159533,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.4933414866150296e-05,
+      "loss": 0.959,
+      "step": 35480
+    },
+    {
+      "epoch": 2.410993341486615,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.4931291615708653e-05,
+      "loss": 1.072,
+      "step": 35485
+    },
+    {
+      "epoch": 2.4113330615572766,
+      "grad_norm": 7.5,
+      "learning_rate": 3.4929168365267024e-05,
+      "loss": 0.9495,
+      "step": 35490
+    },
+    {
+      "epoch": 2.4116727816279386,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.492704511482539e-05,
+      "loss": 0.9668,
+      "step": 35495
+    },
+    {
+      "epoch": 2.4120125016986003,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.4924921864383745e-05,
+      "loss": 1.0567,
+      "step": 35500
+    },
+    {
+      "epoch": 2.412352221769262,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.4922798613942116e-05,
+      "loss": 0.9773,
+      "step": 35505
+    },
+    {
+      "epoch": 2.412691941839924,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.492067536350048e-05,
+      "loss": 0.991,
+      "step": 35510
+    },
+    {
+      "epoch": 2.4130316619105856,
+      "grad_norm": 6.75,
+      "learning_rate": 3.491855211305884e-05,
+      "loss": 0.9768,
+      "step": 35515
+    },
+    {
+      "epoch": 2.4133713819812472,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.491642886261721e-05,
+      "loss": 0.9582,
+      "step": 35520
+    },
+    {
+      "epoch": 2.4137111020519093,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.4914305612175565e-05,
+      "loss": 0.877,
+      "step": 35525
+    },
+    {
+      "epoch": 2.414050822122571,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.491218236173393e-05,
+      "loss": 1.0464,
+      "step": 35530
+    },
+    {
+      "epoch": 2.4143905421932326,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.49100591112923e-05,
+      "loss": 1.0042,
+      "step": 35535
+    },
+    {
+      "epoch": 2.4147302622638946,
+      "grad_norm": 6.625,
+      "learning_rate": 3.490793586085066e-05,
+      "loss": 1.0171,
+      "step": 35540
+    },
+    {
+      "epoch": 2.4150699823345563,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.490581261040902e-05,
+      "loss": 0.9319,
+      "step": 35545
+    },
+    {
+      "epoch": 2.415409702405218,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.490368935996739e-05,
+      "loss": 0.9018,
+      "step": 35550
+    },
+    {
+      "epoch": 2.41574942247588,
+      "grad_norm": 7.5,
+      "learning_rate": 3.490156610952575e-05,
+      "loss": 1.0074,
+      "step": 35555
+    },
+    {
+      "epoch": 2.4160891425465416,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.4899442859084113e-05,
+      "loss": 0.9966,
+      "step": 35560
+    },
+    {
+      "epoch": 2.4164288626172032,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.4897319608642484e-05,
+      "loss": 0.9857,
+      "step": 35565
+    },
+    {
+      "epoch": 2.4167685826878653,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.489519635820084e-05,
+      "loss": 0.947,
+      "step": 35570
+    },
+    {
+      "epoch": 2.417108302758527,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.4893073107759205e-05,
+      "loss": 0.9462,
+      "step": 35575
+    },
+    {
+      "epoch": 2.4174480228291886,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.4890949857317576e-05,
+      "loss": 1.012,
+      "step": 35580
+    },
+    {
+      "epoch": 2.4177877428998507,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.4888826606875933e-05,
+      "loss": 0.9661,
+      "step": 35585
+    },
+    {
+      "epoch": 2.4181274629705123,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.48867033564343e-05,
+      "loss": 0.9616,
+      "step": 35590
+    },
+    {
+      "epoch": 2.418467183041174,
+      "grad_norm": 8.125,
+      "learning_rate": 3.488458010599266e-05,
+      "loss": 0.9562,
+      "step": 35595
+    },
+    {
+      "epoch": 2.418806903111836,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.4882456855551025e-05,
+      "loss": 0.9682,
+      "step": 35600
+    },
+    {
+      "epoch": 2.4191466231824976,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.4880333605109396e-05,
+      "loss": 1.0526,
+      "step": 35605
+    },
+    {
+      "epoch": 2.4194863432531593,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.4878210354667753e-05,
+      "loss": 1.0587,
+      "step": 35610
+    },
+    {
+      "epoch": 2.4198260633238213,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.487608710422612e-05,
+      "loss": 1.0113,
+      "step": 35615
+    },
+    {
+      "epoch": 2.420165783394483,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.487396385378449e-05,
+      "loss": 0.9585,
+      "step": 35620
+    },
+    {
+      "epoch": 2.4205055034651446,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.4871840603342845e-05,
+      "loss": 0.9121,
+      "step": 35625
+    },
+    {
+      "epoch": 2.4208452235358067,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.486971735290121e-05,
+      "loss": 0.9532,
+      "step": 35630
+    },
+    {
+      "epoch": 2.4211849436064683,
+      "grad_norm": 7.125,
+      "learning_rate": 3.486759410245958e-05,
+      "loss": 1.0039,
+      "step": 35635
+    },
+    {
+      "epoch": 2.42152466367713,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.486547085201794e-05,
+      "loss": 1.0099,
+      "step": 35640
+    },
+    {
+      "epoch": 2.421864383747792,
+      "grad_norm": 6.625,
+      "learning_rate": 3.48633476015763e-05,
+      "loss": 0.9748,
+      "step": 35645
+    },
+    {
+      "epoch": 2.4222041038184536,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.486122435113467e-05,
+      "loss": 1.0583,
+      "step": 35650
+    },
+    {
+      "epoch": 2.4225438238891153,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.485910110069303e-05,
+      "loss": 0.8749,
+      "step": 35655
+    },
+    {
+      "epoch": 2.4228835439597773,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.4856977850251394e-05,
+      "loss": 1.0321,
+      "step": 35660
+    },
+    {
+      "epoch": 2.423223264030439,
+      "grad_norm": 6.875,
+      "learning_rate": 3.485485459980976e-05,
+      "loss": 0.9415,
+      "step": 35665
+    },
+    {
+      "epoch": 2.4235629841011006,
+      "grad_norm": 6.75,
+      "learning_rate": 3.485273134936812e-05,
+      "loss": 0.9605,
+      "step": 35670
+    },
+    {
+      "epoch": 2.4239027041717627,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.4850608098926486e-05,
+      "loss": 0.9502,
+      "step": 35675
+    },
+    {
+      "epoch": 2.4242424242424243,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.484848484848485e-05,
+      "loss": 0.9669,
+      "step": 35680
+    },
+    {
+      "epoch": 2.424582144313086,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4846361598043214e-05,
+      "loss": 0.9869,
+      "step": 35685
+    },
+    {
+      "epoch": 2.424921864383748,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.484423834760158e-05,
+      "loss": 0.8892,
+      "step": 35690
+    },
+    {
+      "epoch": 2.4252615844544096,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.484211509715994e-05,
+      "loss": 0.985,
+      "step": 35695
+    },
+    {
+      "epoch": 2.4256013045250713,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4839991846718306e-05,
+      "loss": 1.044,
+      "step": 35700
+    },
+    {
+      "epoch": 2.4259410245957334,
+      "grad_norm": 7.25,
+      "learning_rate": 3.483786859627667e-05,
+      "loss": 0.9399,
+      "step": 35705
+    },
+    {
+      "epoch": 2.426280744666395,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4835745345835034e-05,
+      "loss": 0.973,
+      "step": 35710
+    },
+    {
+      "epoch": 2.4266204647370566,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.48336220953934e-05,
+      "loss": 0.9711,
+      "step": 35715
+    },
+    {
+      "epoch": 2.4269601848077182,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.483149884495176e-05,
+      "loss": 0.9328,
+      "step": 35720
+    },
+    {
+      "epoch": 2.4272999048783803,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.4829375594510126e-05,
+      "loss": 0.94,
+      "step": 35725
+    },
+    {
+      "epoch": 2.427639624949042,
+      "grad_norm": 6.375,
+      "learning_rate": 3.482725234406849e-05,
+      "loss": 1.0215,
+      "step": 35730
+    },
+    {
+      "epoch": 2.4279793450197036,
+      "grad_norm": 6.25,
+      "learning_rate": 3.4825129093626854e-05,
+      "loss": 0.955,
+      "step": 35735
+    },
+    {
+      "epoch": 2.4283190650903657,
+      "grad_norm": 7.625,
+      "learning_rate": 3.482300584318522e-05,
+      "loss": 0.9886,
+      "step": 35740
+    },
+    {
+      "epoch": 2.4286587851610273,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.482088259274358e-05,
+      "loss": 0.9683,
+      "step": 35745
+    },
+    {
+      "epoch": 2.428998505231689,
+      "grad_norm": 7.5,
+      "learning_rate": 3.4818759342301946e-05,
+      "loss": 0.9518,
+      "step": 35750
+    },
+    {
+      "epoch": 2.429338225302351,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.481663609186031e-05,
+      "loss": 1.016,
+      "step": 35755
+    },
+    {
+      "epoch": 2.4296779453730126,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.4814512841418674e-05,
+      "loss": 1.0631,
+      "step": 35760
+    },
+    {
+      "epoch": 2.4300176654436743,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.481238959097704e-05,
+      "loss": 0.9017,
+      "step": 35765
+    },
+    {
+      "epoch": 2.4303573855143363,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.48102663405354e-05,
+      "loss": 0.9922,
+      "step": 35770
+    },
+    {
+      "epoch": 2.430697105584998,
+      "grad_norm": 7.125,
+      "learning_rate": 3.4808143090093766e-05,
+      "loss": 0.8957,
+      "step": 35775
+    },
+    {
+      "epoch": 2.4310368256556596,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.480601983965213e-05,
+      "loss": 0.9782,
+      "step": 35780
+    },
+    {
+      "epoch": 2.4313765457263217,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.4803896589210494e-05,
+      "loss": 1.0034,
+      "step": 35785
+    },
+    {
+      "epoch": 2.4317162657969833,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.480177333876886e-05,
+      "loss": 0.9714,
+      "step": 35790
+    },
+    {
+      "epoch": 2.432055985867645,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.4799650088327215e-05,
+      "loss": 0.9216,
+      "step": 35795
+    },
+    {
+      "epoch": 2.432395705938307,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.4797526837885586e-05,
+      "loss": 0.8532,
+      "step": 35800
+    },
+    {
+      "epoch": 2.4327354260089686,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.479540358744395e-05,
+      "loss": 0.9713,
+      "step": 35805
+    },
+    {
+      "epoch": 2.4330751460796303,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.479328033700231e-05,
+      "loss": 0.955,
+      "step": 35810
+    },
+    {
+      "epoch": 2.4334148661502923,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.479115708656068e-05,
+      "loss": 0.9588,
+      "step": 35815
+    },
+    {
+      "epoch": 2.433754586220954,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.478903383611904e-05,
+      "loss": 1.0366,
+      "step": 35820
+    },
+    {
+      "epoch": 2.4340943062916156,
+      "grad_norm": 7.125,
+      "learning_rate": 3.47869105856774e-05,
+      "loss": 1.0359,
+      "step": 35825
+    },
+    {
+      "epoch": 2.4344340263622772,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.478478733523577e-05,
+      "loss": 0.9323,
+      "step": 35830
+    },
+    {
+      "epoch": 2.4347737464329393,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.4782664084794134e-05,
+      "loss": 0.8981,
+      "step": 35835
+    },
+    {
+      "epoch": 2.435113466503601,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.478054083435249e-05,
+      "loss": 0.8846,
+      "step": 35840
+    },
+    {
+      "epoch": 2.4354531865742626,
+      "grad_norm": 7.25,
+      "learning_rate": 3.477841758391086e-05,
+      "loss": 0.9761,
+      "step": 35845
+    },
+    {
+      "epoch": 2.4357929066449246,
+      "grad_norm": 5.09375,
+      "learning_rate": 3.4776294333469226e-05,
+      "loss": 0.956,
+      "step": 35850
+    },
+    {
+      "epoch": 2.4361326267155863,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.477417108302758e-05,
+      "loss": 0.9995,
+      "step": 35855
+    },
+    {
+      "epoch": 2.436472346786248,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4772047832585954e-05,
+      "loss": 0.9216,
+      "step": 35860
+    },
+    {
+      "epoch": 2.43681206685691,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.476992458214431e-05,
+      "loss": 0.9751,
+      "step": 35865
+    },
+    {
+      "epoch": 2.4371517869275716,
+      "grad_norm": 6.375,
+      "learning_rate": 3.4767801331702675e-05,
+      "loss": 0.9657,
+      "step": 35870
+    },
+    {
+      "epoch": 2.4374915069982332,
+      "grad_norm": 8.75,
+      "learning_rate": 3.4765678081261046e-05,
+      "loss": 0.9596,
+      "step": 35875
+    },
+    {
+      "epoch": 2.4378312270688953,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.47635548308194e-05,
+      "loss": 0.9492,
+      "step": 35880
+    },
+    {
+      "epoch": 2.438170947139557,
+      "grad_norm": 7.375,
+      "learning_rate": 3.476143158037777e-05,
+      "loss": 0.8874,
+      "step": 35885
+    },
+    {
+      "epoch": 2.4385106672102186,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.475930832993614e-05,
+      "loss": 0.9097,
+      "step": 35890
+    },
+    {
+      "epoch": 2.4388503872808807,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.4757185079494495e-05,
+      "loss": 1.0567,
+      "step": 35895
+    },
+    {
+      "epoch": 2.4391901073515423,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.475506182905286e-05,
+      "loss": 0.8995,
+      "step": 35900
+    },
+    {
+      "epoch": 2.439529827422204,
+      "grad_norm": 6.5,
+      "learning_rate": 3.475293857861123e-05,
+      "loss": 0.9741,
+      "step": 35905
+    },
+    {
+      "epoch": 2.439869547492866,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.475081532816959e-05,
+      "loss": 0.9573,
+      "step": 35910
+    },
+    {
+      "epoch": 2.4402092675635276,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.474869207772795e-05,
+      "loss": 0.9898,
+      "step": 35915
+    },
+    {
+      "epoch": 2.4405489876341893,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.474656882728632e-05,
+      "loss": 0.8627,
+      "step": 35920
+    },
+    {
+      "epoch": 2.4408887077048513,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.474444557684468e-05,
+      "loss": 0.9806,
+      "step": 35925
+    },
+    {
+      "epoch": 2.441228427775513,
+      "grad_norm": 6.625,
+      "learning_rate": 3.474232232640304e-05,
+      "loss": 1.0482,
+      "step": 35930
+    },
+    {
+      "epoch": 2.4415681478461746,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.4740199075961414e-05,
+      "loss": 1.0289,
+      "step": 35935
+    },
+    {
+      "epoch": 2.4419078679168367,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.473807582551977e-05,
+      "loss": 1.097,
+      "step": 35940
+    },
+    {
+      "epoch": 2.4422475879874983,
+      "grad_norm": 8.625,
+      "learning_rate": 3.473595257507814e-05,
+      "loss": 1.0024,
+      "step": 35945
+    },
+    {
+      "epoch": 2.44258730805816,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.47338293246365e-05,
+      "loss": 0.9632,
+      "step": 35950
+    },
+    {
+      "epoch": 2.442927028128822,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.473170607419486e-05,
+      "loss": 1.0711,
+      "step": 35955
+    },
+    {
+      "epoch": 2.4432667481994836,
+      "grad_norm": 6.375,
+      "learning_rate": 3.4729582823753234e-05,
+      "loss": 0.8894,
+      "step": 35960
+    },
+    {
+      "epoch": 2.4436064682701453,
+      "grad_norm": 7.75,
+      "learning_rate": 3.472745957331159e-05,
+      "loss": 0.8969,
+      "step": 35965
+    },
+    {
+      "epoch": 2.4439461883408073,
+      "grad_norm": 6.375,
+      "learning_rate": 3.4725336322869955e-05,
+      "loss": 0.9052,
+      "step": 35970
+    },
+    {
+      "epoch": 2.444285908411469,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.4723213072428326e-05,
+      "loss": 0.8939,
+      "step": 35975
+    },
+    {
+      "epoch": 2.4446256284821306,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.472108982198668e-05,
+      "loss": 0.9639,
+      "step": 35980
+    },
+    {
+      "epoch": 2.4449653485527927,
+      "grad_norm": 5.5,
+      "learning_rate": 3.471896657154505e-05,
+      "loss": 0.8765,
+      "step": 35985
+    },
+    {
+      "epoch": 2.4453050686234543,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.471684332110342e-05,
+      "loss": 0.937,
+      "step": 35990
+    },
+    {
+      "epoch": 2.445644788694116,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.4714720070661775e-05,
+      "loss": 0.9565,
+      "step": 35995
+    },
+    {
+      "epoch": 2.445984508764778,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.471259682022014e-05,
+      "loss": 0.9122,
+      "step": 36000
+    },
+    {
+      "epoch": 2.4463242288354397,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.471047356977851e-05,
+      "loss": 0.9547,
+      "step": 36005
+    },
+    {
+      "epoch": 2.4466639489061013,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.470835031933687e-05,
+      "loss": 0.9505,
+      "step": 36010
+    },
+    {
+      "epoch": 2.4470036689767634,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.470622706889523e-05,
+      "loss": 0.9344,
+      "step": 36015
+    },
+    {
+      "epoch": 2.447343389047425,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.4704103818453595e-05,
+      "loss": 1.0087,
+      "step": 36020
+    },
+    {
+      "epoch": 2.4476831091180866,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.470198056801196e-05,
+      "loss": 0.9445,
+      "step": 36025
+    },
+    {
+      "epoch": 2.4480228291887487,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.469985731757032e-05,
+      "loss": 0.9696,
+      "step": 36030
+    },
+    {
+      "epoch": 2.4483625492594103,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.469773406712869e-05,
+      "loss": 0.9983,
+      "step": 36035
+    },
+    {
+      "epoch": 2.448702269330072,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.469561081668705e-05,
+      "loss": 0.9169,
+      "step": 36040
+    },
+    {
+      "epoch": 2.449041989400734,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.4693487566245415e-05,
+      "loss": 1.0158,
+      "step": 36045
+    },
+    {
+      "epoch": 2.4493817094713957,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.469136431580378e-05,
+      "loss": 0.9555,
+      "step": 36050
+    },
+    {
+      "epoch": 2.4497214295420573,
+      "grad_norm": 7.125,
+      "learning_rate": 3.468924106536214e-05,
+      "loss": 0.9484,
+      "step": 36055
+    },
+    {
+      "epoch": 2.4500611496127194,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.468711781492051e-05,
+      "loss": 0.8892,
+      "step": 36060
+    },
+    {
+      "epoch": 2.450400869683381,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.468499456447887e-05,
+      "loss": 0.9601,
+      "step": 36065
+    },
+    {
+      "epoch": 2.4507405897540426,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.4682871314037235e-05,
+      "loss": 0.9258,
+      "step": 36070
+    },
+    {
+      "epoch": 2.4510803098247043,
+      "grad_norm": 6.75,
+      "learning_rate": 3.46807480635956e-05,
+      "loss": 0.9493,
+      "step": 36075
+    },
+    {
+      "epoch": 2.4514200298953663,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.467862481315396e-05,
+      "loss": 0.9983,
+      "step": 36080
+    },
+    {
+      "epoch": 2.451759749966028,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.467650156271233e-05,
+      "loss": 0.9291,
+      "step": 36085
+    },
+    {
+      "epoch": 2.4520994700366896,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.467437831227069e-05,
+      "loss": 1.0227,
+      "step": 36090
+    },
+    {
+      "epoch": 2.4524391901073517,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.4672255061829055e-05,
+      "loss": 0.9123,
+      "step": 36095
+    },
+    {
+      "epoch": 2.4527789101780133,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.467013181138742e-05,
+      "loss": 0.8758,
+      "step": 36100
+    },
+    {
+      "epoch": 2.453118630248675,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.466800856094578e-05,
+      "loss": 0.874,
+      "step": 36105
+    },
+    {
+      "epoch": 2.453458350319337,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.466588531050415e-05,
+      "loss": 1.098,
+      "step": 36110
+    },
+    {
+      "epoch": 2.4537980703899986,
+      "grad_norm": 7.125,
+      "learning_rate": 3.466376206006251e-05,
+      "loss": 0.9541,
+      "step": 36115
+    },
+    {
+      "epoch": 2.4541377904606603,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.4661638809620875e-05,
+      "loss": 0.9802,
+      "step": 36120
+    },
+    {
+      "epoch": 2.4544775105313223,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.465951555917924e-05,
+      "loss": 0.9935,
+      "step": 36125
+    },
+    {
+      "epoch": 2.454817230601984,
+      "grad_norm": 6.25,
+      "learning_rate": 3.46573923087376e-05,
+      "loss": 0.9894,
+      "step": 36130
+    },
+    {
+      "epoch": 2.4551569506726456,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.465526905829597e-05,
+      "loss": 1.0004,
+      "step": 36135
+    },
+    {
+      "epoch": 2.4554966707433077,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.465314580785433e-05,
+      "loss": 0.8183,
+      "step": 36140
+    },
+    {
+      "epoch": 2.4558363908139693,
+      "grad_norm": 7.75,
+      "learning_rate": 3.4651022557412695e-05,
+      "loss": 0.913,
+      "step": 36145
+    },
+    {
+      "epoch": 2.456176110884631,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.464889930697105e-05,
+      "loss": 0.9303,
+      "step": 36150
+    },
+    {
+      "epoch": 2.456515830955293,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.464677605652942e-05,
+      "loss": 1.0489,
+      "step": 36155
+    },
+    {
+      "epoch": 2.4568555510259547,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.464465280608779e-05,
+      "loss": 0.9565,
+      "step": 36160
+    },
+    {
+      "epoch": 2.4571952710966163,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.4642529555646144e-05,
+      "loss": 0.9495,
+      "step": 36165
+    },
+    {
+      "epoch": 2.457534991167278,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.4640406305204515e-05,
+      "loss": 0.8594,
+      "step": 36170
+    },
+    {
+      "epoch": 2.45787471123794,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.463828305476288e-05,
+      "loss": 0.9976,
+      "step": 36175
+    },
+    {
+      "epoch": 2.4582144313086016,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.4636159804321236e-05,
+      "loss": 1.0025,
+      "step": 36180
+    },
+    {
+      "epoch": 2.4585541513792633,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.463403655387961e-05,
+      "loss": 0.9422,
+      "step": 36185
+    },
+    {
+      "epoch": 2.4588938714499253,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.463191330343797e-05,
+      "loss": 0.989,
+      "step": 36190
+    },
+    {
+      "epoch": 2.459233591520587,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.462979005299633e-05,
+      "loss": 0.9829,
+      "step": 36195
+    },
+    {
+      "epoch": 2.4595733115912486,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.46276668025547e-05,
+      "loss": 0.9382,
+      "step": 36200
+    },
+    {
+      "epoch": 2.4599130316619107,
+      "grad_norm": 6.25,
+      "learning_rate": 3.462554355211306e-05,
+      "loss": 1.0071,
+      "step": 36205
+    },
+    {
+      "epoch": 2.4602527517325723,
+      "grad_norm": 7.25,
+      "learning_rate": 3.462342030167142e-05,
+      "loss": 1.0087,
+      "step": 36210
+    },
+    {
+      "epoch": 2.460592471803234,
+      "grad_norm": 6.375,
+      "learning_rate": 3.462129705122979e-05,
+      "loss": 0.9754,
+      "step": 36215
+    },
+    {
+      "epoch": 2.460932191873896,
+      "grad_norm": 8.0,
+      "learning_rate": 3.461917380078815e-05,
+      "loss": 0.9693,
+      "step": 36220
+    },
+    {
+      "epoch": 2.4612719119445576,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.461705055034651e-05,
+      "loss": 0.953,
+      "step": 36225
+    },
+    {
+      "epoch": 2.4616116320152193,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.461492729990488e-05,
+      "loss": 0.947,
+      "step": 36230
+    },
+    {
+      "epoch": 2.4619513520858813,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.461280404946324e-05,
+      "loss": 0.9167,
+      "step": 36235
+    },
+    {
+      "epoch": 2.462291072156543,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.4610680799021604e-05,
+      "loss": 0.9138,
+      "step": 36240
+    },
+    {
+      "epoch": 2.4626307922272046,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.4608557548579975e-05,
+      "loss": 0.9407,
+      "step": 36245
+    },
+    {
+      "epoch": 2.4629705122978667,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.460643429813833e-05,
+      "loss": 0.9712,
+      "step": 36250
+    },
+    {
+      "epoch": 2.4633102323685283,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.4604311047696696e-05,
+      "loss": 0.8838,
+      "step": 36255
+    },
+    {
+      "epoch": 2.46364995243919,
+      "grad_norm": 6.625,
+      "learning_rate": 3.460218779725507e-05,
+      "loss": 0.8936,
+      "step": 36260
+    },
+    {
+      "epoch": 2.463989672509852,
+      "grad_norm": 6.25,
+      "learning_rate": 3.4600064546813424e-05,
+      "loss": 0.9146,
+      "step": 36265
+    },
+    {
+      "epoch": 2.4643293925805136,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.459794129637179e-05,
+      "loss": 1.068,
+      "step": 36270
+    },
+    {
+      "epoch": 2.4646691126511753,
+      "grad_norm": 7.625,
+      "learning_rate": 3.459581804593016e-05,
+      "loss": 1.0077,
+      "step": 36275
+    },
+    {
+      "epoch": 2.4650088327218374,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.4593694795488516e-05,
+      "loss": 0.9312,
+      "step": 36280
+    },
+    {
+      "epoch": 2.465348552792499,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.459157154504689e-05,
+      "loss": 1.0152,
+      "step": 36285
+    },
+    {
+      "epoch": 2.4656882728631606,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.4589448294605244e-05,
+      "loss": 1.0509,
+      "step": 36290
+    },
+    {
+      "epoch": 2.4660279929338227,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.458732504416361e-05,
+      "loss": 0.8565,
+      "step": 36295
+    },
+    {
+      "epoch": 2.4663677130044843,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.458520179372198e-05,
+      "loss": 0.9414,
+      "step": 36300
+    },
+    {
+      "epoch": 2.466707433075146,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.4583078543280336e-05,
+      "loss": 0.9758,
+      "step": 36305
+    },
+    {
+      "epoch": 2.467047153145808,
+      "grad_norm": 6.75,
+      "learning_rate": 3.45809552928387e-05,
+      "loss": 0.9711,
+      "step": 36310
+    },
+    {
+      "epoch": 2.4673868732164697,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.457883204239707e-05,
+      "loss": 0.9751,
+      "step": 36315
+    },
+    {
+      "epoch": 2.4677265932871313,
+      "grad_norm": 6.75,
+      "learning_rate": 3.457670879195543e-05,
+      "loss": 1.0554,
+      "step": 36320
+    },
+    {
+      "epoch": 2.4680663133577934,
+      "grad_norm": 5.75,
+      "learning_rate": 3.457458554151379e-05,
+      "loss": 0.9928,
+      "step": 36325
+    },
+    {
+      "epoch": 2.468406033428455,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.457246229107216e-05,
+      "loss": 0.8958,
+      "step": 36330
+    },
+    {
+      "epoch": 2.4687457534991166,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.457033904063052e-05,
+      "loss": 1.0862,
+      "step": 36335
+    },
+    {
+      "epoch": 2.4690854735697787,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4568215790188884e-05,
+      "loss": 0.8932,
+      "step": 36340
+    },
+    {
+      "epoch": 2.4694251936404403,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.4566092539747255e-05,
+      "loss": 0.9277,
+      "step": 36345
+    },
+    {
+      "epoch": 2.469764913711102,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.456396928930561e-05,
+      "loss": 0.914,
+      "step": 36350
+    },
+    {
+      "epoch": 2.470104633781764,
+      "grad_norm": 7.875,
+      "learning_rate": 3.4561846038863977e-05,
+      "loss": 0.932,
+      "step": 36355
+    },
+    {
+      "epoch": 2.4704443538524257,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.455972278842235e-05,
+      "loss": 0.9961,
+      "step": 36360
+    },
+    {
+      "epoch": 2.4707840739230873,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.4557599537980705e-05,
+      "loss": 0.9453,
+      "step": 36365
+    },
+    {
+      "epoch": 2.4711237939937494,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.455547628753907e-05,
+      "loss": 0.9128,
+      "step": 36370
+    },
+    {
+      "epoch": 2.471463514064411,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.455335303709743e-05,
+      "loss": 0.8364,
+      "step": 36375
+    },
+    {
+      "epoch": 2.4718032341350726,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.4551229786655797e-05,
+      "loss": 0.9295,
+      "step": 36380
+    },
+    {
+      "epoch": 2.4721429542057347,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.454910653621416e-05,
+      "loss": 0.9836,
+      "step": 36385
+    },
+    {
+      "epoch": 2.4724826742763963,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.4546983285772525e-05,
+      "loss": 0.9664,
+      "step": 36390
+    },
+    {
+      "epoch": 2.472822394347058,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.454486003533089e-05,
+      "loss": 1.036,
+      "step": 36395
+    },
+    {
+      "epoch": 2.47316211441772,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.454273678488925e-05,
+      "loss": 1.0007,
+      "step": 36400
+    },
+    {
+      "epoch": 2.4735018344883817,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.4540613534447617e-05,
+      "loss": 0.928,
+      "step": 36405
+    },
+    {
+      "epoch": 2.4738415545590433,
+      "grad_norm": 7.0,
+      "learning_rate": 3.453849028400598e-05,
+      "loss": 1.0007,
+      "step": 36410
+    },
+    {
+      "epoch": 2.474181274629705,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.4536367033564345e-05,
+      "loss": 0.9785,
+      "step": 36415
+    },
+    {
+      "epoch": 2.474520994700367,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.453424378312271e-05,
+      "loss": 0.9163,
+      "step": 36420
+    },
+    {
+      "epoch": 2.4748607147710286,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.453212053268107e-05,
+      "loss": 0.9835,
+      "step": 36425
+    },
+    {
+      "epoch": 2.4752004348416903,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.4529997282239437e-05,
+      "loss": 0.9857,
+      "step": 36430
+    },
+    {
+      "epoch": 2.4755401549123524,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.45278740317978e-05,
+      "loss": 0.9843,
+      "step": 36435
+    },
+    {
+      "epoch": 2.475879874983014,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.4525750781356165e-05,
+      "loss": 0.9846,
+      "step": 36440
+    },
+    {
+      "epoch": 2.4762195950536756,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.452362753091453e-05,
+      "loss": 0.9646,
+      "step": 36445
+    },
+    {
+      "epoch": 2.4765593151243377,
+      "grad_norm": 7.0,
+      "learning_rate": 3.452150428047289e-05,
+      "loss": 0.9967,
+      "step": 36450
+    },
+    {
+      "epoch": 2.4768990351949993,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.4519381030031257e-05,
+      "loss": 0.9129,
+      "step": 36455
+    },
+    {
+      "epoch": 2.477238755265661,
+      "grad_norm": 6.75,
+      "learning_rate": 3.451725777958962e-05,
+      "loss": 1.0191,
+      "step": 36460
+    },
+    {
+      "epoch": 2.477578475336323,
+      "grad_norm": 6.25,
+      "learning_rate": 3.4515134529147985e-05,
+      "loss": 0.9494,
+      "step": 36465
+    },
+    {
+      "epoch": 2.4779181954069847,
+      "grad_norm": 6.875,
+      "learning_rate": 3.451301127870635e-05,
+      "loss": 0.8852,
+      "step": 36470
+    },
+    {
+      "epoch": 2.4782579154776463,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.451088802826471e-05,
+      "loss": 0.9794,
+      "step": 36475
+    },
+    {
+      "epoch": 2.4785976355483084,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.4508764777823077e-05,
+      "loss": 0.9622,
+      "step": 36480
+    },
+    {
+      "epoch": 2.47893735561897,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.450664152738144e-05,
+      "loss": 0.9392,
+      "step": 36485
+    },
+    {
+      "epoch": 2.4792770756896316,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.45045182769398e-05,
+      "loss": 0.8441,
+      "step": 36490
+    },
+    {
+      "epoch": 2.4796167957602937,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.450239502649817e-05,
+      "loss": 1.0301,
+      "step": 36495
+    },
+    {
+      "epoch": 2.4799565158309553,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.450027177605653e-05,
+      "loss": 0.8996,
+      "step": 36500
+    },
+    {
+      "epoch": 2.480296235901617,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.449814852561489e-05,
+      "loss": 1.0691,
+      "step": 36505
+    },
+    {
+      "epoch": 2.4806359559722786,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.449602527517326e-05,
+      "loss": 0.959,
+      "step": 36510
+    },
+    {
+      "epoch": 2.4809756760429407,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.4493902024731625e-05,
+      "loss": 0.9359,
+      "step": 36515
+    },
+    {
+      "epoch": 2.4813153961136023,
+      "grad_norm": 5.875,
+      "learning_rate": 3.449177877428998e-05,
+      "loss": 0.9337,
+      "step": 36520
+    },
+    {
+      "epoch": 2.481655116184264,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.448965552384835e-05,
+      "loss": 0.9931,
+      "step": 36525
+    },
+    {
+      "epoch": 2.481994836254926,
+      "grad_norm": 6.625,
+      "learning_rate": 3.448753227340672e-05,
+      "loss": 0.8959,
+      "step": 36530
+    },
+    {
+      "epoch": 2.4823345563255876,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.4485409022965074e-05,
+      "loss": 0.9677,
+      "step": 36535
+    },
+    {
+      "epoch": 2.4826742763962493,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.4483285772523445e-05,
+      "loss": 0.8879,
+      "step": 36540
+    },
+    {
+      "epoch": 2.4830139964669113,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.448116252208181e-05,
+      "loss": 1.0137,
+      "step": 36545
+    },
+    {
+      "epoch": 2.483353716537573,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.4479039271640166e-05,
+      "loss": 0.9389,
+      "step": 36550
+    },
+    {
+      "epoch": 2.4836934366082346,
+      "grad_norm": 6.875,
+      "learning_rate": 3.447691602119854e-05,
+      "loss": 0.9186,
+      "step": 36555
+    },
+    {
+      "epoch": 2.4840331566788967,
+      "grad_norm": 6.625,
+      "learning_rate": 3.44747927707569e-05,
+      "loss": 0.8883,
+      "step": 36560
+    },
+    {
+      "epoch": 2.4843728767495583,
+      "grad_norm": 7.0,
+      "learning_rate": 3.447266952031526e-05,
+      "loss": 0.9362,
+      "step": 36565
+    },
+    {
+      "epoch": 2.48471259682022,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.447054626987363e-05,
+      "loss": 0.9231,
+      "step": 36570
+    },
+    {
+      "epoch": 2.485052316890882,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.4468423019431986e-05,
+      "loss": 1.0309,
+      "step": 36575
+    },
+    {
+      "epoch": 2.4853920369615436,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.446629976899035e-05,
+      "loss": 1.0205,
+      "step": 36580
+    },
+    {
+      "epoch": 2.4857317570322053,
+      "grad_norm": 6.875,
+      "learning_rate": 3.446417651854872e-05,
+      "loss": 0.9584,
+      "step": 36585
+    },
+    {
+      "epoch": 2.4860714771028674,
+      "grad_norm": 6.875,
+      "learning_rate": 3.446205326810708e-05,
+      "loss": 0.9269,
+      "step": 36590
+    },
+    {
+      "epoch": 2.486411197173529,
+      "grad_norm": 7.0,
+      "learning_rate": 3.445993001766544e-05,
+      "loss": 0.9361,
+      "step": 36595
+    },
+    {
+      "epoch": 2.4867509172441906,
+      "grad_norm": 8.25,
+      "learning_rate": 3.445780676722381e-05,
+      "loss": 0.9166,
+      "step": 36600
+    },
+    {
+      "epoch": 2.4870906373148527,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.445568351678217e-05,
+      "loss": 0.9062,
+      "step": 36605
+    },
+    {
+      "epoch": 2.4874303573855143,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4453560266340534e-05,
+      "loss": 0.9424,
+      "step": 36610
+    },
+    {
+      "epoch": 2.487770077456176,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.4451437015898905e-05,
+      "loss": 0.9297,
+      "step": 36615
+    },
+    {
+      "epoch": 2.488109797526838,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.444931376545726e-05,
+      "loss": 0.9458,
+      "step": 36620
+    },
+    {
+      "epoch": 2.4884495175974997,
+      "grad_norm": 6.375,
+      "learning_rate": 3.444719051501563e-05,
+      "loss": 1.037,
+      "step": 36625
+    },
+    {
+      "epoch": 2.4887892376681613,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.4445067264574e-05,
+      "loss": 0.9175,
+      "step": 36630
+    },
+    {
+      "epoch": 2.4891289577388234,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.4442944014132354e-05,
+      "loss": 1.0175,
+      "step": 36635
+    },
+    {
+      "epoch": 2.489468677809485,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.4440820763690725e-05,
+      "loss": 0.9174,
+      "step": 36640
+    },
+    {
+      "epoch": 2.4898083978801466,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.443869751324908e-05,
+      "loss": 0.9352,
+      "step": 36645
+    },
+    {
+      "epoch": 2.4901481179508087,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.4436574262807446e-05,
+      "loss": 0.9712,
+      "step": 36650
+    },
+    {
+      "epoch": 2.4904878380214703,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.443445101236582e-05,
+      "loss": 0.981,
+      "step": 36655
+    },
+    {
+      "epoch": 2.490827558092132,
+      "grad_norm": 5.75,
+      "learning_rate": 3.4432327761924174e-05,
+      "loss": 0.7853,
+      "step": 36660
+    },
+    {
+      "epoch": 2.491167278162794,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.443020451148254e-05,
+      "loss": 0.9278,
+      "step": 36665
+    },
+    {
+      "epoch": 2.4915069982334557,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.442808126104091e-05,
+      "loss": 0.9875,
+      "step": 36670
+    },
+    {
+      "epoch": 2.4918467183041173,
+      "grad_norm": 6.375,
+      "learning_rate": 3.4425958010599266e-05,
+      "loss": 0.9441,
+      "step": 36675
+    },
+    {
+      "epoch": 2.4921864383747794,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.442383476015763e-05,
+      "loss": 0.9524,
+      "step": 36680
+    },
+    {
+      "epoch": 2.492526158445441,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.4421711509716e-05,
+      "loss": 0.8953,
+      "step": 36685
+    },
+    {
+      "epoch": 2.4928658785161026,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.441958825927436e-05,
+      "loss": 0.8991,
+      "step": 36690
+    },
+    {
+      "epoch": 2.4932055985867647,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.441746500883272e-05,
+      "loss": 1.016,
+      "step": 36695
+    },
+    {
+      "epoch": 2.4935453186574263,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.441534175839109e-05,
+      "loss": 0.9162,
+      "step": 36700
+    },
+    {
+      "epoch": 2.493885038728088,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.441321850794945e-05,
+      "loss": 0.9479,
+      "step": 36705
+    },
+    {
+      "epoch": 2.49422475879875,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.4411095257507814e-05,
+      "loss": 1.0278,
+      "step": 36710
+    },
+    {
+      "epoch": 2.4945644788694117,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.4408972007066185e-05,
+      "loss": 0.9907,
+      "step": 36715
+    },
+    {
+      "epoch": 2.4949041989400733,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.440684875662454e-05,
+      "loss": 0.9772,
+      "step": 36720
+    },
+    {
+      "epoch": 2.4952439190107354,
+      "grad_norm": 6.25,
+      "learning_rate": 3.4404725506182906e-05,
+      "loss": 0.9273,
+      "step": 36725
+    },
+    {
+      "epoch": 2.495583639081397,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.440260225574127e-05,
+      "loss": 0.9061,
+      "step": 36730
+    },
+    {
+      "epoch": 2.4959233591520587,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.4400479005299634e-05,
+      "loss": 0.9795,
+      "step": 36735
+    },
+    {
+      "epoch": 2.4962630792227207,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.4398355754858e-05,
+      "loss": 0.9976,
+      "step": 36740
+    },
+    {
+      "epoch": 2.4966027992933824,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.439623250441636e-05,
+      "loss": 0.923,
+      "step": 36745
+    },
+    {
+      "epoch": 2.496942519364044,
+      "grad_norm": 7.375,
+      "learning_rate": 3.4394109253974726e-05,
+      "loss": 1.0789,
+      "step": 36750
+    },
+    {
+      "epoch": 2.4972822394347056,
+      "grad_norm": 7.5,
+      "learning_rate": 3.439198600353309e-05,
+      "loss": 0.957,
+      "step": 36755
+    },
+    {
+      "epoch": 2.4976219595053677,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.4389862753091454e-05,
+      "loss": 1.0893,
+      "step": 36760
+    },
+    {
+      "epoch": 2.4979616795760293,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.438773950264982e-05,
+      "loss": 0.9928,
+      "step": 36765
+    },
+    {
+      "epoch": 2.498301399646691,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.438561625220818e-05,
+      "loss": 1.0233,
+      "step": 36770
+    },
+    {
+      "epoch": 2.498641119717353,
+      "grad_norm": 6.375,
+      "learning_rate": 3.4383493001766546e-05,
+      "loss": 1.0077,
+      "step": 36775
+    },
+    {
+      "epoch": 2.4989808397880147,
+      "grad_norm": 6.875,
+      "learning_rate": 3.438136975132491e-05,
+      "loss": 0.9551,
+      "step": 36780
+    },
+    {
+      "epoch": 2.4993205598586763,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4379246500883274e-05,
+      "loss": 0.997,
+      "step": 36785
+    },
+    {
+      "epoch": 2.4996602799293384,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.437712325044164e-05,
+      "loss": 0.9596,
+      "step": 36790
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.4375e-05,
+      "loss": 0.8686,
+      "step": 36795
+    },
+    {
+      "epoch": 2.5003397200706616,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4372876749558366e-05,
+      "loss": 0.9523,
+      "step": 36800
+    },
+    {
+      "epoch": 2.5006794401413237,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.437075349911673e-05,
+      "loss": 0.9398,
+      "step": 36805
+    },
+    {
+      "epoch": 2.5010191602119853,
+      "grad_norm": 6.75,
+      "learning_rate": 3.4368630248675094e-05,
+      "loss": 0.9827,
+      "step": 36810
+    },
+    {
+      "epoch": 2.501358880282647,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.436650699823346e-05,
+      "loss": 0.9751,
+      "step": 36815
+    },
+    {
+      "epoch": 2.5016986003533086,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.436438374779182e-05,
+      "loss": 0.9959,
+      "step": 36820
+    },
+    {
+      "epoch": 2.5020383204239707,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.4362260497350186e-05,
+      "loss": 0.9354,
+      "step": 36825
+    },
+    {
+      "epoch": 2.5023780404946323,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.436013724690855e-05,
+      "loss": 1.0514,
+      "step": 36830
+    },
+    {
+      "epoch": 2.502717760565294,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.4358013996466914e-05,
+      "loss": 0.9425,
+      "step": 36835
+    },
+    {
+      "epoch": 2.503057480635956,
+      "grad_norm": 7.375,
+      "learning_rate": 3.435589074602528e-05,
+      "loss": 1.1494,
+      "step": 36840
+    },
+    {
+      "epoch": 2.5033972007066176,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.4353767495583635e-05,
+      "loss": 0.9555,
+      "step": 36845
+    },
+    {
+      "epoch": 2.5037369207772793,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.4351644245142006e-05,
+      "loss": 0.9807,
+      "step": 36850
+    },
+    {
+      "epoch": 2.5040766408479413,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.434952099470037e-05,
+      "loss": 0.9093,
+      "step": 36855
+    },
+    {
+      "epoch": 2.504416360918603,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.434739774425873e-05,
+      "loss": 0.9236,
+      "step": 36860
+    },
+    {
+      "epoch": 2.5047560809892646,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.43452744938171e-05,
+      "loss": 0.956,
+      "step": 36865
+    },
+    {
+      "epoch": 2.5050958010599267,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.434315124337546e-05,
+      "loss": 1.0402,
+      "step": 36870
+    },
+    {
+      "epoch": 2.5054355211305883,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.434102799293382e-05,
+      "loss": 0.9446,
+      "step": 36875
+    },
+    {
+      "epoch": 2.50577524120125,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.433890474249219e-05,
+      "loss": 0.977,
+      "step": 36880
+    },
+    {
+      "epoch": 2.506114961271912,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.4336781492050554e-05,
+      "loss": 1.0477,
+      "step": 36885
+    },
+    {
+      "epoch": 2.5064546813425737,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.433465824160891e-05,
+      "loss": 0.9992,
+      "step": 36890
+    },
+    {
+      "epoch": 2.5067944014132353,
+      "grad_norm": 7.125,
+      "learning_rate": 3.433253499116728e-05,
+      "loss": 0.8626,
+      "step": 36895
+    },
+    {
+      "epoch": 2.5071341214838974,
+      "grad_norm": 6.75,
+      "learning_rate": 3.4330411740725646e-05,
+      "loss": 0.8964,
+      "step": 36900
+    },
+    {
+      "epoch": 2.507473841554559,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.4328288490284e-05,
+      "loss": 0.9846,
+      "step": 36905
+    },
+    {
+      "epoch": 2.5078135616252206,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4326165239842374e-05,
+      "loss": 0.9928,
+      "step": 36910
+    },
+    {
+      "epoch": 2.5081532816958827,
+      "grad_norm": 7.0,
+      "learning_rate": 3.432404198940074e-05,
+      "loss": 0.9294,
+      "step": 36915
+    },
+    {
+      "epoch": 2.5084930017665443,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.4321918738959095e-05,
+      "loss": 0.9675,
+      "step": 36920
+    },
+    {
+      "epoch": 2.508832721837206,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.4319795488517466e-05,
+      "loss": 0.9795,
+      "step": 36925
+    },
+    {
+      "epoch": 2.509172441907868,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.4317672238075823e-05,
+      "loss": 0.8898,
+      "step": 36930
+    },
+    {
+      "epoch": 2.5095121619785297,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.431554898763419e-05,
+      "loss": 1.0617,
+      "step": 36935
+    },
+    {
+      "epoch": 2.5098518820491913,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.431342573719256e-05,
+      "loss": 0.9549,
+      "step": 36940
+    },
+    {
+      "epoch": 2.5101916021198534,
+      "grad_norm": 6.5,
+      "learning_rate": 3.4311302486750915e-05,
+      "loss": 1.0753,
+      "step": 36945
+    },
+    {
+      "epoch": 2.510531322190515,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.430917923630928e-05,
+      "loss": 0.9904,
+      "step": 36950
+    },
+    {
+      "epoch": 2.5108710422611766,
+      "grad_norm": 6.875,
+      "learning_rate": 3.430705598586765e-05,
+      "loss": 0.944,
+      "step": 36955
+    },
+    {
+      "epoch": 2.5112107623318387,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.430493273542601e-05,
+      "loss": 1.0536,
+      "step": 36960
+    },
+    {
+      "epoch": 2.5115504824025003,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.430280948498437e-05,
+      "loss": 0.9642,
+      "step": 36965
+    },
+    {
+      "epoch": 2.511890202473162,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.430068623454274e-05,
+      "loss": 0.9816,
+      "step": 36970
+    },
+    {
+      "epoch": 2.512229922543824,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.42985629841011e-05,
+      "loss": 1.0128,
+      "step": 36975
+    },
+    {
+      "epoch": 2.5125696426144857,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.429643973365947e-05,
+      "loss": 0.9251,
+      "step": 36980
+    },
+    {
+      "epoch": 2.5129093626851473,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.4294316483217834e-05,
+      "loss": 0.9528,
+      "step": 36985
+    },
+    {
+      "epoch": 2.5132490827558094,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.429219323277619e-05,
+      "loss": 0.9466,
+      "step": 36990
+    },
+    {
+      "epoch": 2.513588802826471,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.429006998233456e-05,
+      "loss": 1.0406,
+      "step": 36995
+    },
+    {
+      "epoch": 2.5139285228971326,
+      "grad_norm": 6.0,
+      "learning_rate": 3.428794673189292e-05,
+      "loss": 0.9809,
+      "step": 37000
+    },
+    {
+      "epoch": 2.5142682429677947,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.4285823481451283e-05,
+      "loss": 1.0272,
+      "step": 37005
+    },
+    {
+      "epoch": 2.5146079630384564,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.4283700231009654e-05,
+      "loss": 0.9695,
+      "step": 37010
+    },
+    {
+      "epoch": 2.514947683109118,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.428157698056801e-05,
+      "loss": 0.9407,
+      "step": 37015
+    },
+    {
+      "epoch": 2.51528740317978,
+      "grad_norm": 8.9375,
+      "learning_rate": 3.4279453730126375e-05,
+      "loss": 0.9692,
+      "step": 37020
+    },
+    {
+      "epoch": 2.5156271232504417,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.4277330479684746e-05,
+      "loss": 1.0265,
+      "step": 37025
+    },
+    {
+      "epoch": 2.5159668433211033,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.4275207229243103e-05,
+      "loss": 0.9519,
+      "step": 37030
+    },
+    {
+      "epoch": 2.5163065633917654,
+      "grad_norm": 6.125,
+      "learning_rate": 3.427308397880147e-05,
+      "loss": 0.956,
+      "step": 37035
+    },
+    {
+      "epoch": 2.516646283462427,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.427096072835984e-05,
+      "loss": 1.0162,
+      "step": 37040
+    },
+    {
+      "epoch": 2.5169860035330887,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.4268837477918195e-05,
+      "loss": 0.9371,
+      "step": 37045
+    },
+    {
+      "epoch": 2.5173257236037507,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.426671422747656e-05,
+      "loss": 0.9783,
+      "step": 37050
+    },
+    {
+      "epoch": 2.5176654436744124,
+      "grad_norm": 5.625,
+      "learning_rate": 3.426459097703493e-05,
+      "loss": 0.9451,
+      "step": 37055
+    },
+    {
+      "epoch": 2.518005163745074,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.426246772659329e-05,
+      "loss": 0.9939,
+      "step": 37060
+    },
+    {
+      "epoch": 2.518344883815736,
+      "grad_norm": 7.0,
+      "learning_rate": 3.426034447615165e-05,
+      "loss": 1.0067,
+      "step": 37065
+    },
+    {
+      "epoch": 2.5186846038863977,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4258221225710016e-05,
+      "loss": 0.9752,
+      "step": 37070
+    },
+    {
+      "epoch": 2.5190243239570593,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.425609797526838e-05,
+      "loss": 0.958,
+      "step": 37075
+    },
+    {
+      "epoch": 2.5193640440277214,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.4253974724826744e-05,
+      "loss": 0.9796,
+      "step": 37080
+    },
+    {
+      "epoch": 2.519703764098383,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.425185147438511e-05,
+      "loss": 1.0351,
+      "step": 37085
+    },
+    {
+      "epoch": 2.5200434841690447,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.424972822394347e-05,
+      "loss": 0.9991,
+      "step": 37090
+    },
+    {
+      "epoch": 2.5203832042397067,
+      "grad_norm": 7.5,
+      "learning_rate": 3.4247604973501836e-05,
+      "loss": 0.998,
+      "step": 37095
+    },
+    {
+      "epoch": 2.5207229243103684,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.42454817230602e-05,
+      "loss": 0.9475,
+      "step": 37100
+    },
+    {
+      "epoch": 2.52106264438103,
+      "grad_norm": 6.375,
+      "learning_rate": 3.4243358472618564e-05,
+      "loss": 0.8498,
+      "step": 37105
+    },
+    {
+      "epoch": 2.521402364451692,
+      "grad_norm": 6.75,
+      "learning_rate": 3.424123522217693e-05,
+      "loss": 0.9465,
+      "step": 37110
+    },
+    {
+      "epoch": 2.5217420845223537,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.423911197173529e-05,
+      "loss": 0.9229,
+      "step": 37115
+    },
+    {
+      "epoch": 2.5220818045930153,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.4236988721293656e-05,
+      "loss": 0.9894,
+      "step": 37120
+    },
+    {
+      "epoch": 2.522421524663677,
+      "grad_norm": 7.5,
+      "learning_rate": 3.423486547085202e-05,
+      "loss": 1.0398,
+      "step": 37125
+    },
+    {
+      "epoch": 2.522761244734339,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.4232742220410384e-05,
+      "loss": 0.9252,
+      "step": 37130
+    },
+    {
+      "epoch": 2.5231009648050007,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.423061896996875e-05,
+      "loss": 0.9829,
+      "step": 37135
+    },
+    {
+      "epoch": 2.5234406848756623,
+      "grad_norm": 6.125,
+      "learning_rate": 3.422849571952711e-05,
+      "loss": 1.0169,
+      "step": 37140
+    },
+    {
+      "epoch": 2.5237804049463244,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.4226372469085476e-05,
+      "loss": 0.956,
+      "step": 37145
+    },
+    {
+      "epoch": 2.524120125016986,
+      "grad_norm": 5.875,
+      "learning_rate": 3.422424921864384e-05,
+      "loss": 1.0784,
+      "step": 37150
+    },
+    {
+      "epoch": 2.5244598450876476,
+      "grad_norm": 7.5,
+      "learning_rate": 3.4222125968202204e-05,
+      "loss": 1.0489,
+      "step": 37155
+    },
+    {
+      "epoch": 2.5247995651583093,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.422000271776057e-05,
+      "loss": 0.9362,
+      "step": 37160
+    },
+    {
+      "epoch": 2.5251392852289714,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.421787946731893e-05,
+      "loss": 0.9608,
+      "step": 37165
+    },
+    {
+      "epoch": 2.525479005299633,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.4215756216877296e-05,
+      "loss": 0.8091,
+      "step": 37170
+    },
+    {
+      "epoch": 2.5258187253702946,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.421363296643566e-05,
+      "loss": 0.9011,
+      "step": 37175
+    },
+    {
+      "epoch": 2.5261584454409567,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.4211509715994024e-05,
+      "loss": 1.0382,
+      "step": 37180
+    },
+    {
+      "epoch": 2.5264981655116183,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.420938646555239e-05,
+      "loss": 0.9828,
+      "step": 37185
+    },
+    {
+      "epoch": 2.52683788558228,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.420726321511075e-05,
+      "loss": 0.9857,
+      "step": 37190
+    },
+    {
+      "epoch": 2.527177605652942,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.4205139964669116e-05,
+      "loss": 1.0322,
+      "step": 37195
+    },
+    {
+      "epoch": 2.5275173257236037,
+      "grad_norm": 6.875,
+      "learning_rate": 3.420301671422747e-05,
+      "loss": 0.8545,
+      "step": 37200
+    },
+    {
+      "epoch": 2.5278570457942653,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.4200893463785844e-05,
+      "loss": 1.0378,
+      "step": 37205
+    },
+    {
+      "epoch": 2.5281967658649274,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.419877021334421e-05,
+      "loss": 0.9858,
+      "step": 37210
+    },
+    {
+      "epoch": 2.528536485935589,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.4196646962902565e-05,
+      "loss": 0.8916,
+      "step": 37215
+    },
+    {
+      "epoch": 2.5288762060062506,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.4194523712460936e-05,
+      "loss": 0.9784,
+      "step": 37220
+    },
+    {
+      "epoch": 2.5292159260769127,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.41924004620193e-05,
+      "loss": 0.9669,
+      "step": 37225
+    },
+    {
+      "epoch": 2.5295556461475743,
+      "grad_norm": 7.875,
+      "learning_rate": 3.419027721157766e-05,
+      "loss": 0.9937,
+      "step": 37230
+    },
+    {
+      "epoch": 2.529895366218236,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.418815396113603e-05,
+      "loss": 0.9528,
+      "step": 37235
+    },
+    {
+      "epoch": 2.530235086288898,
+      "grad_norm": 7.375,
+      "learning_rate": 3.418603071069439e-05,
+      "loss": 0.9728,
+      "step": 37240
+    },
+    {
+      "epoch": 2.5305748063595597,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.418390746025275e-05,
+      "loss": 0.84,
+      "step": 37245
+    },
+    {
+      "epoch": 2.5309145264302213,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.418178420981112e-05,
+      "loss": 0.9738,
+      "step": 37250
+    },
+    {
+      "epoch": 2.5312542465008834,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.4179660959369484e-05,
+      "loss": 1.0887,
+      "step": 37255
+    },
+    {
+      "epoch": 2.531593966571545,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.417753770892784e-05,
+      "loss": 0.9802,
+      "step": 37260
+    },
+    {
+      "epoch": 2.5319336866422066,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.417541445848621e-05,
+      "loss": 0.9538,
+      "step": 37265
+    },
+    {
+      "epoch": 2.5322734067128687,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.417329120804457e-05,
+      "loss": 1.0193,
+      "step": 37270
+    },
+    {
+      "epoch": 2.5326131267835303,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.417116795760293e-05,
+      "loss": 0.9613,
+      "step": 37275
+    },
+    {
+      "epoch": 2.532952846854192,
+      "grad_norm": 5.5625,
+      "learning_rate": 3.4169044707161304e-05,
+      "loss": 0.9401,
+      "step": 37280
+    },
+    {
+      "epoch": 2.533292566924854,
+      "grad_norm": 7.125,
+      "learning_rate": 3.416692145671966e-05,
+      "loss": 0.9462,
+      "step": 37285
+    },
+    {
+      "epoch": 2.5336322869955157,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.4164798206278025e-05,
+      "loss": 0.9516,
+      "step": 37290
+    },
+    {
+      "epoch": 2.5339720070661773,
+      "grad_norm": 6.125,
+      "learning_rate": 3.4162674955836396e-05,
+      "loss": 0.9165,
+      "step": 37295
+    },
+    {
+      "epoch": 2.5343117271368394,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.416055170539475e-05,
+      "loss": 0.8759,
+      "step": 37300
+    },
+    {
+      "epoch": 2.534651447207501,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.415842845495312e-05,
+      "loss": 0.9073,
+      "step": 37305
+    },
+    {
+      "epoch": 2.5349911672781626,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.415630520451149e-05,
+      "loss": 0.8773,
+      "step": 37310
+    },
+    {
+      "epoch": 2.5353308873488247,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.4154181954069845e-05,
+      "loss": 0.9928,
+      "step": 37315
+    },
+    {
+      "epoch": 2.5356706074194864,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.4152058703628216e-05,
+      "loss": 0.9873,
+      "step": 37320
+    },
+    {
+      "epoch": 2.536010327490148,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.414993545318658e-05,
+      "loss": 0.8704,
+      "step": 37325
+    },
+    {
+      "epoch": 2.53635004756081,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.414781220274494e-05,
+      "loss": 0.9395,
+      "step": 37330
+    },
+    {
+      "epoch": 2.5366897676314717,
+      "grad_norm": 6.75,
+      "learning_rate": 3.414568895230331e-05,
+      "loss": 0.9917,
+      "step": 37335
+    },
+    {
+      "epoch": 2.5370294877021333,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.414356570186167e-05,
+      "loss": 0.9712,
+      "step": 37340
+    },
+    {
+      "epoch": 2.5373692077727954,
+      "grad_norm": 6.875,
+      "learning_rate": 3.414144245142003e-05,
+      "loss": 0.9553,
+      "step": 37345
+    },
+    {
+      "epoch": 2.537708927843457,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.41393192009784e-05,
+      "loss": 0.9908,
+      "step": 37350
+    },
+    {
+      "epoch": 2.5380486479141187,
+      "grad_norm": 7.5,
+      "learning_rate": 3.413719595053676e-05,
+      "loss": 0.9603,
+      "step": 37355
+    },
+    {
+      "epoch": 2.5383883679847807,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.413507270009512e-05,
+      "loss": 0.9637,
+      "step": 37360
+    },
+    {
+      "epoch": 2.5387280880554424,
+      "grad_norm": 7.25,
+      "learning_rate": 3.413294944965349e-05,
+      "loss": 0.9683,
+      "step": 37365
+    },
+    {
+      "epoch": 2.539067808126104,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.413082619921185e-05,
+      "loss": 0.9847,
+      "step": 37370
+    },
+    {
+      "epoch": 2.539407528196766,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.412870294877021e-05,
+      "loss": 0.9657,
+      "step": 37375
+    },
+    {
+      "epoch": 2.5397472482674277,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4126579698328584e-05,
+      "loss": 1.0349,
+      "step": 37380
+    },
+    {
+      "epoch": 2.5400869683380893,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.412445644788694e-05,
+      "loss": 0.9458,
+      "step": 37385
+    },
+    {
+      "epoch": 2.5404266884087514,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.4122333197445305e-05,
+      "loss": 0.9932,
+      "step": 37390
+    },
+    {
+      "epoch": 2.540766408479413,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.4120209947003676e-05,
+      "loss": 1.0448,
+      "step": 37395
+    },
+    {
+      "epoch": 2.5411061285500747,
+      "grad_norm": 6.875,
+      "learning_rate": 3.411808669656203e-05,
+      "loss": 0.9139,
+      "step": 37400
+    },
+    {
+      "epoch": 2.5414458486207367,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.41159634461204e-05,
+      "loss": 0.9472,
+      "step": 37405
+    },
+    {
+      "epoch": 2.5417855686913984,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.411384019567877e-05,
+      "loss": 1.0211,
+      "step": 37410
+    },
+    {
+      "epoch": 2.54212528876206,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.4111716945237125e-05,
+      "loss": 1.0255,
+      "step": 37415
+    },
+    {
+      "epoch": 2.542465008832722,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.410959369479549e-05,
+      "loss": 0.9803,
+      "step": 37420
+    },
+    {
+      "epoch": 2.5428047289033837,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.410747044435385e-05,
+      "loss": 0.9114,
+      "step": 37425
+    },
+    {
+      "epoch": 2.5431444489740453,
+      "grad_norm": 6.5,
+      "learning_rate": 3.410534719391222e-05,
+      "loss": 0.9976,
+      "step": 37430
+    },
+    {
+      "epoch": 2.5434841690447074,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.410322394347058e-05,
+      "loss": 0.973,
+      "step": 37435
+    },
+    {
+      "epoch": 2.543823889115369,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.4101100693028945e-05,
+      "loss": 0.9776,
+      "step": 37440
+    },
+    {
+      "epoch": 2.5441636091860307,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.409897744258731e-05,
+      "loss": 0.8995,
+      "step": 37445
+    },
+    {
+      "epoch": 2.5445033292566928,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.409685419214567e-05,
+      "loss": 0.9061,
+      "step": 37450
+    },
+    {
+      "epoch": 2.5448430493273544,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.409473094170404e-05,
+      "loss": 0.9292,
+      "step": 37455
+    },
+    {
+      "epoch": 2.545182769398016,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.40926076912624e-05,
+      "loss": 0.9335,
+      "step": 37460
+    },
+    {
+      "epoch": 2.5455224894686777,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.4090484440820765e-05,
+      "loss": 0.9813,
+      "step": 37465
+    },
+    {
+      "epoch": 2.5458622095393397,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.408836119037913e-05,
+      "loss": 1.003,
+      "step": 37470
+    },
+    {
+      "epoch": 2.5462019296100014,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.408623793993749e-05,
+      "loss": 0.9434,
+      "step": 37475
+    },
+    {
+      "epoch": 2.546541649680663,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.408411468949586e-05,
+      "loss": 0.9136,
+      "step": 37480
+    },
+    {
+      "epoch": 2.546881369751325,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.408199143905422e-05,
+      "loss": 0.9625,
+      "step": 37485
+    },
+    {
+      "epoch": 2.5472210898219867,
+      "grad_norm": 6.625,
+      "learning_rate": 3.4079868188612585e-05,
+      "loss": 1.0278,
+      "step": 37490
+    },
+    {
+      "epoch": 2.5475608098926483,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.407774493817095e-05,
+      "loss": 0.8822,
+      "step": 37495
+    },
+    {
+      "epoch": 2.54790052996331,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.407562168772931e-05,
+      "loss": 1.0336,
+      "step": 37500
+    },
+    {
+      "epoch": 2.548240250033972,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.407349843728768e-05,
+      "loss": 1.0547,
+      "step": 37505
+    },
+    {
+      "epoch": 2.5485799701046337,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.407137518684604e-05,
+      "loss": 0.9075,
+      "step": 37510
+    },
+    {
+      "epoch": 2.5489196901752953,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4069251936404405e-05,
+      "loss": 0.9892,
+      "step": 37515
+    },
+    {
+      "epoch": 2.5492594102459574,
+      "grad_norm": 7.0,
+      "learning_rate": 3.406712868596277e-05,
+      "loss": 0.9469,
+      "step": 37520
+    },
+    {
+      "epoch": 2.549599130316619,
+      "grad_norm": 6.375,
+      "learning_rate": 3.406500543552113e-05,
+      "loss": 0.9408,
+      "step": 37525
+    },
+    {
+      "epoch": 2.5499388503872806,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.40628821850795e-05,
+      "loss": 0.9598,
+      "step": 37530
+    },
+    {
+      "epoch": 2.5502785704579427,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.406075893463786e-05,
+      "loss": 1.0129,
+      "step": 37535
+    },
+    {
+      "epoch": 2.5506182905286043,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.4058635684196225e-05,
+      "loss": 0.923,
+      "step": 37540
+    },
+    {
+      "epoch": 2.550958010599266,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.405651243375459e-05,
+      "loss": 1.0136,
+      "step": 37545
+    },
+    {
+      "epoch": 2.551297730669928,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.405438918331295e-05,
+      "loss": 0.8588,
+      "step": 37550
+    },
+    {
+      "epoch": 2.5516374507405897,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.405226593287131e-05,
+      "loss": 0.9636,
+      "step": 37555
+    },
+    {
+      "epoch": 2.5519771708112513,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.405014268242968e-05,
+      "loss": 0.8856,
+      "step": 37560
+    },
+    {
+      "epoch": 2.5523168908819134,
+      "grad_norm": 6.5,
+      "learning_rate": 3.4048019431988045e-05,
+      "loss": 1.0038,
+      "step": 37565
+    },
+    {
+      "epoch": 2.552656610952575,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.40458961815464e-05,
+      "loss": 0.9376,
+      "step": 37570
+    },
+    {
+      "epoch": 2.5529963310232366,
+      "grad_norm": 6.875,
+      "learning_rate": 3.404377293110477e-05,
+      "loss": 0.8982,
+      "step": 37575
+    },
+    {
+      "epoch": 2.5533360510938987,
+      "grad_norm": 7.125,
+      "learning_rate": 3.404164968066314e-05,
+      "loss": 1.0314,
+      "step": 37580
+    },
+    {
+      "epoch": 2.5536757711645603,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.4039526430221494e-05,
+      "loss": 0.9417,
+      "step": 37585
+    },
+    {
+      "epoch": 2.554015491235222,
+      "grad_norm": 6.875,
+      "learning_rate": 3.4037403179779865e-05,
+      "loss": 0.8346,
+      "step": 37590
+    },
+    {
+      "epoch": 2.554355211305884,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.403527992933823e-05,
+      "loss": 0.9668,
+      "step": 37595
+    },
+    {
+      "epoch": 2.5546949313765457,
+      "grad_norm": 6.75,
+      "learning_rate": 3.4033156678896586e-05,
+      "loss": 0.995,
+      "step": 37600
+    },
+    {
+      "epoch": 2.5550346514472073,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.403103342845496e-05,
+      "loss": 1.0343,
+      "step": 37605
+    },
+    {
+      "epoch": 2.5553743715178694,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.402891017801332e-05,
+      "loss": 0.9341,
+      "step": 37610
+    },
+    {
+      "epoch": 2.555714091588531,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.402678692757168e-05,
+      "loss": 0.8579,
+      "step": 37615
+    },
+    {
+      "epoch": 2.5560538116591927,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.402466367713005e-05,
+      "loss": 1.0758,
+      "step": 37620
+    },
+    {
+      "epoch": 2.5563935317298547,
+      "grad_norm": 7.625,
+      "learning_rate": 3.4022540426688406e-05,
+      "loss": 1.0322,
+      "step": 37625
+    },
+    {
+      "epoch": 2.5567332518005164,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.402041717624677e-05,
+      "loss": 0.9736,
+      "step": 37630
+    },
+    {
+      "epoch": 2.557072971871178,
+      "grad_norm": 6.75,
+      "learning_rate": 3.401829392580514e-05,
+      "loss": 1.0561,
+      "step": 37635
+    },
+    {
+      "epoch": 2.55741269194184,
+      "grad_norm": 7.5,
+      "learning_rate": 3.40161706753635e-05,
+      "loss": 0.9419,
+      "step": 37640
+    },
+    {
+      "epoch": 2.5577524120125017,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.401404742492186e-05,
+      "loss": 0.9746,
+      "step": 37645
+    },
+    {
+      "epoch": 2.5580921320831633,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.401192417448023e-05,
+      "loss": 0.8832,
+      "step": 37650
+    },
+    {
+      "epoch": 2.5584318521538254,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.400980092403859e-05,
+      "loss": 0.9447,
+      "step": 37655
+    },
+    {
+      "epoch": 2.558771572224487,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.400767767359696e-05,
+      "loss": 0.9193,
+      "step": 37660
+    },
+    {
+      "epoch": 2.5591112922951487,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.4005554423155325e-05,
+      "loss": 0.9929,
+      "step": 37665
+    },
+    {
+      "epoch": 2.5594510123658107,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.400343117271368e-05,
+      "loss": 1.0423,
+      "step": 37670
+    },
+    {
+      "epoch": 2.5597907324364724,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.400130792227205e-05,
+      "loss": 0.9526,
+      "step": 37675
+    },
+    {
+      "epoch": 2.560130452507134,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.399918467183042e-05,
+      "loss": 0.9445,
+      "step": 37680
+    },
+    {
+      "epoch": 2.560470172577796,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.3997061421388774e-05,
+      "loss": 1.0166,
+      "step": 37685
+    },
+    {
+      "epoch": 2.5608098926484577,
+      "grad_norm": 6.625,
+      "learning_rate": 3.3994938170947145e-05,
+      "loss": 0.9127,
+      "step": 37690
+    },
+    {
+      "epoch": 2.5611496127191193,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.39928149205055e-05,
+      "loss": 0.9689,
+      "step": 37695
+    },
+    {
+      "epoch": 2.5614893327897814,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.3990691670063866e-05,
+      "loss": 0.9337,
+      "step": 37700
+    },
+    {
+      "epoch": 2.561829052860443,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.398856841962224e-05,
+      "loss": 0.9809,
+      "step": 37705
+    },
+    {
+      "epoch": 2.5621687729311047,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.3986445169180594e-05,
+      "loss": 0.8502,
+      "step": 37710
+    },
+    {
+      "epoch": 2.5625084930017668,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.398432191873896e-05,
+      "loss": 1.0465,
+      "step": 37715
+    },
+    {
+      "epoch": 2.5628482130724284,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.398219866829733e-05,
+      "loss": 0.9192,
+      "step": 37720
+    },
+    {
+      "epoch": 2.56318793314309,
+      "grad_norm": 6.375,
+      "learning_rate": 3.3980075417855686e-05,
+      "loss": 0.939,
+      "step": 37725
+    },
+    {
+      "epoch": 2.563527653213752,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.397795216741405e-05,
+      "loss": 0.9419,
+      "step": 37730
+    },
+    {
+      "epoch": 2.5638673732844137,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.397582891697242e-05,
+      "loss": 0.9983,
+      "step": 37735
+    },
+    {
+      "epoch": 2.5642070933550754,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.397370566653078e-05,
+      "loss": 1.0487,
+      "step": 37740
+    },
+    {
+      "epoch": 2.5645468134257374,
+      "grad_norm": 6.875,
+      "learning_rate": 3.397158241608914e-05,
+      "loss": 0.9789,
+      "step": 37745
+    },
+    {
+      "epoch": 2.564886533496399,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.396945916564751e-05,
+      "loss": 0.8884,
+      "step": 37750
+    },
+    {
+      "epoch": 2.5652262535670607,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.396733591520587e-05,
+      "loss": 0.9877,
+      "step": 37755
+    },
+    {
+      "epoch": 2.5655659736377228,
+      "grad_norm": 6.0,
+      "learning_rate": 3.3965212664764234e-05,
+      "loss": 0.9742,
+      "step": 37760
+    },
+    {
+      "epoch": 2.5659056937083844,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3963089414322605e-05,
+      "loss": 1.0018,
+      "step": 37765
+    },
+    {
+      "epoch": 2.566245413779046,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.396096616388096e-05,
+      "loss": 0.9138,
+      "step": 37770
+    },
+    {
+      "epoch": 2.566585133849708,
+      "grad_norm": 8.0,
+      "learning_rate": 3.3958842913439326e-05,
+      "loss": 0.9763,
+      "step": 37775
+    },
+    {
+      "epoch": 2.5669248539203697,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.395671966299769e-05,
+      "loss": 1.0246,
+      "step": 37780
+    },
+    {
+      "epoch": 2.5672645739910314,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3954596412556055e-05,
+      "loss": 0.9547,
+      "step": 37785
+    },
+    {
+      "epoch": 2.5676042940616934,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.395247316211442e-05,
+      "loss": 0.9541,
+      "step": 37790
+    },
+    {
+      "epoch": 2.567944014132355,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.395034991167278e-05,
+      "loss": 0.903,
+      "step": 37795
+    },
+    {
+      "epoch": 2.5682837342030167,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.3948226661231147e-05,
+      "loss": 0.9326,
+      "step": 37800
+    },
+    {
+      "epoch": 2.5686234542736783,
+      "grad_norm": 7.125,
+      "learning_rate": 3.394610341078951e-05,
+      "loss": 0.944,
+      "step": 37805
+    },
+    {
+      "epoch": 2.5689631743443404,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.3943980160347875e-05,
+      "loss": 1.0485,
+      "step": 37810
+    },
+    {
+      "epoch": 2.569302894415002,
+      "grad_norm": 6.75,
+      "learning_rate": 3.394185690990624e-05,
+      "loss": 1.0078,
+      "step": 37815
+    },
+    {
+      "epoch": 2.5696426144856637,
+      "grad_norm": 6.375,
+      "learning_rate": 3.39397336594646e-05,
+      "loss": 1.0375,
+      "step": 37820
+    },
+    {
+      "epoch": 2.5699823345563257,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3937610409022967e-05,
+      "loss": 0.9178,
+      "step": 37825
+    },
+    {
+      "epoch": 2.5703220546269874,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.393548715858133e-05,
+      "loss": 0.8765,
+      "step": 37830
+    },
+    {
+      "epoch": 2.570661774697649,
+      "grad_norm": 7.25,
+      "learning_rate": 3.3933363908139695e-05,
+      "loss": 0.9579,
+      "step": 37835
+    },
+    {
+      "epoch": 2.5710014947683106,
+      "grad_norm": 5.6875,
+      "learning_rate": 3.393124065769806e-05,
+      "loss": 0.9324,
+      "step": 37840
+    },
+    {
+      "epoch": 2.5713412148389727,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.392911740725642e-05,
+      "loss": 0.9473,
+      "step": 37845
+    },
+    {
+      "epoch": 2.5716809349096343,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.3926994156814787e-05,
+      "loss": 0.8936,
+      "step": 37850
+    },
+    {
+      "epoch": 2.572020654980296,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.392487090637315e-05,
+      "loss": 0.9635,
+      "step": 37855
+    },
+    {
+      "epoch": 2.572360375050958,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.3922747655931515e-05,
+      "loss": 1.0208,
+      "step": 37860
+    },
+    {
+      "epoch": 2.5727000951216197,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.392062440548988e-05,
+      "loss": 0.9729,
+      "step": 37865
+    },
+    {
+      "epoch": 2.5730398151922813,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.391850115504824e-05,
+      "loss": 0.966,
+      "step": 37870
+    },
+    {
+      "epoch": 2.5733795352629434,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.3916377904606607e-05,
+      "loss": 0.8055,
+      "step": 37875
+    },
+    {
+      "epoch": 2.573719255333605,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.391425465416497e-05,
+      "loss": 0.9968,
+      "step": 37880
+    },
+    {
+      "epoch": 2.5740589754042666,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.3912131403723335e-05,
+      "loss": 0.9714,
+      "step": 37885
+    },
+    {
+      "epoch": 2.5743986954749287,
+      "grad_norm": 7.375,
+      "learning_rate": 3.39100081532817e-05,
+      "loss": 0.9427,
+      "step": 37890
+    },
+    {
+      "epoch": 2.5747384155455904,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.3907884902840056e-05,
+      "loss": 0.9384,
+      "step": 37895
+    },
+    {
+      "epoch": 2.575078135616252,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.3905761652398427e-05,
+      "loss": 0.9723,
+      "step": 37900
+    },
+    {
+      "epoch": 2.575417855686914,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.390363840195679e-05,
+      "loss": 0.9631,
+      "step": 37905
+    },
+    {
+      "epoch": 2.5757575757575757,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.390151515151515e-05,
+      "loss": 0.948,
+      "step": 37910
+    },
+    {
+      "epoch": 2.5760972958282373,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.389939190107352e-05,
+      "loss": 1.0055,
+      "step": 37915
+    },
+    {
+      "epoch": 2.5764370158988994,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.389726865063188e-05,
+      "loss": 0.9651,
+      "step": 37920
+    },
+    {
+      "epoch": 2.576776735969561,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.389514540019024e-05,
+      "loss": 0.9757,
+      "step": 37925
+    },
+    {
+      "epoch": 2.5771164560402227,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.389302214974861e-05,
+      "loss": 0.9253,
+      "step": 37930
+    },
+    {
+      "epoch": 2.5774561761108847,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.3890898899306975e-05,
+      "loss": 0.8704,
+      "step": 37935
+    },
+    {
+      "epoch": 2.5777958961815464,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.388877564886533e-05,
+      "loss": 1.0088,
+      "step": 37940
+    },
+    {
+      "epoch": 2.578135616252208,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.38866523984237e-05,
+      "loss": 0.9656,
+      "step": 37945
+    },
+    {
+      "epoch": 2.57847533632287,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.388452914798207e-05,
+      "loss": 0.9404,
+      "step": 37950
+    },
+    {
+      "epoch": 2.5788150563935317,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.3882405897540424e-05,
+      "loss": 0.9442,
+      "step": 37955
+    },
+    {
+      "epoch": 2.5791547764641933,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.3880282647098795e-05,
+      "loss": 0.9361,
+      "step": 37960
+    },
+    {
+      "epoch": 2.5794944965348554,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.387815939665716e-05,
+      "loss": 1.0602,
+      "step": 37965
+    },
+    {
+      "epoch": 2.579834216605517,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.3876036146215516e-05,
+      "loss": 0.9038,
+      "step": 37970
+    },
+    {
+      "epoch": 2.5801739366761787,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.387391289577389e-05,
+      "loss": 1.0059,
+      "step": 37975
+    },
+    {
+      "epoch": 2.5805136567468407,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.3871789645332244e-05,
+      "loss": 0.9224,
+      "step": 37980
+    },
+    {
+      "epoch": 2.5808533768175024,
+      "grad_norm": 5.75,
+      "learning_rate": 3.386966639489061e-05,
+      "loss": 0.9392,
+      "step": 37985
+    },
+    {
+      "epoch": 2.581193096888164,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.386754314444898e-05,
+      "loss": 0.9508,
+      "step": 37990
+    },
+    {
+      "epoch": 2.581532816958826,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.3865419894007336e-05,
+      "loss": 0.9862,
+      "step": 37995
+    },
+    {
+      "epoch": 2.5818725370294877,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.386329664356571e-05,
+      "loss": 0.9757,
+      "step": 38000
+    },
+    {
+      "epoch": 2.5822122571001493,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.386117339312407e-05,
+      "loss": 0.9796,
+      "step": 38005
+    },
+    {
+      "epoch": 2.5825519771708114,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.385905014268243e-05,
+      "loss": 1.0727,
+      "step": 38010
+    },
+    {
+      "epoch": 2.582891697241473,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.38569268922408e-05,
+      "loss": 1.0297,
+      "step": 38015
+    },
+    {
+      "epoch": 2.5832314173121347,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.385480364179916e-05,
+      "loss": 1.0021,
+      "step": 38020
+    },
+    {
+      "epoch": 2.5835711373827968,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.385268039135752e-05,
+      "loss": 0.9813,
+      "step": 38025
+    },
+    {
+      "epoch": 2.5839108574534584,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.385055714091589e-05,
+      "loss": 0.9682,
+      "step": 38030
+    },
+    {
+      "epoch": 2.58425057752412,
+      "grad_norm": 7.125,
+      "learning_rate": 3.3848433890474255e-05,
+      "loss": 0.9248,
+      "step": 38035
+    },
+    {
+      "epoch": 2.584590297594782,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.384631064003261e-05,
+      "loss": 1.0491,
+      "step": 38040
+    },
+    {
+      "epoch": 2.5849300176654437,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.384418738959098e-05,
+      "loss": 0.9946,
+      "step": 38045
+    },
+    {
+      "epoch": 2.5852697377361054,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.384206413914934e-05,
+      "loss": 0.9882,
+      "step": 38050
+    },
+    {
+      "epoch": 2.5856094578067674,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3839940888707704e-05,
+      "loss": 1.0595,
+      "step": 38055
+    },
+    {
+      "epoch": 2.585949177877429,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.3837817638266075e-05,
+      "loss": 0.9637,
+      "step": 38060
+    },
+    {
+      "epoch": 2.5862888979480907,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.383569438782443e-05,
+      "loss": 0.9474,
+      "step": 38065
+    },
+    {
+      "epoch": 2.5866286180187528,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.3833571137382796e-05,
+      "loss": 1.0366,
+      "step": 38070
+    },
+    {
+      "epoch": 2.5869683380894144,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.383144788694117e-05,
+      "loss": 0.9612,
+      "step": 38075
+    },
+    {
+      "epoch": 2.587308058160076,
+      "grad_norm": 4.8125,
+      "learning_rate": 3.3829324636499524e-05,
+      "loss": 1.0267,
+      "step": 38080
+    },
+    {
+      "epoch": 2.587647778230738,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.382720138605789e-05,
+      "loss": 0.9852,
+      "step": 38085
+    },
+    {
+      "epoch": 2.5879874983013997,
+      "grad_norm": 6.75,
+      "learning_rate": 3.382507813561626e-05,
+      "loss": 1.0023,
+      "step": 38090
+    },
+    {
+      "epoch": 2.5883272183720614,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.3822954885174616e-05,
+      "loss": 0.9453,
+      "step": 38095
+    },
+    {
+      "epoch": 2.5886669384427234,
+      "grad_norm": 7.125,
+      "learning_rate": 3.382083163473298e-05,
+      "loss": 1.0469,
+      "step": 38100
+    },
+    {
+      "epoch": 2.589006658513385,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.381870838429135e-05,
+      "loss": 0.9617,
+      "step": 38105
+    },
+    {
+      "epoch": 2.5893463785840467,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.381658513384971e-05,
+      "loss": 1.0175,
+      "step": 38110
+    },
+    {
+      "epoch": 2.589686098654709,
+      "grad_norm": 6.75,
+      "learning_rate": 3.381446188340807e-05,
+      "loss": 1.0049,
+      "step": 38115
+    },
+    {
+      "epoch": 2.5900258187253704,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.3812338632966436e-05,
+      "loss": 0.9203,
+      "step": 38120
+    },
+    {
+      "epoch": 2.590365538796032,
+      "grad_norm": 7.625,
+      "learning_rate": 3.38102153825248e-05,
+      "loss": 1.0335,
+      "step": 38125
+    },
+    {
+      "epoch": 2.590705258866694,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.3808092132083164e-05,
+      "loss": 0.9438,
+      "step": 38130
+    },
+    {
+      "epoch": 2.5910449789373557,
+      "grad_norm": 7.75,
+      "learning_rate": 3.380596888164153e-05,
+      "loss": 0.9553,
+      "step": 38135
+    },
+    {
+      "epoch": 2.5913846990080174,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.380384563119989e-05,
+      "loss": 0.9393,
+      "step": 38140
+    },
+    {
+      "epoch": 2.591724419078679,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.3801722380758256e-05,
+      "loss": 0.9518,
+      "step": 38145
+    },
+    {
+      "epoch": 2.592064139149341,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.379959913031662e-05,
+      "loss": 0.8581,
+      "step": 38150
+    },
+    {
+      "epoch": 2.5924038592200027,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3797475879874984e-05,
+      "loss": 1.0507,
+      "step": 38155
+    },
+    {
+      "epoch": 2.5927435792906643,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.379535262943335e-05,
+      "loss": 0.9186,
+      "step": 38160
+    },
+    {
+      "epoch": 2.5930832993613264,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.379322937899171e-05,
+      "loss": 1.0248,
+      "step": 38165
+    },
+    {
+      "epoch": 2.593423019431988,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.3791106128550076e-05,
+      "loss": 0.922,
+      "step": 38170
+    },
+    {
+      "epoch": 2.5937627395026497,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.378898287810844e-05,
+      "loss": 0.9524,
+      "step": 38175
+    },
+    {
+      "epoch": 2.5941024595733113,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.3786859627666804e-05,
+      "loss": 1.0112,
+      "step": 38180
+    },
+    {
+      "epoch": 2.5944421796439734,
+      "grad_norm": 7.25,
+      "learning_rate": 3.378473637722517e-05,
+      "loss": 0.9812,
+      "step": 38185
+    },
+    {
+      "epoch": 2.594781899714635,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.378261312678353e-05,
+      "loss": 0.9316,
+      "step": 38190
+    },
+    {
+      "epoch": 2.5951216197852967,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.3780489876341896e-05,
+      "loss": 1.0192,
+      "step": 38195
+    },
+    {
+      "epoch": 2.5954613398559587,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.377836662590026e-05,
+      "loss": 0.9539,
+      "step": 38200
+    },
+    {
+      "epoch": 2.5958010599266204,
+      "grad_norm": 6.5,
+      "learning_rate": 3.3776243375458624e-05,
+      "loss": 0.9811,
+      "step": 38205
+    },
+    {
+      "epoch": 2.596140779997282,
+      "grad_norm": 6.75,
+      "learning_rate": 3.377412012501699e-05,
+      "loss": 1.035,
+      "step": 38210
+    },
+    {
+      "epoch": 2.596480500067944,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.377199687457535e-05,
+      "loss": 0.9324,
+      "step": 38215
+    },
+    {
+      "epoch": 2.5968202201386057,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.3769873624133716e-05,
+      "loss": 0.883,
+      "step": 38220
+    },
+    {
+      "epoch": 2.5971599402092673,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.376775037369208e-05,
+      "loss": 1.0517,
+      "step": 38225
+    },
+    {
+      "epoch": 2.5974996602799294,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.3765627123250444e-05,
+      "loss": 0.9712,
+      "step": 38230
+    },
+    {
+      "epoch": 2.597839380350591,
+      "grad_norm": 6.5,
+      "learning_rate": 3.376350387280881e-05,
+      "loss": 0.9388,
+      "step": 38235
+    },
+    {
+      "epoch": 2.5981791004212527,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.376138062236717e-05,
+      "loss": 1.0052,
+      "step": 38240
+    },
+    {
+      "epoch": 2.5985188204919147,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.3759257371925536e-05,
+      "loss": 0.9872,
+      "step": 38245
+    },
+    {
+      "epoch": 2.5988585405625764,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.375713412148389e-05,
+      "loss": 0.9797,
+      "step": 38250
+    },
+    {
+      "epoch": 2.599198260633238,
+      "grad_norm": 7.875,
+      "learning_rate": 3.3755010871042264e-05,
+      "loss": 1.037,
+      "step": 38255
+    },
+    {
+      "epoch": 2.5995379807039,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.375288762060063e-05,
+      "loss": 1.0082,
+      "step": 38260
+    },
+    {
+      "epoch": 2.5998777007745617,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.3750764370158985e-05,
+      "loss": 0.9784,
+      "step": 38265
+    },
+    {
+      "epoch": 2.6002174208452233,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.3748641119717356e-05,
+      "loss": 0.9724,
+      "step": 38270
+    },
+    {
+      "epoch": 2.6005571409158854,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.374651786927572e-05,
+      "loss": 0.9581,
+      "step": 38275
+    },
+    {
+      "epoch": 2.600896860986547,
+      "grad_norm": 6.875,
+      "learning_rate": 3.374439461883408e-05,
+      "loss": 0.9326,
+      "step": 38280
+    },
+    {
+      "epoch": 2.6012365810572087,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.374227136839245e-05,
+      "loss": 1.1418,
+      "step": 38285
+    },
+    {
+      "epoch": 2.6015763011278707,
+      "grad_norm": 6.625,
+      "learning_rate": 3.374014811795081e-05,
+      "loss": 0.9564,
+      "step": 38290
+    },
+    {
+      "epoch": 2.6019160211985324,
+      "grad_norm": 6.75,
+      "learning_rate": 3.373802486750917e-05,
+      "loss": 1.0547,
+      "step": 38295
+    },
+    {
+      "epoch": 2.602255741269194,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.373590161706754e-05,
+      "loss": 0.9814,
+      "step": 38300
+    },
+    {
+      "epoch": 2.602595461339856,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.3733778366625904e-05,
+      "loss": 0.9176,
+      "step": 38305
+    },
+    {
+      "epoch": 2.6029351814105177,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.373165511618426e-05,
+      "loss": 0.897,
+      "step": 38310
+    },
+    {
+      "epoch": 2.6032749014811793,
+      "grad_norm": 7.375,
+      "learning_rate": 3.372953186574263e-05,
+      "loss": 0.9829,
+      "step": 38315
+    },
+    {
+      "epoch": 2.6036146215518414,
+      "grad_norm": 6.625,
+      "learning_rate": 3.372740861530099e-05,
+      "loss": 1.025,
+      "step": 38320
+    },
+    {
+      "epoch": 2.603954341622503,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.372528536485935e-05,
+      "loss": 1.0025,
+      "step": 38325
+    },
+    {
+      "epoch": 2.6042940616931647,
+      "grad_norm": 7.125,
+      "learning_rate": 3.3723162114417724e-05,
+      "loss": 1.0852,
+      "step": 38330
+    },
+    {
+      "epoch": 2.6046337817638268,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.372103886397608e-05,
+      "loss": 0.9145,
+      "step": 38335
+    },
+    {
+      "epoch": 2.6049735018344884,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.371891561353445e-05,
+      "loss": 1.0414,
+      "step": 38340
+    },
+    {
+      "epoch": 2.60531322190515,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.3716792363092816e-05,
+      "loss": 0.94,
+      "step": 38345
+    },
+    {
+      "epoch": 2.605652941975812,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3714669112651173e-05,
+      "loss": 0.9068,
+      "step": 38350
+    },
+    {
+      "epoch": 2.6059926620464737,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.3712545862209544e-05,
+      "loss": 0.9889,
+      "step": 38355
+    },
+    {
+      "epoch": 2.6063323821171354,
+      "grad_norm": 7.375,
+      "learning_rate": 3.371042261176791e-05,
+      "loss": 1.0062,
+      "step": 38360
+    },
+    {
+      "epoch": 2.6066721021877974,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.3708299361326265e-05,
+      "loss": 1.0146,
+      "step": 38365
+    },
+    {
+      "epoch": 2.607011822258459,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.3706176110884636e-05,
+      "loss": 1.0264,
+      "step": 38370
+    },
+    {
+      "epoch": 2.6073515423291207,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.3704052860443e-05,
+      "loss": 1.0751,
+      "step": 38375
+    },
+    {
+      "epoch": 2.6076912623997828,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.370192961000136e-05,
+      "loss": 0.9416,
+      "step": 38380
+    },
+    {
+      "epoch": 2.6080309824704444,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.369980635955973e-05,
+      "loss": 0.9323,
+      "step": 38385
+    },
+    {
+      "epoch": 2.608370702541106,
+      "grad_norm": 6.625,
+      "learning_rate": 3.369768310911809e-05,
+      "loss": 0.9402,
+      "step": 38390
+    },
+    {
+      "epoch": 2.608710422611768,
+      "grad_norm": 5.46875,
+      "learning_rate": 3.369555985867645e-05,
+      "loss": 1.0856,
+      "step": 38395
+    },
+    {
+      "epoch": 2.6090501426824297,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.369343660823482e-05,
+      "loss": 0.989,
+      "step": 38400
+    },
+    {
+      "epoch": 2.6093898627530914,
+      "grad_norm": 7.0,
+      "learning_rate": 3.369131335779318e-05,
+      "loss": 0.9517,
+      "step": 38405
+    },
+    {
+      "epoch": 2.6097295828237534,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.368919010735154e-05,
+      "loss": 0.9352,
+      "step": 38410
+    },
+    {
+      "epoch": 2.610069302894415,
+      "grad_norm": 9.125,
+      "learning_rate": 3.368706685690991e-05,
+      "loss": 0.8857,
+      "step": 38415
+    },
+    {
+      "epoch": 2.6104090229650767,
+      "grad_norm": 6.5,
+      "learning_rate": 3.368494360646827e-05,
+      "loss": 0.9911,
+      "step": 38420
+    },
+    {
+      "epoch": 2.610748743035739,
+      "grad_norm": 6.0,
+      "learning_rate": 3.3682820356026633e-05,
+      "loss": 0.9348,
+      "step": 38425
+    },
+    {
+      "epoch": 2.6110884631064004,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.3680697105585004e-05,
+      "loss": 0.9763,
+      "step": 38430
+    },
+    {
+      "epoch": 2.611428183177062,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.367857385514336e-05,
+      "loss": 0.8508,
+      "step": 38435
+    },
+    {
+      "epoch": 2.611767903247724,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.3676450604701725e-05,
+      "loss": 0.8812,
+      "step": 38440
+    },
+    {
+      "epoch": 2.6121076233183858,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3674327354260096e-05,
+      "loss": 0.848,
+      "step": 38445
+    },
+    {
+      "epoch": 2.6124473433890474,
+      "grad_norm": 7.0,
+      "learning_rate": 3.3672204103818453e-05,
+      "loss": 0.9866,
+      "step": 38450
+    },
+    {
+      "epoch": 2.6127870634597095,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.367008085337682e-05,
+      "loss": 0.9961,
+      "step": 38455
+    },
+    {
+      "epoch": 2.613126783530371,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.366795760293519e-05,
+      "loss": 0.9669,
+      "step": 38460
+    },
+    {
+      "epoch": 2.6134665036010327,
+      "grad_norm": 5.75,
+      "learning_rate": 3.3665834352493545e-05,
+      "loss": 0.9989,
+      "step": 38465
+    },
+    {
+      "epoch": 2.613806223671695,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.366371110205191e-05,
+      "loss": 0.9636,
+      "step": 38470
+    },
+    {
+      "epoch": 2.6141459437423564,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3661587851610273e-05,
+      "loss": 0.9701,
+      "step": 38475
+    },
+    {
+      "epoch": 2.614485663813018,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.365946460116864e-05,
+      "loss": 0.9975,
+      "step": 38480
+    },
+    {
+      "epoch": 2.6148253838836797,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.3657341350727e-05,
+      "loss": 1.0161,
+      "step": 38485
+    },
+    {
+      "epoch": 2.6151651039543418,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.3655218100285365e-05,
+      "loss": 0.9153,
+      "step": 38490
+    },
+    {
+      "epoch": 2.6155048240250034,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.365309484984373e-05,
+      "loss": 0.968,
+      "step": 38495
+    },
+    {
+      "epoch": 2.615844544095665,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.3650971599402094e-05,
+      "loss": 0.952,
+      "step": 38500
+    },
+    {
+      "epoch": 2.616184264166327,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.364884834896046e-05,
+      "loss": 0.9414,
+      "step": 38505
+    },
+    {
+      "epoch": 2.6165239842369887,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.364672509851882e-05,
+      "loss": 0.8506,
+      "step": 38510
+    },
+    {
+      "epoch": 2.6168637043076504,
+      "grad_norm": 7.25,
+      "learning_rate": 3.3644601848077186e-05,
+      "loss": 0.8937,
+      "step": 38515
+    },
+    {
+      "epoch": 2.617203424378312,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.364247859763555e-05,
+      "loss": 0.9675,
+      "step": 38520
+    },
+    {
+      "epoch": 2.617543144448974,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.3640355347193914e-05,
+      "loss": 0.9929,
+      "step": 38525
+    },
+    {
+      "epoch": 2.6178828645196357,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.363823209675228e-05,
+      "loss": 1.0229,
+      "step": 38530
+    },
+    {
+      "epoch": 2.6182225845902973,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.363610884631064e-05,
+      "loss": 0.9154,
+      "step": 38535
+    },
+    {
+      "epoch": 2.6185623046609594,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.3633985595869006e-05,
+      "loss": 0.929,
+      "step": 38540
+    },
+    {
+      "epoch": 2.618902024731621,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.363186234542737e-05,
+      "loss": 0.9692,
+      "step": 38545
+    },
+    {
+      "epoch": 2.6192417448022827,
+      "grad_norm": 6.75,
+      "learning_rate": 3.3629739094985734e-05,
+      "loss": 0.8906,
+      "step": 38550
+    },
+    {
+      "epoch": 2.6195814648729447,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.36276158445441e-05,
+      "loss": 0.8928,
+      "step": 38555
+    },
+    {
+      "epoch": 2.6199211849436064,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.362549259410246e-05,
+      "loss": 0.9476,
+      "step": 38560
+    },
+    {
+      "epoch": 2.620260905014268,
+      "grad_norm": 6.125,
+      "learning_rate": 3.3623369343660826e-05,
+      "loss": 0.9099,
+      "step": 38565
+    },
+    {
+      "epoch": 2.62060062508493,
+      "grad_norm": 7.0,
+      "learning_rate": 3.362124609321919e-05,
+      "loss": 0.9678,
+      "step": 38570
+    },
+    {
+      "epoch": 2.6209403451555917,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.3619122842777554e-05,
+      "loss": 0.931,
+      "step": 38575
+    },
+    {
+      "epoch": 2.6212800652262533,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.361699959233592e-05,
+      "loss": 0.9848,
+      "step": 38580
+    },
+    {
+      "epoch": 2.6216197852969154,
+      "grad_norm": 7.625,
+      "learning_rate": 3.361487634189428e-05,
+      "loss": 0.9359,
+      "step": 38585
+    },
+    {
+      "epoch": 2.621959505367577,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.3612753091452646e-05,
+      "loss": 1.0189,
+      "step": 38590
+    },
+    {
+      "epoch": 2.6222992254382387,
+      "grad_norm": 8.5,
+      "learning_rate": 3.361062984101101e-05,
+      "loss": 1.0052,
+      "step": 38595
+    },
+    {
+      "epoch": 2.6226389455089008,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.3608506590569374e-05,
+      "loss": 0.8887,
+      "step": 38600
+    },
+    {
+      "epoch": 2.6229786655795624,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.360638334012773e-05,
+      "loss": 0.9638,
+      "step": 38605
+    },
+    {
+      "epoch": 2.623318385650224,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.36042600896861e-05,
+      "loss": 0.9659,
+      "step": 38610
+    },
+    {
+      "epoch": 2.623658105720886,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.3602136839244466e-05,
+      "loss": 0.9928,
+      "step": 38615
+    },
+    {
+      "epoch": 2.6239978257915477,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.360001358880282e-05,
+      "loss": 0.9155,
+      "step": 38620
+    },
+    {
+      "epoch": 2.6243375458622094,
+      "grad_norm": 8.125,
+      "learning_rate": 3.3597890338361194e-05,
+      "loss": 1.0548,
+      "step": 38625
+    },
+    {
+      "epoch": 2.6246772659328714,
+      "grad_norm": 7.375,
+      "learning_rate": 3.359576708791956e-05,
+      "loss": 1.0822,
+      "step": 38630
+    },
+    {
+      "epoch": 2.625016986003533,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.3593643837477915e-05,
+      "loss": 0.9816,
+      "step": 38635
+    },
+    {
+      "epoch": 2.6253567060741947,
+      "grad_norm": 6.5,
+      "learning_rate": 3.3591520587036286e-05,
+      "loss": 0.9607,
+      "step": 38640
+    },
+    {
+      "epoch": 2.6256964261448568,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.358939733659465e-05,
+      "loss": 1.0506,
+      "step": 38645
+    },
+    {
+      "epoch": 2.6260361462155184,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.358727408615301e-05,
+      "loss": 1.0345,
+      "step": 38650
+    },
+    {
+      "epoch": 2.62637586628618,
+      "grad_norm": 6.375,
+      "learning_rate": 3.358515083571138e-05,
+      "loss": 0.8715,
+      "step": 38655
+    },
+    {
+      "epoch": 2.626715586356842,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.358302758526974e-05,
+      "loss": 0.9806,
+      "step": 38660
+    },
+    {
+      "epoch": 2.6270553064275037,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.35809043348281e-05,
+      "loss": 0.9933,
+      "step": 38665
+    },
+    {
+      "epoch": 2.6273950264981654,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.357878108438647e-05,
+      "loss": 0.9694,
+      "step": 38670
+    },
+    {
+      "epoch": 2.6277347465688274,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.357665783394483e-05,
+      "loss": 0.7689,
+      "step": 38675
+    },
+    {
+      "epoch": 2.628074466639489,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.35745345835032e-05,
+      "loss": 1.0193,
+      "step": 38680
+    },
+    {
+      "epoch": 2.6284141867101507,
+      "grad_norm": 7.375,
+      "learning_rate": 3.357241133306156e-05,
+      "loss": 1.0201,
+      "step": 38685
+    },
+    {
+      "epoch": 2.6287539067808128,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.357028808261992e-05,
+      "loss": 1.0179,
+      "step": 38690
+    },
+    {
+      "epoch": 2.6290936268514744,
+      "grad_norm": 6.5,
+      "learning_rate": 3.356816483217829e-05,
+      "loss": 0.9569,
+      "step": 38695
+    },
+    {
+      "epoch": 2.629433346922136,
+      "grad_norm": 6.875,
+      "learning_rate": 3.3566041581736654e-05,
+      "loss": 1.0505,
+      "step": 38700
+    },
+    {
+      "epoch": 2.629773066992798,
+      "grad_norm": 8.125,
+      "learning_rate": 3.356391833129501e-05,
+      "loss": 0.9653,
+      "step": 38705
+    },
+    {
+      "epoch": 2.6301127870634597,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.356179508085338e-05,
+      "loss": 1.0398,
+      "step": 38710
+    },
+    {
+      "epoch": 2.6304525071341214,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3559671830411746e-05,
+      "loss": 0.9507,
+      "step": 38715
+    },
+    {
+      "epoch": 2.6307922272047835,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.35575485799701e-05,
+      "loss": 0.971,
+      "step": 38720
+    },
+    {
+      "epoch": 2.631131947275445,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3555425329528474e-05,
+      "loss": 1.0353,
+      "step": 38725
+    },
+    {
+      "epoch": 2.6314716673461067,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.355330207908684e-05,
+      "loss": 0.8928,
+      "step": 38730
+    },
+    {
+      "epoch": 2.631811387416769,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3551178828645195e-05,
+      "loss": 1.0141,
+      "step": 38735
+    },
+    {
+      "epoch": 2.6321511074874304,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.3549055578203566e-05,
+      "loss": 0.959,
+      "step": 38740
+    },
+    {
+      "epoch": 2.632490827558092,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.354693232776192e-05,
+      "loss": 1.0404,
+      "step": 38745
+    },
+    {
+      "epoch": 2.632830547628754,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.354480907732029e-05,
+      "loss": 0.9279,
+      "step": 38750
+    },
+    {
+      "epoch": 2.6331702676994158,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.354268582687866e-05,
+      "loss": 0.909,
+      "step": 38755
+    },
+    {
+      "epoch": 2.6335099877700774,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.3540562576437015e-05,
+      "loss": 0.8892,
+      "step": 38760
+    },
+    {
+      "epoch": 2.6338497078407395,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.353843932599538e-05,
+      "loss": 0.9588,
+      "step": 38765
+    },
+    {
+      "epoch": 2.634189427911401,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.353631607555375e-05,
+      "loss": 0.9048,
+      "step": 38770
+    },
+    {
+      "epoch": 2.6345291479820627,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.353419282511211e-05,
+      "loss": 0.9225,
+      "step": 38775
+    },
+    {
+      "epoch": 2.634868868052725,
+      "grad_norm": 6.875,
+      "learning_rate": 3.353206957467047e-05,
+      "loss": 1.0186,
+      "step": 38780
+    },
+    {
+      "epoch": 2.6352085881233864,
+      "grad_norm": 7.375,
+      "learning_rate": 3.352994632422884e-05,
+      "loss": 0.9235,
+      "step": 38785
+    },
+    {
+      "epoch": 2.635548308194048,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.35278230737872e-05,
+      "loss": 0.9496,
+      "step": 38790
+    },
+    {
+      "epoch": 2.63588802826471,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.352569982334556e-05,
+      "loss": 0.9379,
+      "step": 38795
+    },
+    {
+      "epoch": 2.6362277483353718,
+      "grad_norm": 7.125,
+      "learning_rate": 3.3523576572903934e-05,
+      "loss": 0.9714,
+      "step": 38800
+    },
+    {
+      "epoch": 2.6365674684060334,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.352145332246229e-05,
+      "loss": 0.9506,
+      "step": 38805
+    },
+    {
+      "epoch": 2.6369071884766955,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.3519330072020655e-05,
+      "loss": 0.9604,
+      "step": 38810
+    },
+    {
+      "epoch": 2.637246908547357,
+      "grad_norm": 7.0,
+      "learning_rate": 3.3517206821579026e-05,
+      "loss": 0.8965,
+      "step": 38815
+    },
+    {
+      "epoch": 2.6375866286180187,
+      "grad_norm": 7.5,
+      "learning_rate": 3.351508357113738e-05,
+      "loss": 0.9845,
+      "step": 38820
+    },
+    {
+      "epoch": 2.6379263486886804,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.351296032069575e-05,
+      "loss": 0.9302,
+      "step": 38825
+    },
+    {
+      "epoch": 2.6382660687593424,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.351083707025411e-05,
+      "loss": 0.986,
+      "step": 38830
+    },
+    {
+      "epoch": 2.638605788830004,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.3508713819812475e-05,
+      "loss": 1.0021,
+      "step": 38835
+    },
+    {
+      "epoch": 2.6389455089006657,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.350659056937084e-05,
+      "loss": 0.9151,
+      "step": 38840
+    },
+    {
+      "epoch": 2.639285228971328,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.35044673189292e-05,
+      "loss": 1.008,
+      "step": 38845
+    },
+    {
+      "epoch": 2.6396249490419894,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.350234406848757e-05,
+      "loss": 1.0095,
+      "step": 38850
+    },
+    {
+      "epoch": 2.639964669112651,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.350022081804593e-05,
+      "loss": 0.9794,
+      "step": 38855
+    },
+    {
+      "epoch": 2.640304389183313,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.3498097567604295e-05,
+      "loss": 0.9293,
+      "step": 38860
+    },
+    {
+      "epoch": 2.6406441092539747,
+      "grad_norm": 6.75,
+      "learning_rate": 3.349597431716266e-05,
+      "loss": 0.9517,
+      "step": 38865
+    },
+    {
+      "epoch": 2.6409838293246364,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.349385106672102e-05,
+      "loss": 0.9781,
+      "step": 38870
+    },
+    {
+      "epoch": 2.641323549395298,
+      "grad_norm": 8.125,
+      "learning_rate": 3.349172781627939e-05,
+      "loss": 1.0221,
+      "step": 38875
+    },
+    {
+      "epoch": 2.64166326946596,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.348960456583775e-05,
+      "loss": 0.8984,
+      "step": 38880
+    },
+    {
+      "epoch": 2.6420029895366217,
+      "grad_norm": 6.875,
+      "learning_rate": 3.3487481315396115e-05,
+      "loss": 1.0597,
+      "step": 38885
+    },
+    {
+      "epoch": 2.6423427096072833,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.348535806495448e-05,
+      "loss": 1.0124,
+      "step": 38890
+    },
+    {
+      "epoch": 2.6426824296779454,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.348323481451284e-05,
+      "loss": 0.9384,
+      "step": 38895
+    },
+    {
+      "epoch": 2.643022149748607,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.348111156407121e-05,
+      "loss": 0.9752,
+      "step": 38900
+    },
+    {
+      "epoch": 2.6433618698192687,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.347898831362957e-05,
+      "loss": 0.9672,
+      "step": 38905
+    },
+    {
+      "epoch": 2.6437015898899308,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.3476865063187935e-05,
+      "loss": 0.942,
+      "step": 38910
+    },
+    {
+      "epoch": 2.6440413099605924,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.34747418127463e-05,
+      "loss": 0.9813,
+      "step": 38915
+    },
+    {
+      "epoch": 2.644381030031254,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.347261856230466e-05,
+      "loss": 1.0153,
+      "step": 38920
+    },
+    {
+      "epoch": 2.644720750101916,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.347049531186303e-05,
+      "loss": 0.9462,
+      "step": 38925
+    },
+    {
+      "epoch": 2.6450604701725777,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.346837206142139e-05,
+      "loss": 0.9025,
+      "step": 38930
+    },
+    {
+      "epoch": 2.6454001902432394,
+      "grad_norm": 6.5,
+      "learning_rate": 3.3466248810979755e-05,
+      "loss": 0.9405,
+      "step": 38935
+    },
+    {
+      "epoch": 2.6457399103139014,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.346412556053812e-05,
+      "loss": 0.8863,
+      "step": 38940
+    },
+    {
+      "epoch": 2.646079630384563,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.3462002310096476e-05,
+      "loss": 0.995,
+      "step": 38945
+    },
+    {
+      "epoch": 2.6464193504552247,
+      "grad_norm": 7.0,
+      "learning_rate": 3.345987905965485e-05,
+      "loss": 0.9205,
+      "step": 38950
+    },
+    {
+      "epoch": 2.6467590705258868,
+      "grad_norm": 6.5,
+      "learning_rate": 3.345775580921321e-05,
+      "loss": 0.9721,
+      "step": 38955
+    },
+    {
+      "epoch": 2.6470987905965484,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.345563255877157e-05,
+      "loss": 1.0002,
+      "step": 38960
+    },
+    {
+      "epoch": 2.64743851066721,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.345350930832994e-05,
+      "loss": 0.9585,
+      "step": 38965
+    },
+    {
+      "epoch": 2.647778230737872,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.34513860578883e-05,
+      "loss": 0.897,
+      "step": 38970
+    },
+    {
+      "epoch": 2.6481179508085337,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.344926280744666e-05,
+      "loss": 0.9114,
+      "step": 38975
+    },
+    {
+      "epoch": 2.6484576708791954,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.344713955700503e-05,
+      "loss": 1.0111,
+      "step": 38980
+    },
+    {
+      "epoch": 2.6487973909498574,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.3445016306563395e-05,
+      "loss": 0.9208,
+      "step": 38985
+    },
+    {
+      "epoch": 2.649137111020519,
+      "grad_norm": 7.125,
+      "learning_rate": 3.344289305612175e-05,
+      "loss": 0.9307,
+      "step": 38990
+    },
+    {
+      "epoch": 2.6494768310911807,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.344076980568012e-05,
+      "loss": 0.977,
+      "step": 38995
+    },
+    {
+      "epoch": 2.649816551161843,
+      "grad_norm": 6.75,
+      "learning_rate": 3.343864655523849e-05,
+      "loss": 1.0062,
+      "step": 39000
+    },
+    {
+      "epoch": 2.6501562712325044,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.3436523304796844e-05,
+      "loss": 0.9037,
+      "step": 39005
+    },
+    {
+      "epoch": 2.650495991303166,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3434400054355215e-05,
+      "loss": 0.979,
+      "step": 39010
+    },
+    {
+      "epoch": 2.650835711373828,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.343227680391358e-05,
+      "loss": 0.9723,
+      "step": 39015
+    },
+    {
+      "epoch": 2.6511754314444897,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.343015355347194e-05,
+      "loss": 0.9228,
+      "step": 39020
+    },
+    {
+      "epoch": 2.6515151515151514,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.342803030303031e-05,
+      "loss": 0.9459,
+      "step": 39025
+    },
+    {
+      "epoch": 2.6518548715858135,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.3425907052588664e-05,
+      "loss": 0.9616,
+      "step": 39030
+    },
+    {
+      "epoch": 2.652194591656475,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.3423783802147035e-05,
+      "loss": 0.9103,
+      "step": 39035
+    },
+    {
+      "epoch": 2.6525343117271367,
+      "grad_norm": 9.4375,
+      "learning_rate": 3.34216605517054e-05,
+      "loss": 1.0174,
+      "step": 39040
+    },
+    {
+      "epoch": 2.652874031797799,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3419537301263756e-05,
+      "loss": 0.901,
+      "step": 39045
+    },
+    {
+      "epoch": 2.6532137518684604,
+      "grad_norm": 6.75,
+      "learning_rate": 3.341741405082213e-05,
+      "loss": 0.8519,
+      "step": 39050
+    },
+    {
+      "epoch": 2.653553471939122,
+      "grad_norm": 6.625,
+      "learning_rate": 3.341529080038049e-05,
+      "loss": 0.8231,
+      "step": 39055
+    },
+    {
+      "epoch": 2.653893192009784,
+      "grad_norm": 6.375,
+      "learning_rate": 3.341316754993885e-05,
+      "loss": 1.0659,
+      "step": 39060
+    },
+    {
+      "epoch": 2.6542329120804458,
+      "grad_norm": 6.625,
+      "learning_rate": 3.341104429949722e-05,
+      "loss": 0.8945,
+      "step": 39065
+    },
+    {
+      "epoch": 2.6545726321511074,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.340892104905558e-05,
+      "loss": 1.0813,
+      "step": 39070
+    },
+    {
+      "epoch": 2.6549123522217695,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.340679779861394e-05,
+      "loss": 1.0184,
+      "step": 39075
+    },
+    {
+      "epoch": 2.655252072292431,
+      "grad_norm": 6.5,
+      "learning_rate": 3.340467454817231e-05,
+      "loss": 0.9811,
+      "step": 39080
+    },
+    {
+      "epoch": 2.6555917923630927,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.3402551297730675e-05,
+      "loss": 0.9268,
+      "step": 39085
+    },
+    {
+      "epoch": 2.655931512433755,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.340042804728903e-05,
+      "loss": 0.9054,
+      "step": 39090
+    },
+    {
+      "epoch": 2.6562712325044164,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.33983047968474e-05,
+      "loss": 0.9813,
+      "step": 39095
+    },
+    {
+      "epoch": 2.656610952575078,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.339618154640576e-05,
+      "loss": 0.9389,
+      "step": 39100
+    },
+    {
+      "epoch": 2.65695067264574,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.3394058295964124e-05,
+      "loss": 0.8714,
+      "step": 39105
+    },
+    {
+      "epoch": 2.6572903927164018,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3391935045522495e-05,
+      "loss": 0.9114,
+      "step": 39110
+    },
+    {
+      "epoch": 2.6576301127870634,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.338981179508085e-05,
+      "loss": 0.9682,
+      "step": 39115
+    },
+    {
+      "epoch": 2.6579698328577255,
+      "grad_norm": 6.875,
+      "learning_rate": 3.3387688544639216e-05,
+      "loss": 1.0281,
+      "step": 39120
+    },
+    {
+      "epoch": 2.658309552928387,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.338556529419759e-05,
+      "loss": 0.9242,
+      "step": 39125
+    },
+    {
+      "epoch": 2.6586492729990487,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.3383442043755944e-05,
+      "loss": 0.91,
+      "step": 39130
+    },
+    {
+      "epoch": 2.658988993069711,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.338131879331431e-05,
+      "loss": 0.9622,
+      "step": 39135
+    },
+    {
+      "epoch": 2.6593287131403724,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.337919554287268e-05,
+      "loss": 0.9322,
+      "step": 39140
+    },
+    {
+      "epoch": 2.659668433211034,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.3377072292431036e-05,
+      "loss": 1.0291,
+      "step": 39145
+    },
+    {
+      "epoch": 2.660008153281696,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.33749490419894e-05,
+      "loss": 0.9204,
+      "step": 39150
+    },
+    {
+      "epoch": 2.660347873352358,
+      "grad_norm": 7.125,
+      "learning_rate": 3.337282579154777e-05,
+      "loss": 1.0196,
+      "step": 39155
+    },
+    {
+      "epoch": 2.6606875934230194,
+      "grad_norm": 7.75,
+      "learning_rate": 3.337070254110613e-05,
+      "loss": 0.9704,
+      "step": 39160
+    },
+    {
+      "epoch": 2.661027313493681,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.336857929066449e-05,
+      "loss": 0.9639,
+      "step": 39165
+    },
+    {
+      "epoch": 2.661367033564343,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.336645604022286e-05,
+      "loss": 0.9285,
+      "step": 39170
+    },
+    {
+      "epoch": 2.6617067536350048,
+      "grad_norm": 6.75,
+      "learning_rate": 3.336433278978122e-05,
+      "loss": 0.9672,
+      "step": 39175
+    },
+    {
+      "epoch": 2.6620464737056664,
+      "grad_norm": 6.125,
+      "learning_rate": 3.3362209539339584e-05,
+      "loss": 0.8803,
+      "step": 39180
+    },
+    {
+      "epoch": 2.6623861937763285,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.336008628889795e-05,
+      "loss": 0.9121,
+      "step": 39185
+    },
+    {
+      "epoch": 2.66272591384699,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.335796303845631e-05,
+      "loss": 1.0091,
+      "step": 39190
+    },
+    {
+      "epoch": 2.6630656339176517,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.3355839788014676e-05,
+      "loss": 0.861,
+      "step": 39195
+    },
+    {
+      "epoch": 2.663405353988314,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.335371653757304e-05,
+      "loss": 0.9732,
+      "step": 39200
+    },
+    {
+      "epoch": 2.6637450740589754,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.3351593287131404e-05,
+      "loss": 0.9266,
+      "step": 39205
+    },
+    {
+      "epoch": 2.664084794129637,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.334947003668977e-05,
+      "loss": 1.0127,
+      "step": 39210
+    },
+    {
+      "epoch": 2.6644245142002987,
+      "grad_norm": 6.375,
+      "learning_rate": 3.334734678624813e-05,
+      "loss": 0.9259,
+      "step": 39215
+    },
+    {
+      "epoch": 2.6647642342709608,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.3345223535806497e-05,
+      "loss": 0.8924,
+      "step": 39220
+    },
+    {
+      "epoch": 2.6651039543416224,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.334310028536486e-05,
+      "loss": 0.9964,
+      "step": 39225
+    },
+    {
+      "epoch": 2.665443674412284,
+      "grad_norm": 6.5,
+      "learning_rate": 3.3340977034923225e-05,
+      "loss": 0.945,
+      "step": 39230
+    },
+    {
+      "epoch": 2.665783394482946,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.333885378448159e-05,
+      "loss": 0.9087,
+      "step": 39235
+    },
+    {
+      "epoch": 2.6661231145536077,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.333673053403995e-05,
+      "loss": 0.9807,
+      "step": 39240
+    },
+    {
+      "epoch": 2.6664628346242694,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.3334607283598317e-05,
+      "loss": 0.9546,
+      "step": 39245
+    },
+    {
+      "epoch": 2.6668025546949314,
+      "grad_norm": 5.1875,
+      "learning_rate": 3.333248403315668e-05,
+      "loss": 1.0768,
+      "step": 39250
+    },
+    {
+      "epoch": 2.667142274765593,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.3330360782715045e-05,
+      "loss": 0.8602,
+      "step": 39255
+    },
+    {
+      "epoch": 2.6674819948362547,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.332823753227341e-05,
+      "loss": 0.9876,
+      "step": 39260
+    },
+    {
+      "epoch": 2.6678217149069168,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.332611428183177e-05,
+      "loss": 0.924,
+      "step": 39265
+    },
+    {
+      "epoch": 2.6681614349775784,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3323991031390137e-05,
+      "loss": 0.9577,
+      "step": 39270
+    },
+    {
+      "epoch": 2.66850115504824,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.33218677809485e-05,
+      "loss": 1.056,
+      "step": 39275
+    },
+    {
+      "epoch": 2.668840875118902,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.3319744530506865e-05,
+      "loss": 0.9532,
+      "step": 39280
+    },
+    {
+      "epoch": 2.6691805951895637,
+      "grad_norm": 7.125,
+      "learning_rate": 3.331762128006523e-05,
+      "loss": 0.9673,
+      "step": 39285
+    },
+    {
+      "epoch": 2.6695203152602254,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.331549802962359e-05,
+      "loss": 1.0307,
+      "step": 39290
+    },
+    {
+      "epoch": 2.6698600353308874,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.3313374779181957e-05,
+      "loss": 1.07,
+      "step": 39295
+    },
+    {
+      "epoch": 2.670199755401549,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.3311251528740314e-05,
+      "loss": 0.9099,
+      "step": 39300
+    },
+    {
+      "epoch": 2.6705394754722107,
+      "grad_norm": 7.625,
+      "learning_rate": 3.3309128278298685e-05,
+      "loss": 0.9919,
+      "step": 39305
+    },
+    {
+      "epoch": 2.670879195542873,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.330700502785705e-05,
+      "loss": 0.9359,
+      "step": 39310
+    },
+    {
+      "epoch": 2.6712189156135344,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.3304881777415406e-05,
+      "loss": 1.0645,
+      "step": 39315
+    },
+    {
+      "epoch": 2.671558635684196,
+      "grad_norm": 6.75,
+      "learning_rate": 3.3302758526973777e-05,
+      "loss": 1.236,
+      "step": 39320
+    },
+    {
+      "epoch": 2.671898355754858,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.330063527653214e-05,
+      "loss": 0.8997,
+      "step": 39325
+    },
+    {
+      "epoch": 2.6722380758255198,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.32985120260905e-05,
+      "loss": 0.9826,
+      "step": 39330
+    },
+    {
+      "epoch": 2.6725777958961814,
+      "grad_norm": 6.75,
+      "learning_rate": 3.329638877564887e-05,
+      "loss": 0.9257,
+      "step": 39335
+    },
+    {
+      "epoch": 2.6729175159668435,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.329426552520723e-05,
+      "loss": 1.0116,
+      "step": 39340
+    },
+    {
+      "epoch": 2.673257236037505,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.329214227476559e-05,
+      "loss": 0.9188,
+      "step": 39345
+    },
+    {
+      "epoch": 2.6735969561081667,
+      "grad_norm": 7.375,
+      "learning_rate": 3.329001902432396e-05,
+      "loss": 1.0077,
+      "step": 39350
+    },
+    {
+      "epoch": 2.673936676178829,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.3287895773882325e-05,
+      "loss": 1.0537,
+      "step": 39355
+    },
+    {
+      "epoch": 2.6742763962494904,
+      "grad_norm": 7.125,
+      "learning_rate": 3.328577252344069e-05,
+      "loss": 1.0256,
+      "step": 39360
+    },
+    {
+      "epoch": 2.674616116320152,
+      "grad_norm": 7.75,
+      "learning_rate": 3.328364927299905e-05,
+      "loss": 0.8936,
+      "step": 39365
+    },
+    {
+      "epoch": 2.674955836390814,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.328152602255742e-05,
+      "loss": 0.9091,
+      "step": 39370
+    },
+    {
+      "epoch": 2.6752955564614758,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.327940277211578e-05,
+      "loss": 0.994,
+      "step": 39375
+    },
+    {
+      "epoch": 2.6756352765321374,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.3277279521674145e-05,
+      "loss": 0.9961,
+      "step": 39380
+    },
+    {
+      "epoch": 2.6759749966027995,
+      "grad_norm": 6.75,
+      "learning_rate": 3.32751562712325e-05,
+      "loss": 0.9499,
+      "step": 39385
+    },
+    {
+      "epoch": 2.676314716673461,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.327303302079087e-05,
+      "loss": 0.9914,
+      "step": 39390
+    },
+    {
+      "epoch": 2.6766544367441227,
+      "grad_norm": 7.0,
+      "learning_rate": 3.327090977034924e-05,
+      "loss": 1.0773,
+      "step": 39395
+    },
+    {
+      "epoch": 2.676994156814785,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3268786519907594e-05,
+      "loss": 0.9931,
+      "step": 39400
+    },
+    {
+      "epoch": 2.6773338768854464,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.3266663269465965e-05,
+      "loss": 0.997,
+      "step": 39405
+    },
+    {
+      "epoch": 2.677673596956108,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.326454001902433e-05,
+      "loss": 1.043,
+      "step": 39410
+    },
+    {
+      "epoch": 2.67801331702677,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.3262416768582686e-05,
+      "loss": 0.9324,
+      "step": 39415
+    },
+    {
+      "epoch": 2.6783530370974318,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.326029351814106e-05,
+      "loss": 0.957,
+      "step": 39420
+    },
+    {
+      "epoch": 2.6786927571680934,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.325817026769942e-05,
+      "loss": 1.019,
+      "step": 39425
+    },
+    {
+      "epoch": 2.6790324772387555,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.325604701725778e-05,
+      "loss": 1.0253,
+      "step": 39430
+    },
+    {
+      "epoch": 2.679372197309417,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.325392376681615e-05,
+      "loss": 0.9815,
+      "step": 39435
+    },
+    {
+      "epoch": 2.6797119173800787,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.325180051637451e-05,
+      "loss": 1.0197,
+      "step": 39440
+    },
+    {
+      "epoch": 2.680051637450741,
+      "grad_norm": 6.625,
+      "learning_rate": 3.324967726593287e-05,
+      "loss": 1.0444,
+      "step": 39445
+    },
+    {
+      "epoch": 2.6803913575214025,
+      "grad_norm": 6.5,
+      "learning_rate": 3.324755401549124e-05,
+      "loss": 1.2443,
+      "step": 39450
+    },
+    {
+      "epoch": 2.680731077592064,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.32454307650496e-05,
+      "loss": 0.9481,
+      "step": 39455
+    },
+    {
+      "epoch": 2.681070797662726,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.324330751460796e-05,
+      "loss": 0.9722,
+      "step": 39460
+    },
+    {
+      "epoch": 2.681410517733388,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.324118426416633e-05,
+      "loss": 0.9898,
+      "step": 39465
+    },
+    {
+      "epoch": 2.6817502378040494,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.323906101372469e-05,
+      "loss": 0.9801,
+      "step": 39470
+    },
+    {
+      "epoch": 2.6820899578747115,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.3236937763283054e-05,
+      "loss": 0.9688,
+      "step": 39475
+    },
+    {
+      "epoch": 2.682429677945373,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.3234814512841425e-05,
+      "loss": 1.0062,
+      "step": 39480
+    },
+    {
+      "epoch": 2.6827693980160348,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.323269126239978e-05,
+      "loss": 0.9534,
+      "step": 39485
+    },
+    {
+      "epoch": 2.683109118086697,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.3230568011958146e-05,
+      "loss": 0.8851,
+      "step": 39490
+    },
+    {
+      "epoch": 2.6834488381573585,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.322844476151652e-05,
+      "loss": 0.9504,
+      "step": 39495
+    },
+    {
+      "epoch": 2.68378855822802,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3226321511074874e-05,
+      "loss": 0.9928,
+      "step": 39500
+    },
+    {
+      "epoch": 2.6841282782986817,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.322419826063324e-05,
+      "loss": 0.8856,
+      "step": 39505
+    },
+    {
+      "epoch": 2.684467998369344,
+      "grad_norm": 6.625,
+      "learning_rate": 3.322207501019161e-05,
+      "loss": 0.8967,
+      "step": 39510
+    },
+    {
+      "epoch": 2.6848077184400054,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.3219951759749966e-05,
+      "loss": 0.9074,
+      "step": 39515
+    },
+    {
+      "epoch": 2.685147438510667,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.321782850930833e-05,
+      "loss": 0.8863,
+      "step": 39520
+    },
+    {
+      "epoch": 2.685487158581329,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.3215705258866694e-05,
+      "loss": 0.9826,
+      "step": 39525
+    },
+    {
+      "epoch": 2.6858268786519908,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.321358200842506e-05,
+      "loss": 0.9812,
+      "step": 39530
+    },
+    {
+      "epoch": 2.6861665987226524,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.321145875798342e-05,
+      "loss": 1.0299,
+      "step": 39535
+    },
+    {
+      "epoch": 2.6865063187933145,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.3209335507541786e-05,
+      "loss": 0.9865,
+      "step": 39540
+    },
+    {
+      "epoch": 2.686846038863976,
+      "grad_norm": 6.875,
+      "learning_rate": 3.320721225710015e-05,
+      "loss": 0.9616,
+      "step": 39545
+    },
+    {
+      "epoch": 2.6871857589346377,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.3205089006658514e-05,
+      "loss": 0.9297,
+      "step": 39550
+    },
+    {
+      "epoch": 2.6875254790052994,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.320296575621688e-05,
+      "loss": 1.0477,
+      "step": 39555
+    },
+    {
+      "epoch": 2.6878651990759614,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.320084250577524e-05,
+      "loss": 0.9801,
+      "step": 39560
+    },
+    {
+      "epoch": 2.688204919146623,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.3198719255333606e-05,
+      "loss": 0.9451,
+      "step": 39565
+    },
+    {
+      "epoch": 2.6885446392172847,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.319659600489197e-05,
+      "loss": 0.9562,
+      "step": 39570
+    },
+    {
+      "epoch": 2.688884359287947,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.3194472754450334e-05,
+      "loss": 1.0213,
+      "step": 39575
+    },
+    {
+      "epoch": 2.6892240793586084,
+      "grad_norm": 7.0,
+      "learning_rate": 3.31923495040087e-05,
+      "loss": 0.9569,
+      "step": 39580
+    },
+    {
+      "epoch": 2.68956379942927,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.319022625356706e-05,
+      "loss": 0.9771,
+      "step": 39585
+    },
+    {
+      "epoch": 2.689903519499932,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.3188103003125426e-05,
+      "loss": 1.0019,
+      "step": 39590
+    },
+    {
+      "epoch": 2.6902432395705937,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.318597975268379e-05,
+      "loss": 1.0255,
+      "step": 39595
+    },
+    {
+      "epoch": 2.6905829596412554,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.3183856502242154e-05,
+      "loss": 1.0105,
+      "step": 39600
+    },
+    {
+      "epoch": 2.6909226797119175,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.318173325180052e-05,
+      "loss": 0.9736,
+      "step": 39605
+    },
+    {
+      "epoch": 2.691262399782579,
+      "grad_norm": 6.875,
+      "learning_rate": 3.317961000135888e-05,
+      "loss": 0.9247,
+      "step": 39610
+    },
+    {
+      "epoch": 2.6916021198532407,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.3177486750917246e-05,
+      "loss": 0.9929,
+      "step": 39615
+    },
+    {
+      "epoch": 2.691941839923903,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.317536350047561e-05,
+      "loss": 1.0723,
+      "step": 39620
+    },
+    {
+      "epoch": 2.6922815599945644,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3173240250033974e-05,
+      "loss": 0.9515,
+      "step": 39625
+    },
+    {
+      "epoch": 2.692621280065226,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.317111699959234e-05,
+      "loss": 0.9926,
+      "step": 39630
+    },
+    {
+      "epoch": 2.692961000135888,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.31689937491507e-05,
+      "loss": 0.9729,
+      "step": 39635
+    },
+    {
+      "epoch": 2.6933007202065498,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.3166870498709066e-05,
+      "loss": 0.9728,
+      "step": 39640
+    },
+    {
+      "epoch": 2.6936404402772114,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.316474724826743e-05,
+      "loss": 0.9677,
+      "step": 39645
+    },
+    {
+      "epoch": 2.6939801603478735,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.3162623997825794e-05,
+      "loss": 1.0249,
+      "step": 39650
+    },
+    {
+      "epoch": 2.694319880418535,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.316050074738415e-05,
+      "loss": 0.8593,
+      "step": 39655
+    },
+    {
+      "epoch": 2.6946596004891967,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.315837749694252e-05,
+      "loss": 0.976,
+      "step": 39660
+    },
+    {
+      "epoch": 2.694999320559859,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.3156254246500886e-05,
+      "loss": 1.0899,
+      "step": 39665
+    },
+    {
+      "epoch": 2.6953390406305204,
+      "grad_norm": 5.875,
+      "learning_rate": 3.315413099605924e-05,
+      "loss": 1.0178,
+      "step": 39670
+    },
+    {
+      "epoch": 2.695678760701182,
+      "grad_norm": 7.125,
+      "learning_rate": 3.3152007745617614e-05,
+      "loss": 0.9636,
+      "step": 39675
+    },
+    {
+      "epoch": 2.696018480771844,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.314988449517598e-05,
+      "loss": 0.9523,
+      "step": 39680
+    },
+    {
+      "epoch": 2.6963582008425058,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.3147761244734335e-05,
+      "loss": 0.9595,
+      "step": 39685
+    },
+    {
+      "epoch": 2.6966979209131674,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.3145637994292706e-05,
+      "loss": 0.989,
+      "step": 39690
+    },
+    {
+      "epoch": 2.6970376409838295,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.314351474385107e-05,
+      "loss": 1.0608,
+      "step": 39695
+    },
+    {
+      "epoch": 2.697377361054491,
+      "grad_norm": 7.625,
+      "learning_rate": 3.3141391493409434e-05,
+      "loss": 1.0592,
+      "step": 39700
+    },
+    {
+      "epoch": 2.6977170811251527,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.31392682429678e-05,
+      "loss": 0.9338,
+      "step": 39705
+    },
+    {
+      "epoch": 2.698056801195815,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.313714499252616e-05,
+      "loss": 0.9449,
+      "step": 39710
+    },
+    {
+      "epoch": 2.6983965212664764,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.3135021742084526e-05,
+      "loss": 1.0362,
+      "step": 39715
+    },
+    {
+      "epoch": 2.698736241337138,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.313289849164289e-05,
+      "loss": 1.0666,
+      "step": 39720
+    },
+    {
+      "epoch": 2.6990759614078,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.313077524120125e-05,
+      "loss": 1.0336,
+      "step": 39725
+    },
+    {
+      "epoch": 2.699415681478462,
+      "grad_norm": 5.625,
+      "learning_rate": 3.312865199075962e-05,
+      "loss": 0.9319,
+      "step": 39730
+    },
+    {
+      "epoch": 2.6997554015491234,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.312652874031798e-05,
+      "loss": 0.9675,
+      "step": 39735
+    },
+    {
+      "epoch": 2.7000951216197855,
+      "grad_norm": 6.375,
+      "learning_rate": 3.312440548987634e-05,
+      "loss": 0.9361,
+      "step": 39740
+    },
+    {
+      "epoch": 2.700434841690447,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.312228223943471e-05,
+      "loss": 0.9515,
+      "step": 39745
+    },
+    {
+      "epoch": 2.7007745617611087,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.3120158988993074e-05,
+      "loss": 0.9561,
+      "step": 39750
+    },
+    {
+      "epoch": 2.701114281831771,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.311803573855143e-05,
+      "loss": 0.8842,
+      "step": 39755
+    },
+    {
+      "epoch": 2.7014540019024325,
+      "grad_norm": 6.75,
+      "learning_rate": 3.31159124881098e-05,
+      "loss": 0.9601,
+      "step": 39760
+    },
+    {
+      "epoch": 2.701793721973094,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.3113789237668166e-05,
+      "loss": 1.0769,
+      "step": 39765
+    },
+    {
+      "epoch": 2.702133442043756,
+      "grad_norm": 7.125,
+      "learning_rate": 3.311166598722652e-05,
+      "loss": 0.956,
+      "step": 39770
+    },
+    {
+      "epoch": 2.702473162114418,
+      "grad_norm": 7.125,
+      "learning_rate": 3.3109542736784894e-05,
+      "loss": 0.9736,
+      "step": 39775
+    },
+    {
+      "epoch": 2.7028128821850794,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.310741948634326e-05,
+      "loss": 0.9155,
+      "step": 39780
+    },
+    {
+      "epoch": 2.7031526022557415,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.3105296235901615e-05,
+      "loss": 1.0203,
+      "step": 39785
+    },
+    {
+      "epoch": 2.703492322326403,
+      "grad_norm": 6.875,
+      "learning_rate": 3.3103172985459986e-05,
+      "loss": 1.0145,
+      "step": 39790
+    },
+    {
+      "epoch": 2.7038320423970648,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.310104973501835e-05,
+      "loss": 0.9639,
+      "step": 39795
+    },
+    {
+      "epoch": 2.704171762467727,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.309892648457671e-05,
+      "loss": 0.9401,
+      "step": 39800
+    },
+    {
+      "epoch": 2.7045114825383885,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.309680323413508e-05,
+      "loss": 0.9662,
+      "step": 39805
+    },
+    {
+      "epoch": 2.70485120260905,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.3094679983693435e-05,
+      "loss": 1.0111,
+      "step": 39810
+    },
+    {
+      "epoch": 2.705190922679712,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.30925567332518e-05,
+      "loss": 0.9596,
+      "step": 39815
+    },
+    {
+      "epoch": 2.705530642750374,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.309043348281017e-05,
+      "loss": 1.0138,
+      "step": 39820
+    },
+    {
+      "epoch": 2.7058703628210354,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.308831023236853e-05,
+      "loss": 0.8468,
+      "step": 39825
+    },
+    {
+      "epoch": 2.7062100828916975,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.308618698192689e-05,
+      "loss": 0.9267,
+      "step": 39830
+    },
+    {
+      "epoch": 2.706549802962359,
+      "grad_norm": 6.875,
+      "learning_rate": 3.308406373148526e-05,
+      "loss": 1.0014,
+      "step": 39835
+    },
+    {
+      "epoch": 2.7068895230330208,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.308194048104362e-05,
+      "loss": 0.9352,
+      "step": 39840
+    },
+    {
+      "epoch": 2.7072292431036824,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.3079817230601983e-05,
+      "loss": 0.9452,
+      "step": 39845
+    },
+    {
+      "epoch": 2.7075689631743445,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.3077693980160354e-05,
+      "loss": 0.9561,
+      "step": 39850
+    },
+    {
+      "epoch": 2.707908683245006,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.307557072971871e-05,
+      "loss": 0.9759,
+      "step": 39855
+    },
+    {
+      "epoch": 2.7082484033156677,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.3073447479277075e-05,
+      "loss": 0.96,
+      "step": 39860
+    },
+    {
+      "epoch": 2.70858812338633,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.3071324228835446e-05,
+      "loss": 0.8737,
+      "step": 39865
+    },
+    {
+      "epoch": 2.7089278434569914,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.3069200978393803e-05,
+      "loss": 0.9467,
+      "step": 39870
+    },
+    {
+      "epoch": 2.709267563527653,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.306707772795217e-05,
+      "loss": 0.9738,
+      "step": 39875
+    },
+    {
+      "epoch": 2.709607283598315,
+      "grad_norm": 6.5,
+      "learning_rate": 3.306495447751053e-05,
+      "loss": 0.9652,
+      "step": 39880
+    },
+    {
+      "epoch": 2.709947003668977,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.3062831227068895e-05,
+      "loss": 1.0272,
+      "step": 39885
+    },
+    {
+      "epoch": 2.7102867237396384,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.306070797662726e-05,
+      "loss": 0.9902,
+      "step": 39890
+    },
+    {
+      "epoch": 2.7106264438103,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.3058584726185623e-05,
+      "loss": 0.9514,
+      "step": 39895
+    },
+    {
+      "epoch": 2.710966163880962,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.305646147574399e-05,
+      "loss": 0.9845,
+      "step": 39900
+    },
+    {
+      "epoch": 2.7113058839516238,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.305433822530235e-05,
+      "loss": 0.8991,
+      "step": 39905
+    },
+    {
+      "epoch": 2.7116456040222854,
+      "grad_norm": 6.625,
+      "learning_rate": 3.3052214974860715e-05,
+      "loss": 0.9008,
+      "step": 39910
+    },
+    {
+      "epoch": 2.7119853240929475,
+      "grad_norm": 6.625,
+      "learning_rate": 3.305009172441908e-05,
+      "loss": 1.0429,
+      "step": 39915
+    },
+    {
+      "epoch": 2.712325044163609,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3047968473977443e-05,
+      "loss": 1.013,
+      "step": 39920
+    },
+    {
+      "epoch": 2.7126647642342707,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.304584522353581e-05,
+      "loss": 1.0721,
+      "step": 39925
+    },
+    {
+      "epoch": 2.713004484304933,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.304372197309417e-05,
+      "loss": 0.9599,
+      "step": 39930
+    },
+    {
+      "epoch": 2.7133442043755944,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3041598722652536e-05,
+      "loss": 1.015,
+      "step": 39935
+    },
+    {
+      "epoch": 2.713683924446256,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.30394754722109e-05,
+      "loss": 0.9716,
+      "step": 39940
+    },
+    {
+      "epoch": 2.714023644516918,
+      "grad_norm": 6.5,
+      "learning_rate": 3.3037352221769264e-05,
+      "loss": 0.8852,
+      "step": 39945
+    },
+    {
+      "epoch": 2.7143633645875798,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.303522897132763e-05,
+      "loss": 0.9913,
+      "step": 39950
+    },
+    {
+      "epoch": 2.7147030846582414,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.303310572088599e-05,
+      "loss": 0.9487,
+      "step": 39955
+    },
+    {
+      "epoch": 2.7150428047289035,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.3030982470444356e-05,
+      "loss": 1.1521,
+      "step": 39960
+    },
+    {
+      "epoch": 2.715382524799565,
+      "grad_norm": 6.125,
+      "learning_rate": 3.302885922000272e-05,
+      "loss": 0.9427,
+      "step": 39965
+    },
+    {
+      "epoch": 2.7157222448702267,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.3026735969561084e-05,
+      "loss": 0.9162,
+      "step": 39970
+    },
+    {
+      "epoch": 2.716061964940889,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.302461271911945e-05,
+      "loss": 0.9953,
+      "step": 39975
+    },
+    {
+      "epoch": 2.7164016850115504,
+      "grad_norm": 7.25,
+      "learning_rate": 3.302248946867781e-05,
+      "loss": 1.0086,
+      "step": 39980
+    },
+    {
+      "epoch": 2.716741405082212,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.3020366218236176e-05,
+      "loss": 0.8604,
+      "step": 39985
+    },
+    {
+      "epoch": 2.717081125152874,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.301824296779454e-05,
+      "loss": 1.0428,
+      "step": 39990
+    },
+    {
+      "epoch": 2.7174208452235358,
+      "grad_norm": 6.625,
+      "learning_rate": 3.3016119717352904e-05,
+      "loss": 1.0727,
+      "step": 39995
+    },
+    {
+      "epoch": 2.7177605652941974,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.301399646691127e-05,
+      "loss": 1.1012,
+      "step": 40000
+    },
+    {
+      "epoch": 2.7181002853648595,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.301187321646963e-05,
+      "loss": 0.9169,
+      "step": 40005
+    },
+    {
+      "epoch": 2.718440005435521,
+      "grad_norm": 6.875,
+      "learning_rate": 3.300974996602799e-05,
+      "loss": 0.895,
+      "step": 40010
+    },
+    {
+      "epoch": 2.7187797255061827,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.300762671558636e-05,
+      "loss": 1.0392,
+      "step": 40015
+    },
+    {
+      "epoch": 2.719119445576845,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.3005503465144724e-05,
+      "loss": 0.9346,
+      "step": 40020
+    },
+    {
+      "epoch": 2.7194591656475064,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.300338021470308e-05,
+      "loss": 0.9335,
+      "step": 40025
+    },
+    {
+      "epoch": 2.719798885718168,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.300125696426145e-05,
+      "loss": 0.9641,
+      "step": 40030
+    },
+    {
+      "epoch": 2.72013860578883,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2999133713819816e-05,
+      "loss": 0.9589,
+      "step": 40035
+    },
+    {
+      "epoch": 2.720478325859492,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.299701046337818e-05,
+      "loss": 0.9212,
+      "step": 40040
+    },
+    {
+      "epoch": 2.7208180459301534,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.2994887212936544e-05,
+      "loss": 0.981,
+      "step": 40045
+    },
+    {
+      "epoch": 2.7211577660008155,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.299276396249491e-05,
+      "loss": 0.9661,
+      "step": 40050
+    },
+    {
+      "epoch": 2.721497486071477,
+      "grad_norm": 5.65625,
+      "learning_rate": 3.299064071205327e-05,
+      "loss": 0.9807,
+      "step": 40055
+    },
+    {
+      "epoch": 2.7218372061421388,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.2988517461611636e-05,
+      "loss": 0.9564,
+      "step": 40060
+    },
+    {
+      "epoch": 2.722176926212801,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.298639421117e-05,
+      "loss": 0.9553,
+      "step": 40065
+    },
+    {
+      "epoch": 2.7225166462834625,
+      "grad_norm": 7.0,
+      "learning_rate": 3.2984270960728364e-05,
+      "loss": 0.9107,
+      "step": 40070
+    },
+    {
+      "epoch": 2.722856366354124,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.298214771028673e-05,
+      "loss": 0.9973,
+      "step": 40075
+    },
+    {
+      "epoch": 2.723196086424786,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.2980024459845085e-05,
+      "loss": 0.8549,
+      "step": 40080
+    },
+    {
+      "epoch": 2.723535806495448,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.2977901209403456e-05,
+      "loss": 0.9672,
+      "step": 40085
+    },
+    {
+      "epoch": 2.7238755265661094,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.297577795896182e-05,
+      "loss": 0.9287,
+      "step": 40090
+    },
+    {
+      "epoch": 2.7242152466367715,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.297365470852018e-05,
+      "loss": 0.9538,
+      "step": 40095
+    },
+    {
+      "epoch": 2.724554966707433,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.297153145807855e-05,
+      "loss": 0.8827,
+      "step": 40100
+    },
+    {
+      "epoch": 2.7248946867780948,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.296940820763691e-05,
+      "loss": 0.9102,
+      "step": 40105
+    },
+    {
+      "epoch": 2.725234406848757,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.296728495719527e-05,
+      "loss": 0.9553,
+      "step": 40110
+    },
+    {
+      "epoch": 2.7255741269194185,
+      "grad_norm": 7.5,
+      "learning_rate": 3.296516170675364e-05,
+      "loss": 0.9439,
+      "step": 40115
+    },
+    {
+      "epoch": 2.72591384699008,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.2963038456312004e-05,
+      "loss": 0.9314,
+      "step": 40120
+    },
+    {
+      "epoch": 2.726253567060742,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.296091520587036e-05,
+      "loss": 0.9459,
+      "step": 40125
+    },
+    {
+      "epoch": 2.726593287131404,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.295879195542873e-05,
+      "loss": 1.0691,
+      "step": 40130
+    },
+    {
+      "epoch": 2.7269330072020654,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.2956668704987096e-05,
+      "loss": 0.8954,
+      "step": 40135
+    },
+    {
+      "epoch": 2.7272727272727275,
+      "grad_norm": 6.5,
+      "learning_rate": 3.295454545454545e-05,
+      "loss": 0.9565,
+      "step": 40140
+    },
+    {
+      "epoch": 2.727612447343389,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2952422204103824e-05,
+      "loss": 0.9809,
+      "step": 40145
+    },
+    {
+      "epoch": 2.7279521674140508,
+      "grad_norm": 7.125,
+      "learning_rate": 3.295029895366218e-05,
+      "loss": 0.9977,
+      "step": 40150
+    },
+    {
+      "epoch": 2.728291887484713,
+      "grad_norm": 7.0,
+      "learning_rate": 3.2948175703220545e-05,
+      "loss": 1.0728,
+      "step": 40155
+    },
+    {
+      "epoch": 2.7286316075553745,
+      "grad_norm": 7.25,
+      "learning_rate": 3.2946052452778916e-05,
+      "loss": 0.9687,
+      "step": 40160
+    },
+    {
+      "epoch": 2.728971327626036,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.294392920233727e-05,
+      "loss": 0.948,
+      "step": 40165
+    },
+    {
+      "epoch": 2.729311047696698,
+      "grad_norm": 6.75,
+      "learning_rate": 3.294180595189564e-05,
+      "loss": 0.9723,
+      "step": 40170
+    },
+    {
+      "epoch": 2.72965076776736,
+      "grad_norm": 6.75,
+      "learning_rate": 3.293968270145401e-05,
+      "loss": 1.0075,
+      "step": 40175
+    },
+    {
+      "epoch": 2.7299904878380215,
+      "grad_norm": 7.125,
+      "learning_rate": 3.2937559451012365e-05,
+      "loss": 1.0064,
+      "step": 40180
+    },
+    {
+      "epoch": 2.730330207908683,
+      "grad_norm": 6.0,
+      "learning_rate": 3.293543620057073e-05,
+      "loss": 0.9315,
+      "step": 40185
+    },
+    {
+      "epoch": 2.730669927979345,
+      "grad_norm": 6.5,
+      "learning_rate": 3.29333129501291e-05,
+      "loss": 0.9653,
+      "step": 40190
+    },
+    {
+      "epoch": 2.731009648050007,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.293118969968746e-05,
+      "loss": 0.9605,
+      "step": 40195
+    },
+    {
+      "epoch": 2.7313493681206684,
+      "grad_norm": 6.25,
+      "learning_rate": 3.292906644924582e-05,
+      "loss": 0.9894,
+      "step": 40200
+    },
+    {
+      "epoch": 2.7316890881913305,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.292694319880419e-05,
+      "loss": 0.9316,
+      "step": 40205
+    },
+    {
+      "epoch": 2.732028808261992,
+      "grad_norm": 7.125,
+      "learning_rate": 3.292481994836255e-05,
+      "loss": 0.9015,
+      "step": 40210
+    },
+    {
+      "epoch": 2.7323685283326538,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.292269669792091e-05,
+      "loss": 1.0351,
+      "step": 40215
+    },
+    {
+      "epoch": 2.732708248403316,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.2920573447479284e-05,
+      "loss": 1.0691,
+      "step": 40220
+    },
+    {
+      "epoch": 2.7330479684739775,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.291845019703764e-05,
+      "loss": 0.9555,
+      "step": 40225
+    },
+    {
+      "epoch": 2.733387688544639,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.2916326946596005e-05,
+      "loss": 0.8405,
+      "step": 40230
+    },
+    {
+      "epoch": 2.7337274086153007,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.291420369615437e-05,
+      "loss": 0.875,
+      "step": 40235
+    },
+    {
+      "epoch": 2.734067128685963,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.291208044571273e-05,
+      "loss": 1.008,
+      "step": 40240
+    },
+    {
+      "epoch": 2.7344068487566244,
+      "grad_norm": 6.75,
+      "learning_rate": 3.29099571952711e-05,
+      "loss": 0.9467,
+      "step": 40245
+    },
+    {
+      "epoch": 2.734746568827286,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.290783394482946e-05,
+      "loss": 1.0189,
+      "step": 40250
+    },
+    {
+      "epoch": 2.735086288897948,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2905710694387825e-05,
+      "loss": 0.9867,
+      "step": 40255
+    },
+    {
+      "epoch": 2.7354260089686098,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.290358744394619e-05,
+      "loss": 1.0158,
+      "step": 40260
+    },
+    {
+      "epoch": 2.7357657290392714,
+      "grad_norm": 8.0,
+      "learning_rate": 3.290146419350455e-05,
+      "loss": 0.9424,
+      "step": 40265
+    },
+    {
+      "epoch": 2.7361054491099335,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.289934094306292e-05,
+      "loss": 0.9506,
+      "step": 40270
+    },
+    {
+      "epoch": 2.736445169180595,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.289721769262128e-05,
+      "loss": 0.9589,
+      "step": 40275
+    },
+    {
+      "epoch": 2.7367848892512567,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.2895094442179645e-05,
+      "loss": 0.9937,
+      "step": 40280
+    },
+    {
+      "epoch": 2.737124609321919,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.289297119173801e-05,
+      "loss": 0.9608,
+      "step": 40285
+    },
+    {
+      "epoch": 2.7374643293925804,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.289084794129637e-05,
+      "loss": 1.0479,
+      "step": 40290
+    },
+    {
+      "epoch": 2.737804049463242,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.288872469085474e-05,
+      "loss": 0.9493,
+      "step": 40295
+    },
+    {
+      "epoch": 2.738143769533904,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.28866014404131e-05,
+      "loss": 0.952,
+      "step": 40300
+    },
+    {
+      "epoch": 2.738483489604566,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.2884478189971465e-05,
+      "loss": 0.9936,
+      "step": 40305
+    },
+    {
+      "epoch": 2.7388232096752274,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.288235493952983e-05,
+      "loss": 0.9469,
+      "step": 40310
+    },
+    {
+      "epoch": 2.7391629297458895,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.288023168908819e-05,
+      "loss": 1.0032,
+      "step": 40315
+    },
+    {
+      "epoch": 2.739502649816551,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.287810843864656e-05,
+      "loss": 0.9163,
+      "step": 40320
+    },
+    {
+      "epoch": 2.7398423698872127,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.287598518820492e-05,
+      "loss": 0.9026,
+      "step": 40325
+    },
+    {
+      "epoch": 2.740182089957875,
+      "grad_norm": 6.625,
+      "learning_rate": 3.2873861937763285e-05,
+      "loss": 1.0311,
+      "step": 40330
+    },
+    {
+      "epoch": 2.7405218100285365,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.287173868732165e-05,
+      "loss": 0.941,
+      "step": 40335
+    },
+    {
+      "epoch": 2.740861530099198,
+      "grad_norm": 6.5,
+      "learning_rate": 3.286961543688001e-05,
+      "loss": 0.8733,
+      "step": 40340
+    },
+    {
+      "epoch": 2.74120125016986,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.286749218643838e-05,
+      "loss": 1.0004,
+      "step": 40345
+    },
+    {
+      "epoch": 2.741540970240522,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.2865368935996734e-05,
+      "loss": 0.8922,
+      "step": 40350
+    },
+    {
+      "epoch": 2.7418806903111834,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.2863245685555105e-05,
+      "loss": 1.0165,
+      "step": 40355
+    },
+    {
+      "epoch": 2.7422204103818455,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.286112243511347e-05,
+      "loss": 0.9995,
+      "step": 40360
+    },
+    {
+      "epoch": 2.742560130452507,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2858999184671826e-05,
+      "loss": 1.0086,
+      "step": 40365
+    },
+    {
+      "epoch": 2.7428998505231688,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.28568759342302e-05,
+      "loss": 1.1164,
+      "step": 40370
+    },
+    {
+      "epoch": 2.743239570593831,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.285475268378856e-05,
+      "loss": 0.9771,
+      "step": 40375
+    },
+    {
+      "epoch": 2.7435792906644925,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.2852629433346925e-05,
+      "loss": 0.9917,
+      "step": 40380
+    },
+    {
+      "epoch": 2.743919010735154,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.285050618290529e-05,
+      "loss": 0.9172,
+      "step": 40385
+    },
+    {
+      "epoch": 2.744258730805816,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.284838293246365e-05,
+      "loss": 1.0117,
+      "step": 40390
+    },
+    {
+      "epoch": 2.744598450876478,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.284625968202202e-05,
+      "loss": 1.006,
+      "step": 40395
+    },
+    {
+      "epoch": 2.7449381709471394,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.284413643158038e-05,
+      "loss": 0.9061,
+      "step": 40400
+    },
+    {
+      "epoch": 2.7452778910178015,
+      "grad_norm": 6.375,
+      "learning_rate": 3.2842013181138745e-05,
+      "loss": 0.8872,
+      "step": 40405
+    },
+    {
+      "epoch": 2.745617611088463,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.283988993069711e-05,
+      "loss": 0.9806,
+      "step": 40410
+    },
+    {
+      "epoch": 2.7459573311591248,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.283776668025547e-05,
+      "loss": 1.0616,
+      "step": 40415
+    },
+    {
+      "epoch": 2.746297051229787,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.283564342981384e-05,
+      "loss": 1.0293,
+      "step": 40420
+    },
+    {
+      "epoch": 2.7466367713004485,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.28335201793722e-05,
+      "loss": 0.9886,
+      "step": 40425
+    },
+    {
+      "epoch": 2.74697649137111,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.2831396928930565e-05,
+      "loss": 0.9418,
+      "step": 40430
+    },
+    {
+      "epoch": 2.747316211441772,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.282927367848892e-05,
+      "loss": 0.955,
+      "step": 40435
+    },
+    {
+      "epoch": 2.747655931512434,
+      "grad_norm": 8.0,
+      "learning_rate": 3.282715042804729e-05,
+      "loss": 1.0202,
+      "step": 40440
+    },
+    {
+      "epoch": 2.7479956515830954,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.282502717760566e-05,
+      "loss": 0.9685,
+      "step": 40445
+    },
+    {
+      "epoch": 2.7483353716537575,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.2822903927164014e-05,
+      "loss": 0.9756,
+      "step": 40450
+    },
+    {
+      "epoch": 2.748675091724419,
+      "grad_norm": 5.1875,
+      "learning_rate": 3.2820780676722385e-05,
+      "loss": 0.9922,
+      "step": 40455
+    },
+    {
+      "epoch": 2.749014811795081,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.281865742628075e-05,
+      "loss": 0.9538,
+      "step": 40460
+    },
+    {
+      "epoch": 2.749354531865743,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.2816534175839106e-05,
+      "loss": 0.9409,
+      "step": 40465
+    },
+    {
+      "epoch": 2.7496942519364045,
+      "grad_norm": 7.0,
+      "learning_rate": 3.281441092539748e-05,
+      "loss": 0.8935,
+      "step": 40470
+    },
+    {
+      "epoch": 2.750033972007066,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.281228767495584e-05,
+      "loss": 0.9163,
+      "step": 40475
+    },
+    {
+      "epoch": 2.750373692077728,
+      "grad_norm": 5.625,
+      "learning_rate": 3.28101644245142e-05,
+      "loss": 0.898,
+      "step": 40480
+    },
+    {
+      "epoch": 2.75071341214839,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.280804117407257e-05,
+      "loss": 0.971,
+      "step": 40485
+    },
+    {
+      "epoch": 2.7510531322190515,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.280591792363093e-05,
+      "loss": 1.0017,
+      "step": 40490
+    },
+    {
+      "epoch": 2.7513928522897135,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.280379467318929e-05,
+      "loss": 0.9894,
+      "step": 40495
+    },
+    {
+      "epoch": 2.751732572360375,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.280167142274766e-05,
+      "loss": 0.9106,
+      "step": 40500
+    },
+    {
+      "epoch": 2.752072292431037,
+      "grad_norm": 7.375,
+      "learning_rate": 3.279954817230602e-05,
+      "loss": 0.9908,
+      "step": 40505
+    },
+    {
+      "epoch": 2.752412012501699,
+      "grad_norm": 6.875,
+      "learning_rate": 3.279742492186438e-05,
+      "loss": 0.9149,
+      "step": 40510
+    },
+    {
+      "epoch": 2.7527517325723605,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.279530167142275e-05,
+      "loss": 1.008,
+      "step": 40515
+    },
+    {
+      "epoch": 2.753091452643022,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.279317842098111e-05,
+      "loss": 0.9264,
+      "step": 40520
+    },
+    {
+      "epoch": 2.7534311727136838,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.2791055170539474e-05,
+      "loss": 0.9203,
+      "step": 40525
+    },
+    {
+      "epoch": 2.753770892784346,
+      "grad_norm": 8.5,
+      "learning_rate": 3.2788931920097845e-05,
+      "loss": 0.9516,
+      "step": 40530
+    },
+    {
+      "epoch": 2.7541106128550075,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.27868086696562e-05,
+      "loss": 0.9612,
+      "step": 40535
+    },
+    {
+      "epoch": 2.754450332925669,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.2784685419214566e-05,
+      "loss": 1.0627,
+      "step": 40540
+    },
+    {
+      "epoch": 2.754790052996331,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.278256216877294e-05,
+      "loss": 0.9379,
+      "step": 40545
+    },
+    {
+      "epoch": 2.755129773066993,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.2780438918331294e-05,
+      "loss": 0.9812,
+      "step": 40550
+    },
+    {
+      "epoch": 2.7554694931376544,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.277831566788966e-05,
+      "loss": 1.0338,
+      "step": 40555
+    },
+    {
+      "epoch": 2.7558092132083165,
+      "grad_norm": 6.875,
+      "learning_rate": 3.277619241744803e-05,
+      "loss": 1.0186,
+      "step": 40560
+    },
+    {
+      "epoch": 2.756148933278978,
+      "grad_norm": 5.21875,
+      "learning_rate": 3.2774069167006386e-05,
+      "loss": 1.0045,
+      "step": 40565
+    },
+    {
+      "epoch": 2.7564886533496398,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.277194591656475e-05,
+      "loss": 0.9632,
+      "step": 40570
+    },
+    {
+      "epoch": 2.7568283734203014,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.2769822666123114e-05,
+      "loss": 0.9282,
+      "step": 40575
+    },
+    {
+      "epoch": 2.7571680934909635,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.276769941568148e-05,
+      "loss": 1.0978,
+      "step": 40580
+    },
+    {
+      "epoch": 2.757507813561625,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.276557616523984e-05,
+      "loss": 0.9861,
+      "step": 40585
+    },
+    {
+      "epoch": 2.7578475336322867,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.2763452914798206e-05,
+      "loss": 0.8646,
+      "step": 40590
+    },
+    {
+      "epoch": 2.758187253702949,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.276132966435657e-05,
+      "loss": 0.9056,
+      "step": 40595
+    },
+    {
+      "epoch": 2.7585269737736104,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.2759206413914934e-05,
+      "loss": 0.9866,
+      "step": 40600
+    },
+    {
+      "epoch": 2.758866693844272,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.27570831634733e-05,
+      "loss": 0.9645,
+      "step": 40605
+    },
+    {
+      "epoch": 2.759206413914934,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.275495991303166e-05,
+      "loss": 0.9611,
+      "step": 40610
+    },
+    {
+      "epoch": 2.759546133985596,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.2752836662590026e-05,
+      "loss": 0.972,
+      "step": 40615
+    },
+    {
+      "epoch": 2.7598858540562574,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.275071341214839e-05,
+      "loss": 0.9402,
+      "step": 40620
+    },
+    {
+      "epoch": 2.7602255741269195,
+      "grad_norm": 6.5,
+      "learning_rate": 3.2748590161706754e-05,
+      "loss": 0.9616,
+      "step": 40625
+    },
+    {
+      "epoch": 2.760565294197581,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.274646691126512e-05,
+      "loss": 0.9023,
+      "step": 40630
+    },
+    {
+      "epoch": 2.7609050142682428,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.274434366082348e-05,
+      "loss": 0.8913,
+      "step": 40635
+    },
+    {
+      "epoch": 2.761244734338905,
+      "grad_norm": 7.375,
+      "learning_rate": 3.2742220410381847e-05,
+      "loss": 0.9808,
+      "step": 40640
+    },
+    {
+      "epoch": 2.7615844544095665,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.274009715994021e-05,
+      "loss": 0.9844,
+      "step": 40645
+    },
+    {
+      "epoch": 2.761924174480228,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.2737973909498575e-05,
+      "loss": 1.0516,
+      "step": 40650
+    },
+    {
+      "epoch": 2.76226389455089,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.273585065905694e-05,
+      "loss": 1.0439,
+      "step": 40655
+    },
+    {
+      "epoch": 2.762603614621552,
+      "grad_norm": 6.625,
+      "learning_rate": 3.27337274086153e-05,
+      "loss": 0.9436,
+      "step": 40660
+    },
+    {
+      "epoch": 2.7629433346922134,
+      "grad_norm": 6.875,
+      "learning_rate": 3.2731604158173667e-05,
+      "loss": 0.9677,
+      "step": 40665
+    },
+    {
+      "epoch": 2.7632830547628755,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.272948090773203e-05,
+      "loss": 0.9885,
+      "step": 40670
+    },
+    {
+      "epoch": 2.763622774833537,
+      "grad_norm": 6.75,
+      "learning_rate": 3.2727357657290395e-05,
+      "loss": 0.993,
+      "step": 40675
+    },
+    {
+      "epoch": 2.7639624949041988,
+      "grad_norm": 6.375,
+      "learning_rate": 3.272523440684876e-05,
+      "loss": 1.0355,
+      "step": 40680
+    },
+    {
+      "epoch": 2.764302214974861,
+      "grad_norm": 8.375,
+      "learning_rate": 3.272311115640712e-05,
+      "loss": 1.0166,
+      "step": 40685
+    },
+    {
+      "epoch": 2.7646419350455225,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.2720987905965487e-05,
+      "loss": 0.9058,
+      "step": 40690
+    },
+    {
+      "epoch": 2.764981655116184,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.271886465552385e-05,
+      "loss": 0.9876,
+      "step": 40695
+    },
+    {
+      "epoch": 2.765321375186846,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.2716741405082215e-05,
+      "loss": 0.862,
+      "step": 40700
+    },
+    {
+      "epoch": 2.765661095257508,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.271461815464057e-05,
+      "loss": 0.9424,
+      "step": 40705
+    },
+    {
+      "epoch": 2.7660008153281694,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.271249490419894e-05,
+      "loss": 0.8868,
+      "step": 40710
+    },
+    {
+      "epoch": 2.7663405353988315,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.2710371653757307e-05,
+      "loss": 0.9597,
+      "step": 40715
+    },
+    {
+      "epoch": 2.766680255469493,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.270824840331567e-05,
+      "loss": 1.0078,
+      "step": 40720
+    },
+    {
+      "epoch": 2.7670199755401548,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.2706125152874035e-05,
+      "loss": 0.9522,
+      "step": 40725
+    },
+    {
+      "epoch": 2.767359695610817,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.27040019024324e-05,
+      "loss": 0.8892,
+      "step": 40730
+    },
+    {
+      "epoch": 2.7676994156814785,
+      "grad_norm": 6.625,
+      "learning_rate": 3.270187865199076e-05,
+      "loss": 0.9033,
+      "step": 40735
+    },
+    {
+      "epoch": 2.76803913575214,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2699755401549127e-05,
+      "loss": 1.0069,
+      "step": 40740
+    },
+    {
+      "epoch": 2.768378855822802,
+      "grad_norm": 6.25,
+      "learning_rate": 3.269763215110749e-05,
+      "loss": 0.9529,
+      "step": 40745
+    },
+    {
+      "epoch": 2.768718575893464,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.2695508900665855e-05,
+      "loss": 0.9795,
+      "step": 40750
+    },
+    {
+      "epoch": 2.7690582959641254,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.269338565022422e-05,
+      "loss": 0.8991,
+      "step": 40755
+    },
+    {
+      "epoch": 2.7693980160347875,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.269126239978258e-05,
+      "loss": 1.0568,
+      "step": 40760
+    },
+    {
+      "epoch": 2.769737736105449,
+      "grad_norm": 7.125,
+      "learning_rate": 3.2689139149340947e-05,
+      "loss": 1.0075,
+      "step": 40765
+    },
+    {
+      "epoch": 2.770077456176111,
+      "grad_norm": 6.875,
+      "learning_rate": 3.268701589889931e-05,
+      "loss": 0.9592,
+      "step": 40770
+    },
+    {
+      "epoch": 2.770417176246773,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.268489264845767e-05,
+      "loss": 0.9864,
+      "step": 40775
+    },
+    {
+      "epoch": 2.7707568963174345,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.268276939801604e-05,
+      "loss": 1.0051,
+      "step": 40780
+    },
+    {
+      "epoch": 2.771096616388096,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.26806461475744e-05,
+      "loss": 1.0227,
+      "step": 40785
+    },
+    {
+      "epoch": 2.771436336458758,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.267852289713276e-05,
+      "loss": 0.9294,
+      "step": 40790
+    },
+    {
+      "epoch": 2.77177605652942,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.267639964669113e-05,
+      "loss": 0.9427,
+      "step": 40795
+    },
+    {
+      "epoch": 2.7721157766000815,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.2674276396249495e-05,
+      "loss": 0.982,
+      "step": 40800
+    },
+    {
+      "epoch": 2.7724554966707435,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.267215314580785e-05,
+      "loss": 0.9045,
+      "step": 40805
+    },
+    {
+      "epoch": 2.772795216741405,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.267002989536622e-05,
+      "loss": 0.9653,
+      "step": 40810
+    },
+    {
+      "epoch": 2.773134936812067,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.266790664492459e-05,
+      "loss": 1.0076,
+      "step": 40815
+    },
+    {
+      "epoch": 2.773474656882729,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.2665783394482944e-05,
+      "loss": 0.9762,
+      "step": 40820
+    },
+    {
+      "epoch": 2.7738143769533905,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.2663660144041315e-05,
+      "loss": 0.9353,
+      "step": 40825
+    },
+    {
+      "epoch": 2.774154097024052,
+      "grad_norm": 7.125,
+      "learning_rate": 3.266153689359968e-05,
+      "loss": 1.0167,
+      "step": 40830
+    },
+    {
+      "epoch": 2.774493817094714,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.2659413643158036e-05,
+      "loss": 1.0354,
+      "step": 40835
+    },
+    {
+      "epoch": 2.774833537165376,
+      "grad_norm": 8.0,
+      "learning_rate": 3.265729039271641e-05,
+      "loss": 1.0845,
+      "step": 40840
+    },
+    {
+      "epoch": 2.7751732572360375,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.265516714227477e-05,
+      "loss": 0.9759,
+      "step": 40845
+    },
+    {
+      "epoch": 2.7755129773066995,
+      "grad_norm": 6.125,
+      "learning_rate": 3.265304389183313e-05,
+      "loss": 0.9335,
+      "step": 40850
+    },
+    {
+      "epoch": 2.775852697377361,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.26509206413915e-05,
+      "loss": 0.941,
+      "step": 40855
+    },
+    {
+      "epoch": 2.776192417448023,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.2648797390949856e-05,
+      "loss": 0.8989,
+      "step": 40860
+    },
+    {
+      "epoch": 2.7765321375186844,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.264667414050822e-05,
+      "loss": 0.9511,
+      "step": 40865
+    },
+    {
+      "epoch": 2.7768718575893465,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.264455089006659e-05,
+      "loss": 0.9231,
+      "step": 40870
+    },
+    {
+      "epoch": 2.777211577660008,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.264242763962495e-05,
+      "loss": 0.9038,
+      "step": 40875
+    },
+    {
+      "epoch": 2.7775512977306698,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.264030438918331e-05,
+      "loss": 1.0065,
+      "step": 40880
+    },
+    {
+      "epoch": 2.777891017801332,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.263818113874168e-05,
+      "loss": 0.9362,
+      "step": 40885
+    },
+    {
+      "epoch": 2.7782307378719935,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.263605788830004e-05,
+      "loss": 0.9903,
+      "step": 40890
+    },
+    {
+      "epoch": 2.778570457942655,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.2633934637858404e-05,
+      "loss": 0.9999,
+      "step": 40895
+    },
+    {
+      "epoch": 2.778910178013317,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.2631811387416775e-05,
+      "loss": 0.8773,
+      "step": 40900
+    },
+    {
+      "epoch": 2.779249898083979,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.262968813697513e-05,
+      "loss": 0.9615,
+      "step": 40905
+    },
+    {
+      "epoch": 2.7795896181546405,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.2627564886533496e-05,
+      "loss": 0.9181,
+      "step": 40910
+    },
+    {
+      "epoch": 2.779929338225302,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.262544163609187e-05,
+      "loss": 0.9253,
+      "step": 40915
+    },
+    {
+      "epoch": 2.780269058295964,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2623318385650224e-05,
+      "loss": 0.8664,
+      "step": 40920
+    },
+    {
+      "epoch": 2.780608778366626,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.262119513520859e-05,
+      "loss": 1.0509,
+      "step": 40925
+    },
+    {
+      "epoch": 2.7809484984372874,
+      "grad_norm": 6.5,
+      "learning_rate": 3.261907188476695e-05,
+      "loss": 0.9807,
+      "step": 40930
+    },
+    {
+      "epoch": 2.7812882185079495,
+      "grad_norm": 5.3125,
+      "learning_rate": 3.2616948634325316e-05,
+      "loss": 1.0537,
+      "step": 40935
+    },
+    {
+      "epoch": 2.781627938578611,
+      "grad_norm": 6.5,
+      "learning_rate": 3.261482538388368e-05,
+      "loss": 0.9688,
+      "step": 40940
+    },
+    {
+      "epoch": 2.7819676586492728,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.2612702133442044e-05,
+      "loss": 0.9336,
+      "step": 40945
+    },
+    {
+      "epoch": 2.782307378719935,
+      "grad_norm": 8.0,
+      "learning_rate": 3.261057888300041e-05,
+      "loss": 1.0377,
+      "step": 40950
+    },
+    {
+      "epoch": 2.7826470987905965,
+      "grad_norm": 7.125,
+      "learning_rate": 3.260845563255877e-05,
+      "loss": 0.9547,
+      "step": 40955
+    },
+    {
+      "epoch": 2.782986818861258,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.2606332382117136e-05,
+      "loss": 1.0294,
+      "step": 40960
+    },
+    {
+      "epoch": 2.78332653893192,
+      "grad_norm": 7.125,
+      "learning_rate": 3.26042091316755e-05,
+      "loss": 0.8946,
+      "step": 40965
+    },
+    {
+      "epoch": 2.783666259002582,
+      "grad_norm": 7.0,
+      "learning_rate": 3.2602085881233864e-05,
+      "loss": 0.9915,
+      "step": 40970
+    },
+    {
+      "epoch": 2.7840059790732434,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.259996263079223e-05,
+      "loss": 0.9598,
+      "step": 40975
+    },
+    {
+      "epoch": 2.7843456991439055,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.259783938035059e-05,
+      "loss": 0.9718,
+      "step": 40980
+    },
+    {
+      "epoch": 2.784685419214567,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.2595716129908956e-05,
+      "loss": 1.0222,
+      "step": 40985
+    },
+    {
+      "epoch": 2.7850251392852288,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.259359287946732e-05,
+      "loss": 0.9031,
+      "step": 40990
+    },
+    {
+      "epoch": 2.785364859355891,
+      "grad_norm": 7.375,
+      "learning_rate": 3.2591469629025684e-05,
+      "loss": 0.9373,
+      "step": 40995
+    },
+    {
+      "epoch": 2.7857045794265525,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.258934637858405e-05,
+      "loss": 0.935,
+      "step": 41000
+    },
+    {
+      "epoch": 2.786044299497214,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.258722312814241e-05,
+      "loss": 1.087,
+      "step": 41005
+    },
+    {
+      "epoch": 2.786384019567876,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.2585099877700776e-05,
+      "loss": 1.0643,
+      "step": 41010
+    },
+    {
+      "epoch": 2.786723739638538,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.258297662725914e-05,
+      "loss": 0.984,
+      "step": 41015
+    },
+    {
+      "epoch": 2.7870634597091994,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.2580853376817504e-05,
+      "loss": 0.915,
+      "step": 41020
+    },
+    {
+      "epoch": 2.7874031797798615,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.257873012637587e-05,
+      "loss": 1.0607,
+      "step": 41025
+    },
+    {
+      "epoch": 2.787742899850523,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.257660687593423e-05,
+      "loss": 0.9311,
+      "step": 41030
+    },
+    {
+      "epoch": 2.788082619921185,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.2574483625492596e-05,
+      "loss": 1.0043,
+      "step": 41035
+    },
+    {
+      "epoch": 2.788422339991847,
+      "grad_norm": 6.75,
+      "learning_rate": 3.257236037505096e-05,
+      "loss": 1.0263,
+      "step": 41040
+    },
+    {
+      "epoch": 2.7887620600625085,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.2570237124609324e-05,
+      "loss": 0.8925,
+      "step": 41045
+    },
+    {
+      "epoch": 2.78910178013317,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.256811387416769e-05,
+      "loss": 0.9942,
+      "step": 41050
+    },
+    {
+      "epoch": 2.789441500203832,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.256599062372605e-05,
+      "loss": 1.0365,
+      "step": 41055
+    },
+    {
+      "epoch": 2.789781220274494,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.2563867373284416e-05,
+      "loss": 1.0687,
+      "step": 41060
+    },
+    {
+      "epoch": 2.7901209403451555,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.256174412284278e-05,
+      "loss": 0.9557,
+      "step": 41065
+    },
+    {
+      "epoch": 2.7904606604158175,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.2559620872401144e-05,
+      "loss": 1.0846,
+      "step": 41070
+    },
+    {
+      "epoch": 2.790800380486479,
+      "grad_norm": 6.875,
+      "learning_rate": 3.255749762195951e-05,
+      "loss": 0.9797,
+      "step": 41075
+    },
+    {
+      "epoch": 2.791140100557141,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.255537437151787e-05,
+      "loss": 0.9329,
+      "step": 41080
+    },
+    {
+      "epoch": 2.791479820627803,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.2553251121076236e-05,
+      "loss": 0.9499,
+      "step": 41085
+    },
+    {
+      "epoch": 2.7918195406984645,
+      "grad_norm": 5.71875,
+      "learning_rate": 3.25511278706346e-05,
+      "loss": 0.856,
+      "step": 41090
+    },
+    {
+      "epoch": 2.792159260769126,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.2549004620192964e-05,
+      "loss": 1.0216,
+      "step": 41095
+    },
+    {
+      "epoch": 2.792498980839788,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.254688136975133e-05,
+      "loss": 0.9214,
+      "step": 41100
+    },
+    {
+      "epoch": 2.79283870091045,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.254475811930969e-05,
+      "loss": 1.0133,
+      "step": 41105
+    },
+    {
+      "epoch": 2.7931784209811115,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.2542634868868056e-05,
+      "loss": 1.0238,
+      "step": 41110
+    },
+    {
+      "epoch": 2.7935181410517735,
+      "grad_norm": 7.0,
+      "learning_rate": 3.254051161842642e-05,
+      "loss": 0.9704,
+      "step": 41115
+    },
+    {
+      "epoch": 2.793857861122435,
+      "grad_norm": 6.125,
+      "learning_rate": 3.2538388367984784e-05,
+      "loss": 0.9501,
+      "step": 41120
+    },
+    {
+      "epoch": 2.794197581193097,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.253626511754315e-05,
+      "loss": 0.9762,
+      "step": 41125
+    },
+    {
+      "epoch": 2.794537301263759,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.2534141867101505e-05,
+      "loss": 0.8991,
+      "step": 41130
+    },
+    {
+      "epoch": 2.7948770213344205,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2532018616659876e-05,
+      "loss": 0.9383,
+      "step": 41135
+    },
+    {
+      "epoch": 2.795216741405082,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.252989536621824e-05,
+      "loss": 0.9336,
+      "step": 41140
+    },
+    {
+      "epoch": 2.795556461475744,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.25277721157766e-05,
+      "loss": 0.8367,
+      "step": 41145
+    },
+    {
+      "epoch": 2.795896181546406,
+      "grad_norm": 6.5,
+      "learning_rate": 3.252564886533497e-05,
+      "loss": 0.883,
+      "step": 41150
+    },
+    {
+      "epoch": 2.7962359016170675,
+      "grad_norm": 6.375,
+      "learning_rate": 3.252352561489333e-05,
+      "loss": 0.9174,
+      "step": 41155
+    },
+    {
+      "epoch": 2.7965756216877296,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.252140236445169e-05,
+      "loss": 1.0241,
+      "step": 41160
+    },
+    {
+      "epoch": 2.796915341758391,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.251927911401006e-05,
+      "loss": 1.0141,
+      "step": 41165
+    },
+    {
+      "epoch": 2.797255061829053,
+      "grad_norm": 6.75,
+      "learning_rate": 3.2517155863568424e-05,
+      "loss": 1.0825,
+      "step": 41170
+    },
+    {
+      "epoch": 2.797594781899715,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.251503261312678e-05,
+      "loss": 0.8909,
+      "step": 41175
+    },
+    {
+      "epoch": 2.7979345019703765,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.251290936268515e-05,
+      "loss": 0.9687,
+      "step": 41180
+    },
+    {
+      "epoch": 2.798274222041038,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.2510786112243516e-05,
+      "loss": 0.8955,
+      "step": 41185
+    },
+    {
+      "epoch": 2.7986139421117002,
+      "grad_norm": 6.625,
+      "learning_rate": 3.250866286180187e-05,
+      "loss": 1.0081,
+      "step": 41190
+    },
+    {
+      "epoch": 2.798953662182362,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.2506539611360244e-05,
+      "loss": 1.0788,
+      "step": 41195
+    },
+    {
+      "epoch": 2.7992933822530235,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.25044163609186e-05,
+      "loss": 1.0455,
+      "step": 41200
+    },
+    {
+      "epoch": 2.799633102323685,
+      "grad_norm": 7.5,
+      "learning_rate": 3.2502293110476965e-05,
+      "loss": 1.1031,
+      "step": 41205
+    },
+    {
+      "epoch": 2.799972822394347,
+      "grad_norm": 6.875,
+      "learning_rate": 3.2500169860035336e-05,
+      "loss": 1.0136,
+      "step": 41210
+    },
+    {
+      "epoch": 2.800312542465009,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.2498046609593693e-05,
+      "loss": 0.993,
+      "step": 41215
+    },
+    {
+      "epoch": 2.8006522625356705,
+      "grad_norm": 6.25,
+      "learning_rate": 3.249592335915206e-05,
+      "loss": 0.9323,
+      "step": 41220
+    },
+    {
+      "epoch": 2.8009919826063325,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.249380010871043e-05,
+      "loss": 1.0839,
+      "step": 41225
+    },
+    {
+      "epoch": 2.801331702676994,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.2491676858268785e-05,
+      "loss": 0.9545,
+      "step": 41230
+    },
+    {
+      "epoch": 2.801671422747656,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.248955360782715e-05,
+      "loss": 0.9309,
+      "step": 41235
+    },
+    {
+      "epoch": 2.802011142818318,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.248743035738552e-05,
+      "loss": 0.9852,
+      "step": 41240
+    },
+    {
+      "epoch": 2.8023508628889795,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.248530710694388e-05,
+      "loss": 0.9625,
+      "step": 41245
+    },
+    {
+      "epoch": 2.802690582959641,
+      "grad_norm": 7.5,
+      "learning_rate": 3.248318385650224e-05,
+      "loss": 1.0589,
+      "step": 41250
+    },
+    {
+      "epoch": 2.8030303030303028,
+      "grad_norm": 7.125,
+      "learning_rate": 3.248106060606061e-05,
+      "loss": 0.9121,
+      "step": 41255
+    },
+    {
+      "epoch": 2.803370023100965,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.247893735561897e-05,
+      "loss": 1.0766,
+      "step": 41260
+    },
+    {
+      "epoch": 2.8037097431716265,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.2476814105177333e-05,
+      "loss": 0.9976,
+      "step": 41265
+    },
+    {
+      "epoch": 2.804049463242288,
+      "grad_norm": 6.75,
+      "learning_rate": 3.2474690854735704e-05,
+      "loss": 0.9665,
+      "step": 41270
+    },
+    {
+      "epoch": 2.80438918331295,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.247256760429406e-05,
+      "loss": 0.998,
+      "step": 41275
+    },
+    {
+      "epoch": 2.804728903383612,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.2470444353852425e-05,
+      "loss": 0.9705,
+      "step": 41280
+    },
+    {
+      "epoch": 2.8050686234542734,
+      "grad_norm": 6.5,
+      "learning_rate": 3.246832110341079e-05,
+      "loss": 0.9666,
+      "step": 41285
+    },
+    {
+      "epoch": 2.8054083435249355,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.2466197852969153e-05,
+      "loss": 0.9041,
+      "step": 41290
+    },
+    {
+      "epoch": 2.805748063595597,
+      "grad_norm": 6.25,
+      "learning_rate": 3.246407460252752e-05,
+      "loss": 0.871,
+      "step": 41295
+    },
+    {
+      "epoch": 2.8060877836662588,
+      "grad_norm": 6.25,
+      "learning_rate": 3.246195135208588e-05,
+      "loss": 1.0262,
+      "step": 41300
+    },
+    {
+      "epoch": 2.806427503736921,
+      "grad_norm": 6.125,
+      "learning_rate": 3.2459828101644245e-05,
+      "loss": 1.0073,
+      "step": 41305
+    },
+    {
+      "epoch": 2.8067672238075825,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.245770485120261e-05,
+      "loss": 0.9618,
+      "step": 41310
+    },
+    {
+      "epoch": 2.807106943878244,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.2455581600760973e-05,
+      "loss": 0.9977,
+      "step": 41315
+    },
+    {
+      "epoch": 2.807446663948906,
+      "grad_norm": 6.125,
+      "learning_rate": 3.245345835031934e-05,
+      "loss": 0.9619,
+      "step": 41320
+    },
+    {
+      "epoch": 2.807786384019568,
+      "grad_norm": 6.625,
+      "learning_rate": 3.24513350998777e-05,
+      "loss": 1.0012,
+      "step": 41325
+    },
+    {
+      "epoch": 2.8081261040902294,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2449211849436065e-05,
+      "loss": 0.9346,
+      "step": 41330
+    },
+    {
+      "epoch": 2.8084658241608915,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.244708859899443e-05,
+      "loss": 1.0132,
+      "step": 41335
+    },
+    {
+      "epoch": 2.808805544231553,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2444965348552793e-05,
+      "loss": 1.0461,
+      "step": 41340
+    },
+    {
+      "epoch": 2.809145264302215,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.244284209811116e-05,
+      "loss": 0.9527,
+      "step": 41345
+    },
+    {
+      "epoch": 2.809484984372877,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.244071884766952e-05,
+      "loss": 0.9242,
+      "step": 41350
+    },
+    {
+      "epoch": 2.8098247044435385,
+      "grad_norm": 6.25,
+      "learning_rate": 3.2438595597227885e-05,
+      "loss": 0.9962,
+      "step": 41355
+    },
+    {
+      "epoch": 2.8101644245142,
+      "grad_norm": 7.125,
+      "learning_rate": 3.243647234678625e-05,
+      "loss": 1.1026,
+      "step": 41360
+    },
+    {
+      "epoch": 2.810504144584862,
+      "grad_norm": 5.53125,
+      "learning_rate": 3.2434349096344614e-05,
+      "loss": 0.927,
+      "step": 41365
+    },
+    {
+      "epoch": 2.810843864655524,
+      "grad_norm": 6.125,
+      "learning_rate": 3.243222584590298e-05,
+      "loss": 0.9372,
+      "step": 41370
+    },
+    {
+      "epoch": 2.8111835847261855,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.243010259546134e-05,
+      "loss": 0.8808,
+      "step": 41375
+    },
+    {
+      "epoch": 2.8115233047968475,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.2427979345019706e-05,
+      "loss": 1.0112,
+      "step": 41380
+    },
+    {
+      "epoch": 2.811863024867509,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.242585609457807e-05,
+      "loss": 0.9823,
+      "step": 41385
+    },
+    {
+      "epoch": 2.812202744938171,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.2423732844136434e-05,
+      "loss": 1.0079,
+      "step": 41390
+    },
+    {
+      "epoch": 2.812542465008833,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.24216095936948e-05,
+      "loss": 1.1728,
+      "step": 41395
+    },
+    {
+      "epoch": 2.8128821850794945,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.241948634325316e-05,
+      "loss": 0.8892,
+      "step": 41400
+    },
+    {
+      "epoch": 2.813221905150156,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.2417363092811526e-05,
+      "loss": 0.9549,
+      "step": 41405
+    },
+    {
+      "epoch": 2.813561625220818,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.241523984236989e-05,
+      "loss": 1.086,
+      "step": 41410
+    },
+    {
+      "epoch": 2.81390134529148,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2413116591928254e-05,
+      "loss": 1.0031,
+      "step": 41415
+    },
+    {
+      "epoch": 2.8142410653621415,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.241099334148662e-05,
+      "loss": 0.8372,
+      "step": 41420
+    },
+    {
+      "epoch": 2.8145807854328035,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.240887009104498e-05,
+      "loss": 0.9626,
+      "step": 41425
+    },
+    {
+      "epoch": 2.814920505503465,
+      "grad_norm": 8.0,
+      "learning_rate": 3.2406746840603346e-05,
+      "loss": 0.9544,
+      "step": 41430
+    },
+    {
+      "epoch": 2.815260225574127,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.240462359016171e-05,
+      "loss": 0.9677,
+      "step": 41435
+    },
+    {
+      "epoch": 2.815599945644789,
+      "grad_norm": 7.25,
+      "learning_rate": 3.2402500339720074e-05,
+      "loss": 0.9225,
+      "step": 41440
+    },
+    {
+      "epoch": 2.8159396657154505,
+      "grad_norm": 7.0,
+      "learning_rate": 3.240037708927844e-05,
+      "loss": 0.9738,
+      "step": 41445
+    },
+    {
+      "epoch": 2.816279385786112,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.23982538388368e-05,
+      "loss": 1.0372,
+      "step": 41450
+    },
+    {
+      "epoch": 2.816619105856774,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.2396130588395166e-05,
+      "loss": 0.9557,
+      "step": 41455
+    },
+    {
+      "epoch": 2.816958825927436,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.239400733795353e-05,
+      "loss": 0.9264,
+      "step": 41460
+    },
+    {
+      "epoch": 2.8172985459980975,
+      "grad_norm": 7.25,
+      "learning_rate": 3.2391884087511894e-05,
+      "loss": 1.0111,
+      "step": 41465
+    },
+    {
+      "epoch": 2.8176382660687596,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.238976083707026e-05,
+      "loss": 0.9707,
+      "step": 41470
+    },
+    {
+      "epoch": 2.817977986139421,
+      "grad_norm": 6.625,
+      "learning_rate": 3.238763758662862e-05,
+      "loss": 1.0284,
+      "step": 41475
+    },
+    {
+      "epoch": 2.818317706210083,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.2385514336186986e-05,
+      "loss": 1.0131,
+      "step": 41480
+    },
+    {
+      "epoch": 2.818657426280745,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.238339108574534e-05,
+      "loss": 0.9309,
+      "step": 41485
+    },
+    {
+      "epoch": 2.8189971463514065,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.2381267835303714e-05,
+      "loss": 0.9455,
+      "step": 41490
+    },
+    {
+      "epoch": 2.819336866422068,
+      "grad_norm": 6.75,
+      "learning_rate": 3.237914458486208e-05,
+      "loss": 0.9939,
+      "step": 41495
+    },
+    {
+      "epoch": 2.8196765864927302,
+      "grad_norm": 7.0,
+      "learning_rate": 3.2377021334420435e-05,
+      "loss": 0.9049,
+      "step": 41500
+    },
+    {
+      "epoch": 2.820016306563392,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.2374898083978806e-05,
+      "loss": 0.9809,
+      "step": 41505
+    },
+    {
+      "epoch": 2.8203560266340535,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.237277483353717e-05,
+      "loss": 0.9658,
+      "step": 41510
+    },
+    {
+      "epoch": 2.8206957467047156,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.237065158309553e-05,
+      "loss": 0.9116,
+      "step": 41515
+    },
+    {
+      "epoch": 2.821035466775377,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.23685283326539e-05,
+      "loss": 0.9772,
+      "step": 41520
+    },
+    {
+      "epoch": 2.821375186846039,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.236640508221226e-05,
+      "loss": 0.9869,
+      "step": 41525
+    },
+    {
+      "epoch": 2.821714906916701,
+      "grad_norm": 6.375,
+      "learning_rate": 3.236428183177062e-05,
+      "loss": 1.0075,
+      "step": 41530
+    },
+    {
+      "epoch": 2.8220546269873625,
+      "grad_norm": 7.625,
+      "learning_rate": 3.236215858132899e-05,
+      "loss": 1.0158,
+      "step": 41535
+    },
+    {
+      "epoch": 2.822394347058024,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.2360035330887354e-05,
+      "loss": 1.0115,
+      "step": 41540
+    },
+    {
+      "epoch": 2.822734067128686,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.235791208044571e-05,
+      "loss": 0.9724,
+      "step": 41545
+    },
+    {
+      "epoch": 2.823073787199348,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.235578883000408e-05,
+      "loss": 0.9927,
+      "step": 41550
+    },
+    {
+      "epoch": 2.8234135072700095,
+      "grad_norm": 6.25,
+      "learning_rate": 3.235366557956244e-05,
+      "loss": 0.9817,
+      "step": 41555
+    },
+    {
+      "epoch": 2.823753227340671,
+      "grad_norm": 7.25,
+      "learning_rate": 3.23515423291208e-05,
+      "loss": 0.9691,
+      "step": 41560
+    },
+    {
+      "epoch": 2.824092947411333,
+      "grad_norm": 8.25,
+      "learning_rate": 3.2349419078679174e-05,
+      "loss": 1.0439,
+      "step": 41565
+    },
+    {
+      "epoch": 2.824432667481995,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.234729582823753e-05,
+      "loss": 0.9421,
+      "step": 41570
+    },
+    {
+      "epoch": 2.8247723875526565,
+      "grad_norm": 8.25,
+      "learning_rate": 3.2345172577795895e-05,
+      "loss": 0.9096,
+      "step": 41575
+    },
+    {
+      "epoch": 2.8251121076233185,
+      "grad_norm": 6.625,
+      "learning_rate": 3.2343049327354266e-05,
+      "loss": 1.0106,
+      "step": 41580
+    },
+    {
+      "epoch": 2.82545182769398,
+      "grad_norm": 6.5,
+      "learning_rate": 3.234092607691262e-05,
+      "loss": 0.8465,
+      "step": 41585
+    },
+    {
+      "epoch": 2.825791547764642,
+      "grad_norm": 7.5,
+      "learning_rate": 3.233880282647099e-05,
+      "loss": 0.9647,
+      "step": 41590
+    },
+    {
+      "epoch": 2.8261312678353034,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.233667957602936e-05,
+      "loss": 0.9889,
+      "step": 41595
+    },
+    {
+      "epoch": 2.8264709879059655,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.2334556325587715e-05,
+      "loss": 0.9993,
+      "step": 41600
+    },
+    {
+      "epoch": 2.826810707976627,
+      "grad_norm": 5.4375,
+      "learning_rate": 3.233243307514608e-05,
+      "loss": 0.9163,
+      "step": 41605
+    },
+    {
+      "epoch": 2.8271504280472888,
+      "grad_norm": 6.625,
+      "learning_rate": 3.233030982470445e-05,
+      "loss": 0.9288,
+      "step": 41610
+    },
+    {
+      "epoch": 2.827490148117951,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.232818657426281e-05,
+      "loss": 1.0293,
+      "step": 41615
+    },
+    {
+      "epoch": 2.8278298681886125,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.232606332382117e-05,
+      "loss": 0.9575,
+      "step": 41620
+    },
+    {
+      "epoch": 2.828169588259274,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.232394007337954e-05,
+      "loss": 0.8745,
+      "step": 41625
+    },
+    {
+      "epoch": 2.828509308329936,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.23218168229379e-05,
+      "loss": 0.9963,
+      "step": 41630
+    },
+    {
+      "epoch": 2.828849028400598,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.231969357249626e-05,
+      "loss": 0.912,
+      "step": 41635
+    },
+    {
+      "epoch": 2.8291887484712595,
+      "grad_norm": 6.875,
+      "learning_rate": 3.231757032205463e-05,
+      "loss": 0.9836,
+      "step": 41640
+    },
+    {
+      "epoch": 2.8295284685419215,
+      "grad_norm": 7.5,
+      "learning_rate": 3.231544707161299e-05,
+      "loss": 0.8899,
+      "step": 41645
+    },
+    {
+      "epoch": 2.829868188612583,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.2313323821171355e-05,
+      "loss": 1.019,
+      "step": 41650
+    },
+    {
+      "epoch": 2.830207908683245,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.231120057072972e-05,
+      "loss": 1.0644,
+      "step": 41655
+    },
+    {
+      "epoch": 2.830547628753907,
+      "grad_norm": 6.75,
+      "learning_rate": 3.230907732028808e-05,
+      "loss": 0.9688,
+      "step": 41660
+    },
+    {
+      "epoch": 2.8308873488245685,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.230695406984645e-05,
+      "loss": 1.0399,
+      "step": 41665
+    },
+    {
+      "epoch": 2.83122706889523,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.230483081940481e-05,
+      "loss": 0.9638,
+      "step": 41670
+    },
+    {
+      "epoch": 2.831566788965892,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.2302707568963175e-05,
+      "loss": 1.0624,
+      "step": 41675
+    },
+    {
+      "epoch": 2.831906509036554,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.230058431852154e-05,
+      "loss": 1.0235,
+      "step": 41680
+    },
+    {
+      "epoch": 2.8322462291072155,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.22984610680799e-05,
+      "loss": 1.0396,
+      "step": 41685
+    },
+    {
+      "epoch": 2.8325859491778775,
+      "grad_norm": 6.75,
+      "learning_rate": 3.229633781763827e-05,
+      "loss": 1.0041,
+      "step": 41690
+    },
+    {
+      "epoch": 2.832925669248539,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.229421456719663e-05,
+      "loss": 0.9821,
+      "step": 41695
+    },
+    {
+      "epoch": 2.833265389319201,
+      "grad_norm": 7.125,
+      "learning_rate": 3.2292091316754995e-05,
+      "loss": 0.9921,
+      "step": 41700
+    },
+    {
+      "epoch": 2.833605109389863,
+      "grad_norm": 7.5,
+      "learning_rate": 3.228996806631336e-05,
+      "loss": 0.9606,
+      "step": 41705
+    },
+    {
+      "epoch": 2.8339448294605245,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.228784481587172e-05,
+      "loss": 0.9905,
+      "step": 41710
+    },
+    {
+      "epoch": 2.834284549531186,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.228572156543009e-05,
+      "loss": 0.9233,
+      "step": 41715
+    },
+    {
+      "epoch": 2.834624269601848,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.228359831498845e-05,
+      "loss": 0.8949,
+      "step": 41720
+    },
+    {
+      "epoch": 2.83496398967251,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.2281475064546815e-05,
+      "loss": 1.002,
+      "step": 41725
+    },
+    {
+      "epoch": 2.8353037097431715,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.227935181410518e-05,
+      "loss": 0.9773,
+      "step": 41730
+    },
+    {
+      "epoch": 2.8356434298138335,
+      "grad_norm": 6.375,
+      "learning_rate": 3.227722856366354e-05,
+      "loss": 0.9738,
+      "step": 41735
+    },
+    {
+      "epoch": 2.835983149884495,
+      "grad_norm": 6.25,
+      "learning_rate": 3.227510531322191e-05,
+      "loss": 0.9928,
+      "step": 41740
+    },
+    {
+      "epoch": 2.836322869955157,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.227298206278027e-05,
+      "loss": 1.0303,
+      "step": 41745
+    },
+    {
+      "epoch": 2.836662590025819,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.2270858812338635e-05,
+      "loss": 0.9159,
+      "step": 41750
+    },
+    {
+      "epoch": 2.8370023100964805,
+      "grad_norm": 5.875,
+      "learning_rate": 3.2268735561897e-05,
+      "loss": 0.9443,
+      "step": 41755
+    },
+    {
+      "epoch": 2.837342030167142,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.226661231145536e-05,
+      "loss": 0.9958,
+      "step": 41760
+    },
+    {
+      "epoch": 2.8376817502378042,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.226448906101373e-05,
+      "loss": 1.0357,
+      "step": 41765
+    },
+    {
+      "epoch": 2.838021470308466,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.226236581057209e-05,
+      "loss": 0.8862,
+      "step": 41770
+    },
+    {
+      "epoch": 2.8383611903791275,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.2260242560130455e-05,
+      "loss": 0.9706,
+      "step": 41775
+    },
+    {
+      "epoch": 2.8387009104497896,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.225811930968882e-05,
+      "loss": 0.9871,
+      "step": 41780
+    },
+    {
+      "epoch": 2.839040630520451,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.225599605924718e-05,
+      "loss": 0.9246,
+      "step": 41785
+    },
+    {
+      "epoch": 2.839380350591113,
+      "grad_norm": 7.125,
+      "learning_rate": 3.225387280880555e-05,
+      "loss": 0.9717,
+      "step": 41790
+    },
+    {
+      "epoch": 2.839720070661775,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.225174955836391e-05,
+      "loss": 1.0258,
+      "step": 41795
+    },
+    {
+      "epoch": 2.8400597907324365,
+      "grad_norm": 6.75,
+      "learning_rate": 3.2249626307922275e-05,
+      "loss": 0.905,
+      "step": 41800
+    },
+    {
+      "epoch": 2.840399510803098,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.224750305748064e-05,
+      "loss": 1.1027,
+      "step": 41805
+    },
+    {
+      "epoch": 2.8407392308737602,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.2245379807039e-05,
+      "loss": 1.0662,
+      "step": 41810
+    },
+    {
+      "epoch": 2.841078950944422,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.224325655659737e-05,
+      "loss": 0.9962,
+      "step": 41815
+    },
+    {
+      "epoch": 2.8414186710150835,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.224113330615573e-05,
+      "loss": 0.9112,
+      "step": 41820
+    },
+    {
+      "epoch": 2.8417583910857456,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.2239010055714095e-05,
+      "loss": 1.0185,
+      "step": 41825
+    },
+    {
+      "epoch": 2.842098111156407,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.223688680527246e-05,
+      "loss": 0.8752,
+      "step": 41830
+    },
+    {
+      "epoch": 2.842437831227069,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.223476355483082e-05,
+      "loss": 0.952,
+      "step": 41835
+    },
+    {
+      "epoch": 2.842777551297731,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.223264030438918e-05,
+      "loss": 1.0244,
+      "step": 41840
+    },
+    {
+      "epoch": 2.8431172713683925,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.223051705394755e-05,
+      "loss": 0.9043,
+      "step": 41845
+    },
+    {
+      "epoch": 2.843456991439054,
+      "grad_norm": 7.25,
+      "learning_rate": 3.2228393803505915e-05,
+      "loss": 0.9679,
+      "step": 41850
+    },
+    {
+      "epoch": 2.8437967115097162,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.222627055306427e-05,
+      "loss": 0.9392,
+      "step": 41855
+    },
+    {
+      "epoch": 2.844136431580378,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.222414730262264e-05,
+      "loss": 0.8729,
+      "step": 41860
+    },
+    {
+      "epoch": 2.8444761516510395,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.222202405218101e-05,
+      "loss": 0.8803,
+      "step": 41865
+    },
+    {
+      "epoch": 2.8448158717217016,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.2219900801739364e-05,
+      "loss": 1.0358,
+      "step": 41870
+    },
+    {
+      "epoch": 2.845155591792363,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.2217777551297735e-05,
+      "loss": 0.9657,
+      "step": 41875
+    },
+    {
+      "epoch": 2.845495311863025,
+      "grad_norm": 7.0,
+      "learning_rate": 3.22156543008561e-05,
+      "loss": 1.0192,
+      "step": 41880
+    },
+    {
+      "epoch": 2.8458350319336865,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.2213531050414456e-05,
+      "loss": 0.9332,
+      "step": 41885
+    },
+    {
+      "epoch": 2.8461747520043486,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.221140779997283e-05,
+      "loss": 1.0087,
+      "step": 41890
+    },
+    {
+      "epoch": 2.84651447207501,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.220928454953119e-05,
+      "loss": 0.9686,
+      "step": 41895
+    },
+    {
+      "epoch": 2.846854192145672,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.220716129908955e-05,
+      "loss": 0.918,
+      "step": 41900
+    },
+    {
+      "epoch": 2.847193912216334,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.220503804864792e-05,
+      "loss": 0.9176,
+      "step": 41905
+    },
+    {
+      "epoch": 2.8475336322869955,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.2202914798206276e-05,
+      "loss": 1.0132,
+      "step": 41910
+    },
+    {
+      "epoch": 2.847873352357657,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.220079154776464e-05,
+      "loss": 1.042,
+      "step": 41915
+    },
+    {
+      "epoch": 2.8482130724283192,
+      "grad_norm": 6.75,
+      "learning_rate": 3.219866829732301e-05,
+      "loss": 0.979,
+      "step": 41920
+    },
+    {
+      "epoch": 2.848552792498981,
+      "grad_norm": 6.75,
+      "learning_rate": 3.219654504688137e-05,
+      "loss": 0.9315,
+      "step": 41925
+    },
+    {
+      "epoch": 2.8488925125696425,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.219442179643973e-05,
+      "loss": 1.079,
+      "step": 41930
+    },
+    {
+      "epoch": 2.849232232640304,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.21922985459981e-05,
+      "loss": 0.9295,
+      "step": 41935
+    },
+    {
+      "epoch": 2.849571952710966,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.219017529555646e-05,
+      "loss": 0.8928,
+      "step": 41940
+    },
+    {
+      "epoch": 2.849911672781628,
+      "grad_norm": 7.375,
+      "learning_rate": 3.2188052045114824e-05,
+      "loss": 0.967,
+      "step": 41945
+    },
+    {
+      "epoch": 2.8502513928522895,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.2185928794673195e-05,
+      "loss": 0.9344,
+      "step": 41950
+    },
+    {
+      "epoch": 2.8505911129229515,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.218380554423155e-05,
+      "loss": 0.969,
+      "step": 41955
+    },
+    {
+      "epoch": 2.850930832993613,
+      "grad_norm": 6.875,
+      "learning_rate": 3.2181682293789916e-05,
+      "loss": 0.9799,
+      "step": 41960
+    },
+    {
+      "epoch": 2.851270553064275,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.217955904334829e-05,
+      "loss": 0.8763,
+      "step": 41965
+    },
+    {
+      "epoch": 2.851610273134937,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.2177435792906644e-05,
+      "loss": 1.0221,
+      "step": 41970
+    },
+    {
+      "epoch": 2.8519499932055985,
+      "grad_norm": 6.875,
+      "learning_rate": 3.217531254246501e-05,
+      "loss": 0.9924,
+      "step": 41975
+    },
+    {
+      "epoch": 2.85228971327626,
+      "grad_norm": 6.375,
+      "learning_rate": 3.217318929202337e-05,
+      "loss": 0.9435,
+      "step": 41980
+    },
+    {
+      "epoch": 2.852629433346922,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.2171066041581736e-05,
+      "loss": 0.9242,
+      "step": 41985
+    },
+    {
+      "epoch": 2.852969153417584,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.21689427911401e-05,
+      "loss": 0.9552,
+      "step": 41990
+    },
+    {
+      "epoch": 2.8533088734882455,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2166819540698464e-05,
+      "loss": 0.9618,
+      "step": 41995
+    },
+    {
+      "epoch": 2.8536485935589075,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.216469629025683e-05,
+      "loss": 1.0098,
+      "step": 42000
+    },
+    {
+      "epoch": 2.853988313629569,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.216257303981519e-05,
+      "loss": 0.9552,
+      "step": 42005
+    },
+    {
+      "epoch": 2.854328033700231,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.2160449789373556e-05,
+      "loss": 0.9044,
+      "step": 42010
+    },
+    {
+      "epoch": 2.854667753770893,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.215832653893192e-05,
+      "loss": 0.9661,
+      "step": 42015
+    },
+    {
+      "epoch": 2.8550074738415545,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.2156203288490284e-05,
+      "loss": 1.1102,
+      "step": 42020
+    },
+    {
+      "epoch": 2.855347193912216,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.215408003804865e-05,
+      "loss": 0.9308,
+      "step": 42025
+    },
+    {
+      "epoch": 2.855686913982878,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.215195678760701e-05,
+      "loss": 0.964,
+      "step": 42030
+    },
+    {
+      "epoch": 2.85602663405354,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.2149833537165376e-05,
+      "loss": 0.9,
+      "step": 42035
+    },
+    {
+      "epoch": 2.8563663541242015,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.214771028672374e-05,
+      "loss": 0.9863,
+      "step": 42040
+    },
+    {
+      "epoch": 2.8567060741948636,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.2145587036282104e-05,
+      "loss": 0.9906,
+      "step": 42045
+    },
+    {
+      "epoch": 2.857045794265525,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.214346378584047e-05,
+      "loss": 0.8068,
+      "step": 42050
+    },
+    {
+      "epoch": 2.857385514336187,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.214134053539883e-05,
+      "loss": 0.9946,
+      "step": 42055
+    },
+    {
+      "epoch": 2.857725234406849,
+      "grad_norm": 6.25,
+      "learning_rate": 3.2139217284957196e-05,
+      "loss": 0.9774,
+      "step": 42060
+    },
+    {
+      "epoch": 2.8580649544775105,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.213709403451556e-05,
+      "loss": 1.0025,
+      "step": 42065
+    },
+    {
+      "epoch": 2.858404674548172,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.2134970784073924e-05,
+      "loss": 0.987,
+      "step": 42070
+    },
+    {
+      "epoch": 2.8587443946188342,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.213284753363229e-05,
+      "loss": 0.9526,
+      "step": 42075
+    },
+    {
+      "epoch": 2.859084114689496,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.213072428319065e-05,
+      "loss": 0.9303,
+      "step": 42080
+    },
+    {
+      "epoch": 2.8594238347601575,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.2128601032749017e-05,
+      "loss": 0.9929,
+      "step": 42085
+    },
+    {
+      "epoch": 2.8597635548308196,
+      "grad_norm": 7.0,
+      "learning_rate": 3.212647778230738e-05,
+      "loss": 0.9931,
+      "step": 42090
+    },
+    {
+      "epoch": 2.860103274901481,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.2124354531865745e-05,
+      "loss": 0.9134,
+      "step": 42095
+    },
+    {
+      "epoch": 2.860442994972143,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.212223128142411e-05,
+      "loss": 0.9317,
+      "step": 42100
+    },
+    {
+      "epoch": 2.860782715042805,
+      "grad_norm": 5.0625,
+      "learning_rate": 3.212010803098247e-05,
+      "loss": 1.046,
+      "step": 42105
+    },
+    {
+      "epoch": 2.8611224351134665,
+      "grad_norm": 7.25,
+      "learning_rate": 3.2117984780540837e-05,
+      "loss": 1.1204,
+      "step": 42110
+    },
+    {
+      "epoch": 2.861462155184128,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.21158615300992e-05,
+      "loss": 0.9613,
+      "step": 42115
+    },
+    {
+      "epoch": 2.8618018752547902,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.2113738279657565e-05,
+      "loss": 0.9954,
+      "step": 42120
+    },
+    {
+      "epoch": 2.862141595325452,
+      "grad_norm": 7.625,
+      "learning_rate": 3.211161502921593e-05,
+      "loss": 1.0129,
+      "step": 42125
+    },
+    {
+      "epoch": 2.8624813153961135,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.210949177877429e-05,
+      "loss": 0.9826,
+      "step": 42130
+    },
+    {
+      "epoch": 2.8628210354667756,
+      "grad_norm": 7.5,
+      "learning_rate": 3.2107368528332657e-05,
+      "loss": 0.9221,
+      "step": 42135
+    },
+    {
+      "epoch": 2.863160755537437,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.210524527789102e-05,
+      "loss": 0.9286,
+      "step": 42140
+    },
+    {
+      "epoch": 2.863500475608099,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2103122027449385e-05,
+      "loss": 0.9248,
+      "step": 42145
+    },
+    {
+      "epoch": 2.863840195678761,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.210099877700775e-05,
+      "loss": 0.9273,
+      "step": 42150
+    },
+    {
+      "epoch": 2.8641799157494225,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.209887552656611e-05,
+      "loss": 0.9103,
+      "step": 42155
+    },
+    {
+      "epoch": 2.864519635820084,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.2096752276124477e-05,
+      "loss": 0.9174,
+      "step": 42160
+    },
+    {
+      "epoch": 2.8648593558907463,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.209462902568284e-05,
+      "loss": 0.9356,
+      "step": 42165
+    },
+    {
+      "epoch": 2.865199075961408,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.2092505775241205e-05,
+      "loss": 0.9579,
+      "step": 42170
+    },
+    {
+      "epoch": 2.8655387960320695,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.209038252479957e-05,
+      "loss": 1.0007,
+      "step": 42175
+    },
+    {
+      "epoch": 2.8658785161027316,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.2088259274357926e-05,
+      "loss": 0.9505,
+      "step": 42180
+    },
+    {
+      "epoch": 2.866218236173393,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.2086136023916297e-05,
+      "loss": 0.8921,
+      "step": 42185
+    },
+    {
+      "epoch": 2.866557956244055,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.208401277347466e-05,
+      "loss": 1.0639,
+      "step": 42190
+    },
+    {
+      "epoch": 2.866897676314717,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.208188952303302e-05,
+      "loss": 1.0501,
+      "step": 42195
+    },
+    {
+      "epoch": 2.8672373963853786,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.207976627259139e-05,
+      "loss": 0.9752,
+      "step": 42200
+    },
+    {
+      "epoch": 2.86757711645604,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.207764302214975e-05,
+      "loss": 0.9546,
+      "step": 42205
+    },
+    {
+      "epoch": 2.8679168365267023,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.207551977170811e-05,
+      "loss": 0.9424,
+      "step": 42210
+    },
+    {
+      "epoch": 2.868256556597364,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.207339652126648e-05,
+      "loss": 0.9393,
+      "step": 42215
+    },
+    {
+      "epoch": 2.8685962766680255,
+      "grad_norm": 7.125,
+      "learning_rate": 3.2071273270824845e-05,
+      "loss": 0.9942,
+      "step": 42220
+    },
+    {
+      "epoch": 2.868935996738687,
+      "grad_norm": 7.375,
+      "learning_rate": 3.20691500203832e-05,
+      "loss": 1.0095,
+      "step": 42225
+    },
+    {
+      "epoch": 2.8692757168093492,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.206702676994157e-05,
+      "loss": 0.9077,
+      "step": 42230
+    },
+    {
+      "epoch": 2.869615436880011,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.206490351949994e-05,
+      "loss": 1.0223,
+      "step": 42235
+    },
+    {
+      "epoch": 2.8699551569506725,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.2062780269058294e-05,
+      "loss": 1.0249,
+      "step": 42240
+    },
+    {
+      "epoch": 2.8702948770213346,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.2060657018616665e-05,
+      "loss": 0.9461,
+      "step": 42245
+    },
+    {
+      "epoch": 2.870634597091996,
+      "grad_norm": 6.625,
+      "learning_rate": 3.205853376817503e-05,
+      "loss": 0.9329,
+      "step": 42250
+    },
+    {
+      "epoch": 2.870974317162658,
+      "grad_norm": 6.625,
+      "learning_rate": 3.2056410517733386e-05,
+      "loss": 1.0225,
+      "step": 42255
+    },
+    {
+      "epoch": 2.87131403723332,
+      "grad_norm": 7.0,
+      "learning_rate": 3.205428726729176e-05,
+      "loss": 1.0044,
+      "step": 42260
+    },
+    {
+      "epoch": 2.8716537573039815,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.2052164016850114e-05,
+      "loss": 0.9817,
+      "step": 42265
+    },
+    {
+      "epoch": 2.871993477374643,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.205004076640848e-05,
+      "loss": 0.9266,
+      "step": 42270
+    },
+    {
+      "epoch": 2.872333197445305,
+      "grad_norm": 6.875,
+      "learning_rate": 3.204791751596685e-05,
+      "loss": 0.977,
+      "step": 42275
+    },
+    {
+      "epoch": 2.872672917515967,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.2045794265525206e-05,
+      "loss": 0.9788,
+      "step": 42280
+    },
+    {
+      "epoch": 2.8730126375866285,
+      "grad_norm": 6.5,
+      "learning_rate": 3.204367101508357e-05,
+      "loss": 0.9814,
+      "step": 42285
+    },
+    {
+      "epoch": 2.87335235765729,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.204154776464194e-05,
+      "loss": 0.9788,
+      "step": 42290
+    },
+    {
+      "epoch": 2.873692077727952,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.20394245142003e-05,
+      "loss": 0.9776,
+      "step": 42295
+    },
+    {
+      "epoch": 2.874031797798614,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.203730126375866e-05,
+      "loss": 1.0857,
+      "step": 42300
+    },
+    {
+      "epoch": 2.8743715178692755,
+      "grad_norm": 6.625,
+      "learning_rate": 3.203517801331703e-05,
+      "loss": 0.9447,
+      "step": 42305
+    },
+    {
+      "epoch": 2.8747112379399375,
+      "grad_norm": 6.875,
+      "learning_rate": 3.203305476287539e-05,
+      "loss": 0.9606,
+      "step": 42310
+    },
+    {
+      "epoch": 2.875050958010599,
+      "grad_norm": 6.875,
+      "learning_rate": 3.2030931512433754e-05,
+      "loss": 0.9283,
+      "step": 42315
+    },
+    {
+      "epoch": 2.875390678081261,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.2028808261992125e-05,
+      "loss": 0.9508,
+      "step": 42320
+    },
+    {
+      "epoch": 2.875730398151923,
+      "grad_norm": 7.0,
+      "learning_rate": 3.202668501155048e-05,
+      "loss": 0.941,
+      "step": 42325
+    },
+    {
+      "epoch": 2.8760701182225845,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.2024561761108846e-05,
+      "loss": 1.0058,
+      "step": 42330
+    },
+    {
+      "epoch": 2.876409838293246,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.202243851066721e-05,
+      "loss": 0.9162,
+      "step": 42335
+    },
+    {
+      "epoch": 2.876749558363908,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.2020315260225574e-05,
+      "loss": 0.9717,
+      "step": 42340
+    },
+    {
+      "epoch": 2.87708927843457,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.201819200978394e-05,
+      "loss": 0.9333,
+      "step": 42345
+    },
+    {
+      "epoch": 2.8774289985052315,
+      "grad_norm": 6.25,
+      "learning_rate": 3.20160687593423e-05,
+      "loss": 1.0005,
+      "step": 42350
+    },
+    {
+      "epoch": 2.8777687185758936,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.2013945508900666e-05,
+      "loss": 0.9068,
+      "step": 42355
+    },
+    {
+      "epoch": 2.878108438646555,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.201182225845903e-05,
+      "loss": 1.0888,
+      "step": 42360
+    },
+    {
+      "epoch": 2.878448158717217,
+      "grad_norm": 7.625,
+      "learning_rate": 3.2009699008017394e-05,
+      "loss": 1.0595,
+      "step": 42365
+    },
+    {
+      "epoch": 2.878787878787879,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.200757575757576e-05,
+      "loss": 0.9457,
+      "step": 42370
+    },
+    {
+      "epoch": 2.8791275988585405,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.200545250713412e-05,
+      "loss": 1.057,
+      "step": 42375
+    },
+    {
+      "epoch": 2.879467318929202,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.2003329256692486e-05,
+      "loss": 0.9914,
+      "step": 42380
+    },
+    {
+      "epoch": 2.8798070389998642,
+      "grad_norm": 6.75,
+      "learning_rate": 3.200120600625085e-05,
+      "loss": 0.9822,
+      "step": 42385
+    },
+    {
+      "epoch": 2.880146759070526,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.1999082755809214e-05,
+      "loss": 0.9352,
+      "step": 42390
+    },
+    {
+      "epoch": 2.8804864791411875,
+      "grad_norm": 5.5,
+      "learning_rate": 3.199695950536758e-05,
+      "loss": 0.8837,
+      "step": 42395
+    },
+    {
+      "epoch": 2.8808261992118496,
+      "grad_norm": 6.875,
+      "learning_rate": 3.199483625492594e-05,
+      "loss": 1.0376,
+      "step": 42400
+    },
+    {
+      "epoch": 2.881165919282511,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.1992713004484306e-05,
+      "loss": 0.9139,
+      "step": 42405
+    },
+    {
+      "epoch": 2.881505639353173,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.199058975404267e-05,
+      "loss": 0.9869,
+      "step": 42410
+    },
+    {
+      "epoch": 2.881845359423835,
+      "grad_norm": 6.875,
+      "learning_rate": 3.1988466503601034e-05,
+      "loss": 1.0394,
+      "step": 42415
+    },
+    {
+      "epoch": 2.8821850794944965,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.19863432531594e-05,
+      "loss": 0.9871,
+      "step": 42420
+    },
+    {
+      "epoch": 2.882524799565158,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.198422000271776e-05,
+      "loss": 0.9849,
+      "step": 42425
+    },
+    {
+      "epoch": 2.8828645196358202,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.1982096752276126e-05,
+      "loss": 0.959,
+      "step": 42430
+    },
+    {
+      "epoch": 2.883204239706482,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.197997350183449e-05,
+      "loss": 0.9597,
+      "step": 42435
+    },
+    {
+      "epoch": 2.8835439597771435,
+      "grad_norm": 6.5,
+      "learning_rate": 3.1977850251392854e-05,
+      "loss": 1.007,
+      "step": 42440
+    },
+    {
+      "epoch": 2.8838836798478056,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.197572700095122e-05,
+      "loss": 0.9634,
+      "step": 42445
+    },
+    {
+      "epoch": 2.884223399918467,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.197360375050958e-05,
+      "loss": 0.9424,
+      "step": 42450
+    },
+    {
+      "epoch": 2.884563119989129,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.1971480500067946e-05,
+      "loss": 0.9181,
+      "step": 42455
+    },
+    {
+      "epoch": 2.884902840059791,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.196935724962631e-05,
+      "loss": 1.0304,
+      "step": 42460
+    },
+    {
+      "epoch": 2.8852425601304525,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.1967233999184674e-05,
+      "loss": 1.0001,
+      "step": 42465
+    },
+    {
+      "epoch": 2.885582280201114,
+      "grad_norm": 7.5,
+      "learning_rate": 3.196511074874304e-05,
+      "loss": 0.9989,
+      "step": 42470
+    },
+    {
+      "epoch": 2.8859220002717763,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.19629874983014e-05,
+      "loss": 1.0803,
+      "step": 42475
+    },
+    {
+      "epoch": 2.886261720342438,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.1960864247859766e-05,
+      "loss": 1.0455,
+      "step": 42480
+    },
+    {
+      "epoch": 2.8866014404130995,
+      "grad_norm": 5.90625,
+      "learning_rate": 3.195874099741813e-05,
+      "loss": 0.9206,
+      "step": 42485
+    },
+    {
+      "epoch": 2.8869411604837616,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.1956617746976494e-05,
+      "loss": 0.9623,
+      "step": 42490
+    },
+    {
+      "epoch": 2.8872808805544232,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.195449449653486e-05,
+      "loss": 1.0577,
+      "step": 42495
+    },
+    {
+      "epoch": 2.887620600625085,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.195237124609322e-05,
+      "loss": 0.9374,
+      "step": 42500
+    },
+    {
+      "epoch": 2.887960320695747,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.1950247995651586e-05,
+      "loss": 0.9756,
+      "step": 42505
+    },
+    {
+      "epoch": 2.8883000407664086,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.194812474520995e-05,
+      "loss": 0.9694,
+      "step": 42510
+    },
+    {
+      "epoch": 2.88863976083707,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.1946001494768314e-05,
+      "loss": 0.9029,
+      "step": 42515
+    },
+    {
+      "epoch": 2.8889794809077323,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.194387824432668e-05,
+      "loss": 0.9938,
+      "step": 42520
+    },
+    {
+      "epoch": 2.889319200978394,
+      "grad_norm": 6.375,
+      "learning_rate": 3.194175499388504e-05,
+      "loss": 0.8724,
+      "step": 42525
+    },
+    {
+      "epoch": 2.8896589210490555,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.1939631743443406e-05,
+      "loss": 0.8853,
+      "step": 42530
+    },
+    {
+      "epoch": 2.8899986411197176,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.193750849300176e-05,
+      "loss": 0.9941,
+      "step": 42535
+    },
+    {
+      "epoch": 2.8903383611903792,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.1935385242560134e-05,
+      "loss": 0.9564,
+      "step": 42540
+    },
+    {
+      "epoch": 2.890678081261041,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.19332619921185e-05,
+      "loss": 0.9947,
+      "step": 42545
+    },
+    {
+      "epoch": 2.891017801331703,
+      "grad_norm": 8.75,
+      "learning_rate": 3.1931138741676855e-05,
+      "loss": 0.9846,
+      "step": 42550
+    },
+    {
+      "epoch": 2.8913575214023646,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1929015491235226e-05,
+      "loss": 0.8934,
+      "step": 42555
+    },
+    {
+      "epoch": 2.891697241473026,
+      "grad_norm": 6.75,
+      "learning_rate": 3.192689224079359e-05,
+      "loss": 1.0502,
+      "step": 42560
+    },
+    {
+      "epoch": 2.8920369615436883,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.192476899035195e-05,
+      "loss": 1.0197,
+      "step": 42565
+    },
+    {
+      "epoch": 2.89237668161435,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.192264573991032e-05,
+      "loss": 1.3147,
+      "step": 42570
+    },
+    {
+      "epoch": 2.8927164016850115,
+      "grad_norm": 6.25,
+      "learning_rate": 3.192052248946868e-05,
+      "loss": 1.0027,
+      "step": 42575
+    },
+    {
+      "epoch": 2.893056121755673,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.191839923902704e-05,
+      "loss": 0.8879,
+      "step": 42580
+    },
+    {
+      "epoch": 2.8933958418263352,
+      "grad_norm": 6.75,
+      "learning_rate": 3.191627598858541e-05,
+      "loss": 0.9745,
+      "step": 42585
+    },
+    {
+      "epoch": 2.893735561896997,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.1914152738143774e-05,
+      "loss": 0.9439,
+      "step": 42590
+    },
+    {
+      "epoch": 2.8940752819676585,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.191202948770213e-05,
+      "loss": 1.0101,
+      "step": 42595
+    },
+    {
+      "epoch": 2.8944150020383206,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.19099062372605e-05,
+      "loss": 0.9222,
+      "step": 42600
+    },
+    {
+      "epoch": 2.894754722108982,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.190778298681886e-05,
+      "loss": 0.9017,
+      "step": 42605
+    },
+    {
+      "epoch": 2.895094442179644,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.190565973637722e-05,
+      "loss": 0.988,
+      "step": 42610
+    },
+    {
+      "epoch": 2.8954341622503055,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.1903536485935594e-05,
+      "loss": 0.9241,
+      "step": 42615
+    },
+    {
+      "epoch": 2.8957738823209676,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.190141323549395e-05,
+      "loss": 1.0266,
+      "step": 42620
+    },
+    {
+      "epoch": 2.896113602391629,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.1899289985052315e-05,
+      "loss": 0.9495,
+      "step": 42625
+    },
+    {
+      "epoch": 2.896453322462291,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.1897166734610686e-05,
+      "loss": 0.93,
+      "step": 42630
+    },
+    {
+      "epoch": 2.896793042532953,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.189504348416904e-05,
+      "loss": 0.91,
+      "step": 42635
+    },
+    {
+      "epoch": 2.8971327626036145,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.189292023372741e-05,
+      "loss": 1.0078,
+      "step": 42640
+    },
+    {
+      "epoch": 2.897472482674276,
+      "grad_norm": 5.375,
+      "learning_rate": 3.189079698328578e-05,
+      "loss": 1.0725,
+      "step": 42645
+    },
+    {
+      "epoch": 2.8978122027449382,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.1888673732844135e-05,
+      "loss": 0.9608,
+      "step": 42650
+    },
+    {
+      "epoch": 2.8981519228156,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.18865504824025e-05,
+      "loss": 1.0101,
+      "step": 42655
+    },
+    {
+      "epoch": 2.8984916428862615,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.188442723196087e-05,
+      "loss": 0.9698,
+      "step": 42660
+    },
+    {
+      "epoch": 2.8988313629569236,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.188230398151923e-05,
+      "loss": 1.0478,
+      "step": 42665
+    },
+    {
+      "epoch": 2.899171083027585,
+      "grad_norm": 6.125,
+      "learning_rate": 3.188018073107759e-05,
+      "loss": 0.9294,
+      "step": 42670
+    },
+    {
+      "epoch": 2.899510803098247,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.187805748063596e-05,
+      "loss": 0.9947,
+      "step": 42675
+    },
+    {
+      "epoch": 2.899850523168909,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.187593423019432e-05,
+      "loss": 1.0011,
+      "step": 42680
+    },
+    {
+      "epoch": 2.9001902432395705,
+      "grad_norm": 8.0,
+      "learning_rate": 3.1873810979752683e-05,
+      "loss": 0.9349,
+      "step": 42685
+    },
+    {
+      "epoch": 2.900529963310232,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.187168772931105e-05,
+      "loss": 0.9308,
+      "step": 42690
+    },
+    {
+      "epoch": 2.9008696833808942,
+      "grad_norm": 6.625,
+      "learning_rate": 3.186956447886941e-05,
+      "loss": 1.0091,
+      "step": 42695
+    },
+    {
+      "epoch": 2.901209403451556,
+      "grad_norm": 7.0,
+      "learning_rate": 3.1867441228427775e-05,
+      "loss": 0.9527,
+      "step": 42700
+    },
+    {
+      "epoch": 2.9015491235222175,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.186531797798614e-05,
+      "loss": 0.9482,
+      "step": 42705
+    },
+    {
+      "epoch": 2.9018888435928796,
+      "grad_norm": 6.375,
+      "learning_rate": 3.1863194727544503e-05,
+      "loss": 0.9904,
+      "step": 42710
+    },
+    {
+      "epoch": 2.902228563663541,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.186107147710287e-05,
+      "loss": 0.9435,
+      "step": 42715
+    },
+    {
+      "epoch": 2.902568283734203,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.185894822666123e-05,
+      "loss": 0.9411,
+      "step": 42720
+    },
+    {
+      "epoch": 2.902908003804865,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.1856824976219595e-05,
+      "loss": 0.9209,
+      "step": 42725
+    },
+    {
+      "epoch": 2.9032477238755265,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.185470172577796e-05,
+      "loss": 0.9334,
+      "step": 42730
+    },
+    {
+      "epoch": 2.903587443946188,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1852578475336323e-05,
+      "loss": 0.9474,
+      "step": 42735
+    },
+    {
+      "epoch": 2.9039271640168502,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.185045522489469e-05,
+      "loss": 0.8493,
+      "step": 42740
+    },
+    {
+      "epoch": 2.904266884087512,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.184833197445305e-05,
+      "loss": 0.9324,
+      "step": 42745
+    },
+    {
+      "epoch": 2.9046066041581735,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.1846208724011415e-05,
+      "loss": 0.8497,
+      "step": 42750
+    },
+    {
+      "epoch": 2.9049463242288356,
+      "grad_norm": 6.25,
+      "learning_rate": 3.184408547356978e-05,
+      "loss": 0.9523,
+      "step": 42755
+    },
+    {
+      "epoch": 2.905286044299497,
+      "grad_norm": 7.625,
+      "learning_rate": 3.1841962223128143e-05,
+      "loss": 0.9992,
+      "step": 42760
+    },
+    {
+      "epoch": 2.905625764370159,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.183983897268651e-05,
+      "loss": 0.9143,
+      "step": 42765
+    },
+    {
+      "epoch": 2.905965484440821,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.183771572224487e-05,
+      "loss": 0.9855,
+      "step": 42770
+    },
+    {
+      "epoch": 2.9063052045114826,
+      "grad_norm": 6.875,
+      "learning_rate": 3.1835592471803235e-05,
+      "loss": 1.0031,
+      "step": 42775
+    },
+    {
+      "epoch": 2.906644924582144,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.18334692213616e-05,
+      "loss": 0.9938,
+      "step": 42780
+    },
+    {
+      "epoch": 2.9069846446528063,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.1831345970919963e-05,
+      "loss": 0.9191,
+      "step": 42785
+    },
+    {
+      "epoch": 2.907324364723468,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.182922272047833e-05,
+      "loss": 0.9416,
+      "step": 42790
+    },
+    {
+      "epoch": 2.9076640847941295,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.182709947003669e-05,
+      "loss": 0.9252,
+      "step": 42795
+    },
+    {
+      "epoch": 2.9080038048647916,
+      "grad_norm": 7.5,
+      "learning_rate": 3.1824976219595056e-05,
+      "loss": 0.9347,
+      "step": 42800
+    },
+    {
+      "epoch": 2.9083435249354532,
+      "grad_norm": 6.875,
+      "learning_rate": 3.182285296915342e-05,
+      "loss": 0.9413,
+      "step": 42805
+    },
+    {
+      "epoch": 2.908683245006115,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.1820729718711784e-05,
+      "loss": 1.0394,
+      "step": 42810
+    },
+    {
+      "epoch": 2.909022965076777,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.181860646827015e-05,
+      "loss": 1.0168,
+      "step": 42815
+    },
+    {
+      "epoch": 2.9093626851474386,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.181648321782851e-05,
+      "loss": 0.9228,
+      "step": 42820
+    },
+    {
+      "epoch": 2.9097024052181,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.1814359967386876e-05,
+      "loss": 0.9363,
+      "step": 42825
+    },
+    {
+      "epoch": 2.9100421252887623,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.181223671694524e-05,
+      "loss": 0.9513,
+      "step": 42830
+    },
+    {
+      "epoch": 2.910381845359424,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1810113466503604e-05,
+      "loss": 1.0332,
+      "step": 42835
+    },
+    {
+      "epoch": 2.9107215654300855,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.180799021606197e-05,
+      "loss": 0.9551,
+      "step": 42840
+    },
+    {
+      "epoch": 2.9110612855007476,
+      "grad_norm": 6.5,
+      "learning_rate": 3.180586696562033e-05,
+      "loss": 0.9016,
+      "step": 42845
+    },
+    {
+      "epoch": 2.9114010055714092,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.1803743715178696e-05,
+      "loss": 0.9282,
+      "step": 42850
+    },
+    {
+      "epoch": 2.911740725642071,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.180162046473706e-05,
+      "loss": 0.9501,
+      "step": 42855
+    },
+    {
+      "epoch": 2.912080445712733,
+      "grad_norm": 7.75,
+      "learning_rate": 3.1799497214295424e-05,
+      "loss": 1.0154,
+      "step": 42860
+    },
+    {
+      "epoch": 2.9124201657833946,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.179737396385379e-05,
+      "loss": 1.0639,
+      "step": 42865
+    },
+    {
+      "epoch": 2.912759885854056,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.179525071341215e-05,
+      "loss": 0.9073,
+      "step": 42870
+    },
+    {
+      "epoch": 2.9130996059247183,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1793127462970516e-05,
+      "loss": 0.98,
+      "step": 42875
+    },
+    {
+      "epoch": 2.91343932599538,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.179100421252888e-05,
+      "loss": 1.0384,
+      "step": 42880
+    },
+    {
+      "epoch": 2.9137790460660415,
+      "grad_norm": 6.625,
+      "learning_rate": 3.1788880962087244e-05,
+      "loss": 1.1217,
+      "step": 42885
+    },
+    {
+      "epoch": 2.9141187661367036,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.17867577116456e-05,
+      "loss": 1.065,
+      "step": 42890
+    },
+    {
+      "epoch": 2.9144584862073653,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.178463446120397e-05,
+      "loss": 0.9838,
+      "step": 42895
+    },
+    {
+      "epoch": 2.914798206278027,
+      "grad_norm": 7.25,
+      "learning_rate": 3.1782511210762336e-05,
+      "loss": 0.9592,
+      "step": 42900
+    },
+    {
+      "epoch": 2.915137926348689,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.178038796032069e-05,
+      "loss": 0.979,
+      "step": 42905
+    },
+    {
+      "epoch": 2.9154776464193506,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.1778264709879064e-05,
+      "loss": 1.0004,
+      "step": 42910
+    },
+    {
+      "epoch": 2.915817366490012,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.177614145943743e-05,
+      "loss": 1.1002,
+      "step": 42915
+    },
+    {
+      "epoch": 2.916157086560674,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1774018208995785e-05,
+      "loss": 0.9468,
+      "step": 42920
+    },
+    {
+      "epoch": 2.916496806631336,
+      "grad_norm": 8.125,
+      "learning_rate": 3.1771894958554156e-05,
+      "loss": 0.9511,
+      "step": 42925
+    },
+    {
+      "epoch": 2.9168365267019976,
+      "grad_norm": 7.125,
+      "learning_rate": 3.176977170811252e-05,
+      "loss": 0.9109,
+      "step": 42930
+    },
+    {
+      "epoch": 2.917176246772659,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.176764845767088e-05,
+      "loss": 0.9763,
+      "step": 42935
+    },
+    {
+      "epoch": 2.9175159668433213,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.176552520722925e-05,
+      "loss": 1.0498,
+      "step": 42940
+    },
+    {
+      "epoch": 2.917855686913983,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.176340195678761e-05,
+      "loss": 0.994,
+      "step": 42945
+    },
+    {
+      "epoch": 2.9181954069846445,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.176127870634597e-05,
+      "loss": 1.0082,
+      "step": 42950
+    },
+    {
+      "epoch": 2.918535127055306,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.175915545590434e-05,
+      "loss": 0.9963,
+      "step": 42955
+    },
+    {
+      "epoch": 2.9188748471259682,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.17570322054627e-05,
+      "loss": 0.8998,
+      "step": 42960
+    },
+    {
+      "epoch": 2.91921456719663,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.175490895502106e-05,
+      "loss": 0.9393,
+      "step": 42965
+    },
+    {
+      "epoch": 2.9195542872672915,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.175278570457943e-05,
+      "loss": 0.9044,
+      "step": 42970
+    },
+    {
+      "epoch": 2.9198940073379536,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.175066245413779e-05,
+      "loss": 0.9278,
+      "step": 42975
+    },
+    {
+      "epoch": 2.920233727408615,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.174853920369615e-05,
+      "loss": 0.9656,
+      "step": 42980
+    },
+    {
+      "epoch": 2.920573447479277,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.1746415953254524e-05,
+      "loss": 0.9299,
+      "step": 42985
+    },
+    {
+      "epoch": 2.920913167549939,
+      "grad_norm": 6.5,
+      "learning_rate": 3.174429270281288e-05,
+      "loss": 0.9808,
+      "step": 42990
+    },
+    {
+      "epoch": 2.9212528876206005,
+      "grad_norm": 7.75,
+      "learning_rate": 3.1742169452371245e-05,
+      "loss": 0.9975,
+      "step": 42995
+    },
+    {
+      "epoch": 2.921592607691262,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.1740046201929616e-05,
+      "loss": 0.9979,
+      "step": 43000
+    },
+    {
+      "epoch": 2.9219323277619242,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.173792295148797e-05,
+      "loss": 1.0424,
+      "step": 43005
+    },
+    {
+      "epoch": 2.922272047832586,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.173579970104634e-05,
+      "loss": 1.0135,
+      "step": 43010
+    },
+    {
+      "epoch": 2.9226117679032475,
+      "grad_norm": 6.25,
+      "learning_rate": 3.173367645060471e-05,
+      "loss": 0.8587,
+      "step": 43015
+    },
+    {
+      "epoch": 2.9229514879739096,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.1731553200163065e-05,
+      "loss": 0.9526,
+      "step": 43020
+    },
+    {
+      "epoch": 2.923291208044571,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.172942994972143e-05,
+      "loss": 0.9657,
+      "step": 43025
+    },
+    {
+      "epoch": 2.923630928115233,
+      "grad_norm": 6.875,
+      "learning_rate": 3.172730669927979e-05,
+      "loss": 1.039,
+      "step": 43030
+    },
+    {
+      "epoch": 2.923970648185895,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.172518344883816e-05,
+      "loss": 0.9141,
+      "step": 43035
+    },
+    {
+      "epoch": 2.9243103682565565,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.172306019839652e-05,
+      "loss": 1.0317,
+      "step": 43040
+    },
+    {
+      "epoch": 2.924650088327218,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.1720936947954885e-05,
+      "loss": 1.0176,
+      "step": 43045
+    },
+    {
+      "epoch": 2.9249898083978803,
+      "grad_norm": 6.25,
+      "learning_rate": 3.171881369751325e-05,
+      "loss": 0.9863,
+      "step": 43050
+    },
+    {
+      "epoch": 2.925329528468542,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.171669044707161e-05,
+      "loss": 0.9575,
+      "step": 43055
+    },
+    {
+      "epoch": 2.9256692485392035,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.171456719662998e-05,
+      "loss": 1.0967,
+      "step": 43060
+    },
+    {
+      "epoch": 2.9260089686098656,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.171244394618834e-05,
+      "loss": 1.0346,
+      "step": 43065
+    },
+    {
+      "epoch": 2.926348688680527,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.1710320695746705e-05,
+      "loss": 0.9469,
+      "step": 43070
+    },
+    {
+      "epoch": 2.926688408751189,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.170819744530507e-05,
+      "loss": 0.9651,
+      "step": 43075
+    },
+    {
+      "epoch": 2.927028128821851,
+      "grad_norm": 5.28125,
+      "learning_rate": 3.170607419486343e-05,
+      "loss": 0.9317,
+      "step": 43080
+    },
+    {
+      "epoch": 2.9273678488925126,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.17039509444218e-05,
+      "loss": 1.0247,
+      "step": 43085
+    },
+    {
+      "epoch": 2.927707568963174,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.170182769398016e-05,
+      "loss": 0.9132,
+      "step": 43090
+    },
+    {
+      "epoch": 2.9280472890338363,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.1699704443538525e-05,
+      "loss": 1.0023,
+      "step": 43095
+    },
+    {
+      "epoch": 2.928387009104498,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.1697581193096896e-05,
+      "loss": 0.9532,
+      "step": 43100
+    },
+    {
+      "epoch": 2.9287267291751595,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.169545794265525e-05,
+      "loss": 1.0169,
+      "step": 43105
+    },
+    {
+      "epoch": 2.9290664492458216,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.169333469221362e-05,
+      "loss": 1.0342,
+      "step": 43110
+    },
+    {
+      "epoch": 2.9294061693164832,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.169121144177198e-05,
+      "loss": 1.0212,
+      "step": 43115
+    },
+    {
+      "epoch": 2.929745889387145,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.1689088191330345e-05,
+      "loss": 0.8894,
+      "step": 43120
+    },
+    {
+      "epoch": 2.930085609457807,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.168696494088871e-05,
+      "loss": 1.0214,
+      "step": 43125
+    },
+    {
+      "epoch": 2.9304253295284686,
+      "grad_norm": 6.75,
+      "learning_rate": 3.168484169044707e-05,
+      "loss": 0.8932,
+      "step": 43130
+    },
+    {
+      "epoch": 2.93076504959913,
+      "grad_norm": 7.0,
+      "learning_rate": 3.168271844000544e-05,
+      "loss": 0.8595,
+      "step": 43135
+    },
+    {
+      "epoch": 2.9311047696697923,
+      "grad_norm": 7.125,
+      "learning_rate": 3.16805951895638e-05,
+      "loss": 0.9494,
+      "step": 43140
+    },
+    {
+      "epoch": 2.931444489740454,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.1678471939122165e-05,
+      "loss": 1.0389,
+      "step": 43145
+    },
+    {
+      "epoch": 2.9317842098111155,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.167634868868053e-05,
+      "loss": 1.0216,
+      "step": 43150
+    },
+    {
+      "epoch": 2.9321239298817776,
+      "grad_norm": 7.5,
+      "learning_rate": 3.167422543823889e-05,
+      "loss": 1.0346,
+      "step": 43155
+    },
+    {
+      "epoch": 2.9324636499524392,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.167210218779726e-05,
+      "loss": 0.9898,
+      "step": 43160
+    },
+    {
+      "epoch": 2.932803370023101,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.166997893735562e-05,
+      "loss": 1.0156,
+      "step": 43165
+    },
+    {
+      "epoch": 2.933143090093763,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.1667855686913985e-05,
+      "loss": 0.8825,
+      "step": 43170
+    },
+    {
+      "epoch": 2.9334828101644246,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.166573243647235e-05,
+      "loss": 0.9126,
+      "step": 43175
+    },
+    {
+      "epoch": 2.933822530235086,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.166360918603071e-05,
+      "loss": 1.036,
+      "step": 43180
+    },
+    {
+      "epoch": 2.9341622503057483,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.166148593558908e-05,
+      "loss": 1.0196,
+      "step": 43185
+    },
+    {
+      "epoch": 2.93450197037641,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.165936268514744e-05,
+      "loss": 0.8849,
+      "step": 43190
+    },
+    {
+      "epoch": 2.9348416904470715,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.1657239434705805e-05,
+      "loss": 0.8992,
+      "step": 43195
+    },
+    {
+      "epoch": 2.9351814105177336,
+      "grad_norm": 6.875,
+      "learning_rate": 3.165511618426417e-05,
+      "loss": 1.007,
+      "step": 43200
+    },
+    {
+      "epoch": 2.9355211305883953,
+      "grad_norm": 7.25,
+      "learning_rate": 3.165299293382253e-05,
+      "loss": 1.0391,
+      "step": 43205
+    },
+    {
+      "epoch": 2.935860850659057,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.16508696833809e-05,
+      "loss": 1.0344,
+      "step": 43210
+    },
+    {
+      "epoch": 2.936200570729719,
+      "grad_norm": 7.0,
+      "learning_rate": 3.164874643293926e-05,
+      "loss": 0.9921,
+      "step": 43215
+    },
+    {
+      "epoch": 2.9365402908003806,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.1646623182497625e-05,
+      "loss": 1.0051,
+      "step": 43220
+    },
+    {
+      "epoch": 2.9368800108710422,
+      "grad_norm": 6.25,
+      "learning_rate": 3.164449993205599e-05,
+      "loss": 1.0774,
+      "step": 43225
+    },
+    {
+      "epoch": 2.9372197309417043,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.1642376681614346e-05,
+      "loss": 0.936,
+      "step": 43230
+    },
+    {
+      "epoch": 2.937559451012366,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.164025343117272e-05,
+      "loss": 0.9683,
+      "step": 43235
+    },
+    {
+      "epoch": 2.9378991710830276,
+      "grad_norm": 7.0,
+      "learning_rate": 3.163813018073108e-05,
+      "loss": 1.0118,
+      "step": 43240
+    },
+    {
+      "epoch": 2.9382388911536896,
+      "grad_norm": 7.0,
+      "learning_rate": 3.163600693028944e-05,
+      "loss": 0.9455,
+      "step": 43245
+    },
+    {
+      "epoch": 2.9385786112243513,
+      "grad_norm": 7.0,
+      "learning_rate": 3.163388367984781e-05,
+      "loss": 0.9822,
+      "step": 43250
+    },
+    {
+      "epoch": 2.938918331295013,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.163176042940617e-05,
+      "loss": 0.9658,
+      "step": 43255
+    },
+    {
+      "epoch": 2.9392580513656745,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.162963717896453e-05,
+      "loss": 0.9803,
+      "step": 43260
+    },
+    {
+      "epoch": 2.9395977714363366,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.16275139285229e-05,
+      "loss": 0.9815,
+      "step": 43265
+    },
+    {
+      "epoch": 2.9399374915069982,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.1625390678081265e-05,
+      "loss": 1.0596,
+      "step": 43270
+    },
+    {
+      "epoch": 2.94027721157766,
+      "grad_norm": 6.25,
+      "learning_rate": 3.162326742763962e-05,
+      "loss": 1.0153,
+      "step": 43275
+    },
+    {
+      "epoch": 2.940616931648322,
+      "grad_norm": 9.0,
+      "learning_rate": 3.162114417719799e-05,
+      "loss": 0.9717,
+      "step": 43280
+    },
+    {
+      "epoch": 2.9409566517189836,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.161902092675636e-05,
+      "loss": 0.9156,
+      "step": 43285
+    },
+    {
+      "epoch": 2.941296371789645,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.1616897676314714e-05,
+      "loss": 1.0026,
+      "step": 43290
+    },
+    {
+      "epoch": 2.941636091860307,
+      "grad_norm": 6.75,
+      "learning_rate": 3.1614774425873085e-05,
+      "loss": 1.0465,
+      "step": 43295
+    },
+    {
+      "epoch": 2.941975811930969,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.161265117543145e-05,
+      "loss": 0.9702,
+      "step": 43300
+    },
+    {
+      "epoch": 2.9423155320016305,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.1610527924989806e-05,
+      "loss": 0.9348,
+      "step": 43305
+    },
+    {
+      "epoch": 2.942655252072292,
+      "grad_norm": 7.25,
+      "learning_rate": 3.160840467454818e-05,
+      "loss": 0.8961,
+      "step": 43310
+    },
+    {
+      "epoch": 2.9429949721429542,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.1606281424106534e-05,
+      "loss": 0.9492,
+      "step": 43315
+    },
+    {
+      "epoch": 2.943334692213616,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.16041581736649e-05,
+      "loss": 1.0237,
+      "step": 43320
+    },
+    {
+      "epoch": 2.9436744122842775,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.160203492322327e-05,
+      "loss": 0.9005,
+      "step": 43325
+    },
+    {
+      "epoch": 2.9440141323549396,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1599911672781626e-05,
+      "loss": 0.9629,
+      "step": 43330
+    },
+    {
+      "epoch": 2.944353852425601,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.159778842233999e-05,
+      "loss": 1.0455,
+      "step": 43335
+    },
+    {
+      "epoch": 2.944693572496263,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.159566517189836e-05,
+      "loss": 1.096,
+      "step": 43340
+    },
+    {
+      "epoch": 2.945033292566925,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.159354192145672e-05,
+      "loss": 0.9407,
+      "step": 43345
+    },
+    {
+      "epoch": 2.9453730126375866,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.159141867101508e-05,
+      "loss": 0.9305,
+      "step": 43350
+    },
+    {
+      "epoch": 2.945712732708248,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.158929542057345e-05,
+      "loss": 1.0124,
+      "step": 43355
+    },
+    {
+      "epoch": 2.9460524527789103,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.158717217013181e-05,
+      "loss": 0.9278,
+      "step": 43360
+    },
+    {
+      "epoch": 2.946392172849572,
+      "grad_norm": 6.25,
+      "learning_rate": 3.1585048919690174e-05,
+      "loss": 0.9489,
+      "step": 43365
+    },
+    {
+      "epoch": 2.9467318929202335,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.1582925669248545e-05,
+      "loss": 1.0258,
+      "step": 43370
+    },
+    {
+      "epoch": 2.9470716129908956,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.15808024188069e-05,
+      "loss": 0.9395,
+      "step": 43375
+    },
+    {
+      "epoch": 2.9474113330615572,
+      "grad_norm": 6.625,
+      "learning_rate": 3.1578679168365266e-05,
+      "loss": 0.9539,
+      "step": 43380
+    },
+    {
+      "epoch": 2.947751053132219,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.157655591792363e-05,
+      "loss": 0.9542,
+      "step": 43385
+    },
+    {
+      "epoch": 2.948090773202881,
+      "grad_norm": 7.75,
+      "learning_rate": 3.1574432667481994e-05,
+      "loss": 0.8688,
+      "step": 43390
+    },
+    {
+      "epoch": 2.9484304932735426,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.157230941704036e-05,
+      "loss": 0.9767,
+      "step": 43395
+    },
+    {
+      "epoch": 2.948770213344204,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.157018616659872e-05,
+      "loss": 0.9978,
+      "step": 43400
+    },
+    {
+      "epoch": 2.9491099334148663,
+      "grad_norm": 6.625,
+      "learning_rate": 3.1568062916157086e-05,
+      "loss": 1.0021,
+      "step": 43405
+    },
+    {
+      "epoch": 2.949449653485528,
+      "grad_norm": 7.125,
+      "learning_rate": 3.156593966571545e-05,
+      "loss": 0.9007,
+      "step": 43410
+    },
+    {
+      "epoch": 2.9497893735561895,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.1563816415273814e-05,
+      "loss": 0.9972,
+      "step": 43415
+    },
+    {
+      "epoch": 2.9501290936268516,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.156169316483218e-05,
+      "loss": 0.9122,
+      "step": 43420
+    },
+    {
+      "epoch": 2.9504688136975132,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.155956991439054e-05,
+      "loss": 0.9794,
+      "step": 43425
+    },
+    {
+      "epoch": 2.950808533768175,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.1557446663948906e-05,
+      "loss": 0.9337,
+      "step": 43430
+    },
+    {
+      "epoch": 2.951148253838837,
+      "grad_norm": 6.5,
+      "learning_rate": 3.155532341350727e-05,
+      "loss": 0.9386,
+      "step": 43435
+    },
+    {
+      "epoch": 2.9514879739094986,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.155320016306564e-05,
+      "loss": 1.0641,
+      "step": 43440
+    },
+    {
+      "epoch": 2.95182769398016,
+      "grad_norm": 6.75,
+      "learning_rate": 3.1551076912624e-05,
+      "loss": 1.0137,
+      "step": 43445
+    },
+    {
+      "epoch": 2.9521674140508223,
+      "grad_norm": 6.625,
+      "learning_rate": 3.154895366218236e-05,
+      "loss": 0.9017,
+      "step": 43450
+    },
+    {
+      "epoch": 2.952507134121484,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.1546830411740726e-05,
+      "loss": 0.9216,
+      "step": 43455
+    },
+    {
+      "epoch": 2.9528468541921455,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.154470716129909e-05,
+      "loss": 1.0254,
+      "step": 43460
+    },
+    {
+      "epoch": 2.9531865742628076,
+      "grad_norm": 6.0,
+      "learning_rate": 3.1542583910857454e-05,
+      "loss": 0.9258,
+      "step": 43465
+    },
+    {
+      "epoch": 2.9535262943334692,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.154046066041582e-05,
+      "loss": 0.9737,
+      "step": 43470
+    },
+    {
+      "epoch": 2.953866014404131,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.153833740997418e-05,
+      "loss": 0.9716,
+      "step": 43475
+    },
+    {
+      "epoch": 2.954205734474793,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.1536214159532546e-05,
+      "loss": 0.9139,
+      "step": 43480
+    },
+    {
+      "epoch": 2.9545454545454546,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.153409090909091e-05,
+      "loss": 0.9534,
+      "step": 43485
+    },
+    {
+      "epoch": 2.954885174616116,
+      "grad_norm": 7.25,
+      "learning_rate": 3.1531967658649274e-05,
+      "loss": 1.0377,
+      "step": 43490
+    },
+    {
+      "epoch": 2.9552248946867783,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.152984440820764e-05,
+      "loss": 0.9126,
+      "step": 43495
+    },
+    {
+      "epoch": 2.95556461475744,
+      "grad_norm": 6.0,
+      "learning_rate": 3.1527721157766e-05,
+      "loss": 0.976,
+      "step": 43500
+    },
+    {
+      "epoch": 2.9559043348281016,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.1525597907324367e-05,
+      "loss": 0.9042,
+      "step": 43505
+    },
+    {
+      "epoch": 2.9562440548987636,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.152347465688273e-05,
+      "loss": 1.014,
+      "step": 43510
+    },
+    {
+      "epoch": 2.9565837749694253,
+      "grad_norm": 5.5,
+      "learning_rate": 3.1521351406441095e-05,
+      "loss": 0.8167,
+      "step": 43515
+    },
+    {
+      "epoch": 2.956923495040087,
+      "grad_norm": 6.75,
+      "learning_rate": 3.151922815599946e-05,
+      "loss": 0.9033,
+      "step": 43520
+    },
+    {
+      "epoch": 2.957263215110749,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.151710490555782e-05,
+      "loss": 0.9603,
+      "step": 43525
+    },
+    {
+      "epoch": 2.9576029351814106,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.1514981655116187e-05,
+      "loss": 1.0259,
+      "step": 43530
+    },
+    {
+      "epoch": 2.9579426552520722,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.151285840467455e-05,
+      "loss": 0.909,
+      "step": 43535
+    },
+    {
+      "epoch": 2.9582823753227343,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.1510735154232915e-05,
+      "loss": 0.8859,
+      "step": 43540
+    },
+    {
+      "epoch": 2.958622095393396,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.150861190379128e-05,
+      "loss": 1.0073,
+      "step": 43545
+    },
+    {
+      "epoch": 2.9589618154640576,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.150648865334964e-05,
+      "loss": 0.9769,
+      "step": 43550
+    },
+    {
+      "epoch": 2.9593015355347196,
+      "grad_norm": 7.0,
+      "learning_rate": 3.1504365402908007e-05,
+      "loss": 0.8803,
+      "step": 43555
+    },
+    {
+      "epoch": 2.9596412556053813,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.150224215246637e-05,
+      "loss": 0.9124,
+      "step": 43560
+    },
+    {
+      "epoch": 2.959980975676043,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.1500118902024735e-05,
+      "loss": 0.8887,
+      "step": 43565
+    },
+    {
+      "epoch": 2.960320695746705,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.14979956515831e-05,
+      "loss": 0.9592,
+      "step": 43570
+    },
+    {
+      "epoch": 2.9606604158173666,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.149587240114146e-05,
+      "loss": 0.9913,
+      "step": 43575
+    },
+    {
+      "epoch": 2.9610001358880282,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.1493749150699827e-05,
+      "loss": 0.9969,
+      "step": 43580
+    },
+    {
+      "epoch": 2.9613398559586903,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.1491625900258184e-05,
+      "loss": 0.9257,
+      "step": 43585
+    },
+    {
+      "epoch": 2.961679576029352,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1489502649816555e-05,
+      "loss": 1.1025,
+      "step": 43590
+    },
+    {
+      "epoch": 2.9620192961000136,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.148737939937492e-05,
+      "loss": 0.9918,
+      "step": 43595
+    },
+    {
+      "epoch": 2.962359016170675,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.1485256148933276e-05,
+      "loss": 0.9942,
+      "step": 43600
+    },
+    {
+      "epoch": 2.9626987362413373,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.1483132898491647e-05,
+      "loss": 0.982,
+      "step": 43605
+    },
+    {
+      "epoch": 2.963038456311999,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.148100964805001e-05,
+      "loss": 0.9405,
+      "step": 43610
+    },
+    {
+      "epoch": 2.9633781763826605,
+      "grad_norm": 5.8125,
+      "learning_rate": 3.147888639760837e-05,
+      "loss": 0.8559,
+      "step": 43615
+    },
+    {
+      "epoch": 2.9637178964533226,
+      "grad_norm": 7.125,
+      "learning_rate": 3.147676314716674e-05,
+      "loss": 0.9305,
+      "step": 43620
+    },
+    {
+      "epoch": 2.9640576165239843,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.14746398967251e-05,
+      "loss": 0.9673,
+      "step": 43625
+    },
+    {
+      "epoch": 2.964397336594646,
+      "grad_norm": 7.25,
+      "learning_rate": 3.147251664628346e-05,
+      "loss": 0.957,
+      "step": 43630
+    },
+    {
+      "epoch": 2.9647370566653075,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.147039339584183e-05,
+      "loss": 0.9075,
+      "step": 43635
+    },
+    {
+      "epoch": 2.9650767767359696,
+      "grad_norm": 6.5,
+      "learning_rate": 3.1468270145400195e-05,
+      "loss": 0.8951,
+      "step": 43640
+    },
+    {
+      "epoch": 2.965416496806631,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.146614689495855e-05,
+      "loss": 0.9511,
+      "step": 43645
+    },
+    {
+      "epoch": 2.965756216877293,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.146402364451692e-05,
+      "loss": 1.0135,
+      "step": 43650
+    },
+    {
+      "epoch": 2.966095936947955,
+      "grad_norm": 6.5,
+      "learning_rate": 3.146190039407528e-05,
+      "loss": 0.9959,
+      "step": 43655
+    },
+    {
+      "epoch": 2.9664356570186166,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.1459777143633644e-05,
+      "loss": 0.9818,
+      "step": 43660
+    },
+    {
+      "epoch": 2.966775377089278,
+      "grad_norm": 6.875,
+      "learning_rate": 3.1457653893192015e-05,
+      "loss": 0.9771,
+      "step": 43665
+    },
+    {
+      "epoch": 2.9671150971599403,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.145553064275037e-05,
+      "loss": 0.9822,
+      "step": 43670
+    },
+    {
+      "epoch": 2.967454817230602,
+      "grad_norm": 6.75,
+      "learning_rate": 3.1453407392308736e-05,
+      "loss": 0.9256,
+      "step": 43675
+    },
+    {
+      "epoch": 2.9677945373012635,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.145128414186711e-05,
+      "loss": 0.9072,
+      "step": 43680
+    },
+    {
+      "epoch": 2.9681342573719256,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.1449160891425464e-05,
+      "loss": 1.003,
+      "step": 43685
+    },
+    {
+      "epoch": 2.9684739774425872,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.144703764098383e-05,
+      "loss": 0.9063,
+      "step": 43690
+    },
+    {
+      "epoch": 2.968813697513249,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.14449143905422e-05,
+      "loss": 0.9834,
+      "step": 43695
+    },
+    {
+      "epoch": 2.969153417583911,
+      "grad_norm": 6.5,
+      "learning_rate": 3.1442791140100556e-05,
+      "loss": 1.0278,
+      "step": 43700
+    },
+    {
+      "epoch": 2.9694931376545726,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.144066788965892e-05,
+      "loss": 0.8345,
+      "step": 43705
+    },
+    {
+      "epoch": 2.969832857725234,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.143854463921729e-05,
+      "loss": 0.9985,
+      "step": 43710
+    },
+    {
+      "epoch": 2.9701725777958963,
+      "grad_norm": 7.0,
+      "learning_rate": 3.143642138877565e-05,
+      "loss": 0.9212,
+      "step": 43715
+    },
+    {
+      "epoch": 2.970512297866558,
+      "grad_norm": 6.25,
+      "learning_rate": 3.143429813833401e-05,
+      "loss": 0.963,
+      "step": 43720
+    },
+    {
+      "epoch": 2.9708520179372195,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.143217488789238e-05,
+      "loss": 0.9255,
+      "step": 43725
+    },
+    {
+      "epoch": 2.9711917380078816,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.143005163745074e-05,
+      "loss": 0.9598,
+      "step": 43730
+    },
+    {
+      "epoch": 2.9715314580785432,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.1427928387009104e-05,
+      "loss": 0.9691,
+      "step": 43735
+    },
+    {
+      "epoch": 2.971871178149205,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.142580513656747e-05,
+      "loss": 1.0147,
+      "step": 43740
+    },
+    {
+      "epoch": 2.972210898219867,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.142368188612583e-05,
+      "loss": 1.0388,
+      "step": 43745
+    },
+    {
+      "epoch": 2.9725506182905286,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.1421558635684196e-05,
+      "loss": 0.9757,
+      "step": 43750
+    },
+    {
+      "epoch": 2.97289033836119,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.141943538524256e-05,
+      "loss": 0.9704,
+      "step": 43755
+    },
+    {
+      "epoch": 2.9732300584318523,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.1417312134800924e-05,
+      "loss": 1.0282,
+      "step": 43760
+    },
+    {
+      "epoch": 2.973569778502514,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.141518888435929e-05,
+      "loss": 0.8942,
+      "step": 43765
+    },
+    {
+      "epoch": 2.9739094985731755,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.141306563391765e-05,
+      "loss": 1.0104,
+      "step": 43770
+    },
+    {
+      "epoch": 2.9742492186438376,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.1410942383476016e-05,
+      "loss": 0.9851,
+      "step": 43775
+    },
+    {
+      "epoch": 2.9745889387144993,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.140881913303439e-05,
+      "loss": 0.9958,
+      "step": 43780
+    },
+    {
+      "epoch": 2.974928658785161,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.1406695882592744e-05,
+      "loss": 1.0243,
+      "step": 43785
+    },
+    {
+      "epoch": 2.975268378855823,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.140457263215111e-05,
+      "loss": 0.9265,
+      "step": 43790
+    },
+    {
+      "epoch": 2.9756080989264846,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.140244938170948e-05,
+      "loss": 0.9849,
+      "step": 43795
+    },
+    {
+      "epoch": 2.975947818997146,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.1400326131267836e-05,
+      "loss": 1.0186,
+      "step": 43800
+    },
+    {
+      "epoch": 2.9762875390678083,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.13982028808262e-05,
+      "loss": 1.0529,
+      "step": 43805
+    },
+    {
+      "epoch": 2.97662725913847,
+      "grad_norm": 6.625,
+      "learning_rate": 3.1396079630384564e-05,
+      "loss": 0.9194,
+      "step": 43810
+    },
+    {
+      "epoch": 2.9769669792091316,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.139395637994293e-05,
+      "loss": 0.9408,
+      "step": 43815
+    },
+    {
+      "epoch": 2.9773066992797936,
+      "grad_norm": 7.25,
+      "learning_rate": 3.139183312950129e-05,
+      "loss": 1.0708,
+      "step": 43820
+    },
+    {
+      "epoch": 2.9776464193504553,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.1389709879059656e-05,
+      "loss": 1.0125,
+      "step": 43825
+    },
+    {
+      "epoch": 2.977986139421117,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.138758662861802e-05,
+      "loss": 0.8885,
+      "step": 43830
+    },
+    {
+      "epoch": 2.978325859491779,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1385463378176384e-05,
+      "loss": 0.9387,
+      "step": 43835
+    },
+    {
+      "epoch": 2.9786655795624406,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.138334012773475e-05,
+      "loss": 0.911,
+      "step": 43840
+    },
+    {
+      "epoch": 2.9790052996331022,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.138121687729311e-05,
+      "loss": 1.0345,
+      "step": 43845
+    },
+    {
+      "epoch": 2.9793450197037643,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.1379093626851476e-05,
+      "loss": 1.0195,
+      "step": 43850
+    },
+    {
+      "epoch": 2.979684739774426,
+      "grad_norm": 5.3125,
+      "learning_rate": 3.137697037640984e-05,
+      "loss": 0.896,
+      "step": 43855
+    },
+    {
+      "epoch": 2.9800244598450876,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.1374847125968204e-05,
+      "loss": 0.9693,
+      "step": 43860
+    },
+    {
+      "epoch": 2.9803641799157496,
+      "grad_norm": 5.5,
+      "learning_rate": 3.137272387552657e-05,
+      "loss": 0.9187,
+      "step": 43865
+    },
+    {
+      "epoch": 2.9807038999864113,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.137060062508493e-05,
+      "loss": 0.8797,
+      "step": 43870
+    },
+    {
+      "epoch": 2.981043620057073,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.1368477374643296e-05,
+      "loss": 0.9163,
+      "step": 43875
+    },
+    {
+      "epoch": 2.981383340127735,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.136635412420166e-05,
+      "loss": 1.1003,
+      "step": 43880
+    },
+    {
+      "epoch": 2.9817230601983966,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1364230873760024e-05,
+      "loss": 0.9243,
+      "step": 43885
+    },
+    {
+      "epoch": 2.9820627802690582,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.136210762331839e-05,
+      "loss": 0.8869,
+      "step": 43890
+    },
+    {
+      "epoch": 2.9824025003397203,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.135998437287675e-05,
+      "loss": 0.9796,
+      "step": 43895
+    },
+    {
+      "epoch": 2.982742220410382,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.1357861122435116e-05,
+      "loss": 1.0421,
+      "step": 43900
+    },
+    {
+      "epoch": 2.9830819404810436,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.135573787199348e-05,
+      "loss": 0.9529,
+      "step": 43905
+    },
+    {
+      "epoch": 2.9834216605517057,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.1353614621551844e-05,
+      "loss": 1.0288,
+      "step": 43910
+    },
+    {
+      "epoch": 2.9837613806223673,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.135149137111021e-05,
+      "loss": 0.9728,
+      "step": 43915
+    },
+    {
+      "epoch": 2.984101100693029,
+      "grad_norm": 7.75,
+      "learning_rate": 3.134936812066857e-05,
+      "loss": 1.0363,
+      "step": 43920
+    },
+    {
+      "epoch": 2.984440820763691,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.1347244870226936e-05,
+      "loss": 0.9939,
+      "step": 43925
+    },
+    {
+      "epoch": 2.9847805408343526,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.13451216197853e-05,
+      "loss": 1.0183,
+      "step": 43930
+    },
+    {
+      "epoch": 2.9851202609050143,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.1342998369343664e-05,
+      "loss": 0.8951,
+      "step": 43935
+    },
+    {
+      "epoch": 2.985459980975676,
+      "grad_norm": 6.03125,
+      "learning_rate": 3.134087511890202e-05,
+      "loss": 0.854,
+      "step": 43940
+    },
+    {
+      "epoch": 2.985799701046338,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.133875186846039e-05,
+      "loss": 0.9211,
+      "step": 43945
+    },
+    {
+      "epoch": 2.9861394211169996,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.1336628618018756e-05,
+      "loss": 0.9544,
+      "step": 43950
+    },
+    {
+      "epoch": 2.9864791411876612,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.133450536757711e-05,
+      "loss": 0.9105,
+      "step": 43955
+    },
+    {
+      "epoch": 2.9868188612583233,
+      "grad_norm": 6.125,
+      "learning_rate": 3.1332382117135484e-05,
+      "loss": 0.9202,
+      "step": 43960
+    },
+    {
+      "epoch": 2.987158581328985,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.133025886669385e-05,
+      "loss": 0.9537,
+      "step": 43965
+    },
+    {
+      "epoch": 2.9874983013996466,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.1328135616252205e-05,
+      "loss": 0.9757,
+      "step": 43970
+    },
+    {
+      "epoch": 2.987838021470308,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.1326012365810576e-05,
+      "loss": 0.9884,
+      "step": 43975
+    },
+    {
+      "epoch": 2.9881777415409703,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.132388911536894e-05,
+      "loss": 1.0425,
+      "step": 43980
+    },
+    {
+      "epoch": 2.988517461611632,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.13217658649273e-05,
+      "loss": 0.9519,
+      "step": 43985
+    },
+    {
+      "epoch": 2.9888571816822935,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.131964261448567e-05,
+      "loss": 0.9988,
+      "step": 43990
+    },
+    {
+      "epoch": 2.9891969017529556,
+      "grad_norm": 7.75,
+      "learning_rate": 3.131751936404403e-05,
+      "loss": 0.9591,
+      "step": 43995
+    },
+    {
+      "epoch": 2.9895366218236172,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.131539611360239e-05,
+      "loss": 0.9383,
+      "step": 44000
+    },
+    {
+      "epoch": 2.989876341894279,
+      "grad_norm": 4.84375,
+      "learning_rate": 3.131327286316076e-05,
+      "loss": 0.8908,
+      "step": 44005
+    },
+    {
+      "epoch": 2.990216061964941,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.131114961271912e-05,
+      "loss": 0.9686,
+      "step": 44010
+    },
+    {
+      "epoch": 2.9905557820356026,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.130902636227748e-05,
+      "loss": 0.999,
+      "step": 44015
+    },
+    {
+      "epoch": 2.990895502106264,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.130690311183585e-05,
+      "loss": 0.9884,
+      "step": 44020
+    },
+    {
+      "epoch": 2.9912352221769263,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.130477986139421e-05,
+      "loss": 0.9577,
+      "step": 44025
+    },
+    {
+      "epoch": 2.991574942247588,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.130265661095257e-05,
+      "loss": 0.9687,
+      "step": 44030
+    },
+    {
+      "epoch": 2.9919146623182495,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.1300533360510944e-05,
+      "loss": 0.9278,
+      "step": 44035
+    },
+    {
+      "epoch": 2.9922543823889116,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.12984101100693e-05,
+      "loss": 0.9476,
+      "step": 44040
+    },
+    {
+      "epoch": 2.9925941024595732,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.1296286859627665e-05,
+      "loss": 1.0965,
+      "step": 44045
+    },
+    {
+      "epoch": 2.992933822530235,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.1294163609186036e-05,
+      "loss": 1.0056,
+      "step": 44050
+    },
+    {
+      "epoch": 2.993273542600897,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.129204035874439e-05,
+      "loss": 0.9746,
+      "step": 44055
+    },
+    {
+      "epoch": 2.9936132626715586,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.128991710830276e-05,
+      "loss": 1.0817,
+      "step": 44060
+    },
+    {
+      "epoch": 2.99395298274222,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.128779385786113e-05,
+      "loss": 0.9432,
+      "step": 44065
+    },
+    {
+      "epoch": 2.9942927028128823,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.1285670607419485e-05,
+      "loss": 0.9654,
+      "step": 44070
+    },
+    {
+      "epoch": 2.994632422883544,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.128354735697785e-05,
+      "loss": 0.9466,
+      "step": 44075
+    },
+    {
+      "epoch": 2.9949721429542056,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.128142410653622e-05,
+      "loss": 0.868,
+      "step": 44080
+    },
+    {
+      "epoch": 2.9953118630248676,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.127930085609458e-05,
+      "loss": 0.9258,
+      "step": 44085
+    },
+    {
+      "epoch": 2.9956515830955293,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.127717760565294e-05,
+      "loss": 0.8946,
+      "step": 44090
+    },
+    {
+      "epoch": 2.995991303166191,
+      "grad_norm": 6.5,
+      "learning_rate": 3.1275054355211305e-05,
+      "loss": 1.0496,
+      "step": 44095
+    },
+    {
+      "epoch": 2.996331023236853,
+      "grad_norm": 7.0,
+      "learning_rate": 3.127293110476967e-05,
+      "loss": 0.9819,
+      "step": 44100
+    },
+    {
+      "epoch": 2.9966707433075146,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.1270807854328033e-05,
+      "loss": 1.0118,
+      "step": 44105
+    },
+    {
+      "epoch": 2.9970104633781762,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.12686846038864e-05,
+      "loss": 0.9975,
+      "step": 44110
+    },
+    {
+      "epoch": 2.9973501834488383,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.126656135344476e-05,
+      "loss": 0.8977,
+      "step": 44115
+    },
+    {
+      "epoch": 2.9976899035195,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.126443810300313e-05,
+      "loss": 1.0502,
+      "step": 44120
+    },
+    {
+      "epoch": 2.9980296235901616,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.126231485256149e-05,
+      "loss": 1.0326,
+      "step": 44125
+    },
+    {
+      "epoch": 2.9983693436608236,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.1260191602119853e-05,
+      "loss": 1.0118,
+      "step": 44130
+    },
+    {
+      "epoch": 2.9987090637314853,
+      "grad_norm": 6.21875,
+      "learning_rate": 3.1258068351678224e-05,
+      "loss": 0.9512,
+      "step": 44135
+    },
+    {
+      "epoch": 2.999048783802147,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.125594510123658e-05,
+      "loss": 0.858,
+      "step": 44140
+    },
+    {
+      "epoch": 2.999388503872809,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.1253821850794945e-05,
+      "loss": 1.0889,
+      "step": 44145
+    },
+    {
+      "epoch": 2.9997282239434706,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.1251698600353316e-05,
+      "loss": 0.9707,
+      "step": 44150
+    },
+    {
+      "epoch": 3.0,
+      "eval_loss": 1.662725806236267,
+      "eval_runtime": 146.1731,
+      "eval_samples_per_second": 70.594,
+      "eval_steps_per_second": 8.825,
+      "step": 44154
+    },
+    {
+      "epoch": 3.0000679440141322,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.1249575349911673e-05,
+      "loss": 0.9545,
+      "step": 44155
+    },
+    {
+      "epoch": 3.0004076640847943,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.124745209947004e-05,
+      "loss": 0.8022,
+      "step": 44160
+    },
+    {
+      "epoch": 3.000747384155456,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.12453288490284e-05,
+      "loss": 0.8995,
+      "step": 44165
+    },
+    {
+      "epoch": 3.0010871042261176,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.1243205598586765e-05,
+      "loss": 0.9118,
+      "step": 44170
+    },
+    {
+      "epoch": 3.0014268242967796,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.124108234814513e-05,
+      "loss": 0.9024,
+      "step": 44175
+    },
+    {
+      "epoch": 3.0017665443674413,
+      "grad_norm": 8.125,
+      "learning_rate": 3.1238959097703493e-05,
+      "loss": 0.8555,
+      "step": 44180
+    },
+    {
+      "epoch": 3.002106264438103,
+      "grad_norm": 7.5,
+      "learning_rate": 3.123683584726186e-05,
+      "loss": 0.9164,
+      "step": 44185
+    },
+    {
+      "epoch": 3.002445984508765,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.123471259682022e-05,
+      "loss": 0.9441,
+      "step": 44190
+    },
+    {
+      "epoch": 3.0027857045794266,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.1232589346378585e-05,
+      "loss": 0.8581,
+      "step": 44195
+    },
+    {
+      "epoch": 3.0031254246500882,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.123046609593695e-05,
+      "loss": 0.9517,
+      "step": 44200
+    },
+    {
+      "epoch": 3.0034651447207503,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1228342845495313e-05,
+      "loss": 0.8725,
+      "step": 44205
+    },
+    {
+      "epoch": 3.003804864791412,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.122621959505368e-05,
+      "loss": 0.8679,
+      "step": 44210
+    },
+    {
+      "epoch": 3.0041445848620736,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.122409634461204e-05,
+      "loss": 0.9015,
+      "step": 44215
+    },
+    {
+      "epoch": 3.004484304932735,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.1221973094170406e-05,
+      "loss": 0.9393,
+      "step": 44220
+    },
+    {
+      "epoch": 3.0048240250033973,
+      "grad_norm": 7.375,
+      "learning_rate": 3.121984984372877e-05,
+      "loss": 0.8948,
+      "step": 44225
+    },
+    {
+      "epoch": 3.005163745074059,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.1217726593287134e-05,
+      "loss": 0.8733,
+      "step": 44230
+    },
+    {
+      "epoch": 3.0055034651447206,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.12156033428455e-05,
+      "loss": 0.8172,
+      "step": 44235
+    },
+    {
+      "epoch": 3.0058431852153826,
+      "grad_norm": 6.875,
+      "learning_rate": 3.121348009240386e-05,
+      "loss": 0.8414,
+      "step": 44240
+    },
+    {
+      "epoch": 3.0061829052860443,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.1211356841962226e-05,
+      "loss": 0.956,
+      "step": 44245
+    },
+    {
+      "epoch": 3.006522625356706,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.120923359152059e-05,
+      "loss": 0.9086,
+      "step": 44250
+    },
+    {
+      "epoch": 3.006862345427368,
+      "grad_norm": 6.75,
+      "learning_rate": 3.1207110341078954e-05,
+      "loss": 0.893,
+      "step": 44255
+    },
+    {
+      "epoch": 3.0072020654980296,
+      "grad_norm": 6.375,
+      "learning_rate": 3.120498709063732e-05,
+      "loss": 1.0635,
+      "step": 44260
+    },
+    {
+      "epoch": 3.0075417855686912,
+      "grad_norm": 7.125,
+      "learning_rate": 3.120286384019568e-05,
+      "loss": 0.9986,
+      "step": 44265
+    },
+    {
+      "epoch": 3.0078815056393533,
+      "grad_norm": 7.625,
+      "learning_rate": 3.1200740589754046e-05,
+      "loss": 0.9173,
+      "step": 44270
+    },
+    {
+      "epoch": 3.008221225710015,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.119861733931241e-05,
+      "loss": 0.8698,
+      "step": 44275
+    },
+    {
+      "epoch": 3.0085609457806766,
+      "grad_norm": 7.5,
+      "learning_rate": 3.1196494088870774e-05,
+      "loss": 0.9533,
+      "step": 44280
+    },
+    {
+      "epoch": 3.0089006658513386,
+      "grad_norm": 8.0,
+      "learning_rate": 3.119437083842914e-05,
+      "loss": 0.8807,
+      "step": 44285
+    },
+    {
+      "epoch": 3.0092403859220003,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.11922475879875e-05,
+      "loss": 0.9547,
+      "step": 44290
+    },
+    {
+      "epoch": 3.009580105992662,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.119012433754586e-05,
+      "loss": 0.8923,
+      "step": 44295
+    },
+    {
+      "epoch": 3.009919826063324,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.118800108710423e-05,
+      "loss": 0.9432,
+      "step": 44300
+    },
+    {
+      "epoch": 3.0102595461339856,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.1185877836662594e-05,
+      "loss": 0.905,
+      "step": 44305
+    },
+    {
+      "epoch": 3.0105992662046472,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.118375458622095e-05,
+      "loss": 0.9423,
+      "step": 44310
+    },
+    {
+      "epoch": 3.0109389862753093,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.118163133577932e-05,
+      "loss": 0.7977,
+      "step": 44315
+    },
+    {
+      "epoch": 3.011278706345971,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.1179508085337686e-05,
+      "loss": 0.8204,
+      "step": 44320
+    },
+    {
+      "epoch": 3.0116184264166326,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.117738483489604e-05,
+      "loss": 0.8749,
+      "step": 44325
+    },
+    {
+      "epoch": 3.0119581464872947,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.1175261584454414e-05,
+      "loss": 0.9861,
+      "step": 44330
+    },
+    {
+      "epoch": 3.0122978665579563,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.117313833401278e-05,
+      "loss": 0.9603,
+      "step": 44335
+    },
+    {
+      "epoch": 3.012637586628618,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.1171015083571135e-05,
+      "loss": 0.8504,
+      "step": 44340
+    },
+    {
+      "epoch": 3.01297730669928,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.1168891833129506e-05,
+      "loss": 0.959,
+      "step": 44345
+    },
+    {
+      "epoch": 3.0133170267699416,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.116676858268787e-05,
+      "loss": 0.8939,
+      "step": 44350
+    },
+    {
+      "epoch": 3.0136567468406033,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.116464533224623e-05,
+      "loss": 0.9404,
+      "step": 44355
+    },
+    {
+      "epoch": 3.0139964669112653,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.11625220818046e-05,
+      "loss": 0.9477,
+      "step": 44360
+    },
+    {
+      "epoch": 3.014336186981927,
+      "grad_norm": 6.875,
+      "learning_rate": 3.1160398831362955e-05,
+      "loss": 0.9174,
+      "step": 44365
+    },
+    {
+      "epoch": 3.0146759070525886,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.115827558092132e-05,
+      "loss": 0.8598,
+      "step": 44370
+    },
+    {
+      "epoch": 3.01501562712325,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.115615233047969e-05,
+      "loss": 0.8851,
+      "step": 44375
+    },
+    {
+      "epoch": 3.0153553471939123,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.115402908003805e-05,
+      "loss": 0.884,
+      "step": 44380
+    },
+    {
+      "epoch": 3.015695067264574,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.115190582959641e-05,
+      "loss": 0.8724,
+      "step": 44385
+    },
+    {
+      "epoch": 3.0160347873352356,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.114978257915478e-05,
+      "loss": 0.8653,
+      "step": 44390
+    },
+    {
+      "epoch": 3.0163745074058976,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.114765932871314e-05,
+      "loss": 0.9578,
+      "step": 44395
+    },
+    {
+      "epoch": 3.0167142274765593,
+      "grad_norm": 7.375,
+      "learning_rate": 3.11455360782715e-05,
+      "loss": 0.9018,
+      "step": 44400
+    },
+    {
+      "epoch": 3.017053947547221,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.1143412827829874e-05,
+      "loss": 0.7452,
+      "step": 44405
+    },
+    {
+      "epoch": 3.017393667617883,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.114128957738823e-05,
+      "loss": 0.8637,
+      "step": 44410
+    },
+    {
+      "epoch": 3.0177333876885446,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.1139166326946595e-05,
+      "loss": 0.8566,
+      "step": 44415
+    },
+    {
+      "epoch": 3.0180731077592062,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.1137043076504966e-05,
+      "loss": 0.9163,
+      "step": 44420
+    },
+    {
+      "epoch": 3.0184128278298683,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.113491982606332e-05,
+      "loss": 0.9732,
+      "step": 44425
+    },
+    {
+      "epoch": 3.01875254790053,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.113279657562169e-05,
+      "loss": 0.9004,
+      "step": 44430
+    },
+    {
+      "epoch": 3.0190922679711916,
+      "grad_norm": 7.5,
+      "learning_rate": 3.113067332518005e-05,
+      "loss": 0.8473,
+      "step": 44435
+    },
+    {
+      "epoch": 3.0194319880418536,
+      "grad_norm": 7.125,
+      "learning_rate": 3.1128550074738415e-05,
+      "loss": 1.0238,
+      "step": 44440
+    },
+    {
+      "epoch": 3.0197717081125153,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.112642682429678e-05,
+      "loss": 0.9068,
+      "step": 44445
+    },
+    {
+      "epoch": 3.020111428183177,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.112430357385514e-05,
+      "loss": 0.8736,
+      "step": 44450
+    },
+    {
+      "epoch": 3.020451148253839,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.112218032341351e-05,
+      "loss": 0.8708,
+      "step": 44455
+    },
+    {
+      "epoch": 3.0207908683245006,
+      "grad_norm": 6.875,
+      "learning_rate": 3.112005707297187e-05,
+      "loss": 0.8247,
+      "step": 44460
+    },
+    {
+      "epoch": 3.0211305883951622,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.1117933822530235e-05,
+      "loss": 0.9506,
+      "step": 44465
+    },
+    {
+      "epoch": 3.0214703084658243,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.11158105720886e-05,
+      "loss": 0.8827,
+      "step": 44470
+    },
+    {
+      "epoch": 3.021810028536486,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.111368732164697e-05,
+      "loss": 0.8839,
+      "step": 44475
+    },
+    {
+      "epoch": 3.0221497486071476,
+      "grad_norm": 6.125,
+      "learning_rate": 3.111156407120533e-05,
+      "loss": 0.9551,
+      "step": 44480
+    },
+    {
+      "epoch": 3.0224894686778097,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.110944082076369e-05,
+      "loss": 0.9365,
+      "step": 44485
+    },
+    {
+      "epoch": 3.0228291887484713,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.110731757032206e-05,
+      "loss": 0.8352,
+      "step": 44490
+    },
+    {
+      "epoch": 3.023168908819133,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.110519431988042e-05,
+      "loss": 0.8874,
+      "step": 44495
+    },
+    {
+      "epoch": 3.023508628889795,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.110307106943878e-05,
+      "loss": 0.8633,
+      "step": 44500
+    },
+    {
+      "epoch": 3.0238483489604566,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.1100947818997154e-05,
+      "loss": 0.7853,
+      "step": 44505
+    },
+    {
+      "epoch": 3.0241880690311183,
+      "grad_norm": 6.5,
+      "learning_rate": 3.109882456855551e-05,
+      "loss": 0.9185,
+      "step": 44510
+    },
+    {
+      "epoch": 3.0245277891017803,
+      "grad_norm": 8.0,
+      "learning_rate": 3.1096701318113875e-05,
+      "loss": 0.9052,
+      "step": 44515
+    },
+    {
+      "epoch": 3.024867509172442,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.109457806767224e-05,
+      "loss": 0.8962,
+      "step": 44520
+    },
+    {
+      "epoch": 3.0252072292431036,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.10924548172306e-05,
+      "loss": 0.9114,
+      "step": 44525
+    },
+    {
+      "epoch": 3.0255469493137657,
+      "grad_norm": 7.75,
+      "learning_rate": 3.109033156678897e-05,
+      "loss": 0.8947,
+      "step": 44530
+    },
+    {
+      "epoch": 3.0258866693844273,
+      "grad_norm": 7.5,
+      "learning_rate": 3.108820831634733e-05,
+      "loss": 0.8329,
+      "step": 44535
+    },
+    {
+      "epoch": 3.026226389455089,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.1086085065905695e-05,
+      "loss": 1.0162,
+      "step": 44540
+    },
+    {
+      "epoch": 3.026566109525751,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.108396181546406e-05,
+      "loss": 0.8905,
+      "step": 44545
+    },
+    {
+      "epoch": 3.0269058295964126,
+      "grad_norm": 7.375,
+      "learning_rate": 3.108183856502242e-05,
+      "loss": 1.0276,
+      "step": 44550
+    },
+    {
+      "epoch": 3.0272455496670743,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.107971531458079e-05,
+      "loss": 0.8652,
+      "step": 44555
+    },
+    {
+      "epoch": 3.027585269737736,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.107759206413915e-05,
+      "loss": 0.8895,
+      "step": 44560
+    },
+    {
+      "epoch": 3.027924989808398,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.1075468813697515e-05,
+      "loss": 0.8978,
+      "step": 44565
+    },
+    {
+      "epoch": 3.0282647098790596,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.107334556325588e-05,
+      "loss": 0.849,
+      "step": 44570
+    },
+    {
+      "epoch": 3.0286044299497212,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.107122231281424e-05,
+      "loss": 0.9061,
+      "step": 44575
+    },
+    {
+      "epoch": 3.0289441500203833,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.106909906237261e-05,
+      "loss": 0.8487,
+      "step": 44580
+    },
+    {
+      "epoch": 3.029283870091045,
+      "grad_norm": 9.4375,
+      "learning_rate": 3.106697581193097e-05,
+      "loss": 0.9558,
+      "step": 44585
+    },
+    {
+      "epoch": 3.0296235901617066,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.1064852561489335e-05,
+      "loss": 0.9075,
+      "step": 44590
+    },
+    {
+      "epoch": 3.0299633102323686,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.10627293110477e-05,
+      "loss": 0.9195,
+      "step": 44595
+    },
+    {
+      "epoch": 3.0303030303030303,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.106060606060606e-05,
+      "loss": 0.9289,
+      "step": 44600
+    },
+    {
+      "epoch": 3.030642750373692,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.105848281016443e-05,
+      "loss": 0.8192,
+      "step": 44605
+    },
+    {
+      "epoch": 3.030982470444354,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.105635955972279e-05,
+      "loss": 0.8699,
+      "step": 44610
+    },
+    {
+      "epoch": 3.0313221905150156,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.1054236309281155e-05,
+      "loss": 0.8453,
+      "step": 44615
+    },
+    {
+      "epoch": 3.0316619105856772,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.105211305883952e-05,
+      "loss": 0.8853,
+      "step": 44620
+    },
+    {
+      "epoch": 3.0320016306563393,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.104998980839788e-05,
+      "loss": 0.9232,
+      "step": 44625
+    },
+    {
+      "epoch": 3.032341350727001,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.104786655795625e-05,
+      "loss": 0.8819,
+      "step": 44630
+    },
+    {
+      "epoch": 3.0326810707976626,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.1045743307514604e-05,
+      "loss": 0.8586,
+      "step": 44635
+    },
+    {
+      "epoch": 3.0330207908683247,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.1043620057072975e-05,
+      "loss": 0.9448,
+      "step": 44640
+    },
+    {
+      "epoch": 3.0333605109389863,
+      "grad_norm": 5.75,
+      "learning_rate": 3.104149680663134e-05,
+      "loss": 0.9542,
+      "step": 44645
+    },
+    {
+      "epoch": 3.033700231009648,
+      "grad_norm": 6.75,
+      "learning_rate": 3.1039373556189696e-05,
+      "loss": 0.8436,
+      "step": 44650
+    },
+    {
+      "epoch": 3.03403995108031,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.103725030574807e-05,
+      "loss": 0.8912,
+      "step": 44655
+    },
+    {
+      "epoch": 3.0343796711509716,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.103512705530643e-05,
+      "loss": 0.9085,
+      "step": 44660
+    },
+    {
+      "epoch": 3.0347193912216333,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.103300380486479e-05,
+      "loss": 0.9441,
+      "step": 44665
+    },
+    {
+      "epoch": 3.0350591112922953,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.103088055442316e-05,
+      "loss": 0.9007,
+      "step": 44670
+    },
+    {
+      "epoch": 3.035398831362957,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.102875730398152e-05,
+      "loss": 0.9474,
+      "step": 44675
+    },
+    {
+      "epoch": 3.0357385514336186,
+      "grad_norm": 6.625,
+      "learning_rate": 3.102663405353988e-05,
+      "loss": 0.8957,
+      "step": 44680
+    },
+    {
+      "epoch": 3.0360782715042807,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.102451080309825e-05,
+      "loss": 1.0049,
+      "step": 44685
+    },
+    {
+      "epoch": 3.0364179915749423,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.1022387552656615e-05,
+      "loss": 0.8408,
+      "step": 44690
+    },
+    {
+      "epoch": 3.036757711645604,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.102026430221497e-05,
+      "loss": 0.8618,
+      "step": 44695
+    },
+    {
+      "epoch": 3.037097431716266,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.101814105177334e-05,
+      "loss": 0.9059,
+      "step": 44700
+    },
+    {
+      "epoch": 3.0374371517869276,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.101601780133171e-05,
+      "loss": 0.7637,
+      "step": 44705
+    },
+    {
+      "epoch": 3.0377768718575893,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.1013894550890064e-05,
+      "loss": 0.915,
+      "step": 44710
+    },
+    {
+      "epoch": 3.038116591928251,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.1011771300448435e-05,
+      "loss": 0.8718,
+      "step": 44715
+    },
+    {
+      "epoch": 3.038456311998913,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.100964805000679e-05,
+      "loss": 0.8839,
+      "step": 44720
+    },
+    {
+      "epoch": 3.0387960320695746,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.1007524799565156e-05,
+      "loss": 0.9063,
+      "step": 44725
+    },
+    {
+      "epoch": 3.0391357521402362,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.100540154912353e-05,
+      "loss": 0.9675,
+      "step": 44730
+    },
+    {
+      "epoch": 3.0394754722108983,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.1003278298681884e-05,
+      "loss": 0.895,
+      "step": 44735
+    },
+    {
+      "epoch": 3.03981519228156,
+      "grad_norm": 7.25,
+      "learning_rate": 3.100115504824025e-05,
+      "loss": 0.8515,
+      "step": 44740
+    },
+    {
+      "epoch": 3.0401549123522216,
+      "grad_norm": 7.5,
+      "learning_rate": 3.099903179779862e-05,
+      "loss": 0.8434,
+      "step": 44745
+    },
+    {
+      "epoch": 3.0404946324228836,
+      "grad_norm": 7.75,
+      "learning_rate": 3.0996908547356976e-05,
+      "loss": 0.9617,
+      "step": 44750
+    },
+    {
+      "epoch": 3.0408343524935453,
+      "grad_norm": 7.375,
+      "learning_rate": 3.099478529691534e-05,
+      "loss": 0.904,
+      "step": 44755
+    },
+    {
+      "epoch": 3.041174072564207,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.099266204647371e-05,
+      "loss": 0.9053,
+      "step": 44760
+    },
+    {
+      "epoch": 3.041513792634869,
+      "grad_norm": 8.125,
+      "learning_rate": 3.099053879603207e-05,
+      "loss": 0.8355,
+      "step": 44765
+    },
+    {
+      "epoch": 3.0418535127055306,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.098841554559043e-05,
+      "loss": 0.8282,
+      "step": 44770
+    },
+    {
+      "epoch": 3.0421932327761922,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.09862922951488e-05,
+      "loss": 0.9297,
+      "step": 44775
+    },
+    {
+      "epoch": 3.0425329528468543,
+      "grad_norm": 9.0,
+      "learning_rate": 3.098416904470716e-05,
+      "loss": 0.8856,
+      "step": 44780
+    },
+    {
+      "epoch": 3.042872672917516,
+      "grad_norm": 7.875,
+      "learning_rate": 3.0982045794265524e-05,
+      "loss": 0.919,
+      "step": 44785
+    },
+    {
+      "epoch": 3.0432123929881776,
+      "grad_norm": 7.125,
+      "learning_rate": 3.097992254382389e-05,
+      "loss": 0.9785,
+      "step": 44790
+    },
+    {
+      "epoch": 3.0435521130588397,
+      "grad_norm": 6.5,
+      "learning_rate": 3.097779929338225e-05,
+      "loss": 0.8987,
+      "step": 44795
+    },
+    {
+      "epoch": 3.0438918331295013,
+      "grad_norm": 6.28125,
+      "learning_rate": 3.0975676042940616e-05,
+      "loss": 0.9094,
+      "step": 44800
+    },
+    {
+      "epoch": 3.044231553200163,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.097355279249898e-05,
+      "loss": 0.8966,
+      "step": 44805
+    },
+    {
+      "epoch": 3.044571273270825,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0971429542057344e-05,
+      "loss": 0.8479,
+      "step": 44810
+    },
+    {
+      "epoch": 3.0449109933414866,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0969306291615715e-05,
+      "loss": 0.8533,
+      "step": 44815
+    },
+    {
+      "epoch": 3.0452507134121483,
+      "grad_norm": 6.625,
+      "learning_rate": 3.096718304117407e-05,
+      "loss": 0.8578,
+      "step": 44820
+    },
+    {
+      "epoch": 3.0455904334828103,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.0965059790732436e-05,
+      "loss": 0.9431,
+      "step": 44825
+    },
+    {
+      "epoch": 3.045930153553472,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.096293654029081e-05,
+      "loss": 0.8973,
+      "step": 44830
+    },
+    {
+      "epoch": 3.0462698736241336,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.0960813289849164e-05,
+      "loss": 0.8021,
+      "step": 44835
+    },
+    {
+      "epoch": 3.0466095936947957,
+      "grad_norm": 6.875,
+      "learning_rate": 3.095869003940753e-05,
+      "loss": 0.8226,
+      "step": 44840
+    },
+    {
+      "epoch": 3.0469493137654573,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.09565667889659e-05,
+      "loss": 0.8841,
+      "step": 44845
+    },
+    {
+      "epoch": 3.047289033836119,
+      "grad_norm": 6.75,
+      "learning_rate": 3.0954443538524256e-05,
+      "loss": 0.9871,
+      "step": 44850
+    },
+    {
+      "epoch": 3.047628753906781,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.095232028808262e-05,
+      "loss": 0.9904,
+      "step": 44855
+    },
+    {
+      "epoch": 3.0479684739774426,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0950197037640984e-05,
+      "loss": 0.8785,
+      "step": 44860
+    },
+    {
+      "epoch": 3.0483081940481043,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.094807378719935e-05,
+      "loss": 0.8495,
+      "step": 44865
+    },
+    {
+      "epoch": 3.0486479141187663,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.094595053675771e-05,
+      "loss": 0.962,
+      "step": 44870
+    },
+    {
+      "epoch": 3.048987634189428,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.0943827286316076e-05,
+      "loss": 0.9377,
+      "step": 44875
+    },
+    {
+      "epoch": 3.0493273542600896,
+      "grad_norm": 6.75,
+      "learning_rate": 3.094170403587444e-05,
+      "loss": 0.8818,
+      "step": 44880
+    },
+    {
+      "epoch": 3.0496670743307517,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.0939580785432804e-05,
+      "loss": 0.8888,
+      "step": 44885
+    },
+    {
+      "epoch": 3.0500067944014133,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.093745753499117e-05,
+      "loss": 0.9901,
+      "step": 44890
+    },
+    {
+      "epoch": 3.050346514472075,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.093533428454953e-05,
+      "loss": 0.9835,
+      "step": 44895
+    },
+    {
+      "epoch": 3.0506862345427366,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.0933211034107896e-05,
+      "loss": 0.793,
+      "step": 44900
+    },
+    {
+      "epoch": 3.0510259546133986,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.093108778366626e-05,
+      "loss": 0.86,
+      "step": 44905
+    },
+    {
+      "epoch": 3.0513656746840603,
+      "grad_norm": 7.5,
+      "learning_rate": 3.0928964533224624e-05,
+      "loss": 0.825,
+      "step": 44910
+    },
+    {
+      "epoch": 3.051705394754722,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.092684128278299e-05,
+      "loss": 0.9635,
+      "step": 44915
+    },
+    {
+      "epoch": 3.052045114825384,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.092471803234135e-05,
+      "loss": 0.9351,
+      "step": 44920
+    },
+    {
+      "epoch": 3.0523848348960456,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.0922594781899716e-05,
+      "loss": 0.8637,
+      "step": 44925
+    },
+    {
+      "epoch": 3.0527245549667072,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.092047153145808e-05,
+      "loss": 0.8808,
+      "step": 44930
+    },
+    {
+      "epoch": 3.0530642750373693,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.0918348281016445e-05,
+      "loss": 0.8186,
+      "step": 44935
+    },
+    {
+      "epoch": 3.053403995108031,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.091622503057481e-05,
+      "loss": 0.8385,
+      "step": 44940
+    },
+    {
+      "epoch": 3.0537437151786926,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.091410178013317e-05,
+      "loss": 0.9217,
+      "step": 44945
+    },
+    {
+      "epoch": 3.0540834352493547,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0911978529691537e-05,
+      "loss": 0.9188,
+      "step": 44950
+    },
+    {
+      "epoch": 3.0544231553200163,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.09098552792499e-05,
+      "loss": 0.9084,
+      "step": 44955
+    },
+    {
+      "epoch": 3.054762875390678,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0907732028808265e-05,
+      "loss": 0.877,
+      "step": 44960
+    },
+    {
+      "epoch": 3.05510259546134,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.090560877836663e-05,
+      "loss": 0.9798,
+      "step": 44965
+    },
+    {
+      "epoch": 3.0554423155320016,
+      "grad_norm": 6.5,
+      "learning_rate": 3.090348552792499e-05,
+      "loss": 0.8801,
+      "step": 44970
+    },
+    {
+      "epoch": 3.0557820356026633,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.0901362277483357e-05,
+      "loss": 0.8735,
+      "step": 44975
+    },
+    {
+      "epoch": 3.0561217556733253,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.089923902704172e-05,
+      "loss": 0.9305,
+      "step": 44980
+    },
+    {
+      "epoch": 3.056461475743987,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0897115776600085e-05,
+      "loss": 0.9316,
+      "step": 44985
+    },
+    {
+      "epoch": 3.0568011958146486,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.089499252615844e-05,
+      "loss": 0.9591,
+      "step": 44990
+    },
+    {
+      "epoch": 3.0571409158853107,
+      "grad_norm": 7.375,
+      "learning_rate": 3.089286927571681e-05,
+      "loss": 0.8202,
+      "step": 44995
+    },
+    {
+      "epoch": 3.0574806359559723,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0890746025275177e-05,
+      "loss": 0.8583,
+      "step": 45000
+    },
+    {
+      "epoch": 3.057820356026634,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.0888622774833534e-05,
+      "loss": 0.9723,
+      "step": 45005
+    },
+    {
+      "epoch": 3.058160076097296,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0886499524391905e-05,
+      "loss": 0.8112,
+      "step": 45010
+    },
+    {
+      "epoch": 3.0584997961679576,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.088437627395027e-05,
+      "loss": 0.9187,
+      "step": 45015
+    },
+    {
+      "epoch": 3.0588395162386193,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.0882253023508626e-05,
+      "loss": 0.8383,
+      "step": 45020
+    },
+    {
+      "epoch": 3.0591792363092813,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0880129773066997e-05,
+      "loss": 0.9068,
+      "step": 45025
+    },
+    {
+      "epoch": 3.059518956379943,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.087800652262536e-05,
+      "loss": 0.9098,
+      "step": 45030
+    },
+    {
+      "epoch": 3.0598586764506046,
+      "grad_norm": 6.625,
+      "learning_rate": 3.087588327218372e-05,
+      "loss": 0.8856,
+      "step": 45035
+    },
+    {
+      "epoch": 3.0601983965212667,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.087376002174209e-05,
+      "loss": 0.8505,
+      "step": 45040
+    },
+    {
+      "epoch": 3.0605381165919283,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.087163677130045e-05,
+      "loss": 0.8743,
+      "step": 45045
+    },
+    {
+      "epoch": 3.06087783666259,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.086951352085881e-05,
+      "loss": 0.9103,
+      "step": 45050
+    },
+    {
+      "epoch": 3.0612175567332516,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.086739027041718e-05,
+      "loss": 0.9289,
+      "step": 45055
+    },
+    {
+      "epoch": 3.0615572768039137,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.086526701997554e-05,
+      "loss": 0.9228,
+      "step": 45060
+    },
+    {
+      "epoch": 3.0618969968745753,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.08631437695339e-05,
+      "loss": 0.9821,
+      "step": 45065
+    },
+    {
+      "epoch": 3.062236716945237,
+      "grad_norm": 6.5,
+      "learning_rate": 3.086102051909227e-05,
+      "loss": 1.0075,
+      "step": 45070
+    },
+    {
+      "epoch": 3.062576437015899,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.085889726865063e-05,
+      "loss": 0.9322,
+      "step": 45075
+    },
+    {
+      "epoch": 3.0629161570865606,
+      "grad_norm": 6.5,
+      "learning_rate": 3.0856774018208994e-05,
+      "loss": 0.9377,
+      "step": 45080
+    },
+    {
+      "epoch": 3.0632558771572223,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0854650767767365e-05,
+      "loss": 0.9113,
+      "step": 45085
+    },
+    {
+      "epoch": 3.0635955972278843,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.085252751732572e-05,
+      "loss": 0.8233,
+      "step": 45090
+    },
+    {
+      "epoch": 3.063935317298546,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.0850404266884086e-05,
+      "loss": 0.9602,
+      "step": 45095
+    },
+    {
+      "epoch": 3.0642750373692076,
+      "grad_norm": 8.125,
+      "learning_rate": 3.084828101644246e-05,
+      "loss": 0.9868,
+      "step": 45100
+    },
+    {
+      "epoch": 3.0646147574398697,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.0846157766000814e-05,
+      "loss": 0.8494,
+      "step": 45105
+    },
+    {
+      "epoch": 3.0649544775105313,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.084403451555918e-05,
+      "loss": 0.8946,
+      "step": 45110
+    },
+    {
+      "epoch": 3.065294197581193,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.084191126511755e-05,
+      "loss": 0.9311,
+      "step": 45115
+    },
+    {
+      "epoch": 3.065633917651855,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0839788014675906e-05,
+      "loss": 0.8846,
+      "step": 45120
+    },
+    {
+      "epoch": 3.0659736377225166,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.083766476423427e-05,
+      "loss": 0.8701,
+      "step": 45125
+    },
+    {
+      "epoch": 3.0663133577931783,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.083554151379264e-05,
+      "loss": 0.9077,
+      "step": 45130
+    },
+    {
+      "epoch": 3.0666530778638403,
+      "grad_norm": 7.5,
+      "learning_rate": 3.0833418263351e-05,
+      "loss": 0.8968,
+      "step": 45135
+    },
+    {
+      "epoch": 3.066992797934502,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.083129501290936e-05,
+      "loss": 0.8876,
+      "step": 45140
+    },
+    {
+      "epoch": 3.0673325180051636,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.0829171762467726e-05,
+      "loss": 0.824,
+      "step": 45145
+    },
+    {
+      "epoch": 3.0676722380758257,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.082704851202609e-05,
+      "loss": 0.8978,
+      "step": 45150
+    },
+    {
+      "epoch": 3.0680119581464873,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.082492526158446e-05,
+      "loss": 0.8624,
+      "step": 45155
+    },
+    {
+      "epoch": 3.068351678217149,
+      "grad_norm": 6.875,
+      "learning_rate": 3.082280201114282e-05,
+      "loss": 0.8854,
+      "step": 45160
+    },
+    {
+      "epoch": 3.068691398287811,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.082067876070118e-05,
+      "loss": 0.9434,
+      "step": 45165
+    },
+    {
+      "epoch": 3.0690311183584726,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.081855551025955e-05,
+      "loss": 0.8087,
+      "step": 45170
+    },
+    {
+      "epoch": 3.0693708384291343,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.081643225981791e-05,
+      "loss": 0.8408,
+      "step": 45175
+    },
+    {
+      "epoch": 3.0697105584997963,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0814309009376274e-05,
+      "loss": 0.9055,
+      "step": 45180
+    },
+    {
+      "epoch": 3.070050278570458,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.0812185758934645e-05,
+      "loss": 0.9889,
+      "step": 45185
+    },
+    {
+      "epoch": 3.0703899986411196,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0810062508493e-05,
+      "loss": 0.9897,
+      "step": 45190
+    },
+    {
+      "epoch": 3.0707297187117817,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.0807939258051366e-05,
+      "loss": 0.9048,
+      "step": 45195
+    },
+    {
+      "epoch": 3.0710694387824433,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.080581600760974e-05,
+      "loss": 0.8791,
+      "step": 45200
+    },
+    {
+      "epoch": 3.071409158853105,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0803692757168094e-05,
+      "loss": 0.8251,
+      "step": 45205
+    },
+    {
+      "epoch": 3.071748878923767,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.080156950672646e-05,
+      "loss": 0.8154,
+      "step": 45210
+    },
+    {
+      "epoch": 3.0720885989944287,
+      "grad_norm": 7.5,
+      "learning_rate": 3.079944625628482e-05,
+      "loss": 0.9354,
+      "step": 45215
+    },
+    {
+      "epoch": 3.0724283190650903,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0797323005843186e-05,
+      "loss": 0.9397,
+      "step": 45220
+    },
+    {
+      "epoch": 3.0727680391357524,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.079519975540155e-05,
+      "loss": 0.9395,
+      "step": 45225
+    },
+    {
+      "epoch": 3.073107759206414,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.0793076504959914e-05,
+      "loss": 0.8958,
+      "step": 45230
+    },
+    {
+      "epoch": 3.0734474792770756,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.079095325451828e-05,
+      "loss": 0.7935,
+      "step": 45235
+    },
+    {
+      "epoch": 3.0737871993477373,
+      "grad_norm": 7.75,
+      "learning_rate": 3.078883000407664e-05,
+      "loss": 0.9324,
+      "step": 45240
+    },
+    {
+      "epoch": 3.0741269194183993,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0786706753635006e-05,
+      "loss": 0.8664,
+      "step": 45245
+    },
+    {
+      "epoch": 3.074466639489061,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.078458350319337e-05,
+      "loss": 0.961,
+      "step": 45250
+    },
+    {
+      "epoch": 3.0748063595597226,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.0782460252751734e-05,
+      "loss": 0.9001,
+      "step": 45255
+    },
+    {
+      "epoch": 3.0751460796303847,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.07803370023101e-05,
+      "loss": 0.8482,
+      "step": 45260
+    },
+    {
+      "epoch": 3.0754857997010463,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.077821375186846e-05,
+      "loss": 0.9133,
+      "step": 45265
+    },
+    {
+      "epoch": 3.075825519771708,
+      "grad_norm": 6.5,
+      "learning_rate": 3.0776090501426826e-05,
+      "loss": 0.7331,
+      "step": 45270
+    },
+    {
+      "epoch": 3.07616523984237,
+      "grad_norm": 7.375,
+      "learning_rate": 3.077396725098519e-05,
+      "loss": 0.8294,
+      "step": 45275
+    },
+    {
+      "epoch": 3.0765049599130316,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0771844000543554e-05,
+      "loss": 0.7546,
+      "step": 45280
+    },
+    {
+      "epoch": 3.0768446799836933,
+      "grad_norm": 6.625,
+      "learning_rate": 3.076972075010192e-05,
+      "loss": 0.8935,
+      "step": 45285
+    },
+    {
+      "epoch": 3.0771844000543553,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.076759749966028e-05,
+      "loss": 0.859,
+      "step": 45290
+    },
+    {
+      "epoch": 3.077524120125017,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.0765474249218646e-05,
+      "loss": 0.9379,
+      "step": 45295
+    },
+    {
+      "epoch": 3.0778638401956786,
+      "grad_norm": 6.75,
+      "learning_rate": 3.076335099877701e-05,
+      "loss": 0.9398,
+      "step": 45300
+    },
+    {
+      "epoch": 3.0782035602663407,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0761227748335374e-05,
+      "loss": 0.9002,
+      "step": 45305
+    },
+    {
+      "epoch": 3.0785432803370023,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.075910449789374e-05,
+      "loss": 0.8425,
+      "step": 45310
+    },
+    {
+      "epoch": 3.078883000407664,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.07569812474521e-05,
+      "loss": 0.8519,
+      "step": 45315
+    },
+    {
+      "epoch": 3.079222720478326,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0754857997010466e-05,
+      "loss": 0.8816,
+      "step": 45320
+    },
+    {
+      "epoch": 3.0795624405489876,
+      "grad_norm": 7.25,
+      "learning_rate": 3.075273474656883e-05,
+      "loss": 0.9333,
+      "step": 45325
+    },
+    {
+      "epoch": 3.0799021606196493,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.0750611496127194e-05,
+      "loss": 0.9063,
+      "step": 45330
+    },
+    {
+      "epoch": 3.0802418806903114,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.074848824568556e-05,
+      "loss": 0.9721,
+      "step": 45335
+    },
+    {
+      "epoch": 3.080581600760973,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.074636499524392e-05,
+      "loss": 0.8297,
+      "step": 45340
+    },
+    {
+      "epoch": 3.0809213208316346,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.074424174480228e-05,
+      "loss": 0.9562,
+      "step": 45345
+    },
+    {
+      "epoch": 3.0812610409022967,
+      "grad_norm": 9.1875,
+      "learning_rate": 3.074211849436065e-05,
+      "loss": 0.8295,
+      "step": 45350
+    },
+    {
+      "epoch": 3.0816007609729583,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0739995243919014e-05,
+      "loss": 0.855,
+      "step": 45355
+    },
+    {
+      "epoch": 3.08194048104362,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.073787199347737e-05,
+      "loss": 0.9309,
+      "step": 45360
+    },
+    {
+      "epoch": 3.082280201114282,
+      "grad_norm": 7.0,
+      "learning_rate": 3.073574874303574e-05,
+      "loss": 0.9224,
+      "step": 45365
+    },
+    {
+      "epoch": 3.0826199211849437,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.0733625492594106e-05,
+      "loss": 0.9283,
+      "step": 45370
+    },
+    {
+      "epoch": 3.0829596412556053,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.073150224215246e-05,
+      "loss": 0.8436,
+      "step": 45375
+    },
+    {
+      "epoch": 3.0832993613262674,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0729378991710834e-05,
+      "loss": 0.9117,
+      "step": 45380
+    },
+    {
+      "epoch": 3.083639081396929,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.07272557412692e-05,
+      "loss": 0.9902,
+      "step": 45385
+    },
+    {
+      "epoch": 3.0839788014675906,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.0725132490827555e-05,
+      "loss": 0.9848,
+      "step": 45390
+    },
+    {
+      "epoch": 3.0843185215382523,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.0723009240385926e-05,
+      "loss": 0.8296,
+      "step": 45395
+    },
+    {
+      "epoch": 3.0846582416089143,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.072088598994429e-05,
+      "loss": 0.8446,
+      "step": 45400
+    },
+    {
+      "epoch": 3.084997961679576,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.071876273950265e-05,
+      "loss": 0.8802,
+      "step": 45405
+    },
+    {
+      "epoch": 3.0853376817502376,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.071663948906102e-05,
+      "loss": 0.8518,
+      "step": 45410
+    },
+    {
+      "epoch": 3.0856774018208997,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.0714516238619375e-05,
+      "loss": 0.8539,
+      "step": 45415
+    },
+    {
+      "epoch": 3.0860171218915613,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.071239298817774e-05,
+      "loss": 0.9234,
+      "step": 45420
+    },
+    {
+      "epoch": 3.086356841962223,
+      "grad_norm": 7.125,
+      "learning_rate": 3.071026973773611e-05,
+      "loss": 0.8948,
+      "step": 45425
+    },
+    {
+      "epoch": 3.086696562032885,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.070814648729447e-05,
+      "loss": 0.8453,
+      "step": 45430
+    },
+    {
+      "epoch": 3.0870362821035466,
+      "grad_norm": 7.125,
+      "learning_rate": 3.070602323685283e-05,
+      "loss": 0.9213,
+      "step": 45435
+    },
+    {
+      "epoch": 3.0873760021742083,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.07038999864112e-05,
+      "loss": 0.9166,
+      "step": 45440
+    },
+    {
+      "epoch": 3.0877157222448703,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.070177673596956e-05,
+      "loss": 0.8824,
+      "step": 45445
+    },
+    {
+      "epoch": 3.088055442315532,
+      "grad_norm": 7.125,
+      "learning_rate": 3.069965348552792e-05,
+      "loss": 0.9169,
+      "step": 45450
+    },
+    {
+      "epoch": 3.0883951623861936,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.0697530235086294e-05,
+      "loss": 0.9653,
+      "step": 45455
+    },
+    {
+      "epoch": 3.0887348824568557,
+      "grad_norm": 7.0,
+      "learning_rate": 3.069540698464465e-05,
+      "loss": 0.9549,
+      "step": 45460
+    },
+    {
+      "epoch": 3.0890746025275173,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0693283734203015e-05,
+      "loss": 0.8868,
+      "step": 45465
+    },
+    {
+      "epoch": 3.089414322598179,
+      "grad_norm": 6.875,
+      "learning_rate": 3.0691160483761386e-05,
+      "loss": 0.7644,
+      "step": 45470
+    },
+    {
+      "epoch": 3.089754042668841,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.068903723331974e-05,
+      "loss": 0.8993,
+      "step": 45475
+    },
+    {
+      "epoch": 3.0900937627395026,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.068691398287811e-05,
+      "loss": 0.8201,
+      "step": 45480
+    },
+    {
+      "epoch": 3.0904334828101643,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.068479073243647e-05,
+      "loss": 0.8465,
+      "step": 45485
+    },
+    {
+      "epoch": 3.0907732028808264,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0682667481994835e-05,
+      "loss": 0.9418,
+      "step": 45490
+    },
+    {
+      "epoch": 3.091112922951488,
+      "grad_norm": 8.0,
+      "learning_rate": 3.0680544231553206e-05,
+      "loss": 0.8864,
+      "step": 45495
+    },
+    {
+      "epoch": 3.0914526430221496,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.067842098111156e-05,
+      "loss": 0.7679,
+      "step": 45500
+    },
+    {
+      "epoch": 3.0917923630928117,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.067629773066993e-05,
+      "loss": 0.8865,
+      "step": 45505
+    },
+    {
+      "epoch": 3.0921320831634733,
+      "grad_norm": 7.625,
+      "learning_rate": 3.06741744802283e-05,
+      "loss": 0.8868,
+      "step": 45510
+    },
+    {
+      "epoch": 3.092471803234135,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.0672051229786655e-05,
+      "loss": 0.91,
+      "step": 45515
+    },
+    {
+      "epoch": 3.092811523304797,
+      "grad_norm": 9.4375,
+      "learning_rate": 3.066992797934502e-05,
+      "loss": 0.8824,
+      "step": 45520
+    },
+    {
+      "epoch": 3.0931512433754587,
+      "grad_norm": 7.25,
+      "learning_rate": 3.066780472890339e-05,
+      "loss": 0.9448,
+      "step": 45525
+    },
+    {
+      "epoch": 3.0934909634461203,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.066568147846175e-05,
+      "loss": 0.845,
+      "step": 45530
+    },
+    {
+      "epoch": 3.0938306835167824,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.066355822802011e-05,
+      "loss": 0.8203,
+      "step": 45535
+    },
+    {
+      "epoch": 3.094170403587444,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.066143497757848e-05,
+      "loss": 0.9021,
+      "step": 45540
+    },
+    {
+      "epoch": 3.0945101236581056,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.065931172713684e-05,
+      "loss": 0.9561,
+      "step": 45545
+    },
+    {
+      "epoch": 3.0948498437287677,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0657188476695203e-05,
+      "loss": 0.9255,
+      "step": 45550
+    },
+    {
+      "epoch": 3.0951895637994293,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.0655065226253574e-05,
+      "loss": 0.8835,
+      "step": 45555
+    },
+    {
+      "epoch": 3.095529283870091,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.065294197581193e-05,
+      "loss": 0.8277,
+      "step": 45560
+    },
+    {
+      "epoch": 3.095869003940753,
+      "grad_norm": 6.625,
+      "learning_rate": 3.0650818725370295e-05,
+      "loss": 0.8936,
+      "step": 45565
+    },
+    {
+      "epoch": 3.0962087240114147,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.064869547492866e-05,
+      "loss": 0.9961,
+      "step": 45570
+    },
+    {
+      "epoch": 3.0965484440820763,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.0646572224487023e-05,
+      "loss": 0.9157,
+      "step": 45575
+    },
+    {
+      "epoch": 3.096888164152738,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.064444897404539e-05,
+      "loss": 0.8279,
+      "step": 45580
+    },
+    {
+      "epoch": 3.0972278842234,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.064232572360375e-05,
+      "loss": 1.0429,
+      "step": 45585
+    },
+    {
+      "epoch": 3.0975676042940616,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.0640202473162115e-05,
+      "loss": 1.0309,
+      "step": 45590
+    },
+    {
+      "epoch": 3.0979073243647233,
+      "grad_norm": 8.25,
+      "learning_rate": 3.063807922272048e-05,
+      "loss": 0.8857,
+      "step": 45595
+    },
+    {
+      "epoch": 3.0982470444353853,
+      "grad_norm": 6.125,
+      "learning_rate": 3.0635955972278843e-05,
+      "loss": 0.9122,
+      "step": 45600
+    },
+    {
+      "epoch": 3.098586764506047,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.063383272183721e-05,
+      "loss": 0.874,
+      "step": 45605
+    },
+    {
+      "epoch": 3.0989264845767086,
+      "grad_norm": 8.75,
+      "learning_rate": 3.063170947139557e-05,
+      "loss": 0.9993,
+      "step": 45610
+    },
+    {
+      "epoch": 3.0992662046473707,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0629586220953935e-05,
+      "loss": 0.876,
+      "step": 45615
+    },
+    {
+      "epoch": 3.0996059247180323,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.06274629705123e-05,
+      "loss": 0.9355,
+      "step": 45620
+    },
+    {
+      "epoch": 3.099945644788694,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.0625339720070663e-05,
+      "loss": 0.8383,
+      "step": 45625
+    },
+    {
+      "epoch": 3.100285364859356,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.062321646962903e-05,
+      "loss": 0.9765,
+      "step": 45630
+    },
+    {
+      "epoch": 3.1006250849300176,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.062109321918739e-05,
+      "loss": 0.9045,
+      "step": 45635
+    },
+    {
+      "epoch": 3.1009648050006793,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0618969968745755e-05,
+      "loss": 0.8366,
+      "step": 45640
+    },
+    {
+      "epoch": 3.1013045250713414,
+      "grad_norm": 7.25,
+      "learning_rate": 3.061684671830412e-05,
+      "loss": 0.8759,
+      "step": 45645
+    },
+    {
+      "epoch": 3.101644245142003,
+      "grad_norm": 6.75,
+      "learning_rate": 3.0614723467862484e-05,
+      "loss": 0.952,
+      "step": 45650
+    },
+    {
+      "epoch": 3.1019839652126646,
+      "grad_norm": 6.75,
+      "learning_rate": 3.061260021742085e-05,
+      "loss": 0.8208,
+      "step": 45655
+    },
+    {
+      "epoch": 3.1023236852833267,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.061047696697921e-05,
+      "loss": 0.9122,
+      "step": 45660
+    },
+    {
+      "epoch": 3.1026634053539883,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.0608353716537576e-05,
+      "loss": 0.9291,
+      "step": 45665
+    },
+    {
+      "epoch": 3.10300312542465,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.060623046609594e-05,
+      "loss": 0.8199,
+      "step": 45670
+    },
+    {
+      "epoch": 3.103342845495312,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.0604107215654304e-05,
+      "loss": 0.8623,
+      "step": 45675
+    },
+    {
+      "epoch": 3.1036825655659737,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.060198396521267e-05,
+      "loss": 0.9239,
+      "step": 45680
+    },
+    {
+      "epoch": 3.1040222856366353,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0599860714771025e-05,
+      "loss": 0.9267,
+      "step": 45685
+    },
+    {
+      "epoch": 3.1043620057072974,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.0597737464329396e-05,
+      "loss": 0.9001,
+      "step": 45690
+    },
+    {
+      "epoch": 3.104701725777959,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.059561421388776e-05,
+      "loss": 0.8214,
+      "step": 45695
+    },
+    {
+      "epoch": 3.1050414458486206,
+      "grad_norm": 9.0,
+      "learning_rate": 3.059349096344612e-05,
+      "loss": 0.8751,
+      "step": 45700
+    },
+    {
+      "epoch": 3.1053811659192827,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.059136771300449e-05,
+      "loss": 0.8219,
+      "step": 45705
+    },
+    {
+      "epoch": 3.1057208859899443,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.058924446256285e-05,
+      "loss": 0.9242,
+      "step": 45710
+    },
+    {
+      "epoch": 3.106060606060606,
+      "grad_norm": 7.75,
+      "learning_rate": 3.058712121212121e-05,
+      "loss": 0.8373,
+      "step": 45715
+    },
+    {
+      "epoch": 3.106400326131268,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.058499796167958e-05,
+      "loss": 0.8318,
+      "step": 45720
+    },
+    {
+      "epoch": 3.1067400462019297,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.0582874711237944e-05,
+      "loss": 0.9162,
+      "step": 45725
+    },
+    {
+      "epoch": 3.1070797662725913,
+      "grad_norm": 7.125,
+      "learning_rate": 3.05807514607963e-05,
+      "loss": 0.9609,
+      "step": 45730
+    },
+    {
+      "epoch": 3.107419486343253,
+      "grad_norm": 6.75,
+      "learning_rate": 3.057862821035467e-05,
+      "loss": 0.8967,
+      "step": 45735
+    },
+    {
+      "epoch": 3.107759206413915,
+      "grad_norm": 8.0,
+      "learning_rate": 3.0576504959913036e-05,
+      "loss": 0.8609,
+      "step": 45740
+    },
+    {
+      "epoch": 3.1080989264845766,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.057438170947139e-05,
+      "loss": 0.8755,
+      "step": 45745
+    },
+    {
+      "epoch": 3.1084386465552383,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.0572258459029764e-05,
+      "loss": 0.9085,
+      "step": 45750
+    },
+    {
+      "epoch": 3.1087783666259003,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.057013520858813e-05,
+      "loss": 0.8952,
+      "step": 45755
+    },
+    {
+      "epoch": 3.109118086696562,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.0568011958146485e-05,
+      "loss": 0.9458,
+      "step": 45760
+    },
+    {
+      "epoch": 3.1094578067672236,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.0565888707704856e-05,
+      "loss": 0.8552,
+      "step": 45765
+    },
+    {
+      "epoch": 3.1097975268378857,
+      "grad_norm": 6.75,
+      "learning_rate": 3.056376545726321e-05,
+      "loss": 0.8412,
+      "step": 45770
+    },
+    {
+      "epoch": 3.1101372469085473,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.056164220682158e-05,
+      "loss": 0.939,
+      "step": 45775
+    },
+    {
+      "epoch": 3.110476966979209,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.055951895637995e-05,
+      "loss": 0.9571,
+      "step": 45780
+    },
+    {
+      "epoch": 3.110816687049871,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.0557395705938305e-05,
+      "loss": 0.9239,
+      "step": 45785
+    },
+    {
+      "epoch": 3.1111564071205327,
+      "grad_norm": 7.25,
+      "learning_rate": 3.055527245549667e-05,
+      "loss": 0.9331,
+      "step": 45790
+    },
+    {
+      "epoch": 3.1114961271911943,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.055314920505504e-05,
+      "loss": 0.8202,
+      "step": 45795
+    },
+    {
+      "epoch": 3.1118358472618564,
+      "grad_norm": 6.875,
+      "learning_rate": 3.05510259546134e-05,
+      "loss": 0.8978,
+      "step": 45800
+    },
+    {
+      "epoch": 3.112175567332518,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.054890270417176e-05,
+      "loss": 0.9119,
+      "step": 45805
+    },
+    {
+      "epoch": 3.1125152874031796,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.054677945373013e-05,
+      "loss": 0.775,
+      "step": 45810
+    },
+    {
+      "epoch": 3.1128550074738417,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.054465620328849e-05,
+      "loss": 0.8642,
+      "step": 45815
+    },
+    {
+      "epoch": 3.1131947275445033,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.054253295284685e-05,
+      "loss": 0.9031,
+      "step": 45820
+    },
+    {
+      "epoch": 3.113534447615165,
+      "grad_norm": 6.15625,
+      "learning_rate": 3.0540409702405224e-05,
+      "loss": 0.948,
+      "step": 45825
+    },
+    {
+      "epoch": 3.113874167685827,
+      "grad_norm": 7.0,
+      "learning_rate": 3.053828645196358e-05,
+      "loss": 0.9329,
+      "step": 45830
+    },
+    {
+      "epoch": 3.1142138877564887,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.053616320152195e-05,
+      "loss": 0.8676,
+      "step": 45835
+    },
+    {
+      "epoch": 3.1145536078271503,
+      "grad_norm": 8.125,
+      "learning_rate": 3.053403995108031e-05,
+      "loss": 1.0187,
+      "step": 45840
+    },
+    {
+      "epoch": 3.1148933278978124,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.053191670063867e-05,
+      "loss": 0.8404,
+      "step": 45845
+    },
+    {
+      "epoch": 3.115233047968474,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0529793450197044e-05,
+      "loss": 0.9463,
+      "step": 45850
+    },
+    {
+      "epoch": 3.1155727680391356,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.05276701997554e-05,
+      "loss": 0.8738,
+      "step": 45855
+    },
+    {
+      "epoch": 3.1159124881097977,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.0525546949313765e-05,
+      "loss": 0.8872,
+      "step": 45860
+    },
+    {
+      "epoch": 3.1162522081804593,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.0523423698872136e-05,
+      "loss": 0.7688,
+      "step": 45865
+    },
+    {
+      "epoch": 3.116591928251121,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.052130044843049e-05,
+      "loss": 0.8876,
+      "step": 45870
+    },
+    {
+      "epoch": 3.116931648321783,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.051917719798886e-05,
+      "loss": 0.8849,
+      "step": 45875
+    },
+    {
+      "epoch": 3.1172713683924447,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0517053947547224e-05,
+      "loss": 0.9671,
+      "step": 45880
+    },
+    {
+      "epoch": 3.1176110884631063,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0514930697105588e-05,
+      "loss": 0.8669,
+      "step": 45885
+    },
+    {
+      "epoch": 3.1179508085337684,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.051280744666395e-05,
+      "loss": 0.8979,
+      "step": 45890
+    },
+    {
+      "epoch": 3.11829052860443,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.0510684196222316e-05,
+      "loss": 0.8984,
+      "step": 45895
+    },
+    {
+      "epoch": 3.1186302486750916,
+      "grad_norm": 6.625,
+      "learning_rate": 3.050856094578068e-05,
+      "loss": 0.9275,
+      "step": 45900
+    },
+    {
+      "epoch": 3.1189699687457537,
+      "grad_norm": 7.75,
+      "learning_rate": 3.050643769533904e-05,
+      "loss": 0.9502,
+      "step": 45905
+    },
+    {
+      "epoch": 3.1193096888164153,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.050431444489741e-05,
+      "loss": 0.8754,
+      "step": 45910
+    },
+    {
+      "epoch": 3.119649408887077,
+      "grad_norm": 5.78125,
+      "learning_rate": 3.050219119445577e-05,
+      "loss": 0.8455,
+      "step": 45915
+    },
+    {
+      "epoch": 3.1199891289577386,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0500067944014133e-05,
+      "loss": 0.8811,
+      "step": 45920
+    },
+    {
+      "epoch": 3.1203288490284007,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.04979446935725e-05,
+      "loss": 0.8826,
+      "step": 45925
+    },
+    {
+      "epoch": 3.1206685690990623,
+      "grad_norm": 8.625,
+      "learning_rate": 3.049582144313086e-05,
+      "loss": 0.9217,
+      "step": 45930
+    },
+    {
+      "epoch": 3.121008289169724,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.0493698192689225e-05,
+      "loss": 0.939,
+      "step": 45935
+    },
+    {
+      "epoch": 3.121348009240386,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0491574942247592e-05,
+      "loss": 0.9473,
+      "step": 45940
+    },
+    {
+      "epoch": 3.1216877293110477,
+      "grad_norm": 6.875,
+      "learning_rate": 3.0489451691805953e-05,
+      "loss": 0.8786,
+      "step": 45945
+    },
+    {
+      "epoch": 3.1220274493817093,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.0487328441364317e-05,
+      "loss": 0.8769,
+      "step": 45950
+    },
+    {
+      "epoch": 3.1223671694523714,
+      "grad_norm": 5.9375,
+      "learning_rate": 3.0485205190922684e-05,
+      "loss": 0.8572,
+      "step": 45955
+    },
+    {
+      "epoch": 3.122706889523033,
+      "grad_norm": 6.75,
+      "learning_rate": 3.0483081940481045e-05,
+      "loss": 0.8438,
+      "step": 45960
+    },
+    {
+      "epoch": 3.1230466095936946,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.048095869003941e-05,
+      "loss": 0.895,
+      "step": 45965
+    },
+    {
+      "epoch": 3.1233863296643567,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0478835439597776e-05,
+      "loss": 0.9052,
+      "step": 45970
+    },
+    {
+      "epoch": 3.1237260497350183,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0476712189156137e-05,
+      "loss": 0.8279,
+      "step": 45975
+    },
+    {
+      "epoch": 3.12406576980568,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.0474588938714498e-05,
+      "loss": 0.9763,
+      "step": 45980
+    },
+    {
+      "epoch": 3.124405489876342,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.0472465688272865e-05,
+      "loss": 0.8904,
+      "step": 45985
+    },
+    {
+      "epoch": 3.1247452099470037,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.047034243783123e-05,
+      "loss": 0.8295,
+      "step": 45990
+    },
+    {
+      "epoch": 3.1250849300176653,
+      "grad_norm": 6.75,
+      "learning_rate": 3.046821918738959e-05,
+      "loss": 0.9202,
+      "step": 45995
+    },
+    {
+      "epoch": 3.1254246500883274,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.0466095936947957e-05,
+      "loss": 0.8238,
+      "step": 46000
+    },
+    {
+      "epoch": 3.125764370158989,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.046397268650632e-05,
+      "loss": 0.8551,
+      "step": 46005
+    },
+    {
+      "epoch": 3.1261040902296506,
+      "grad_norm": 6.75,
+      "learning_rate": 3.046184943606468e-05,
+      "loss": 0.9195,
+      "step": 46010
+    },
+    {
+      "epoch": 3.1264438103003127,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.045972618562305e-05,
+      "loss": 0.8527,
+      "step": 46015
+    },
+    {
+      "epoch": 3.1267835303709743,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.0457602935181413e-05,
+      "loss": 0.8882,
+      "step": 46020
+    },
+    {
+      "epoch": 3.127123250441636,
+      "grad_norm": 8.5,
+      "learning_rate": 3.0455479684739774e-05,
+      "loss": 0.8802,
+      "step": 46025
+    },
+    {
+      "epoch": 3.127462970512298,
+      "grad_norm": 6.5,
+      "learning_rate": 3.045335643429814e-05,
+      "loss": 0.8859,
+      "step": 46030
+    },
+    {
+      "epoch": 3.1278026905829597,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.0451233183856505e-05,
+      "loss": 0.8253,
+      "step": 46035
+    },
+    {
+      "epoch": 3.1281424106536213,
+      "grad_norm": 7.125,
+      "learning_rate": 3.0449109933414866e-05,
+      "loss": 0.8828,
+      "step": 46040
+    },
+    {
+      "epoch": 3.1284821307242834,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.0446986682973233e-05,
+      "loss": 0.9006,
+      "step": 46045
+    },
+    {
+      "epoch": 3.128821850794945,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0444863432531594e-05,
+      "loss": 0.8816,
+      "step": 46050
+    },
+    {
+      "epoch": 3.1291615708656066,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.0442740182089958e-05,
+      "loss": 0.821,
+      "step": 46055
+    },
+    {
+      "epoch": 3.1295012909362687,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.0440616931648325e-05,
+      "loss": 0.9299,
+      "step": 46060
+    },
+    {
+      "epoch": 3.1298410110069304,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.0438493681206686e-05,
+      "loss": 0.8356,
+      "step": 46065
+    },
+    {
+      "epoch": 3.130180731077592,
+      "grad_norm": 7.125,
+      "learning_rate": 3.043637043076505e-05,
+      "loss": 0.8146,
+      "step": 46070
+    },
+    {
+      "epoch": 3.1305204511482536,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.0434247180323417e-05,
+      "loss": 0.8844,
+      "step": 46075
+    },
+    {
+      "epoch": 3.1308601712189157,
+      "grad_norm": 7.5,
+      "learning_rate": 3.0432123929881778e-05,
+      "loss": 0.8973,
+      "step": 46080
+    },
+    {
+      "epoch": 3.1311998912895773,
+      "grad_norm": 6.75,
+      "learning_rate": 3.043000067944014e-05,
+      "loss": 0.8961,
+      "step": 46085
+    },
+    {
+      "epoch": 3.131539611360239,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.042787742899851e-05,
+      "loss": 0.9396,
+      "step": 46090
+    },
+    {
+      "epoch": 3.131879331430901,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.042575417855687e-05,
+      "loss": 0.9129,
+      "step": 46095
+    },
+    {
+      "epoch": 3.1322190515015627,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.0423630928115234e-05,
+      "loss": 0.9529,
+      "step": 46100
+    },
+    {
+      "epoch": 3.1325587715722243,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.04215076776736e-05,
+      "loss": 1.0235,
+      "step": 46105
+    },
+    {
+      "epoch": 3.1328984916428864,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.041938442723196e-05,
+      "loss": 0.8669,
+      "step": 46110
+    },
+    {
+      "epoch": 3.133238211713548,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.0417261176790322e-05,
+      "loss": 0.8564,
+      "step": 46115
+    },
+    {
+      "epoch": 3.1335779317842096,
+      "grad_norm": 7.25,
+      "learning_rate": 3.041513792634869e-05,
+      "loss": 0.8834,
+      "step": 46120
+    },
+    {
+      "epoch": 3.1339176518548717,
+      "grad_norm": 7.125,
+      "learning_rate": 3.0413014675907054e-05,
+      "loss": 0.9777,
+      "step": 46125
+    },
+    {
+      "epoch": 3.1342573719255333,
+      "grad_norm": 6.75,
+      "learning_rate": 3.0410891425465414e-05,
+      "loss": 0.8812,
+      "step": 46130
+    },
+    {
+      "epoch": 3.134597091996195,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0408768175023782e-05,
+      "loss": 0.9511,
+      "step": 46135
+    },
+    {
+      "epoch": 3.134936812066857,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0406644924582146e-05,
+      "loss": 0.9136,
+      "step": 46140
+    },
+    {
+      "epoch": 3.1352765321375187,
+      "grad_norm": 7.875,
+      "learning_rate": 3.0404521674140506e-05,
+      "loss": 0.9342,
+      "step": 46145
+    },
+    {
+      "epoch": 3.1356162522081803,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.0402398423698874e-05,
+      "loss": 0.8853,
+      "step": 46150
+    },
+    {
+      "epoch": 3.1359559722788424,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0400275173257238e-05,
+      "loss": 0.9592,
+      "step": 46155
+    },
+    {
+      "epoch": 3.136295692349504,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.03981519228156e-05,
+      "loss": 0.832,
+      "step": 46160
+    },
+    {
+      "epoch": 3.1366354124201656,
+      "grad_norm": 7.75,
+      "learning_rate": 3.0396028672373966e-05,
+      "loss": 0.8721,
+      "step": 46165
+    },
+    {
+      "epoch": 3.1369751324908277,
+      "grad_norm": 7.625,
+      "learning_rate": 3.039390542193233e-05,
+      "loss": 0.9995,
+      "step": 46170
+    },
+    {
+      "epoch": 3.1373148525614893,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.0391782171490697e-05,
+      "loss": 0.8976,
+      "step": 46175
+    },
+    {
+      "epoch": 3.137654572632151,
+      "grad_norm": 8.125,
+      "learning_rate": 3.0389658921049058e-05,
+      "loss": 0.8786,
+      "step": 46180
+    },
+    {
+      "epoch": 3.137994292702813,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.038753567060742e-05,
+      "loss": 0.9324,
+      "step": 46185
+    },
+    {
+      "epoch": 3.1383340127734747,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.038541242016579e-05,
+      "loss": 0.9272,
+      "step": 46190
+    },
+    {
+      "epoch": 3.1386737328441363,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.038328916972415e-05,
+      "loss": 0.8863,
+      "step": 46195
+    },
+    {
+      "epoch": 3.1390134529147984,
+      "grad_norm": 6.875,
+      "learning_rate": 3.038116591928251e-05,
+      "loss": 0.8737,
+      "step": 46200
+    },
+    {
+      "epoch": 3.13935317298546,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0379042668840878e-05,
+      "loss": 0.8122,
+      "step": 46205
+    },
+    {
+      "epoch": 3.1396928930561216,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.0376919418399242e-05,
+      "loss": 0.9261,
+      "step": 46210
+    },
+    {
+      "epoch": 3.1400326131267837,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0374796167957602e-05,
+      "loss": 0.9365,
+      "step": 46215
+    },
+    {
+      "epoch": 3.1403723331974454,
+      "grad_norm": 7.0,
+      "learning_rate": 3.037267291751597e-05,
+      "loss": 0.8432,
+      "step": 46220
+    },
+    {
+      "epoch": 3.140712053268107,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.0370549667074334e-05,
+      "loss": 0.8322,
+      "step": 46225
+    },
+    {
+      "epoch": 3.141051773338769,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.0368426416632694e-05,
+      "loss": 0.8931,
+      "step": 46230
+    },
+    {
+      "epoch": 3.1413914934094307,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0366303166191062e-05,
+      "loss": 0.8954,
+      "step": 46235
+    },
+    {
+      "epoch": 3.1417312134800923,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.0364179915749426e-05,
+      "loss": 0.9358,
+      "step": 46240
+    },
+    {
+      "epoch": 3.1420709335507544,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0362056665307786e-05,
+      "loss": 0.8864,
+      "step": 46245
+    },
+    {
+      "epoch": 3.142410653621416,
+      "grad_norm": 6.75,
+      "learning_rate": 3.0359933414866154e-05,
+      "loss": 0.9699,
+      "step": 46250
+    },
+    {
+      "epoch": 3.1427503736920777,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.0357810164424514e-05,
+      "loss": 0.9497,
+      "step": 46255
+    },
+    {
+      "epoch": 3.1430900937627397,
+      "grad_norm": 7.25,
+      "learning_rate": 3.035568691398288e-05,
+      "loss": 0.8849,
+      "step": 46260
+    },
+    {
+      "epoch": 3.1434298138334014,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.0353563663541246e-05,
+      "loss": 0.8247,
+      "step": 46265
+    },
+    {
+      "epoch": 3.143769533904063,
+      "grad_norm": 7.5,
+      "learning_rate": 3.0351440413099606e-05,
+      "loss": 1.1597,
+      "step": 46270
+    },
+    {
+      "epoch": 3.1441092539747246,
+      "grad_norm": 6.4375,
+      "learning_rate": 3.034931716265797e-05,
+      "loss": 1.0337,
+      "step": 46275
+    },
+    {
+      "epoch": 3.1444489740453867,
+      "grad_norm": 6.625,
+      "learning_rate": 3.0347193912216338e-05,
+      "loss": 0.8488,
+      "step": 46280
+    },
+    {
+      "epoch": 3.1447886941160483,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.03450706617747e-05,
+      "loss": 0.8508,
+      "step": 46285
+    },
+    {
+      "epoch": 3.14512841418671,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.0342947411333062e-05,
+      "loss": 0.9575,
+      "step": 46290
+    },
+    {
+      "epoch": 3.145468134257372,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.034082416089143e-05,
+      "loss": 0.8311,
+      "step": 46295
+    },
+    {
+      "epoch": 3.1458078543280337,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.033870091044979e-05,
+      "loss": 1.028,
+      "step": 46300
+    },
+    {
+      "epoch": 3.1461475743986953,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0336577660008154e-05,
+      "loss": 0.8952,
+      "step": 46305
+    },
+    {
+      "epoch": 3.1464872944693574,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0334454409566522e-05,
+      "loss": 0.8343,
+      "step": 46310
+    },
+    {
+      "epoch": 3.146827014540019,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0332331159124882e-05,
+      "loss": 0.8794,
+      "step": 46315
+    },
+    {
+      "epoch": 3.1471667346106806,
+      "grad_norm": 5.84375,
+      "learning_rate": 3.0330207908683243e-05,
+      "loss": 0.8292,
+      "step": 46320
+    },
+    {
+      "epoch": 3.1475064546813427,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.0328084658241614e-05,
+      "loss": 0.9076,
+      "step": 46325
+    },
+    {
+      "epoch": 3.1478461747520043,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0325961407799974e-05,
+      "loss": 0.9588,
+      "step": 46330
+    },
+    {
+      "epoch": 3.148185894822666,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.0323838157358335e-05,
+      "loss": 0.9067,
+      "step": 46335
+    },
+    {
+      "epoch": 3.148525614893328,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.0321714906916702e-05,
+      "loss": 0.8919,
+      "step": 46340
+    },
+    {
+      "epoch": 3.1488653349639897,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.0319591656475066e-05,
+      "loss": 0.916,
+      "step": 46345
+    },
+    {
+      "epoch": 3.1492050550346513,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0317468406033427e-05,
+      "loss": 0.9097,
+      "step": 46350
+    },
+    {
+      "epoch": 3.1495447751053134,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.0315345155591794e-05,
+      "loss": 0.9454,
+      "step": 46355
+    },
+    {
+      "epoch": 3.149884495175975,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.031322190515016e-05,
+      "loss": 0.8732,
+      "step": 46360
+    },
+    {
+      "epoch": 3.1502242152466366,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.031109865470852e-05,
+      "loss": 0.8753,
+      "step": 46365
+    },
+    {
+      "epoch": 3.1505639353172987,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.0308975404266887e-05,
+      "loss": 0.9187,
+      "step": 46370
+    },
+    {
+      "epoch": 3.1509036553879604,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.030685215382525e-05,
+      "loss": 0.8341,
+      "step": 46375
+    },
+    {
+      "epoch": 3.151243375458622,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.030472890338361e-05,
+      "loss": 0.9146,
+      "step": 46380
+    },
+    {
+      "epoch": 3.151583095529284,
+      "grad_norm": 6.5,
+      "learning_rate": 3.030260565294198e-05,
+      "loss": 0.929,
+      "step": 46385
+    },
+    {
+      "epoch": 3.1519228155999457,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0300482402500343e-05,
+      "loss": 0.9281,
+      "step": 46390
+    },
+    {
+      "epoch": 3.1522625356706073,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0298359152058703e-05,
+      "loss": 0.9234,
+      "step": 46395
+    },
+    {
+      "epoch": 3.1526022557412694,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.029623590161707e-05,
+      "loss": 0.9289,
+      "step": 46400
+    },
+    {
+      "epoch": 3.152941975811931,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.029411265117543e-05,
+      "loss": 0.8363,
+      "step": 46405
+    },
+    {
+      "epoch": 3.1532816958825927,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.0291989400733795e-05,
+      "loss": 1.0266,
+      "step": 46410
+    },
+    {
+      "epoch": 3.1536214159532543,
+      "grad_norm": 7.125,
+      "learning_rate": 3.0289866150292163e-05,
+      "loss": 0.8174,
+      "step": 46415
+    },
+    {
+      "epoch": 3.1539611360239164,
+      "grad_norm": 7.75,
+      "learning_rate": 3.0287742899850523e-05,
+      "loss": 0.901,
+      "step": 46420
+    },
+    {
+      "epoch": 3.154300856094578,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0285619649408887e-05,
+      "loss": 0.9868,
+      "step": 46425
+    },
+    {
+      "epoch": 3.1546405761652396,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.0283496398967255e-05,
+      "loss": 0.8841,
+      "step": 46430
+    },
+    {
+      "epoch": 3.1549802962359017,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.0281373148525615e-05,
+      "loss": 0.9034,
+      "step": 46435
+    },
+    {
+      "epoch": 3.1553200163065633,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.027924989808398e-05,
+      "loss": 0.8618,
+      "step": 46440
+    },
+    {
+      "epoch": 3.155659736377225,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0277126647642347e-05,
+      "loss": 1.038,
+      "step": 46445
+    },
+    {
+      "epoch": 3.155999456447887,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0275003397200707e-05,
+      "loss": 0.939,
+      "step": 46450
+    },
+    {
+      "epoch": 3.1563391765185487,
+      "grad_norm": 6.875,
+      "learning_rate": 3.0272880146759068e-05,
+      "loss": 0.9601,
+      "step": 46455
+    },
+    {
+      "epoch": 3.1566788965892103,
+      "grad_norm": 6.09375,
+      "learning_rate": 3.027075689631744e-05,
+      "loss": 0.7885,
+      "step": 46460
+    },
+    {
+      "epoch": 3.1570186166598724,
+      "grad_norm": 6.875,
+      "learning_rate": 3.02686336458758e-05,
+      "loss": 0.9395,
+      "step": 46465
+    },
+    {
+      "epoch": 3.157358336730534,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.026651039543416e-05,
+      "loss": 0.9591,
+      "step": 46470
+    },
+    {
+      "epoch": 3.1576980568011956,
+      "grad_norm": 6.0625,
+      "learning_rate": 3.0264387144992527e-05,
+      "loss": 0.9081,
+      "step": 46475
+    },
+    {
+      "epoch": 3.1580377768718577,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.026226389455089e-05,
+      "loss": 0.9066,
+      "step": 46480
+    },
+    {
+      "epoch": 3.1583774969425193,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.0260140644109252e-05,
+      "loss": 0.8507,
+      "step": 46485
+    },
+    {
+      "epoch": 3.158717217013181,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.025801739366762e-05,
+      "loss": 1.0006,
+      "step": 46490
+    },
+    {
+      "epoch": 3.159056937083843,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.0255894143225983e-05,
+      "loss": 0.8818,
+      "step": 46495
+    },
+    {
+      "epoch": 3.1593966571545047,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.0253770892784344e-05,
+      "loss": 0.919,
+      "step": 46500
+    },
+    {
+      "epoch": 3.1597363772251663,
+      "grad_norm": 7.125,
+      "learning_rate": 3.025164764234271e-05,
+      "loss": 0.9778,
+      "step": 46505
+    },
+    {
+      "epoch": 3.1600760972958284,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0249524391901075e-05,
+      "loss": 0.8863,
+      "step": 46510
+    },
+    {
+      "epoch": 3.16041581736649,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0247401141459443e-05,
+      "loss": 0.9046,
+      "step": 46515
+    },
+    {
+      "epoch": 3.1607555374371517,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.0245277891017803e-05,
+      "loss": 0.9499,
+      "step": 46520
+    },
+    {
+      "epoch": 3.1610952575078137,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0243154640576167e-05,
+      "loss": 0.8881,
+      "step": 46525
+    },
+    {
+      "epoch": 3.1614349775784754,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0241031390134535e-05,
+      "loss": 0.8335,
+      "step": 46530
+    },
+    {
+      "epoch": 3.161774697649137,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.0238908139692895e-05,
+      "loss": 1.0174,
+      "step": 46535
+    },
+    {
+      "epoch": 3.162114417719799,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0236784889251256e-05,
+      "loss": 0.8501,
+      "step": 46540
+    },
+    {
+      "epoch": 3.1624541377904607,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.0234661638809623e-05,
+      "loss": 0.9056,
+      "step": 46545
+    },
+    {
+      "epoch": 3.1627938578611223,
+      "grad_norm": 8.5,
+      "learning_rate": 3.0232538388367987e-05,
+      "loss": 0.9051,
+      "step": 46550
+    },
+    {
+      "epoch": 3.1631335779317844,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.0230415137926348e-05,
+      "loss": 0.8348,
+      "step": 46555
+    },
+    {
+      "epoch": 3.163473298002446,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.0228291887484715e-05,
+      "loss": 0.9216,
+      "step": 46560
+    },
+    {
+      "epoch": 3.1638130180731077,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.022616863704308e-05,
+      "loss": 0.9958,
+      "step": 46565
+    },
+    {
+      "epoch": 3.1641527381437697,
+      "grad_norm": 6.625,
+      "learning_rate": 3.022404538660144e-05,
+      "loss": 0.8637,
+      "step": 46570
+    },
+    {
+      "epoch": 3.1644924582144314,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0221922136159807e-05,
+      "loss": 0.9738,
+      "step": 46575
+    },
+    {
+      "epoch": 3.164832178285093,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.021979888571817e-05,
+      "loss": 0.9266,
+      "step": 46580
+    },
+    {
+      "epoch": 3.165171898355755,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0217675635276532e-05,
+      "loss": 0.7882,
+      "step": 46585
+    },
+    {
+      "epoch": 3.1655116184264167,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.02155523848349e-05,
+      "loss": 0.898,
+      "step": 46590
+    },
+    {
+      "epoch": 3.1658513384970783,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0213429134393263e-05,
+      "loss": 0.9271,
+      "step": 46595
+    },
+    {
+      "epoch": 3.1661910585677404,
+      "grad_norm": 7.75,
+      "learning_rate": 3.0211305883951624e-05,
+      "loss": 0.8622,
+      "step": 46600
+    },
+    {
+      "epoch": 3.166530778638402,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.020918263350999e-05,
+      "loss": 0.8664,
+      "step": 46605
+    },
+    {
+      "epoch": 3.1668704987090637,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.0207059383068352e-05,
+      "loss": 0.9246,
+      "step": 46610
+    },
+    {
+      "epoch": 3.1672102187797253,
+      "grad_norm": 7.75,
+      "learning_rate": 3.0204936132626716e-05,
+      "loss": 1.0062,
+      "step": 46615
+    },
+    {
+      "epoch": 3.1675499388503874,
+      "grad_norm": 7.875,
+      "learning_rate": 3.0202812882185083e-05,
+      "loss": 0.9033,
+      "step": 46620
+    },
+    {
+      "epoch": 3.167889658921049,
+      "grad_norm": 6.875,
+      "learning_rate": 3.0200689631743444e-05,
+      "loss": 0.9551,
+      "step": 46625
+    },
+    {
+      "epoch": 3.1682293789917106,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.0198566381301808e-05,
+      "loss": 0.8344,
+      "step": 46630
+    },
+    {
+      "epoch": 3.1685690990623727,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.0196443130860175e-05,
+      "loss": 0.9111,
+      "step": 46635
+    },
+    {
+      "epoch": 3.1689088191330343,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.0194319880418536e-05,
+      "loss": 0.8974,
+      "step": 46640
+    },
+    {
+      "epoch": 3.169248539203696,
+      "grad_norm": 6.3125,
+      "learning_rate": 3.01921966299769e-05,
+      "loss": 0.9646,
+      "step": 46645
+    },
+    {
+      "epoch": 3.169588259274358,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.0190073379535267e-05,
+      "loss": 0.887,
+      "step": 46650
+    },
+    {
+      "epoch": 3.1699279793450197,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.0187950129093628e-05,
+      "loss": 0.8651,
+      "step": 46655
+    },
+    {
+      "epoch": 3.1702676994156813,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0185826878651992e-05,
+      "loss": 0.8478,
+      "step": 46660
+    },
+    {
+      "epoch": 3.1706074194863434,
+      "grad_norm": 6.8125,
+      "learning_rate": 3.018370362821036e-05,
+      "loss": 0.9768,
+      "step": 46665
+    },
+    {
+      "epoch": 3.170947139557005,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.018158037776872e-05,
+      "loss": 0.863,
+      "step": 46670
+    },
+    {
+      "epoch": 3.1712868596276667,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.017945712732708e-05,
+      "loss": 0.8922,
+      "step": 46675
+    },
+    {
+      "epoch": 3.1716265796983287,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.017733387688545e-05,
+      "loss": 0.8956,
+      "step": 46680
+    },
+    {
+      "epoch": 3.1719662997689904,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.0175210626443812e-05,
+      "loss": 0.8477,
+      "step": 46685
+    },
+    {
+      "epoch": 3.172306019839652,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0173087376002173e-05,
+      "loss": 0.9386,
+      "step": 46690
+    },
+    {
+      "epoch": 3.172645739910314,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.017096412556054e-05,
+      "loss": 1.0428,
+      "step": 46695
+    },
+    {
+      "epoch": 3.1729854599809757,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0168840875118904e-05,
+      "loss": 0.9021,
+      "step": 46700
+    },
+    {
+      "epoch": 3.1733251800516373,
+      "grad_norm": 8.375,
+      "learning_rate": 3.0166717624677265e-05,
+      "loss": 0.8939,
+      "step": 46705
+    },
+    {
+      "epoch": 3.1736649001222994,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0164594374235632e-05,
+      "loss": 0.9119,
+      "step": 46710
+    },
+    {
+      "epoch": 3.174004620192961,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.0162471123793996e-05,
+      "loss": 0.9741,
+      "step": 46715
+    },
+    {
+      "epoch": 3.1743443402636227,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0160347873352357e-05,
+      "loss": 0.9794,
+      "step": 46720
+    },
+    {
+      "epoch": 3.1746840603342847,
+      "grad_norm": 9.125,
+      "learning_rate": 3.0158224622910724e-05,
+      "loss": 0.899,
+      "step": 46725
+    },
+    {
+      "epoch": 3.1750237804049464,
+      "grad_norm": 8.0,
+      "learning_rate": 3.0156101372469088e-05,
+      "loss": 0.9588,
+      "step": 46730
+    },
+    {
+      "epoch": 3.175363500475608,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.015397812202745e-05,
+      "loss": 0.9398,
+      "step": 46735
+    },
+    {
+      "epoch": 3.17570322054627,
+      "grad_norm": 7.875,
+      "learning_rate": 3.0151854871585816e-05,
+      "loss": 0.9023,
+      "step": 46740
+    },
+    {
+      "epoch": 3.1760429406169317,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.0149731621144177e-05,
+      "loss": 0.8998,
+      "step": 46745
+    },
+    {
+      "epoch": 3.1763826606875933,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.014760837070254e-05,
+      "loss": 0.9697,
+      "step": 46750
+    },
+    {
+      "epoch": 3.176722380758255,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0145485120260908e-05,
+      "loss": 0.8844,
+      "step": 46755
+    },
+    {
+      "epoch": 3.177062100828917,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.014336186981927e-05,
+      "loss": 0.8957,
+      "step": 46760
+    },
+    {
+      "epoch": 3.1774018208995787,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0141238619377633e-05,
+      "loss": 0.9028,
+      "step": 46765
+    },
+    {
+      "epoch": 3.1777415409702403,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0139115368936e-05,
+      "loss": 0.9175,
+      "step": 46770
+    },
+    {
+      "epoch": 3.1780812610409024,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.013699211849436e-05,
+      "loss": 0.8995,
+      "step": 46775
+    },
+    {
+      "epoch": 3.178420981111564,
+      "grad_norm": 6.875,
+      "learning_rate": 3.0134868868052725e-05,
+      "loss": 0.8901,
+      "step": 46780
+    },
+    {
+      "epoch": 3.1787607011822256,
+      "grad_norm": 7.25,
+      "learning_rate": 3.0132745617611092e-05,
+      "loss": 0.837,
+      "step": 46785
+    },
+    {
+      "epoch": 3.1791004212528877,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0130622367169453e-05,
+      "loss": 0.8502,
+      "step": 46790
+    },
+    {
+      "epoch": 3.1794401413235494,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.0128499116727817e-05,
+      "loss": 0.8741,
+      "step": 46795
+    },
+    {
+      "epoch": 3.179779861394211,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0126375866286184e-05,
+      "loss": 0.8806,
+      "step": 46800
+    },
+    {
+      "epoch": 3.180119581464873,
+      "grad_norm": 6.125,
+      "learning_rate": 3.0124252615844545e-05,
+      "loss": 0.837,
+      "step": 46805
+    },
+    {
+      "epoch": 3.1804593015355347,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.0122129365402905e-05,
+      "loss": 0.9875,
+      "step": 46810
+    },
+    {
+      "epoch": 3.1807990216061963,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.0120006114961276e-05,
+      "loss": 0.8794,
+      "step": 46815
+    },
+    {
+      "epoch": 3.1811387416768584,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.0117882864519637e-05,
+      "loss": 0.9437,
+      "step": 46820
+    },
+    {
+      "epoch": 3.18147846174752,
+      "grad_norm": 7.375,
+      "learning_rate": 3.0115759614077997e-05,
+      "loss": 0.9152,
+      "step": 46825
+    },
+    {
+      "epoch": 3.1818181818181817,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.0113636363636365e-05,
+      "loss": 0.9002,
+      "step": 46830
+    },
+    {
+      "epoch": 3.1821579018888437,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.011151311319473e-05,
+      "loss": 0.8727,
+      "step": 46835
+    },
+    {
+      "epoch": 3.1824976219595054,
+      "grad_norm": 6.875,
+      "learning_rate": 3.010938986275309e-05,
+      "loss": 0.9652,
+      "step": 46840
+    },
+    {
+      "epoch": 3.182837342030167,
+      "grad_norm": 8.0,
+      "learning_rate": 3.0107266612311457e-05,
+      "loss": 0.9111,
+      "step": 46845
+    },
+    {
+      "epoch": 3.183177062100829,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.010514336186982e-05,
+      "loss": 0.8983,
+      "step": 46850
+    },
+    {
+      "epoch": 3.1835167821714907,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.0103020111428188e-05,
+      "loss": 0.9855,
+      "step": 46855
+    },
+    {
+      "epoch": 3.1838565022421523,
+      "grad_norm": 7.625,
+      "learning_rate": 3.010089686098655e-05,
+      "loss": 0.9307,
+      "step": 46860
+    },
+    {
+      "epoch": 3.1841962223128144,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.0098773610544913e-05,
+      "loss": 0.8707,
+      "step": 46865
+    },
+    {
+      "epoch": 3.184535942383476,
+      "grad_norm": 6.875,
+      "learning_rate": 3.009665036010328e-05,
+      "loss": 1.0244,
+      "step": 46870
+    },
+    {
+      "epoch": 3.1848756624541377,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.009452710966164e-05,
+      "loss": 0.9554,
+      "step": 46875
+    },
+    {
+      "epoch": 3.1852153825247997,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.0092403859220005e-05,
+      "loss": 0.9316,
+      "step": 46880
+    },
+    {
+      "epoch": 3.1855551025954614,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.0090280608778372e-05,
+      "loss": 0.9469,
+      "step": 46885
+    },
+    {
+      "epoch": 3.185894822666123,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0088157358336733e-05,
+      "loss": 0.968,
+      "step": 46890
+    },
+    {
+      "epoch": 3.186234542736785,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.0086034107895093e-05,
+      "loss": 0.8513,
+      "step": 46895
+    },
+    {
+      "epoch": 3.1865742628074467,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.008391085745346e-05,
+      "loss": 0.8829,
+      "step": 46900
+    },
+    {
+      "epoch": 3.1869139828781083,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.0081787607011825e-05,
+      "loss": 0.8441,
+      "step": 46905
+    },
+    {
+      "epoch": 3.1872537029487704,
+      "grad_norm": 7.125,
+      "learning_rate": 3.0079664356570185e-05,
+      "loss": 0.7917,
+      "step": 46910
+    },
+    {
+      "epoch": 3.187593423019432,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.0077541106128553e-05,
+      "loss": 0.9115,
+      "step": 46915
+    },
+    {
+      "epoch": 3.1879331430900937,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0075417855686917e-05,
+      "loss": 0.9343,
+      "step": 46920
+    },
+    {
+      "epoch": 3.1882728631607558,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.0073294605245277e-05,
+      "loss": 0.8927,
+      "step": 46925
+    },
+    {
+      "epoch": 3.1886125832314174,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.0071171354803645e-05,
+      "loss": 0.8458,
+      "step": 46930
+    },
+    {
+      "epoch": 3.188952303302079,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.006904810436201e-05,
+      "loss": 0.8842,
+      "step": 46935
+    },
+    {
+      "epoch": 3.189292023372741,
+      "grad_norm": 7.0625,
+      "learning_rate": 3.006692485392037e-05,
+      "loss": 0.8585,
+      "step": 46940
+    },
+    {
+      "epoch": 3.1896317434434027,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.0064801603478737e-05,
+      "loss": 0.8895,
+      "step": 46945
+    },
+    {
+      "epoch": 3.1899714635140644,
+      "grad_norm": 7.625,
+      "learning_rate": 3.00626783530371e-05,
+      "loss": 1.0132,
+      "step": 46950
+    },
+    {
+      "epoch": 3.190311183584726,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.006055510259546e-05,
+      "loss": 0.9673,
+      "step": 46955
+    },
+    {
+      "epoch": 3.190650903655388,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.005843185215383e-05,
+      "loss": 0.8969,
+      "step": 46960
+    },
+    {
+      "epoch": 3.1909906237260497,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.005630860171219e-05,
+      "loss": 0.8555,
+      "step": 46965
+    },
+    {
+      "epoch": 3.1913303437967113,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.0054185351270553e-05,
+      "loss": 0.8487,
+      "step": 46970
+    },
+    {
+      "epoch": 3.1916700638673734,
+      "grad_norm": 7.125,
+      "learning_rate": 3.005206210082892e-05,
+      "loss": 0.8535,
+      "step": 46975
+    },
+    {
+      "epoch": 3.192009783938035,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.004993885038728e-05,
+      "loss": 0.929,
+      "step": 46980
+    },
+    {
+      "epoch": 3.1923495040086967,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0047815599945645e-05,
+      "loss": 0.8105,
+      "step": 46985
+    },
+    {
+      "epoch": 3.1926892240793587,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.0045692349504013e-05,
+      "loss": 0.9235,
+      "step": 46990
+    },
+    {
+      "epoch": 3.1930289441500204,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0043569099062373e-05,
+      "loss": 0.864,
+      "step": 46995
+    },
+    {
+      "epoch": 3.193368664220682,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0041445848620737e-05,
+      "loss": 0.8829,
+      "step": 47000
+    },
+    {
+      "epoch": 3.193708384291344,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.0039322598179105e-05,
+      "loss": 0.9657,
+      "step": 47005
+    },
+    {
+      "epoch": 3.1940481043620057,
+      "grad_norm": 6.40625,
+      "learning_rate": 3.0037199347737465e-05,
+      "loss": 0.8273,
+      "step": 47010
+    },
+    {
+      "epoch": 3.1943878244326673,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.003507609729583e-05,
+      "loss": 0.8495,
+      "step": 47015
+    },
+    {
+      "epoch": 3.1947275445033294,
+      "grad_norm": 5.96875,
+      "learning_rate": 3.0032952846854197e-05,
+      "loss": 0.8979,
+      "step": 47020
+    },
+    {
+      "epoch": 3.195067264573991,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.0030829596412557e-05,
+      "loss": 0.9638,
+      "step": 47025
+    },
+    {
+      "epoch": 3.1954069846446527,
+      "grad_norm": 6.53125,
+      "learning_rate": 3.0028706345970918e-05,
+      "loss": 0.8526,
+      "step": 47030
+    },
+    {
+      "epoch": 3.1957467047153147,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.0026583095529285e-05,
+      "loss": 0.9188,
+      "step": 47035
+    },
+    {
+      "epoch": 3.1960864247859764,
+      "grad_norm": 6.1875,
+      "learning_rate": 3.002445984508765e-05,
+      "loss": 0.8753,
+      "step": 47040
+    },
+    {
+      "epoch": 3.196426144856638,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.002233659464601e-05,
+      "loss": 0.803,
+      "step": 47045
+    },
+    {
+      "epoch": 3.1967658649273,
+      "grad_norm": 6.6875,
+      "learning_rate": 3.0020213344204377e-05,
+      "loss": 0.905,
+      "step": 47050
+    },
+    {
+      "epoch": 3.1971055849979617,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.001809009376274e-05,
+      "loss": 0.8625,
+      "step": 47055
+    },
+    {
+      "epoch": 3.1974453050686233,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.0015966843321102e-05,
+      "loss": 0.8725,
+      "step": 47060
+    },
+    {
+      "epoch": 3.1977850251392854,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.001384359287947e-05,
+      "loss": 0.9124,
+      "step": 47065
+    },
+    {
+      "epoch": 3.198124745209947,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.0011720342437833e-05,
+      "loss": 0.8753,
+      "step": 47070
+    },
+    {
+      "epoch": 3.1984644652806087,
+      "grad_norm": 7.0,
+      "learning_rate": 3.0009597091996194e-05,
+      "loss": 0.9243,
+      "step": 47075
+    },
+    {
+      "epoch": 3.1988041853512708,
+      "grad_norm": 6.875,
+      "learning_rate": 3.000747384155456e-05,
+      "loss": 0.8752,
+      "step": 47080
+    },
+    {
+      "epoch": 3.1991439054219324,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0005350591112926e-05,
+      "loss": 0.9061,
+      "step": 47085
+    },
+    {
+      "epoch": 3.199483625492594,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.0003227340671286e-05,
+      "loss": 0.8985,
+      "step": 47090
+    },
+    {
+      "epoch": 3.1998233455632556,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.0001104090229654e-05,
+      "loss": 0.8727,
+      "step": 47095
+    },
+    {
+      "epoch": 3.2001630656339177,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.9998980839788014e-05,
+      "loss": 0.9254,
+      "step": 47100
+    },
+    {
+      "epoch": 3.2005027857045794,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9996857589346378e-05,
+      "loss": 0.9754,
+      "step": 47105
+    },
+    {
+      "epoch": 3.200842505775241,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9994734338904746e-05,
+      "loss": 0.8811,
+      "step": 47110
+    },
+    {
+      "epoch": 3.201182225845903,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9992611088463106e-05,
+      "loss": 0.8804,
+      "step": 47115
+    },
+    {
+      "epoch": 3.2015219459165647,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.999048783802147e-05,
+      "loss": 0.9153,
+      "step": 47120
+    },
+    {
+      "epoch": 3.2018616659872263,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9988364587579838e-05,
+      "loss": 0.853,
+      "step": 47125
+    },
+    {
+      "epoch": 3.2022013860578884,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9986241337138198e-05,
+      "loss": 0.9564,
+      "step": 47130
+    },
+    {
+      "epoch": 3.20254110612855,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9984118086696562e-05,
+      "loss": 0.8668,
+      "step": 47135
+    },
+    {
+      "epoch": 3.2028808261992117,
+      "grad_norm": 7.125,
+      "learning_rate": 2.998199483625493e-05,
+      "loss": 0.9227,
+      "step": 47140
+    },
+    {
+      "epoch": 3.2032205462698737,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.997987158581329e-05,
+      "loss": 0.9729,
+      "step": 47145
+    },
+    {
+      "epoch": 3.2035602663405354,
+      "grad_norm": 6.5,
+      "learning_rate": 2.9977748335371654e-05,
+      "loss": 0.8685,
+      "step": 47150
+    },
+    {
+      "epoch": 3.203899986411197,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.997562508493002e-05,
+      "loss": 1.0082,
+      "step": 47155
+    },
+    {
+      "epoch": 3.204239706481859,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.9973501834488382e-05,
+      "loss": 0.9097,
+      "step": 47160
+    },
+    {
+      "epoch": 3.2045794265525207,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9971378584046743e-05,
+      "loss": 0.9196,
+      "step": 47165
+    },
+    {
+      "epoch": 3.2049191466231823,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.996925533360511e-05,
+      "loss": 0.9404,
+      "step": 47170
+    },
+    {
+      "epoch": 3.2052588666938444,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.9967132083163474e-05,
+      "loss": 0.9601,
+      "step": 47175
+    },
+    {
+      "epoch": 3.205598586764506,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.9965008832721835e-05,
+      "loss": 0.9506,
+      "step": 47180
+    },
+    {
+      "epoch": 3.2059383068351677,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9962885582280202e-05,
+      "loss": 0.9616,
+      "step": 47185
+    },
+    {
+      "epoch": 3.2062780269058297,
+      "grad_norm": 7.25,
+      "learning_rate": 2.9960762331838566e-05,
+      "loss": 0.8292,
+      "step": 47190
+    },
+    {
+      "epoch": 3.2066177469764914,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9958639081396934e-05,
+      "loss": 0.9778,
+      "step": 47195
+    },
+    {
+      "epoch": 3.206957467047153,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9956515830955294e-05,
+      "loss": 0.8946,
+      "step": 47200
+    },
+    {
+      "epoch": 3.207297187117815,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9954392580513658e-05,
+      "loss": 0.8925,
+      "step": 47205
+    },
+    {
+      "epoch": 3.2076369071884767,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.9952269330072026e-05,
+      "loss": 0.9899,
+      "step": 47210
+    },
+    {
+      "epoch": 3.2079766272591383,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9950146079630386e-05,
+      "loss": 0.7948,
+      "step": 47215
+    },
+    {
+      "epoch": 3.2083163473298004,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.994802282918875e-05,
+      "loss": 0.8638,
+      "step": 47220
+    },
+    {
+      "epoch": 3.208656067400462,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.9945899578747118e-05,
+      "loss": 0.9827,
+      "step": 47225
+    },
+    {
+      "epoch": 3.2089957874711237,
+      "grad_norm": 6.875,
+      "learning_rate": 2.9943776328305478e-05,
+      "loss": 0.831,
+      "step": 47230
+    },
+    {
+      "epoch": 3.2093355075417858,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.994165307786384e-05,
+      "loss": 0.8979,
+      "step": 47235
+    },
+    {
+      "epoch": 3.2096752276124474,
+      "grad_norm": 6.5,
+      "learning_rate": 2.993952982742221e-05,
+      "loss": 0.8873,
+      "step": 47240
+    },
+    {
+      "epoch": 3.210014947683109,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.993740657698057e-05,
+      "loss": 0.9492,
+      "step": 47245
+    },
+    {
+      "epoch": 3.210354667753771,
+      "grad_norm": 6.375,
+      "learning_rate": 2.993528332653893e-05,
+      "loss": 0.9192,
+      "step": 47250
+    },
+    {
+      "epoch": 3.2106943878244327,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9933160076097298e-05,
+      "loss": 0.8284,
+      "step": 47255
+    },
+    {
+      "epoch": 3.2110341078950944,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9931036825655662e-05,
+      "loss": 0.8941,
+      "step": 47260
+    },
+    {
+      "epoch": 3.2113738279657564,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9928913575214023e-05,
+      "loss": 0.8382,
+      "step": 47265
+    },
+    {
+      "epoch": 3.211713548036418,
+      "grad_norm": 7.0,
+      "learning_rate": 2.992679032477239e-05,
+      "loss": 0.9322,
+      "step": 47270
+    },
+    {
+      "epoch": 3.2120532681070797,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9924667074330754e-05,
+      "loss": 0.8035,
+      "step": 47275
+    },
+    {
+      "epoch": 3.2123929881777418,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9922543823889115e-05,
+      "loss": 0.9477,
+      "step": 47280
+    },
+    {
+      "epoch": 3.2127327082484034,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.9920420573447482e-05,
+      "loss": 0.8576,
+      "step": 47285
+    },
+    {
+      "epoch": 3.213072428319065,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9918297323005846e-05,
+      "loss": 0.925,
+      "step": 47290
+    },
+    {
+      "epoch": 3.2134121483897267,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.9916174072564207e-05,
+      "loss": 0.8828,
+      "step": 47295
+    },
+    {
+      "epoch": 3.2137518684603887,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.9914050822122574e-05,
+      "loss": 0.875,
+      "step": 47300
+    },
+    {
+      "epoch": 3.2140915885310504,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.9911927571680938e-05,
+      "loss": 0.9166,
+      "step": 47305
+    },
+    {
+      "epoch": 3.214431308601712,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.99098043212393e-05,
+      "loss": 0.9325,
+      "step": 47310
+    },
+    {
+      "epoch": 3.214771028672374,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9907681070797666e-05,
+      "loss": 0.8832,
+      "step": 47315
+    },
+    {
+      "epoch": 3.2151107487430357,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9905557820356027e-05,
+      "loss": 1.1529,
+      "step": 47320
+    },
+    {
+      "epoch": 3.2154504688136973,
+      "grad_norm": 6.5,
+      "learning_rate": 2.990343456991439e-05,
+      "loss": 0.895,
+      "step": 47325
+    },
+    {
+      "epoch": 3.2157901888843594,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9901311319472758e-05,
+      "loss": 0.9141,
+      "step": 47330
+    },
+    {
+      "epoch": 3.216129908955021,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.989918806903112e-05,
+      "loss": 0.9824,
+      "step": 47335
+    },
+    {
+      "epoch": 3.2164696290256827,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.9897064818589483e-05,
+      "loss": 0.9145,
+      "step": 47340
+    },
+    {
+      "epoch": 3.2168093490963448,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.989494156814785e-05,
+      "loss": 0.8124,
+      "step": 47345
+    },
+    {
+      "epoch": 3.2171490691670064,
+      "grad_norm": 7.375,
+      "learning_rate": 2.989281831770621e-05,
+      "loss": 0.9669,
+      "step": 47350
+    },
+    {
+      "epoch": 3.217488789237668,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9890695067264575e-05,
+      "loss": 0.9457,
+      "step": 47355
+    },
+    {
+      "epoch": 3.21782850930833,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9888571816822942e-05,
+      "loss": 0.9078,
+      "step": 47360
+    },
+    {
+      "epoch": 3.2181682293789917,
+      "grad_norm": 8.625,
+      "learning_rate": 2.9886448566381303e-05,
+      "loss": 0.8268,
+      "step": 47365
+    },
+    {
+      "epoch": 3.2185079494496533,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9884325315939664e-05,
+      "loss": 0.907,
+      "step": 47370
+    },
+    {
+      "epoch": 3.2188476695203154,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9882202065498034e-05,
+      "loss": 0.8293,
+      "step": 47375
+    },
+    {
+      "epoch": 3.219187389590977,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9880078815056395e-05,
+      "loss": 0.9759,
+      "step": 47380
+    },
+    {
+      "epoch": 3.2195271096616387,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9877955564614756e-05,
+      "loss": 0.8794,
+      "step": 47385
+    },
+    {
+      "epoch": 3.2198668297323008,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9875832314173123e-05,
+      "loss": 0.9978,
+      "step": 47390
+    },
+    {
+      "epoch": 3.2202065498029624,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.9873709063731487e-05,
+      "loss": 0.9289,
+      "step": 47395
+    },
+    {
+      "epoch": 3.220546269873624,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9871585813289848e-05,
+      "loss": 0.8814,
+      "step": 47400
+    },
+    {
+      "epoch": 3.220885989944286,
+      "grad_norm": 7.25,
+      "learning_rate": 2.9869462562848215e-05,
+      "loss": 0.8815,
+      "step": 47405
+    },
+    {
+      "epoch": 3.2212257100149477,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.986733931240658e-05,
+      "loss": 0.8099,
+      "step": 47410
+    },
+    {
+      "epoch": 3.2215654300856094,
+      "grad_norm": 8.5,
+      "learning_rate": 2.986521606196494e-05,
+      "loss": 0.8075,
+      "step": 47415
+    },
+    {
+      "epoch": 3.2219051501562714,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9863092811523307e-05,
+      "loss": 0.9475,
+      "step": 47420
+    },
+    {
+      "epoch": 3.222244870226933,
+      "grad_norm": 7.75,
+      "learning_rate": 2.986096956108167e-05,
+      "loss": 0.8478,
+      "step": 47425
+    },
+    {
+      "epoch": 3.2225845902975947,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.985884631064003e-05,
+      "loss": 0.9462,
+      "step": 47430
+    },
+    {
+      "epoch": 3.2229243103682563,
+      "grad_norm": 5.875,
+      "learning_rate": 2.98567230601984e-05,
+      "loss": 0.9977,
+      "step": 47435
+    },
+    {
+      "epoch": 3.2232640304389184,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9854599809756763e-05,
+      "loss": 0.8461,
+      "step": 47440
+    },
+    {
+      "epoch": 3.22360375050958,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9852476559315124e-05,
+      "loss": 0.9582,
+      "step": 47445
+    },
+    {
+      "epoch": 3.2239434705802417,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.985035330887349e-05,
+      "loss": 0.8878,
+      "step": 47450
+    },
+    {
+      "epoch": 3.2242831906509037,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.984823005843185e-05,
+      "loss": 0.9514,
+      "step": 47455
+    },
+    {
+      "epoch": 3.2246229107215654,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9846106807990216e-05,
+      "loss": 0.826,
+      "step": 47460
+    },
+    {
+      "epoch": 3.224962630792227,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.9843983557548583e-05,
+      "loss": 0.8994,
+      "step": 47465
+    },
+    {
+      "epoch": 3.225302350862889,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9841860307106944e-05,
+      "loss": 0.8818,
+      "step": 47470
+    },
+    {
+      "epoch": 3.2256420709335507,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9839737056665308e-05,
+      "loss": 0.9747,
+      "step": 47475
+    },
+    {
+      "epoch": 3.2259817910042123,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9837613806223675e-05,
+      "loss": 0.9345,
+      "step": 47480
+    },
+    {
+      "epoch": 3.2263215110748744,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9835490555782036e-05,
+      "loss": 0.8704,
+      "step": 47485
+    },
+    {
+      "epoch": 3.226661231145536,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.98333673053404e-05,
+      "loss": 0.9302,
+      "step": 47490
+    },
+    {
+      "epoch": 3.2270009512161977,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9831244054898767e-05,
+      "loss": 0.9205,
+      "step": 47495
+    },
+    {
+      "epoch": 3.2273406712868598,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.9829120804457128e-05,
+      "loss": 0.8687,
+      "step": 47500
+    },
+    {
+      "epoch": 3.2276803913575214,
+      "grad_norm": 6.625,
+      "learning_rate": 2.982699755401549e-05,
+      "loss": 0.947,
+      "step": 47505
+    },
+    {
+      "epoch": 3.228020111428183,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.982487430357386e-05,
+      "loss": 0.9404,
+      "step": 47510
+    },
+    {
+      "epoch": 3.228359831498845,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.982275105313222e-05,
+      "loss": 0.9407,
+      "step": 47515
+    },
+    {
+      "epoch": 3.2286995515695067,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.982062780269058e-05,
+      "loss": 0.8777,
+      "step": 47520
+    },
+    {
+      "epoch": 3.2290392716401684,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.9818504552248948e-05,
+      "loss": 0.814,
+      "step": 47525
+    },
+    {
+      "epoch": 3.2293789917108304,
+      "grad_norm": 7.125,
+      "learning_rate": 2.981638130180731e-05,
+      "loss": 0.9196,
+      "step": 47530
+    },
+    {
+      "epoch": 3.229718711781492,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.981425805136568e-05,
+      "loss": 0.871,
+      "step": 47535
+    },
+    {
+      "epoch": 3.2300584318521537,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.981213480092404e-05,
+      "loss": 0.8408,
+      "step": 47540
+    },
+    {
+      "epoch": 3.2303981519228158,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9810011550482404e-05,
+      "loss": 0.9337,
+      "step": 47545
+    },
+    {
+      "epoch": 3.2307378719934774,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.980788830004077e-05,
+      "loss": 0.8943,
+      "step": 47550
+    },
+    {
+      "epoch": 3.231077592064139,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9805765049599132e-05,
+      "loss": 0.8837,
+      "step": 47555
+    },
+    {
+      "epoch": 3.231417312134801,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.9803641799157496e-05,
+      "loss": 0.9365,
+      "step": 47560
+    },
+    {
+      "epoch": 3.2317570322054627,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9801518548715863e-05,
+      "loss": 0.9623,
+      "step": 47565
+    },
+    {
+      "epoch": 3.2320967522761244,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9799395298274224e-05,
+      "loss": 0.881,
+      "step": 47570
+    },
+    {
+      "epoch": 3.2324364723467864,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9797272047832588e-05,
+      "loss": 0.8857,
+      "step": 47575
+    },
+    {
+      "epoch": 3.232776192417448,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.9795148797390955e-05,
+      "loss": 0.8307,
+      "step": 47580
+    },
+    {
+      "epoch": 3.2331159124881097,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.9793025546949316e-05,
+      "loss": 0.8464,
+      "step": 47585
+    },
+    {
+      "epoch": 3.2334556325587718,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.9790902296507676e-05,
+      "loss": 0.9178,
+      "step": 47590
+    },
+    {
+      "epoch": 3.2337953526294334,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9788779046066044e-05,
+      "loss": 0.9161,
+      "step": 47595
+    },
+    {
+      "epoch": 3.234135072700095,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.9786655795624408e-05,
+      "loss": 0.9703,
+      "step": 47600
+    },
+    {
+      "epoch": 3.234474792770757,
+      "grad_norm": 7.0,
+      "learning_rate": 2.978453254518277e-05,
+      "loss": 0.9648,
+      "step": 47605
+    },
+    {
+      "epoch": 3.2348145128414187,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9782409294741136e-05,
+      "loss": 0.9344,
+      "step": 47610
+    },
+    {
+      "epoch": 3.2351542329120804,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.97802860442995e-05,
+      "loss": 0.8824,
+      "step": 47615
+    },
+    {
+      "epoch": 3.2354939529827424,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.977816279385786e-05,
+      "loss": 0.9262,
+      "step": 47620
+    },
+    {
+      "epoch": 3.235833673053404,
+      "grad_norm": 6.75,
+      "learning_rate": 2.9776039543416228e-05,
+      "loss": 0.8713,
+      "step": 47625
+    },
+    {
+      "epoch": 3.2361733931240657,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.9773916292974592e-05,
+      "loss": 0.9221,
+      "step": 47630
+    },
+    {
+      "epoch": 3.2365131131947273,
+      "grad_norm": 6.625,
+      "learning_rate": 2.9771793042532952e-05,
+      "loss": 0.8579,
+      "step": 47635
+    },
+    {
+      "epoch": 3.2368528332653894,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.976966979209132e-05,
+      "loss": 0.8863,
+      "step": 47640
+    },
+    {
+      "epoch": 3.237192553336051,
+      "grad_norm": 5.90625,
+      "learning_rate": 2.9767546541649684e-05,
+      "loss": 0.9333,
+      "step": 47645
+    },
+    {
+      "epoch": 3.2375322734067127,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9765423291208044e-05,
+      "loss": 0.8871,
+      "step": 47650
+    },
+    {
+      "epoch": 3.2378719934773748,
+      "grad_norm": 6.75,
+      "learning_rate": 2.9763300040766412e-05,
+      "loss": 0.8187,
+      "step": 47655
+    },
+    {
+      "epoch": 3.2382117135480364,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.9761176790324772e-05,
+      "loss": 0.9448,
+      "step": 47660
+    },
+    {
+      "epoch": 3.238551433618698,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9759053539883136e-05,
+      "loss": 0.9933,
+      "step": 47665
+    },
+    {
+      "epoch": 3.23889115368936,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9756930289441504e-05,
+      "loss": 0.9289,
+      "step": 47670
+    },
+    {
+      "epoch": 3.2392308737600217,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9754807038999864e-05,
+      "loss": 0.8502,
+      "step": 47675
+    },
+    {
+      "epoch": 3.2395705938306834,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.975268378855823e-05,
+      "loss": 0.8859,
+      "step": 47680
+    },
+    {
+      "epoch": 3.2399103139013454,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.9750560538116596e-05,
+      "loss": 1.0225,
+      "step": 47685
+    },
+    {
+      "epoch": 3.240250033972007,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9748437287674956e-05,
+      "loss": 0.9391,
+      "step": 47690
+    },
+    {
+      "epoch": 3.2405897540426687,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.974631403723332e-05,
+      "loss": 0.988,
+      "step": 47695
+    },
+    {
+      "epoch": 3.2409294741133308,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.9744190786791688e-05,
+      "loss": 0.8694,
+      "step": 47700
+    },
+    {
+      "epoch": 3.2412691941839924,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.974206753635005e-05,
+      "loss": 0.903,
+      "step": 47705
+    },
+    {
+      "epoch": 3.241608914254654,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9739944285908412e-05,
+      "loss": 0.9775,
+      "step": 47710
+    },
+    {
+      "epoch": 3.241948634325316,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.973782103546678e-05,
+      "loss": 0.9045,
+      "step": 47715
+    },
+    {
+      "epoch": 3.2422883543959777,
+      "grad_norm": 7.25,
+      "learning_rate": 2.973569778502514e-05,
+      "loss": 0.9682,
+      "step": 47720
+    },
+    {
+      "epoch": 3.2426280744666394,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.97335745345835e-05,
+      "loss": 0.841,
+      "step": 47725
+    },
+    {
+      "epoch": 3.2429677945373014,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9731451284141872e-05,
+      "loss": 0.8005,
+      "step": 47730
+    },
+    {
+      "epoch": 3.243307514607963,
+      "grad_norm": 6.875,
+      "learning_rate": 2.9729328033700232e-05,
+      "loss": 0.979,
+      "step": 47735
+    },
+    {
+      "epoch": 3.2436472346786247,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.9727204783258593e-05,
+      "loss": 0.8787,
+      "step": 47740
+    },
+    {
+      "epoch": 3.2439869547492868,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.972508153281696e-05,
+      "loss": 0.9527,
+      "step": 47745
+    },
+    {
+      "epoch": 3.2443266748199484,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.9722958282375324e-05,
+      "loss": 0.8484,
+      "step": 47750
+    },
+    {
+      "epoch": 3.24466639489061,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.9720835031933685e-05,
+      "loss": 0.9785,
+      "step": 47755
+    },
+    {
+      "epoch": 3.245006114961272,
+      "grad_norm": 7.5,
+      "learning_rate": 2.9718711781492052e-05,
+      "loss": 0.8647,
+      "step": 47760
+    },
+    {
+      "epoch": 3.2453458350319337,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9716588531050416e-05,
+      "loss": 0.8562,
+      "step": 47765
+    },
+    {
+      "epoch": 3.2456855551025954,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9714465280608777e-05,
+      "loss": 0.9619,
+      "step": 47770
+    },
+    {
+      "epoch": 3.246025275173257,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.9712342030167144e-05,
+      "loss": 0.8583,
+      "step": 47775
+    },
+    {
+      "epoch": 3.246364995243919,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.971021877972551e-05,
+      "loss": 0.9155,
+      "step": 47780
+    },
+    {
+      "epoch": 3.2467047153145807,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.970809552928387e-05,
+      "loss": 0.8771,
+      "step": 47785
+    },
+    {
+      "epoch": 3.2470444353852423,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9705972278842236e-05,
+      "loss": 0.9104,
+      "step": 47790
+    },
+    {
+      "epoch": 3.2473841554559044,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.9703849028400597e-05,
+      "loss": 1.0835,
+      "step": 47795
+    },
+    {
+      "epoch": 3.247723875526566,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.970172577795896e-05,
+      "loss": 0.941,
+      "step": 47800
+    },
+    {
+      "epoch": 3.2480635955972277,
+      "grad_norm": 8.125,
+      "learning_rate": 2.969960252751733e-05,
+      "loss": 0.836,
+      "step": 47805
+    },
+    {
+      "epoch": 3.2484033156678898,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.969747927707569e-05,
+      "loss": 0.8704,
+      "step": 47810
+    },
+    {
+      "epoch": 3.2487430357385514,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.9695356026634053e-05,
+      "loss": 0.885,
+      "step": 47815
+    },
+    {
+      "epoch": 3.249082755809213,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.969323277619242e-05,
+      "loss": 0.8571,
+      "step": 47820
+    },
+    {
+      "epoch": 3.249422475879875,
+      "grad_norm": 9.125,
+      "learning_rate": 2.969110952575078e-05,
+      "loss": 0.8948,
+      "step": 47825
+    },
+    {
+      "epoch": 3.2497621959505367,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.9688986275309145e-05,
+      "loss": 0.8319,
+      "step": 47830
+    },
+    {
+      "epoch": 3.2501019160211984,
+      "grad_norm": 8.625,
+      "learning_rate": 2.9686863024867513e-05,
+      "loss": 0.9673,
+      "step": 47835
+    },
+    {
+      "epoch": 3.2504416360918604,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.9684739774425873e-05,
+      "loss": 0.8891,
+      "step": 47840
+    },
+    {
+      "epoch": 3.250781356162522,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.9682616523984237e-05,
+      "loss": 0.8366,
+      "step": 47845
+    },
+    {
+      "epoch": 3.2511210762331837,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.9680493273542605e-05,
+      "loss": 0.9254,
+      "step": 47850
+    },
+    {
+      "epoch": 3.2514607963038458,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.9678370023100965e-05,
+      "loss": 0.8286,
+      "step": 47855
+    },
+    {
+      "epoch": 3.2518005163745074,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.9676246772659326e-05,
+      "loss": 0.9058,
+      "step": 47860
+    },
+    {
+      "epoch": 3.252140236445169,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9674123522217697e-05,
+      "loss": 0.8607,
+      "step": 47865
+    },
+    {
+      "epoch": 3.252479956515831,
+      "grad_norm": 7.5,
+      "learning_rate": 2.9672000271776057e-05,
+      "loss": 0.8711,
+      "step": 47870
+    },
+    {
+      "epoch": 3.2528196765864927,
+      "grad_norm": 7.25,
+      "learning_rate": 2.9669877021334425e-05,
+      "loss": 0.8416,
+      "step": 47875
+    },
+    {
+      "epoch": 3.2531593966571544,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.9667753770892785e-05,
+      "loss": 0.8089,
+      "step": 47880
+    },
+    {
+      "epoch": 3.2534991167278164,
+      "grad_norm": 8.25,
+      "learning_rate": 2.966563052045115e-05,
+      "loss": 0.8638,
+      "step": 47885
+    },
+    {
+      "epoch": 3.253838836798478,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9663507270009517e-05,
+      "loss": 0.9269,
+      "step": 47890
+    },
+    {
+      "epoch": 3.2541785568691397,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9661384019567877e-05,
+      "loss": 0.8924,
+      "step": 47895
+    },
+    {
+      "epoch": 3.254518276939802,
+      "grad_norm": 7.75,
+      "learning_rate": 2.965926076912624e-05,
+      "loss": 0.9233,
+      "step": 47900
+    },
+    {
+      "epoch": 3.2548579970104634,
+      "grad_norm": 7.0,
+      "learning_rate": 2.965713751868461e-05,
+      "loss": 0.9306,
+      "step": 47905
+    },
+    {
+      "epoch": 3.255197717081125,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.965501426824297e-05,
+      "loss": 0.8939,
+      "step": 47910
+    },
+    {
+      "epoch": 3.255537437151787,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9652891017801333e-05,
+      "loss": 0.9572,
+      "step": 47915
+    },
+    {
+      "epoch": 3.2558771572224487,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.96507677673597e-05,
+      "loss": 0.967,
+      "step": 47920
+    },
+    {
+      "epoch": 3.2562168772931104,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.964864451691806e-05,
+      "loss": 1.0123,
+      "step": 47925
+    },
+    {
+      "epoch": 3.2565565973637725,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.9646521266476425e-05,
+      "loss": 0.9764,
+      "step": 47930
+    },
+    {
+      "epoch": 3.256896317434434,
+      "grad_norm": 8.75,
+      "learning_rate": 2.9644398016034793e-05,
+      "loss": 0.8697,
+      "step": 47935
+    },
+    {
+      "epoch": 3.2572360375050957,
+      "grad_norm": 6.0,
+      "learning_rate": 2.9642274765593153e-05,
+      "loss": 0.8837,
+      "step": 47940
+    },
+    {
+      "epoch": 3.257575757575758,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.9640151515151514e-05,
+      "loss": 0.9493,
+      "step": 47945
+    },
+    {
+      "epoch": 3.2579154776464194,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.963802826470988e-05,
+      "loss": 0.9941,
+      "step": 47950
+    },
+    {
+      "epoch": 3.258255197717081,
+      "grad_norm": 6.375,
+      "learning_rate": 2.9635905014268245e-05,
+      "loss": 0.938,
+      "step": 47955
+    },
+    {
+      "epoch": 3.258594917787743,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9633781763826606e-05,
+      "loss": 0.9433,
+      "step": 47960
+    },
+    {
+      "epoch": 3.2589346378584048,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.9631658513384973e-05,
+      "loss": 0.9048,
+      "step": 47965
+    },
+    {
+      "epoch": 3.2592743579290664,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.9629535262943337e-05,
+      "loss": 0.9074,
+      "step": 47970
+    },
+    {
+      "epoch": 3.2596140779997285,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.9627412012501698e-05,
+      "loss": 0.8816,
+      "step": 47975
+    },
+    {
+      "epoch": 3.25995379807039,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.9625288762060065e-05,
+      "loss": 0.9177,
+      "step": 47980
+    },
+    {
+      "epoch": 3.2602935181410517,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.962316551161843e-05,
+      "loss": 0.9753,
+      "step": 47985
+    },
+    {
+      "epoch": 3.2606332382117134,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.962104226117679e-05,
+      "loss": 0.8597,
+      "step": 47990
+    },
+    {
+      "epoch": 3.2609729582823754,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9618919010735157e-05,
+      "loss": 0.9451,
+      "step": 47995
+    },
+    {
+      "epoch": 3.261312678353037,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.961679576029352e-05,
+      "loss": 0.9953,
+      "step": 48000
+    },
+    {
+      "epoch": 3.2616523984236987,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.9614672509851882e-05,
+      "loss": 0.9327,
+      "step": 48005
+    },
+    {
+      "epoch": 3.2619921184943608,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.961254925941025e-05,
+      "loss": 0.9788,
+      "step": 48010
+    },
+    {
+      "epoch": 3.2623318385650224,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.961042600896861e-05,
+      "loss": 0.9067,
+      "step": 48015
+    },
+    {
+      "epoch": 3.262671558635684,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9608302758526974e-05,
+      "loss": 0.8882,
+      "step": 48020
+    },
+    {
+      "epoch": 3.263011278706346,
+      "grad_norm": 8.0,
+      "learning_rate": 2.960617950808534e-05,
+      "loss": 0.9368,
+      "step": 48025
+    },
+    {
+      "epoch": 3.2633509987770077,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9604056257643702e-05,
+      "loss": 0.8723,
+      "step": 48030
+    },
+    {
+      "epoch": 3.2636907188476694,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.9601933007202066e-05,
+      "loss": 0.8839,
+      "step": 48035
+    },
+    {
+      "epoch": 3.2640304389183314,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.9599809756760433e-05,
+      "loss": 0.9008,
+      "step": 48040
+    },
+    {
+      "epoch": 3.264370158988993,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9597686506318794e-05,
+      "loss": 0.8544,
+      "step": 48045
+    },
+    {
+      "epoch": 3.2647098790596547,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.9595563255877158e-05,
+      "loss": 0.9051,
+      "step": 48050
+    },
+    {
+      "epoch": 3.265049599130317,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.9593440005435525e-05,
+      "loss": 0.8098,
+      "step": 48055
+    },
+    {
+      "epoch": 3.2653893192009784,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.9591316754993886e-05,
+      "loss": 0.8286,
+      "step": 48060
+    },
+    {
+      "epoch": 3.26572903927164,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.958919350455225e-05,
+      "loss": 0.9471,
+      "step": 48065
+    },
+    {
+      "epoch": 3.266068759342302,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.9587070254110617e-05,
+      "loss": 0.9058,
+      "step": 48070
+    },
+    {
+      "epoch": 3.2664084794129638,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9584947003668978e-05,
+      "loss": 0.952,
+      "step": 48075
+    },
+    {
+      "epoch": 3.2667481994836254,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.958282375322734e-05,
+      "loss": 0.9498,
+      "step": 48080
+    },
+    {
+      "epoch": 3.2670879195542875,
+      "grad_norm": 6.875,
+      "learning_rate": 2.9580700502785706e-05,
+      "loss": 0.8522,
+      "step": 48085
+    },
+    {
+      "epoch": 3.267427639624949,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.957857725234407e-05,
+      "loss": 0.8693,
+      "step": 48090
+    },
+    {
+      "epoch": 3.2677673596956107,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.957645400190243e-05,
+      "loss": 0.9025,
+      "step": 48095
+    },
+    {
+      "epoch": 3.2681070797662723,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.9574330751460798e-05,
+      "loss": 0.8971,
+      "step": 48100
+    },
+    {
+      "epoch": 3.2684467998369344,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.9572207501019162e-05,
+      "loss": 0.9048,
+      "step": 48105
+    },
+    {
+      "epoch": 3.268786519907596,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9570084250577523e-05,
+      "loss": 0.8981,
+      "step": 48110
+    },
+    {
+      "epoch": 3.2691262399782577,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.956796100013589e-05,
+      "loss": 0.9063,
+      "step": 48115
+    },
+    {
+      "epoch": 3.2694659600489198,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.9565837749694254e-05,
+      "loss": 0.9623,
+      "step": 48120
+    },
+    {
+      "epoch": 3.2698056801195814,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9563714499252615e-05,
+      "loss": 0.8734,
+      "step": 48125
+    },
+    {
+      "epoch": 3.270145400190243,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.9561591248810982e-05,
+      "loss": 0.836,
+      "step": 48130
+    },
+    {
+      "epoch": 3.270485120260905,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9559467998369346e-05,
+      "loss": 0.8482,
+      "step": 48135
+    },
+    {
+      "epoch": 3.2708248403315667,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9557344747927707e-05,
+      "loss": 0.9279,
+      "step": 48140
+    },
+    {
+      "epoch": 3.2711645604022284,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.9555221497486074e-05,
+      "loss": 0.8444,
+      "step": 48145
+    },
+    {
+      "epoch": 3.2715042804728904,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.9553098247044435e-05,
+      "loss": 0.8706,
+      "step": 48150
+    },
+    {
+      "epoch": 3.271844000543552,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.95509749966028e-05,
+      "loss": 1.0108,
+      "step": 48155
+    },
+    {
+      "epoch": 3.2721837206142137,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.9548851746161166e-05,
+      "loss": 0.7906,
+      "step": 48160
+    },
+    {
+      "epoch": 3.2725234406848758,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9546728495719527e-05,
+      "loss": 0.9077,
+      "step": 48165
+    },
+    {
+      "epoch": 3.2728631607555374,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.954460524527789e-05,
+      "loss": 0.9246,
+      "step": 48170
+    },
+    {
+      "epoch": 3.273202880826199,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.9542481994836258e-05,
+      "loss": 0.8322,
+      "step": 48175
+    },
+    {
+      "epoch": 3.273542600896861,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.954035874439462e-05,
+      "loss": 0.9272,
+      "step": 48180
+    },
+    {
+      "epoch": 3.2738823209675227,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.9538235493952983e-05,
+      "loss": 0.8346,
+      "step": 48185
+    },
+    {
+      "epoch": 3.2742220410381844,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.953611224351135e-05,
+      "loss": 0.9473,
+      "step": 48190
+    },
+    {
+      "epoch": 3.2745617611088464,
+      "grad_norm": 7.0,
+      "learning_rate": 2.953398899306971e-05,
+      "loss": 0.8676,
+      "step": 48195
+    },
+    {
+      "epoch": 3.274901481179508,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9531865742628075e-05,
+      "loss": 0.8968,
+      "step": 48200
+    },
+    {
+      "epoch": 3.2752412012501697,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9529742492186442e-05,
+      "loss": 0.9093,
+      "step": 48205
+    },
+    {
+      "epoch": 3.275580921320832,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.9527619241744803e-05,
+      "loss": 0.8615,
+      "step": 48210
+    },
+    {
+      "epoch": 3.2759206413914934,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.952549599130317e-05,
+      "loss": 0.8574,
+      "step": 48215
+    },
+    {
+      "epoch": 3.276260361462155,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9523372740861534e-05,
+      "loss": 0.9284,
+      "step": 48220
+    },
+    {
+      "epoch": 3.276600081532817,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9521249490419895e-05,
+      "loss": 0.9031,
+      "step": 48225
+    },
+    {
+      "epoch": 3.2769398016034788,
+      "grad_norm": 6.5,
+      "learning_rate": 2.9519126239978262e-05,
+      "loss": 0.8628,
+      "step": 48230
+    },
+    {
+      "epoch": 3.2772795216741404,
+      "grad_norm": 7.5,
+      "learning_rate": 2.9517002989536623e-05,
+      "loss": 0.8464,
+      "step": 48235
+    },
+    {
+      "epoch": 3.2776192417448025,
+      "grad_norm": 7.875,
+      "learning_rate": 2.9514879739094987e-05,
+      "loss": 1.0173,
+      "step": 48240
+    },
+    {
+      "epoch": 3.277958961815464,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9512756488653354e-05,
+      "loss": 0.952,
+      "step": 48245
+    },
+    {
+      "epoch": 3.2782986818861257,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9510633238211715e-05,
+      "loss": 0.9269,
+      "step": 48250
+    },
+    {
+      "epoch": 3.278638401956788,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.950850998777008e-05,
+      "loss": 0.9277,
+      "step": 48255
+    },
+    {
+      "epoch": 3.2789781220274494,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.9506386737328446e-05,
+      "loss": 0.9362,
+      "step": 48260
+    },
+    {
+      "epoch": 3.279317842098111,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.9504263486886807e-05,
+      "loss": 0.8769,
+      "step": 48265
+    },
+    {
+      "epoch": 3.279657562168773,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.950214023644517e-05,
+      "loss": 0.8885,
+      "step": 48270
+    },
+    {
+      "epoch": 3.2799972822394348,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.9500016986003538e-05,
+      "loss": 0.8573,
+      "step": 48275
+    },
+    {
+      "epoch": 3.2803370023100964,
+      "grad_norm": 6.75,
+      "learning_rate": 2.94978937355619e-05,
+      "loss": 0.8975,
+      "step": 48280
+    },
+    {
+      "epoch": 3.2806767223807585,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.949577048512026e-05,
+      "loss": 0.8452,
+      "step": 48285
+    },
+    {
+      "epoch": 3.28101644245142,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.949364723467863e-05,
+      "loss": 0.9397,
+      "step": 48290
+    },
+    {
+      "epoch": 3.2813561625220817,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.949152398423699e-05,
+      "loss": 0.9246,
+      "step": 48295
+    },
+    {
+      "epoch": 3.281695882592744,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.948940073379535e-05,
+      "loss": 0.9181,
+      "step": 48300
+    },
+    {
+      "epoch": 3.2820356026634054,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.948727748335372e-05,
+      "loss": 0.8414,
+      "step": 48305
+    },
+    {
+      "epoch": 3.282375322734067,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9485154232912083e-05,
+      "loss": 0.9486,
+      "step": 48310
+    },
+    {
+      "epoch": 3.282715042804729,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.9483030982470443e-05,
+      "loss": 0.9407,
+      "step": 48315
+    },
+    {
+      "epoch": 3.2830547628753908,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.948090773202881e-05,
+      "loss": 0.8775,
+      "step": 48320
+    },
+    {
+      "epoch": 3.2833944829460524,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.9478784481587175e-05,
+      "loss": 0.8194,
+      "step": 48325
+    },
+    {
+      "epoch": 3.283734203016714,
+      "grad_norm": 5.875,
+      "learning_rate": 2.9476661231145535e-05,
+      "loss": 0.9279,
+      "step": 48330
+    },
+    {
+      "epoch": 3.284073923087376,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9474537980703903e-05,
+      "loss": 0.8707,
+      "step": 48335
+    },
+    {
+      "epoch": 3.2844136431580377,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.9472414730262267e-05,
+      "loss": 0.8836,
+      "step": 48340
+    },
+    {
+      "epoch": 3.2847533632286994,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.9470291479820627e-05,
+      "loss": 0.8147,
+      "step": 48345
+    },
+    {
+      "epoch": 3.2850930832993614,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9468168229378995e-05,
+      "loss": 0.8585,
+      "step": 48350
+    },
+    {
+      "epoch": 3.285432803370023,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.946604497893736e-05,
+      "loss": 0.8575,
+      "step": 48355
+    },
+    {
+      "epoch": 3.2857725234406847,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.946392172849572e-05,
+      "loss": 0.9357,
+      "step": 48360
+    },
+    {
+      "epoch": 3.286112243511347,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9461798478054087e-05,
+      "loss": 0.863,
+      "step": 48365
+    },
+    {
+      "epoch": 3.2864519635820084,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9459675227612447e-05,
+      "loss": 0.9336,
+      "step": 48370
+    },
+    {
+      "epoch": 3.28679168365267,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.945755197717081e-05,
+      "loss": 0.9719,
+      "step": 48375
+    },
+    {
+      "epoch": 3.287131403723332,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.945542872672918e-05,
+      "loss": 0.9653,
+      "step": 48380
+    },
+    {
+      "epoch": 3.2874711237939938,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.945330547628754e-05,
+      "loss": 0.8763,
+      "step": 48385
+    },
+    {
+      "epoch": 3.2878108438646554,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9451182225845903e-05,
+      "loss": 0.8657,
+      "step": 48390
+    },
+    {
+      "epoch": 3.2881505639353175,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.944905897540427e-05,
+      "loss": 0.9205,
+      "step": 48395
+    },
+    {
+      "epoch": 3.288490284005979,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.944693572496263e-05,
+      "loss": 0.8545,
+      "step": 48400
+    },
+    {
+      "epoch": 3.2888300040766407,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.9444812474520995e-05,
+      "loss": 0.9471,
+      "step": 48405
+    },
+    {
+      "epoch": 3.289169724147303,
+      "grad_norm": 7.875,
+      "learning_rate": 2.9442689224079363e-05,
+      "loss": 0.9,
+      "step": 48410
+    },
+    {
+      "epoch": 3.2895094442179644,
+      "grad_norm": 6.75,
+      "learning_rate": 2.9440565973637723e-05,
+      "loss": 0.9239,
+      "step": 48415
+    },
+    {
+      "epoch": 3.289849164288626,
+      "grad_norm": 8.375,
+      "learning_rate": 2.9438442723196087e-05,
+      "loss": 0.8617,
+      "step": 48420
+    },
+    {
+      "epoch": 3.290188884359288,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.9436319472754455e-05,
+      "loss": 0.8838,
+      "step": 48425
+    },
+    {
+      "epoch": 3.2905286044299498,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9434196222312815e-05,
+      "loss": 0.9239,
+      "step": 48430
+    },
+    {
+      "epoch": 3.2908683245006114,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9432072971871176e-05,
+      "loss": 0.9564,
+      "step": 48435
+    },
+    {
+      "epoch": 3.291208044571273,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.9429949721429543e-05,
+      "loss": 1.012,
+      "step": 48440
+    },
+    {
+      "epoch": 3.291547764641935,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.9427826470987907e-05,
+      "loss": 0.9878,
+      "step": 48445
+    },
+    {
+      "epoch": 3.2918874847125967,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9425703220546268e-05,
+      "loss": 0.9817,
+      "step": 48450
+    },
+    {
+      "epoch": 3.2922272047832584,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.9423579970104635e-05,
+      "loss": 0.9579,
+      "step": 48455
+    },
+    {
+      "epoch": 3.2925669248539204,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9421456719663e-05,
+      "loss": 0.8511,
+      "step": 48460
+    },
+    {
+      "epoch": 3.292906644924582,
+      "grad_norm": 9.25,
+      "learning_rate": 2.941933346922136e-05,
+      "loss": 0.8522,
+      "step": 48465
+    },
+    {
+      "epoch": 3.2932463649952437,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.9417210218779727e-05,
+      "loss": 0.8646,
+      "step": 48470
+    },
+    {
+      "epoch": 3.2935860850659058,
+      "grad_norm": 8.375,
+      "learning_rate": 2.941508696833809e-05,
+      "loss": 0.9122,
+      "step": 48475
+    },
+    {
+      "epoch": 3.2939258051365674,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9412963717896452e-05,
+      "loss": 0.9219,
+      "step": 48480
+    },
+    {
+      "epoch": 3.294265525207229,
+      "grad_norm": 6.5,
+      "learning_rate": 2.941084046745482e-05,
+      "loss": 0.801,
+      "step": 48485
+    },
+    {
+      "epoch": 3.294605245277891,
+      "grad_norm": 9.125,
+      "learning_rate": 2.9408717217013183e-05,
+      "loss": 0.9166,
+      "step": 48490
+    },
+    {
+      "epoch": 3.2949449653485527,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9406593966571544e-05,
+      "loss": 0.9424,
+      "step": 48495
+    },
+    {
+      "epoch": 3.2952846854192144,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.940447071612991e-05,
+      "loss": 0.8994,
+      "step": 48500
+    },
+    {
+      "epoch": 3.2956244054898765,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.9402347465688272e-05,
+      "loss": 0.9483,
+      "step": 48505
+    },
+    {
+      "epoch": 3.295964125560538,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9400224215246636e-05,
+      "loss": 0.8961,
+      "step": 48510
+    },
+    {
+      "epoch": 3.2963038456311997,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9398100964805004e-05,
+      "loss": 0.9482,
+      "step": 48515
+    },
+    {
+      "epoch": 3.296643565701862,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9395977714363364e-05,
+      "loss": 0.9021,
+      "step": 48520
+    },
+    {
+      "epoch": 3.2969832857725234,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.9393854463921728e-05,
+      "loss": 0.9506,
+      "step": 48525
+    },
+    {
+      "epoch": 3.297323005843185,
+      "grad_norm": 6.625,
+      "learning_rate": 2.9391731213480096e-05,
+      "loss": 0.879,
+      "step": 48530
+    },
+    {
+      "epoch": 3.297662725913847,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.9389607963038456e-05,
+      "loss": 0.9343,
+      "step": 48535
+    },
+    {
+      "epoch": 3.2980024459845088,
+      "grad_norm": 7.125,
+      "learning_rate": 2.938748471259682e-05,
+      "loss": 0.8735,
+      "step": 48540
+    },
+    {
+      "epoch": 3.2983421660551704,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.9385361462155188e-05,
+      "loss": 0.973,
+      "step": 48545
+    },
+    {
+      "epoch": 3.2986818861258325,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.9383238211713548e-05,
+      "loss": 0.8883,
+      "step": 48550
+    },
+    {
+      "epoch": 3.299021606196494,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.9381114961271916e-05,
+      "loss": 0.905,
+      "step": 48555
+    },
+    {
+      "epoch": 3.2993613262671557,
+      "grad_norm": 7.75,
+      "learning_rate": 2.937899171083028e-05,
+      "loss": 0.968,
+      "step": 48560
+    },
+    {
+      "epoch": 3.299701046337818,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.937686846038864e-05,
+      "loss": 0.8599,
+      "step": 48565
+    },
+    {
+      "epoch": 3.3000407664084794,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.9374745209947008e-05,
+      "loss": 0.9269,
+      "step": 48570
+    },
+    {
+      "epoch": 3.300380486479141,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9372621959505368e-05,
+      "loss": 0.8086,
+      "step": 48575
+    },
+    {
+      "epoch": 3.300720206549803,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.9370498709063732e-05,
+      "loss": 0.8186,
+      "step": 48580
+    },
+    {
+      "epoch": 3.3010599266204648,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.93683754586221e-05,
+      "loss": 0.8974,
+      "step": 48585
+    },
+    {
+      "epoch": 3.3013996466911264,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.936625220818046e-05,
+      "loss": 0.8457,
+      "step": 48590
+    },
+    {
+      "epoch": 3.3017393667617885,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9364128957738824e-05,
+      "loss": 0.9467,
+      "step": 48595
+    },
+    {
+      "epoch": 3.30207908683245,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.936200570729719e-05,
+      "loss": 0.9072,
+      "step": 48600
+    },
+    {
+      "epoch": 3.3024188069031117,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.9359882456855552e-05,
+      "loss": 0.8511,
+      "step": 48605
+    },
+    {
+      "epoch": 3.302758526973774,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9357759206413916e-05,
+      "loss": 0.9845,
+      "step": 48610
+    },
+    {
+      "epoch": 3.3030982470444354,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.9355635955972284e-05,
+      "loss": 0.8346,
+      "step": 48615
+    },
+    {
+      "epoch": 3.303437967115097,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.9353512705530644e-05,
+      "loss": 0.887,
+      "step": 48620
+    },
+    {
+      "epoch": 3.303777687185759,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.9351389455089008e-05,
+      "loss": 0.8493,
+      "step": 48625
+    },
+    {
+      "epoch": 3.304117407256421,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.9349266204647376e-05,
+      "loss": 0.9028,
+      "step": 48630
+    },
+    {
+      "epoch": 3.3044571273270824,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9347142954205736e-05,
+      "loss": 0.9962,
+      "step": 48635
+    },
+    {
+      "epoch": 3.3047968473977445,
+      "grad_norm": 7.625,
+      "learning_rate": 2.9345019703764097e-05,
+      "loss": 0.869,
+      "step": 48640
+    },
+    {
+      "epoch": 3.305136567468406,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9342896453322468e-05,
+      "loss": 0.9683,
+      "step": 48645
+    },
+    {
+      "epoch": 3.3054762875390677,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9340773202880828e-05,
+      "loss": 1.0045,
+      "step": 48650
+    },
+    {
+      "epoch": 3.30581600760973,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.933864995243919e-05,
+      "loss": 0.9149,
+      "step": 48655
+    },
+    {
+      "epoch": 3.3061557276803915,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9336526701997556e-05,
+      "loss": 1.0177,
+      "step": 48660
+    },
+    {
+      "epoch": 3.306495447751053,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.933440345155592e-05,
+      "loss": 0.8526,
+      "step": 48665
+    },
+    {
+      "epoch": 3.3068351678217147,
+      "grad_norm": 7.25,
+      "learning_rate": 2.933228020111428e-05,
+      "loss": 0.8182,
+      "step": 48670
+    },
+    {
+      "epoch": 3.307174887892377,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.9330156950672648e-05,
+      "loss": 1.0022,
+      "step": 48675
+    },
+    {
+      "epoch": 3.3075146079630384,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.9328033700231012e-05,
+      "loss": 0.9286,
+      "step": 48680
+    },
+    {
+      "epoch": 3.3078543280337,
+      "grad_norm": 6.125,
+      "learning_rate": 2.9325910449789373e-05,
+      "loss": 0.8375,
+      "step": 48685
+    },
+    {
+      "epoch": 3.308194048104362,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.932378719934774e-05,
+      "loss": 0.8634,
+      "step": 48690
+    },
+    {
+      "epoch": 3.3085337681750238,
+      "grad_norm": 8.0,
+      "learning_rate": 2.9321663948906104e-05,
+      "loss": 0.8173,
+      "step": 48695
+    },
+    {
+      "epoch": 3.3088734882456854,
+      "grad_norm": 6.375,
+      "learning_rate": 2.9319540698464465e-05,
+      "loss": 0.8776,
+      "step": 48700
+    },
+    {
+      "epoch": 3.3092132083163475,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9317417448022832e-05,
+      "loss": 0.8135,
+      "step": 48705
+    },
+    {
+      "epoch": 3.309552928387009,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9315294197581193e-05,
+      "loss": 0.8084,
+      "step": 48710
+    },
+    {
+      "epoch": 3.3098926484576707,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.9313170947139557e-05,
+      "loss": 0.9619,
+      "step": 48715
+    },
+    {
+      "epoch": 3.310232368528333,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.9311047696697924e-05,
+      "loss": 0.9144,
+      "step": 48720
+    },
+    {
+      "epoch": 3.3105720885989944,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.9308924446256285e-05,
+      "loss": 0.8416,
+      "step": 48725
+    },
+    {
+      "epoch": 3.310911808669656,
+      "grad_norm": 6.375,
+      "learning_rate": 2.930680119581465e-05,
+      "loss": 0.806,
+      "step": 48730
+    },
+    {
+      "epoch": 3.311251528740318,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.9304677945373016e-05,
+      "loss": 0.8317,
+      "step": 48735
+    },
+    {
+      "epoch": 3.3115912488109798,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.9302554694931377e-05,
+      "loss": 0.8219,
+      "step": 48740
+    },
+    {
+      "epoch": 3.3119309688816414,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.930043144448974e-05,
+      "loss": 0.8503,
+      "step": 48745
+    },
+    {
+      "epoch": 3.3122706889523035,
+      "grad_norm": 7.125,
+      "learning_rate": 2.9298308194048108e-05,
+      "loss": 0.9115,
+      "step": 48750
+    },
+    {
+      "epoch": 3.312610409022965,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.929618494360647e-05,
+      "loss": 0.8929,
+      "step": 48755
+    },
+    {
+      "epoch": 3.3129501290936267,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.9294061693164833e-05,
+      "loss": 0.9204,
+      "step": 48760
+    },
+    {
+      "epoch": 3.313289849164289,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.92919384427232e-05,
+      "loss": 0.9451,
+      "step": 48765
+    },
+    {
+      "epoch": 3.3136295692349504,
+      "grad_norm": 8.625,
+      "learning_rate": 2.928981519228156e-05,
+      "loss": 0.7915,
+      "step": 48770
+    },
+    {
+      "epoch": 3.313969289305612,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.928769194183992e-05,
+      "loss": 0.8731,
+      "step": 48775
+    },
+    {
+      "epoch": 3.3143090093762737,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.9285568691398292e-05,
+      "loss": 0.8719,
+      "step": 48780
+    },
+    {
+      "epoch": 3.314648729446936,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.9283445440956653e-05,
+      "loss": 0.9086,
+      "step": 48785
+    },
+    {
+      "epoch": 3.3149884495175974,
+      "grad_norm": 8.25,
+      "learning_rate": 2.9281322190515014e-05,
+      "loss": 0.9474,
+      "step": 48790
+    },
+    {
+      "epoch": 3.315328169588259,
+      "grad_norm": 7.625,
+      "learning_rate": 2.927919894007338e-05,
+      "loss": 0.8893,
+      "step": 48795
+    },
+    {
+      "epoch": 3.315667889658921,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9277075689631745e-05,
+      "loss": 0.8768,
+      "step": 48800
+    },
+    {
+      "epoch": 3.3160076097295828,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9274952439190106e-05,
+      "loss": 0.8769,
+      "step": 48805
+    },
+    {
+      "epoch": 3.3163473298002444,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9272829188748473e-05,
+      "loss": 0.9437,
+      "step": 48810
+    },
+    {
+      "epoch": 3.3166870498709065,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9270705938306837e-05,
+      "loss": 0.855,
+      "step": 48815
+    },
+    {
+      "epoch": 3.317026769941568,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9268582687865198e-05,
+      "loss": 0.8316,
+      "step": 48820
+    },
+    {
+      "epoch": 3.3173664900122297,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.9266459437423565e-05,
+      "loss": 0.9481,
+      "step": 48825
+    },
+    {
+      "epoch": 3.317706210082892,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.926433618698193e-05,
+      "loss": 0.9745,
+      "step": 48830
+    },
+    {
+      "epoch": 3.3180459301535534,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.926221293654029e-05,
+      "loss": 0.9732,
+      "step": 48835
+    },
+    {
+      "epoch": 3.318385650224215,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.9260089686098657e-05,
+      "loss": 0.9044,
+      "step": 48840
+    },
+    {
+      "epoch": 3.318725370294877,
+      "grad_norm": 7.125,
+      "learning_rate": 2.925796643565702e-05,
+      "loss": 0.9033,
+      "step": 48845
+    },
+    {
+      "epoch": 3.3190650903655388,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.925584318521538e-05,
+      "loss": 0.8306,
+      "step": 48850
+    },
+    {
+      "epoch": 3.3194048104362004,
+      "grad_norm": 7.125,
+      "learning_rate": 2.925371993477375e-05,
+      "loss": 0.88,
+      "step": 48855
+    },
+    {
+      "epoch": 3.3197445305068625,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.925159668433211e-05,
+      "loss": 0.8567,
+      "step": 48860
+    },
+    {
+      "epoch": 3.320084250577524,
+      "grad_norm": 6.875,
+      "learning_rate": 2.9249473433890474e-05,
+      "loss": 0.9233,
+      "step": 48865
+    },
+    {
+      "epoch": 3.3204239706481857,
+      "grad_norm": 5.71875,
+      "learning_rate": 2.924735018344884e-05,
+      "loss": 1.0031,
+      "step": 48870
+    },
+    {
+      "epoch": 3.320763690718848,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.92452269330072e-05,
+      "loss": 0.9459,
+      "step": 48875
+    },
+    {
+      "epoch": 3.3211034107895094,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9243103682565566e-05,
+      "loss": 0.8351,
+      "step": 48880
+    },
+    {
+      "epoch": 3.321443130860171,
+      "grad_norm": 6.0625,
+      "learning_rate": 2.9240980432123933e-05,
+      "loss": 0.9215,
+      "step": 48885
+    },
+    {
+      "epoch": 3.321782850930833,
+      "grad_norm": 7.25,
+      "learning_rate": 2.9238857181682294e-05,
+      "loss": 0.9766,
+      "step": 48890
+    },
+    {
+      "epoch": 3.3221225710014948,
+      "grad_norm": 7.375,
+      "learning_rate": 2.923673393124066e-05,
+      "loss": 0.8963,
+      "step": 48895
+    },
+    {
+      "epoch": 3.3224622910721564,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.9234610680799025e-05,
+      "loss": 1.0066,
+      "step": 48900
+    },
+    {
+      "epoch": 3.3228020111428185,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9232487430357386e-05,
+      "loss": 0.8536,
+      "step": 48905
+    },
+    {
+      "epoch": 3.32314173121348,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.9230364179915753e-05,
+      "loss": 0.983,
+      "step": 48910
+    },
+    {
+      "epoch": 3.3234814512841417,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.9228240929474117e-05,
+      "loss": 0.923,
+      "step": 48915
+    },
+    {
+      "epoch": 3.323821171354804,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9226117679032478e-05,
+      "loss": 0.8777,
+      "step": 48920
+    },
+    {
+      "epoch": 3.3241608914254654,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.9223994428590845e-05,
+      "loss": 0.9406,
+      "step": 48925
+    },
+    {
+      "epoch": 3.324500611496127,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9221871178149206e-05,
+      "loss": 0.9197,
+      "step": 48930
+    },
+    {
+      "epoch": 3.324840331566789,
+      "grad_norm": 7.25,
+      "learning_rate": 2.921974792770757e-05,
+      "loss": 1.0427,
+      "step": 48935
+    },
+    {
+      "epoch": 3.325180051637451,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.9217624677265937e-05,
+      "loss": 0.8697,
+      "step": 48940
+    },
+    {
+      "epoch": 3.3255197717081124,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9215501426824298e-05,
+      "loss": 0.9967,
+      "step": 48945
+    },
+    {
+      "epoch": 3.3258594917787745,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.921337817638266e-05,
+      "loss": 0.9168,
+      "step": 48950
+    },
+    {
+      "epoch": 3.326199211849436,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.921125492594103e-05,
+      "loss": 0.9045,
+      "step": 48955
+    },
+    {
+      "epoch": 3.3265389319200978,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.920913167549939e-05,
+      "loss": 0.825,
+      "step": 48960
+    },
+    {
+      "epoch": 3.32687865199076,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.9207008425057754e-05,
+      "loss": 0.9541,
+      "step": 48965
+    },
+    {
+      "epoch": 3.3272183720614215,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.920488517461612e-05,
+      "loss": 0.9096,
+      "step": 48970
+    },
+    {
+      "epoch": 3.327558092132083,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.920276192417448e-05,
+      "loss": 0.9137,
+      "step": 48975
+    },
+    {
+      "epoch": 3.327897812202745,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.9200638673732846e-05,
+      "loss": 0.8358,
+      "step": 48980
+    },
+    {
+      "epoch": 3.328237532273407,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.9198515423291213e-05,
+      "loss": 0.9041,
+      "step": 48985
+    },
+    {
+      "epoch": 3.3285772523440684,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.9196392172849574e-05,
+      "loss": 0.8126,
+      "step": 48990
+    },
+    {
+      "epoch": 3.3289169724147305,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.9194268922407934e-05,
+      "loss": 0.9043,
+      "step": 48995
+    },
+    {
+      "epoch": 3.329256692485392,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9192145671966302e-05,
+      "loss": 0.8775,
+      "step": 49000
+    },
+    {
+      "epoch": 3.3295964125560538,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.9190022421524666e-05,
+      "loss": 0.9572,
+      "step": 49005
+    },
+    {
+      "epoch": 3.3299361326267154,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.9187899171083026e-05,
+      "loss": 0.9386,
+      "step": 49010
+    },
+    {
+      "epoch": 3.3302758526973775,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.9185775920641394e-05,
+      "loss": 0.9174,
+      "step": 49015
+    },
+    {
+      "epoch": 3.330615572768039,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9183652670199758e-05,
+      "loss": 0.9904,
+      "step": 49020
+    },
+    {
+      "epoch": 3.3309552928387007,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.918152941975812e-05,
+      "loss": 0.9125,
+      "step": 49025
+    },
+    {
+      "epoch": 3.331295012909363,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.9179406169316486e-05,
+      "loss": 0.89,
+      "step": 49030
+    },
+    {
+      "epoch": 3.3316347329800244,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.917728291887485e-05,
+      "loss": 0.9721,
+      "step": 49035
+    },
+    {
+      "epoch": 3.331974453050686,
+      "grad_norm": 7.375,
+      "learning_rate": 2.917515966843321e-05,
+      "loss": 0.9621,
+      "step": 49040
+    },
+    {
+      "epoch": 3.332314173121348,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9173036417991578e-05,
+      "loss": 0.8582,
+      "step": 49045
+    },
+    {
+      "epoch": 3.3326538931920098,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9170913167549942e-05,
+      "loss": 0.9106,
+      "step": 49050
+    },
+    {
+      "epoch": 3.3329936132626714,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9168789917108302e-05,
+      "loss": 0.911,
+      "step": 49055
+    },
+    {
+      "epoch": 3.3333333333333335,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 0.8476,
+      "step": 49060
+    },
+    {
+      "epoch": 3.333673053403995,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.916454341622503e-05,
+      "loss": 0.8166,
+      "step": 49065
+    },
+    {
+      "epoch": 3.3340127734746567,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.9162420165783394e-05,
+      "loss": 0.9363,
+      "step": 49070
+    },
+    {
+      "epoch": 3.334352493545319,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9160296915341762e-05,
+      "loss": 0.9407,
+      "step": 49075
+    },
+    {
+      "epoch": 3.3346922136159804,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.9158173664900122e-05,
+      "loss": 0.8927,
+      "step": 49080
+    },
+    {
+      "epoch": 3.335031933686642,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.9156050414458486e-05,
+      "loss": 0.8704,
+      "step": 49085
+    },
+    {
+      "epoch": 3.335371653757304,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.9153927164016854e-05,
+      "loss": 0.9736,
+      "step": 49090
+    },
+    {
+      "epoch": 3.335711373827966,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9151803913575214e-05,
+      "loss": 0.9758,
+      "step": 49095
+    },
+    {
+      "epoch": 3.3360510938986274,
+      "grad_norm": 7.375,
+      "learning_rate": 2.914968066313358e-05,
+      "loss": 0.9406,
+      "step": 49100
+    },
+    {
+      "epoch": 3.3363908139692895,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.9147557412691946e-05,
+      "loss": 0.9028,
+      "step": 49105
+    },
+    {
+      "epoch": 3.336730534039951,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.9145434162250306e-05,
+      "loss": 0.9579,
+      "step": 49110
+    },
+    {
+      "epoch": 3.3370702541106128,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.914331091180867e-05,
+      "loss": 0.9386,
+      "step": 49115
+    },
+    {
+      "epoch": 3.3374099741812744,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.9141187661367038e-05,
+      "loss": 1.2178,
+      "step": 49120
+    },
+    {
+      "epoch": 3.3377496942519365,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.91390644109254e-05,
+      "loss": 0.9398,
+      "step": 49125
+    },
+    {
+      "epoch": 3.338089414322598,
+      "grad_norm": 8.125,
+      "learning_rate": 2.913694116048376e-05,
+      "loss": 0.9044,
+      "step": 49130
+    },
+    {
+      "epoch": 3.3384291343932597,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.913481791004213e-05,
+      "loss": 0.843,
+      "step": 49135
+    },
+    {
+      "epoch": 3.338768854463922,
+      "grad_norm": 9.75,
+      "learning_rate": 2.913269465960049e-05,
+      "loss": 0.9462,
+      "step": 49140
+    },
+    {
+      "epoch": 3.3391085745345834,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.913057140915885e-05,
+      "loss": 0.9234,
+      "step": 49145
+    },
+    {
+      "epoch": 3.339448294605245,
+      "grad_norm": 8.625,
+      "learning_rate": 2.912844815871722e-05,
+      "loss": 0.8486,
+      "step": 49150
+    },
+    {
+      "epoch": 3.339788014675907,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.9126324908275582e-05,
+      "loss": 0.8493,
+      "step": 49155
+    },
+    {
+      "epoch": 3.3401277347465688,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.9124201657833943e-05,
+      "loss": 1.0004,
+      "step": 49160
+    },
+    {
+      "epoch": 3.3404674548172304,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.912207840739231e-05,
+      "loss": 0.9274,
+      "step": 49165
+    },
+    {
+      "epoch": 3.3408071748878925,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.9119955156950674e-05,
+      "loss": 0.8914,
+      "step": 49170
+    },
+    {
+      "epoch": 3.341146894958554,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.9117831906509035e-05,
+      "loss": 0.8553,
+      "step": 49175
+    },
+    {
+      "epoch": 3.3414866150292157,
+      "grad_norm": 8.0,
+      "learning_rate": 2.9115708656067402e-05,
+      "loss": 0.9395,
+      "step": 49180
+    },
+    {
+      "epoch": 3.341826335099878,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.9113585405625766e-05,
+      "loss": 0.8326,
+      "step": 49185
+    },
+    {
+      "epoch": 3.3421660551705394,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.9111462155184127e-05,
+      "loss": 0.9332,
+      "step": 49190
+    },
+    {
+      "epoch": 3.342505775241201,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.9109338904742494e-05,
+      "loss": 0.8445,
+      "step": 49195
+    },
+    {
+      "epoch": 3.342845495311863,
+      "grad_norm": 8.5,
+      "learning_rate": 2.9107215654300855e-05,
+      "loss": 0.8554,
+      "step": 49200
+    },
+    {
+      "epoch": 3.3431852153825248,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.910509240385922e-05,
+      "loss": 0.8555,
+      "step": 49205
+    },
+    {
+      "epoch": 3.3435249354531864,
+      "grad_norm": 9.0,
+      "learning_rate": 2.9102969153417586e-05,
+      "loss": 0.9499,
+      "step": 49210
+    },
+    {
+      "epoch": 3.3438646555238485,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.9100845902975947e-05,
+      "loss": 0.8341,
+      "step": 49215
+    },
+    {
+      "epoch": 3.34420437559451,
+      "grad_norm": 7.875,
+      "learning_rate": 2.909872265253431e-05,
+      "loss": 0.8454,
+      "step": 49220
+    },
+    {
+      "epoch": 3.3445440956651717,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.909659940209268e-05,
+      "loss": 0.894,
+      "step": 49225
+    },
+    {
+      "epoch": 3.344883815735834,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.909447615165104e-05,
+      "loss": 0.8962,
+      "step": 49230
+    },
+    {
+      "epoch": 3.3452235358064955,
+      "grad_norm": 9.375,
+      "learning_rate": 2.9092352901209407e-05,
+      "loss": 0.9072,
+      "step": 49235
+    },
+    {
+      "epoch": 3.345563255877157,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.909022965076777e-05,
+      "loss": 0.9154,
+      "step": 49240
+    },
+    {
+      "epoch": 3.345902975947819,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.908810640032613e-05,
+      "loss": 1.0233,
+      "step": 49245
+    },
+    {
+      "epoch": 3.346242696018481,
+      "grad_norm": 7.5,
+      "learning_rate": 2.90859831498845e-05,
+      "loss": 0.8491,
+      "step": 49250
+    },
+    {
+      "epoch": 3.3465824160891424,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.9083859899442863e-05,
+      "loss": 0.8813,
+      "step": 49255
+    },
+    {
+      "epoch": 3.3469221361598045,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.9081736649001223e-05,
+      "loss": 0.8803,
+      "step": 49260
+    },
+    {
+      "epoch": 3.347261856230466,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.907961339855959e-05,
+      "loss": 0.9566,
+      "step": 49265
+    },
+    {
+      "epoch": 3.3476015763011278,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9077490148117955e-05,
+      "loss": 0.9542,
+      "step": 49270
+    },
+    {
+      "epoch": 3.34794129637179,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.9075366897676315e-05,
+      "loss": 0.9956,
+      "step": 49275
+    },
+    {
+      "epoch": 3.3482810164424515,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.9073243647234683e-05,
+      "loss": 0.9057,
+      "step": 49280
+    },
+    {
+      "epoch": 3.348620736513113,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9071120396793043e-05,
+      "loss": 0.8735,
+      "step": 49285
+    },
+    {
+      "epoch": 3.348960456583775,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.9068997146351407e-05,
+      "loss": 0.9139,
+      "step": 49290
+    },
+    {
+      "epoch": 3.349300176654437,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9066873895909775e-05,
+      "loss": 0.9812,
+      "step": 49295
+    },
+    {
+      "epoch": 3.3496398967250984,
+      "grad_norm": 7.875,
+      "learning_rate": 2.9064750645468135e-05,
+      "loss": 0.851,
+      "step": 49300
+    },
+    {
+      "epoch": 3.3499796167957605,
+      "grad_norm": 6.875,
+      "learning_rate": 2.90626273950265e-05,
+      "loss": 0.8964,
+      "step": 49305
+    },
+    {
+      "epoch": 3.350319336866422,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.9060504144584867e-05,
+      "loss": 0.8476,
+      "step": 49310
+    },
+    {
+      "epoch": 3.3506590569370838,
+      "grad_norm": 21.125,
+      "learning_rate": 2.9058380894143227e-05,
+      "loss": 0.8828,
+      "step": 49315
+    },
+    {
+      "epoch": 3.350998777007746,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.905625764370159e-05,
+      "loss": 0.9612,
+      "step": 49320
+    },
+    {
+      "epoch": 3.3513384970784075,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.905413439325996e-05,
+      "loss": 0.9428,
+      "step": 49325
+    },
+    {
+      "epoch": 3.351678217149069,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.905201114281832e-05,
+      "loss": 0.8895,
+      "step": 49330
+    },
+    {
+      "epoch": 3.352017937219731,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.9049887892376683e-05,
+      "loss": 0.8999,
+      "step": 49335
+    },
+    {
+      "epoch": 3.352357657290393,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.904776464193505e-05,
+      "loss": 0.8863,
+      "step": 49340
+    },
+    {
+      "epoch": 3.3526973773610544,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.904564139149341e-05,
+      "loss": 0.8803,
+      "step": 49345
+    },
+    {
+      "epoch": 3.353037097431716,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.9043518141051772e-05,
+      "loss": 0.8366,
+      "step": 49350
+    },
+    {
+      "epoch": 3.353376817502378,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.904139489061014e-05,
+      "loss": 0.924,
+      "step": 49355
+    },
+    {
+      "epoch": 3.35371653757304,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.9039271640168503e-05,
+      "loss": 0.7938,
+      "step": 49360
+    },
+    {
+      "epoch": 3.3540562576437014,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.9037148389726864e-05,
+      "loss": 0.9227,
+      "step": 49365
+    },
+    {
+      "epoch": 3.3543959777143635,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.903502513928523e-05,
+      "loss": 0.925,
+      "step": 49370
+    },
+    {
+      "epoch": 3.354735697785025,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.9032901888843595e-05,
+      "loss": 0.9038,
+      "step": 49375
+    },
+    {
+      "epoch": 3.3550754178556867,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9030778638401956e-05,
+      "loss": 0.9044,
+      "step": 49380
+    },
+    {
+      "epoch": 3.355415137926349,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.9028655387960323e-05,
+      "loss": 0.876,
+      "step": 49385
+    },
+    {
+      "epoch": 3.3557548579970105,
+      "grad_norm": 7.0,
+      "learning_rate": 2.9026532137518687e-05,
+      "loss": 0.9155,
+      "step": 49390
+    },
+    {
+      "epoch": 3.356094578067672,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9024408887077048e-05,
+      "loss": 0.9592,
+      "step": 49395
+    },
+    {
+      "epoch": 3.356434298138334,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.9022285636635415e-05,
+      "loss": 0.8604,
+      "step": 49400
+    },
+    {
+      "epoch": 3.356774018208996,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.902016238619378e-05,
+      "loss": 0.9143,
+      "step": 49405
+    },
+    {
+      "epoch": 3.3571137382796574,
+      "grad_norm": 7.875,
+      "learning_rate": 2.901803913575214e-05,
+      "loss": 0.9417,
+      "step": 49410
+    },
+    {
+      "epoch": 3.3574534583503195,
+      "grad_norm": 7.25,
+      "learning_rate": 2.9015915885310507e-05,
+      "loss": 0.9187,
+      "step": 49415
+    },
+    {
+      "epoch": 3.357793178420981,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9013792634868868e-05,
+      "loss": 0.8473,
+      "step": 49420
+    },
+    {
+      "epoch": 3.3581328984916428,
+      "grad_norm": 7.5,
+      "learning_rate": 2.9011669384427232e-05,
+      "loss": 1.0255,
+      "step": 49425
+    },
+    {
+      "epoch": 3.358472618562305,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.90095461339856e-05,
+      "loss": 0.8291,
+      "step": 49430
+    },
+    {
+      "epoch": 3.3588123386329665,
+      "grad_norm": 7.25,
+      "learning_rate": 2.900742288354396e-05,
+      "loss": 0.8055,
+      "step": 49435
+    },
+    {
+      "epoch": 3.359152058703628,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.9005299633102324e-05,
+      "loss": 0.8612,
+      "step": 49440
+    },
+    {
+      "epoch": 3.35949177877429,
+      "grad_norm": 7.375,
+      "learning_rate": 2.900317638266069e-05,
+      "loss": 0.8096,
+      "step": 49445
+    },
+    {
+      "epoch": 3.359831498844952,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.9001053132219052e-05,
+      "loss": 0.801,
+      "step": 49450
+    },
+    {
+      "epoch": 3.3601712189156134,
+      "grad_norm": 7.0,
+      "learning_rate": 2.8998929881777416e-05,
+      "loss": 0.9369,
+      "step": 49455
+    },
+    {
+      "epoch": 3.360510938986275,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.8996806631335783e-05,
+      "loss": 0.8485,
+      "step": 49460
+    },
+    {
+      "epoch": 3.360850659056937,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8994683380894144e-05,
+      "loss": 0.7864,
+      "step": 49465
+    },
+    {
+      "epoch": 3.3611903791275988,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.8992560130452508e-05,
+      "loss": 0.8627,
+      "step": 49470
+    },
+    {
+      "epoch": 3.3615300991982604,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8990436880010875e-05,
+      "loss": 0.9291,
+      "step": 49475
+    },
+    {
+      "epoch": 3.3618698192689225,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.8988313629569236e-05,
+      "loss": 0.8774,
+      "step": 49480
+    },
+    {
+      "epoch": 3.362209539339584,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.8986190379127597e-05,
+      "loss": 0.8483,
+      "step": 49485
+    },
+    {
+      "epoch": 3.3625492594102457,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8984067128685964e-05,
+      "loss": 0.9213,
+      "step": 49490
+    },
+    {
+      "epoch": 3.362888979480908,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8981943878244328e-05,
+      "loss": 0.8402,
+      "step": 49495
+    },
+    {
+      "epoch": 3.3632286995515694,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.897982062780269e-05,
+      "loss": 0.8537,
+      "step": 49500
+    },
+    {
+      "epoch": 3.363568419622231,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8977697377361056e-05,
+      "loss": 0.9726,
+      "step": 49505
+    },
+    {
+      "epoch": 3.363908139692893,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.897557412691942e-05,
+      "loss": 0.8223,
+      "step": 49510
+    },
+    {
+      "epoch": 3.364247859763555,
+      "grad_norm": 8.125,
+      "learning_rate": 2.897345087647778e-05,
+      "loss": 0.8871,
+      "step": 49515
+    },
+    {
+      "epoch": 3.3645875798342164,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8971327626036148e-05,
+      "loss": 0.7448,
+      "step": 49520
+    },
+    {
+      "epoch": 3.3649272999048785,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8969204375594512e-05,
+      "loss": 0.8306,
+      "step": 49525
+    },
+    {
+      "epoch": 3.36526701997554,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.8967081125152873e-05,
+      "loss": 0.8943,
+      "step": 49530
+    },
+    {
+      "epoch": 3.3656067400462018,
+      "grad_norm": 7.5,
+      "learning_rate": 2.896495787471124e-05,
+      "loss": 0.9472,
+      "step": 49535
+    },
+    {
+      "epoch": 3.365946460116864,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8962834624269604e-05,
+      "loss": 0.9294,
+      "step": 49540
+    },
+    {
+      "epoch": 3.3662861801875255,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8960711373827965e-05,
+      "loss": 0.8538,
+      "step": 49545
+    },
+    {
+      "epoch": 3.366625900258187,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8958588123386332e-05,
+      "loss": 0.91,
+      "step": 49550
+    },
+    {
+      "epoch": 3.366965620328849,
+      "grad_norm": 6.75,
+      "learning_rate": 2.8956464872944693e-05,
+      "loss": 1.0003,
+      "step": 49555
+    },
+    {
+      "epoch": 3.367305340399511,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8954341622503057e-05,
+      "loss": 0.8169,
+      "step": 49560
+    },
+    {
+      "epoch": 3.3676450604701724,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.8952218372061424e-05,
+      "loss": 0.8448,
+      "step": 49565
+    },
+    {
+      "epoch": 3.3679847805408345,
+      "grad_norm": 6.75,
+      "learning_rate": 2.8950095121619785e-05,
+      "loss": 0.8303,
+      "step": 49570
+    },
+    {
+      "epoch": 3.368324500611496,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8947971871178152e-05,
+      "loss": 0.8604,
+      "step": 49575
+    },
+    {
+      "epoch": 3.3686642206821578,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8945848620736516e-05,
+      "loss": 0.9118,
+      "step": 49580
+    },
+    {
+      "epoch": 3.36900394075282,
+      "grad_norm": 7.0,
+      "learning_rate": 2.8943725370294877e-05,
+      "loss": 0.8079,
+      "step": 49585
+    },
+    {
+      "epoch": 3.3693436608234815,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8941602119853244e-05,
+      "loss": 0.9056,
+      "step": 49590
+    },
+    {
+      "epoch": 3.369683380894143,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8939478869411608e-05,
+      "loss": 0.9141,
+      "step": 49595
+    },
+    {
+      "epoch": 3.370023100964805,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.893735561896997e-05,
+      "loss": 0.9157,
+      "step": 49600
+    },
+    {
+      "epoch": 3.370362821035467,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8935232368528336e-05,
+      "loss": 0.9213,
+      "step": 49605
+    },
+    {
+      "epoch": 3.3707025411061284,
+      "grad_norm": 8.0,
+      "learning_rate": 2.89331091180867e-05,
+      "loss": 0.9044,
+      "step": 49610
+    },
+    {
+      "epoch": 3.3710422611767905,
+      "grad_norm": 6.625,
+      "learning_rate": 2.893098586764506e-05,
+      "loss": 0.9458,
+      "step": 49615
+    },
+    {
+      "epoch": 3.371381981247452,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.8928862617203428e-05,
+      "loss": 0.8665,
+      "step": 49620
+    },
+    {
+      "epoch": 3.3717217013181138,
+      "grad_norm": 7.875,
+      "learning_rate": 2.892673936676179e-05,
+      "loss": 0.8749,
+      "step": 49625
+    },
+    {
+      "epoch": 3.372061421388776,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8924616116320153e-05,
+      "loss": 1.0398,
+      "step": 49630
+    },
+    {
+      "epoch": 3.3724011414594375,
+      "grad_norm": 7.375,
+      "learning_rate": 2.892249286587852e-05,
+      "loss": 0.8717,
+      "step": 49635
+    },
+    {
+      "epoch": 3.372740861530099,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.892036961543688e-05,
+      "loss": 0.8334,
+      "step": 49640
+    },
+    {
+      "epoch": 3.373080581600761,
+      "grad_norm": 7.875,
+      "learning_rate": 2.8918246364995245e-05,
+      "loss": 0.9244,
+      "step": 49645
+    },
+    {
+      "epoch": 3.373420301671423,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.8916123114553612e-05,
+      "loss": 0.8858,
+      "step": 49650
+    },
+    {
+      "epoch": 3.3737600217420844,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.8913999864111973e-05,
+      "loss": 0.8823,
+      "step": 49655
+    },
+    {
+      "epoch": 3.3740997418127465,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.8911876613670337e-05,
+      "loss": 0.8562,
+      "step": 49660
+    },
+    {
+      "epoch": 3.374439461883408,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8909753363228704e-05,
+      "loss": 0.8484,
+      "step": 49665
+    },
+    {
+      "epoch": 3.37477918195407,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8907630112787065e-05,
+      "loss": 0.9054,
+      "step": 49670
+    },
+    {
+      "epoch": 3.375118902024732,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.890550686234543e-05,
+      "loss": 0.8436,
+      "step": 49675
+    },
+    {
+      "epoch": 3.3754586220953935,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8903383611903796e-05,
+      "loss": 0.934,
+      "step": 49680
+    },
+    {
+      "epoch": 3.375798342166055,
+      "grad_norm": 7.75,
+      "learning_rate": 2.8901260361462157e-05,
+      "loss": 0.9434,
+      "step": 49685
+    },
+    {
+      "epoch": 3.3761380622367168,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8899137111020517e-05,
+      "loss": 0.9086,
+      "step": 49690
+    },
+    {
+      "epoch": 3.376477782307379,
+      "grad_norm": 5.15625,
+      "learning_rate": 2.8897013860578888e-05,
+      "loss": 0.8729,
+      "step": 49695
+    },
+    {
+      "epoch": 3.3768175023780405,
+      "grad_norm": 7.875,
+      "learning_rate": 2.889489061013725e-05,
+      "loss": 0.9048,
+      "step": 49700
+    },
+    {
+      "epoch": 3.377157222448702,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.889276735969561e-05,
+      "loss": 0.9723,
+      "step": 49705
+    },
+    {
+      "epoch": 3.377496942519364,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8890644109253977e-05,
+      "loss": 0.8416,
+      "step": 49710
+    },
+    {
+      "epoch": 3.377836662590026,
+      "grad_norm": 8.375,
+      "learning_rate": 2.888852085881234e-05,
+      "loss": 0.9269,
+      "step": 49715
+    },
+    {
+      "epoch": 3.3781763826606874,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.88863976083707e-05,
+      "loss": 0.9449,
+      "step": 49720
+    },
+    {
+      "epoch": 3.3785161027313495,
+      "grad_norm": 7.0,
+      "learning_rate": 2.888427435792907e-05,
+      "loss": 0.9301,
+      "step": 49725
+    },
+    {
+      "epoch": 3.378855822802011,
+      "grad_norm": 7.5,
+      "learning_rate": 2.8882151107487433e-05,
+      "loss": 0.908,
+      "step": 49730
+    },
+    {
+      "epoch": 3.3791955428726728,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8880027857045793e-05,
+      "loss": 0.8825,
+      "step": 49735
+    },
+    {
+      "epoch": 3.379535262943335,
+      "grad_norm": 7.25,
+      "learning_rate": 2.887790460660416e-05,
+      "loss": 0.972,
+      "step": 49740
+    },
+    {
+      "epoch": 3.3798749830139965,
+      "grad_norm": 7.25,
+      "learning_rate": 2.8875781356162525e-05,
+      "loss": 1.0044,
+      "step": 49745
+    },
+    {
+      "epoch": 3.380214703084658,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8873658105720885e-05,
+      "loss": 0.9337,
+      "step": 49750
+    },
+    {
+      "epoch": 3.38055442315532,
+      "grad_norm": 7.0,
+      "learning_rate": 2.8871534855279253e-05,
+      "loss": 0.8793,
+      "step": 49755
+    },
+    {
+      "epoch": 3.380894143225982,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.8869411604837617e-05,
+      "loss": 0.8437,
+      "step": 49760
+    },
+    {
+      "epoch": 3.3812338632966434,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8867288354395977e-05,
+      "loss": 0.8775,
+      "step": 49765
+    },
+    {
+      "epoch": 3.3815735833673055,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8865165103954345e-05,
+      "loss": 0.9983,
+      "step": 49770
+    },
+    {
+      "epoch": 3.381913303437967,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8863041853512705e-05,
+      "loss": 0.9296,
+      "step": 49775
+    },
+    {
+      "epoch": 3.3822530235086288,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.886091860307107e-05,
+      "loss": 0.9143,
+      "step": 49780
+    },
+    {
+      "epoch": 3.382592743579291,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.8858795352629437e-05,
+      "loss": 0.8781,
+      "step": 49785
+    },
+    {
+      "epoch": 3.3829324636499525,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.8856672102187797e-05,
+      "loss": 0.7901,
+      "step": 49790
+    },
+    {
+      "epoch": 3.383272183720614,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.885454885174616e-05,
+      "loss": 0.9687,
+      "step": 49795
+    },
+    {
+      "epoch": 3.3836119037912757,
+      "grad_norm": 7.125,
+      "learning_rate": 2.885242560130453e-05,
+      "loss": 0.8753,
+      "step": 49800
+    },
+    {
+      "epoch": 3.383951623861938,
+      "grad_norm": 8.125,
+      "learning_rate": 2.885030235086289e-05,
+      "loss": 0.8744,
+      "step": 49805
+    },
+    {
+      "epoch": 3.3842913439325994,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8848179100421253e-05,
+      "loss": 0.9778,
+      "step": 49810
+    },
+    {
+      "epoch": 3.384631064003261,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.884605584997962e-05,
+      "loss": 0.9479,
+      "step": 49815
+    },
+    {
+      "epoch": 3.384970784073923,
+      "grad_norm": 6.5,
+      "learning_rate": 2.884393259953798e-05,
+      "loss": 0.8003,
+      "step": 49820
+    },
+    {
+      "epoch": 3.385310504144585,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.8841809349096342e-05,
+      "loss": 0.9565,
+      "step": 49825
+    },
+    {
+      "epoch": 3.3856502242152464,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8839686098654713e-05,
+      "loss": 0.9436,
+      "step": 49830
+    },
+    {
+      "epoch": 3.3859899442859085,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8837562848213073e-05,
+      "loss": 0.9679,
+      "step": 49835
+    },
+    {
+      "epoch": 3.38632966435657,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8835439597771434e-05,
+      "loss": 0.8813,
+      "step": 49840
+    },
+    {
+      "epoch": 3.3866693844272318,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.88333163473298e-05,
+      "loss": 0.8972,
+      "step": 49845
+    },
+    {
+      "epoch": 3.387009104497894,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.8831193096888165e-05,
+      "loss": 0.9091,
+      "step": 49850
+    },
+    {
+      "epoch": 3.3873488245685555,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8829069846446526e-05,
+      "loss": 0.9171,
+      "step": 49855
+    },
+    {
+      "epoch": 3.387688544639217,
+      "grad_norm": 6.75,
+      "learning_rate": 2.8826946596004893e-05,
+      "loss": 0.9081,
+      "step": 49860
+    },
+    {
+      "epoch": 3.388028264709879,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.8824823345563257e-05,
+      "loss": 0.878,
+      "step": 49865
+    },
+    {
+      "epoch": 3.388367984780541,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.8822700095121618e-05,
+      "loss": 0.9444,
+      "step": 49870
+    },
+    {
+      "epoch": 3.3887077048512024,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.8820576844679985e-05,
+      "loss": 0.9526,
+      "step": 49875
+    },
+    {
+      "epoch": 3.3890474249218645,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.881845359423835e-05,
+      "loss": 0.8068,
+      "step": 49880
+    },
+    {
+      "epoch": 3.389387144992526,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.881633034379671e-05,
+      "loss": 0.9169,
+      "step": 49885
+    },
+    {
+      "epoch": 3.3897268650631878,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8814207093355077e-05,
+      "loss": 0.8857,
+      "step": 49890
+    },
+    {
+      "epoch": 3.39006658513385,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.881208384291344e-05,
+      "loss": 0.9917,
+      "step": 49895
+    },
+    {
+      "epoch": 3.3904063052045115,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8809960592471802e-05,
+      "loss": 0.8847,
+      "step": 49900
+    },
+    {
+      "epoch": 3.390746025275173,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.880783734203017e-05,
+      "loss": 0.8976,
+      "step": 49905
+    },
+    {
+      "epoch": 3.391085745345835,
+      "grad_norm": 8.0,
+      "learning_rate": 2.880571409158853e-05,
+      "loss": 0.9681,
+      "step": 49910
+    },
+    {
+      "epoch": 3.391425465416497,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.8803590841146897e-05,
+      "loss": 0.8801,
+      "step": 49915
+    },
+    {
+      "epoch": 3.3917651854871584,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.880146759070526e-05,
+      "loss": 0.8893,
+      "step": 49920
+    },
+    {
+      "epoch": 3.3921049055578205,
+      "grad_norm": 6.625,
+      "learning_rate": 2.8799344340263622e-05,
+      "loss": 0.8785,
+      "step": 49925
+    },
+    {
+      "epoch": 3.392444625628482,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.879722108982199e-05,
+      "loss": 0.9124,
+      "step": 49930
+    },
+    {
+      "epoch": 3.3927843456991438,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8795097839380353e-05,
+      "loss": 0.8083,
+      "step": 49935
+    },
+    {
+      "epoch": 3.393124065769806,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.8792974588938714e-05,
+      "loss": 0.8848,
+      "step": 49940
+    },
+    {
+      "epoch": 3.3934637858404675,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.879085133849708e-05,
+      "loss": 0.8845,
+      "step": 49945
+    },
+    {
+      "epoch": 3.393803505911129,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.8788728088055446e-05,
+      "loss": 0.8844,
+      "step": 49950
+    },
+    {
+      "epoch": 3.394143225981791,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8786604837613806e-05,
+      "loss": 0.9592,
+      "step": 49955
+    },
+    {
+      "epoch": 3.394482946052453,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.8784481587172174e-05,
+      "loss": 1.0551,
+      "step": 49960
+    },
+    {
+      "epoch": 3.3948226661231145,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8782358336730538e-05,
+      "loss": 0.9433,
+      "step": 49965
+    },
+    {
+      "epoch": 3.3951623861937765,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.8780235086288898e-05,
+      "loss": 0.8412,
+      "step": 49970
+    },
+    {
+      "epoch": 3.395502106264438,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8778111835847266e-05,
+      "loss": 0.8665,
+      "step": 49975
+    },
+    {
+      "epoch": 3.3958418263351,
+      "grad_norm": 6.09375,
+      "learning_rate": 2.8775988585405626e-05,
+      "loss": 0.9048,
+      "step": 49980
+    },
+    {
+      "epoch": 3.396181546405762,
+      "grad_norm": 7.0,
+      "learning_rate": 2.877386533496399e-05,
+      "loss": 0.8836,
+      "step": 49985
+    },
+    {
+      "epoch": 3.3965212664764235,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8771742084522358e-05,
+      "loss": 0.9542,
+      "step": 49990
+    },
+    {
+      "epoch": 3.396860986547085,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8769618834080718e-05,
+      "loss": 0.9358,
+      "step": 49995
+    },
+    {
+      "epoch": 3.397200706617747,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.8767495583639082e-05,
+      "loss": 0.9473,
+      "step": 50000
+    },
+    {
+      "epoch": 3.397540426688409,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.876537233319745e-05,
+      "loss": 1.0096,
+      "step": 50005
+    },
+    {
+      "epoch": 3.3978801467590705,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.876324908275581e-05,
+      "loss": 0.8169,
+      "step": 50010
+    },
+    {
+      "epoch": 3.3982198668297325,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8761125832314174e-05,
+      "loss": 1.0047,
+      "step": 50015
+    },
+    {
+      "epoch": 3.398559586900394,
+      "grad_norm": 8.875,
+      "learning_rate": 2.875900258187254e-05,
+      "loss": 0.924,
+      "step": 50020
+    },
+    {
+      "epoch": 3.398899306971056,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8756879331430902e-05,
+      "loss": 0.9382,
+      "step": 50025
+    },
+    {
+      "epoch": 3.3992390270417174,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8754756080989266e-05,
+      "loss": 0.9619,
+      "step": 50030
+    },
+    {
+      "epoch": 3.3995787471123795,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8752632830547634e-05,
+      "loss": 0.9043,
+      "step": 50035
+    },
+    {
+      "epoch": 3.399918467183041,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8750509580105994e-05,
+      "loss": 1.0246,
+      "step": 50040
+    },
+    {
+      "epoch": 3.4002581872537028,
+      "grad_norm": 6.875,
+      "learning_rate": 2.8748386329664355e-05,
+      "loss": 0.8497,
+      "step": 50045
+    },
+    {
+      "epoch": 3.400597907324365,
+      "grad_norm": 7.5,
+      "learning_rate": 2.8746263079222722e-05,
+      "loss": 0.8576,
+      "step": 50050
+    },
+    {
+      "epoch": 3.4009376273950265,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8744139828781086e-05,
+      "loss": 0.9443,
+      "step": 50055
+    },
+    {
+      "epoch": 3.401277347465688,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8742016578339447e-05,
+      "loss": 0.9388,
+      "step": 50060
+    },
+    {
+      "epoch": 3.40161706753635,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8739893327897814e-05,
+      "loss": 0.8975,
+      "step": 50065
+    },
+    {
+      "epoch": 3.401956787607012,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.8737770077456178e-05,
+      "loss": 0.9759,
+      "step": 50070
+    },
+    {
+      "epoch": 3.4022965076776734,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.873564682701454e-05,
+      "loss": 0.8867,
+      "step": 50075
+    },
+    {
+      "epoch": 3.4026362277483355,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.8733523576572906e-05,
+      "loss": 0.7312,
+      "step": 50080
+    },
+    {
+      "epoch": 3.402975947818997,
+      "grad_norm": 7.25,
+      "learning_rate": 2.873140032613127e-05,
+      "loss": 0.8876,
+      "step": 50085
+    },
+    {
+      "epoch": 3.403315667889659,
+      "grad_norm": 7.5,
+      "learning_rate": 2.872927707568963e-05,
+      "loss": 0.8796,
+      "step": 50090
+    },
+    {
+      "epoch": 3.403655387960321,
+      "grad_norm": 7.25,
+      "learning_rate": 2.8727153825247998e-05,
+      "loss": 0.933,
+      "step": 50095
+    },
+    {
+      "epoch": 3.4039951080309825,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.8725030574806362e-05,
+      "loss": 0.8482,
+      "step": 50100
+    },
+    {
+      "epoch": 3.404334828101644,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8722907324364723e-05,
+      "loss": 0.8203,
+      "step": 50105
+    },
+    {
+      "epoch": 3.404674548172306,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.872078407392309e-05,
+      "loss": 0.9682,
+      "step": 50110
+    },
+    {
+      "epoch": 3.405014268242968,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.871866082348145e-05,
+      "loss": 0.9909,
+      "step": 50115
+    },
+    {
+      "epoch": 3.4053539883136295,
+      "grad_norm": 7.0,
+      "learning_rate": 2.8716537573039815e-05,
+      "loss": 0.8846,
+      "step": 50120
+    },
+    {
+      "epoch": 3.4056937083842915,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8714414322598182e-05,
+      "loss": 0.9139,
+      "step": 50125
+    },
+    {
+      "epoch": 3.406033428454953,
+      "grad_norm": 6.375,
+      "learning_rate": 2.8712291072156543e-05,
+      "loss": 0.9327,
+      "step": 50130
+    },
+    {
+      "epoch": 3.406373148525615,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8710167821714907e-05,
+      "loss": 0.8196,
+      "step": 50135
+    },
+    {
+      "epoch": 3.4067128685962764,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8708044571273274e-05,
+      "loss": 0.9566,
+      "step": 50140
+    },
+    {
+      "epoch": 3.4070525886669385,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8705921320831635e-05,
+      "loss": 0.9157,
+      "step": 50145
+    },
+    {
+      "epoch": 3.4073923087376,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.870379807039e-05,
+      "loss": 0.9349,
+      "step": 50150
+    },
+    {
+      "epoch": 3.4077320288082618,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8701674819948366e-05,
+      "loss": 0.9072,
+      "step": 50155
+    },
+    {
+      "epoch": 3.408071748878924,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8699551569506727e-05,
+      "loss": 1.0085,
+      "step": 50160
+    },
+    {
+      "epoch": 3.4084114689495855,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.869742831906509e-05,
+      "loss": 0.9611,
+      "step": 50165
+    },
+    {
+      "epoch": 3.408751189020247,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8695305068623458e-05,
+      "loss": 0.8875,
+      "step": 50170
+    },
+    {
+      "epoch": 3.409090909090909,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.869318181818182e-05,
+      "loss": 0.8925,
+      "step": 50175
+    },
+    {
+      "epoch": 3.409430629161571,
+      "grad_norm": 7.875,
+      "learning_rate": 2.869105856774018e-05,
+      "loss": 0.8916,
+      "step": 50180
+    },
+    {
+      "epoch": 3.4097703492322324,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.868893531729855e-05,
+      "loss": 0.9577,
+      "step": 50185
+    },
+    {
+      "epoch": 3.4101100693028945,
+      "grad_norm": 7.5,
+      "learning_rate": 2.868681206685691e-05,
+      "loss": 0.9064,
+      "step": 50190
+    },
+    {
+      "epoch": 3.410449789373556,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.868468881641527e-05,
+      "loss": 0.9459,
+      "step": 50195
+    },
+    {
+      "epoch": 3.4107895094442178,
+      "grad_norm": 7.5,
+      "learning_rate": 2.868256556597364e-05,
+      "loss": 0.9744,
+      "step": 50200
+    },
+    {
+      "epoch": 3.41112922951488,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.8680442315532003e-05,
+      "loss": 0.9825,
+      "step": 50205
+    },
+    {
+      "epoch": 3.4114689495855415,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8678319065090364e-05,
+      "loss": 0.8391,
+      "step": 50210
+    },
+    {
+      "epoch": 3.411808669656203,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.867619581464873e-05,
+      "loss": 0.9743,
+      "step": 50215
+    },
+    {
+      "epoch": 3.412148389726865,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8674072564207095e-05,
+      "loss": 0.9152,
+      "step": 50220
+    },
+    {
+      "epoch": 3.412488109797527,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.8671949313765456e-05,
+      "loss": 0.9335,
+      "step": 50225
+    },
+    {
+      "epoch": 3.4128278298681884,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.8669826063323823e-05,
+      "loss": 0.9359,
+      "step": 50230
+    },
+    {
+      "epoch": 3.4131675499388505,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.8667702812882187e-05,
+      "loss": 0.8657,
+      "step": 50235
+    },
+    {
+      "epoch": 3.413507270009512,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.8665579562440548e-05,
+      "loss": 0.8465,
+      "step": 50240
+    },
+    {
+      "epoch": 3.413846990080174,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8663456311998915e-05,
+      "loss": 0.8478,
+      "step": 50245
+    },
+    {
+      "epoch": 3.414186710150836,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.8661333061557276e-05,
+      "loss": 0.8589,
+      "step": 50250
+    },
+    {
+      "epoch": 3.4145264302214975,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.8659209811115646e-05,
+      "loss": 0.9075,
+      "step": 50255
+    },
+    {
+      "epoch": 3.414866150292159,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8657086560674007e-05,
+      "loss": 0.9432,
+      "step": 50260
+    },
+    {
+      "epoch": 3.415205870362821,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.8654963310232368e-05,
+      "loss": 1.0046,
+      "step": 50265
+    },
+    {
+      "epoch": 3.415545590433483,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.8652840059790735e-05,
+      "loss": 0.9431,
+      "step": 50270
+    },
+    {
+      "epoch": 3.4158853105041445,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.86507168093491e-05,
+      "loss": 0.8637,
+      "step": 50275
+    },
+    {
+      "epoch": 3.4162250305748065,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.864859355890746e-05,
+      "loss": 0.9512,
+      "step": 50280
+    },
+    {
+      "epoch": 3.416564750645468,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8646470308465827e-05,
+      "loss": 0.8354,
+      "step": 50285
+    },
+    {
+      "epoch": 3.41690447071613,
+      "grad_norm": 7.0,
+      "learning_rate": 2.864434705802419e-05,
+      "loss": 0.9274,
+      "step": 50290
+    },
+    {
+      "epoch": 3.417244190786792,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.864222380758255e-05,
+      "loss": 0.9868,
+      "step": 50295
+    },
+    {
+      "epoch": 3.4175839108574535,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.864010055714092e-05,
+      "loss": 0.9081,
+      "step": 50300
+    },
+    {
+      "epoch": 3.417923630928115,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8637977306699283e-05,
+      "loss": 0.9015,
+      "step": 50305
+    },
+    {
+      "epoch": 3.418263350998777,
+      "grad_norm": 7.75,
+      "learning_rate": 2.8635854056257644e-05,
+      "loss": 0.9395,
+      "step": 50310
+    },
+    {
+      "epoch": 3.418603071069439,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.863373080581601e-05,
+      "loss": 0.8616,
+      "step": 50315
+    },
+    {
+      "epoch": 3.4189427911401005,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8631607555374375e-05,
+      "loss": 0.9444,
+      "step": 50320
+    },
+    {
+      "epoch": 3.4192825112107625,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8629484304932736e-05,
+      "loss": 0.8856,
+      "step": 50325
+    },
+    {
+      "epoch": 3.419622231281424,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8627361054491103e-05,
+      "loss": 0.9599,
+      "step": 50330
+    },
+    {
+      "epoch": 3.419961951352086,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8625237804049464e-05,
+      "loss": 0.8937,
+      "step": 50335
+    },
+    {
+      "epoch": 3.420301671422748,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8623114553607828e-05,
+      "loss": 0.8989,
+      "step": 50340
+    },
+    {
+      "epoch": 3.4206413914934095,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.8620991303166195e-05,
+      "loss": 1.0124,
+      "step": 50345
+    },
+    {
+      "epoch": 3.420981111564071,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8618868052724556e-05,
+      "loss": 0.8716,
+      "step": 50350
+    },
+    {
+      "epoch": 3.421320831634733,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.861674480228292e-05,
+      "loss": 0.9306,
+      "step": 50355
+    },
+    {
+      "epoch": 3.421660551705395,
+      "grad_norm": 7.0,
+      "learning_rate": 2.8614621551841287e-05,
+      "loss": 0.9105,
+      "step": 50360
+    },
+    {
+      "epoch": 3.4220002717760565,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8612498301399648e-05,
+      "loss": 0.8154,
+      "step": 50365
+    },
+    {
+      "epoch": 3.422339991846718,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.861037505095801e-05,
+      "loss": 0.9431,
+      "step": 50370
+    },
+    {
+      "epoch": 3.42267971191738,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.860825180051638e-05,
+      "loss": 0.9046,
+      "step": 50375
+    },
+    {
+      "epoch": 3.423019431988042,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.860612855007474e-05,
+      "loss": 0.81,
+      "step": 50380
+    },
+    {
+      "epoch": 3.4233591520587034,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.8604005299633104e-05,
+      "loss": 0.9564,
+      "step": 50385
+    },
+    {
+      "epoch": 3.4236988721293655,
+      "grad_norm": 8.125,
+      "learning_rate": 2.860188204919147e-05,
+      "loss": 0.8328,
+      "step": 50390
+    },
+    {
+      "epoch": 3.424038592200027,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.859975879874983e-05,
+      "loss": 0.9186,
+      "step": 50395
+    },
+    {
+      "epoch": 3.424378312270689,
+      "grad_norm": 6.75,
+      "learning_rate": 2.8597635548308192e-05,
+      "loss": 0.933,
+      "step": 50400
+    },
+    {
+      "epoch": 3.424718032341351,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.859551229786656e-05,
+      "loss": 0.867,
+      "step": 50405
+    },
+    {
+      "epoch": 3.4250577524120125,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8593389047424924e-05,
+      "loss": 1.053,
+      "step": 50410
+    },
+    {
+      "epoch": 3.425397472482674,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8591265796983284e-05,
+      "loss": 0.8307,
+      "step": 50415
+    },
+    {
+      "epoch": 3.425737192553336,
+      "grad_norm": 6.75,
+      "learning_rate": 2.8589142546541652e-05,
+      "loss": 0.8801,
+      "step": 50420
+    },
+    {
+      "epoch": 3.426076912623998,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8587019296100016e-05,
+      "loss": 0.9401,
+      "step": 50425
+    },
+    {
+      "epoch": 3.4264166326946595,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8584896045658376e-05,
+      "loss": 0.8907,
+      "step": 50430
+    },
+    {
+      "epoch": 3.4267563527653215,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.8582772795216744e-05,
+      "loss": 0.9286,
+      "step": 50435
+    },
+    {
+      "epoch": 3.427096072835983,
+      "grad_norm": 8.375,
+      "learning_rate": 2.8580649544775108e-05,
+      "loss": 0.9056,
+      "step": 50440
+    },
+    {
+      "epoch": 3.427435792906645,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.857852629433347e-05,
+      "loss": 0.8739,
+      "step": 50445
+    },
+    {
+      "epoch": 3.427775512977307,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.8576403043891836e-05,
+      "loss": 0.9103,
+      "step": 50450
+    },
+    {
+      "epoch": 3.4281152330479685,
+      "grad_norm": 7.875,
+      "learning_rate": 2.85742797934502e-05,
+      "loss": 0.8958,
+      "step": 50455
+    },
+    {
+      "epoch": 3.42845495311863,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.857215654300856e-05,
+      "loss": 0.8739,
+      "step": 50460
+    },
+    {
+      "epoch": 3.428794673189292,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8570033292566928e-05,
+      "loss": 0.9103,
+      "step": 50465
+    },
+    {
+      "epoch": 3.429134393259954,
+      "grad_norm": 8.25,
+      "learning_rate": 2.856791004212529e-05,
+      "loss": 0.8498,
+      "step": 50470
+    },
+    {
+      "epoch": 3.4294741133306155,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8565786791683652e-05,
+      "loss": 0.8346,
+      "step": 50475
+    },
+    {
+      "epoch": 3.429813833401277,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.856366354124202e-05,
+      "loss": 0.8866,
+      "step": 50480
+    },
+    {
+      "epoch": 3.430153553471939,
+      "grad_norm": 7.75,
+      "learning_rate": 2.856154029080038e-05,
+      "loss": 0.9547,
+      "step": 50485
+    },
+    {
+      "epoch": 3.430493273542601,
+      "grad_norm": 7.25,
+      "learning_rate": 2.8559417040358744e-05,
+      "loss": 0.94,
+      "step": 50490
+    },
+    {
+      "epoch": 3.4308329936132624,
+      "grad_norm": 6.875,
+      "learning_rate": 2.8557293789917112e-05,
+      "loss": 0.9006,
+      "step": 50495
+    },
+    {
+      "epoch": 3.4311727136839245,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8555170539475472e-05,
+      "loss": 0.973,
+      "step": 50500
+    },
+    {
+      "epoch": 3.431512433754586,
+      "grad_norm": 7.75,
+      "learning_rate": 2.8553047289033836e-05,
+      "loss": 0.9726,
+      "step": 50505
+    },
+    {
+      "epoch": 3.4318521538252478,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.8550924038592204e-05,
+      "loss": 0.9325,
+      "step": 50510
+    },
+    {
+      "epoch": 3.43219187389591,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8548800788150564e-05,
+      "loss": 0.8743,
+      "step": 50515
+    },
+    {
+      "epoch": 3.4325315939665715,
+      "grad_norm": 7.625,
+      "learning_rate": 2.854667753770893e-05,
+      "loss": 0.8827,
+      "step": 50520
+    },
+    {
+      "epoch": 3.432871314037233,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8544554287267296e-05,
+      "loss": 0.8992,
+      "step": 50525
+    },
+    {
+      "epoch": 3.433211034107895,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.8542431036825656e-05,
+      "loss": 0.9095,
+      "step": 50530
+    },
+    {
+      "epoch": 3.433550754178557,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8540307786384017e-05,
+      "loss": 0.8576,
+      "step": 50535
+    },
+    {
+      "epoch": 3.4338904742492184,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8538184535942384e-05,
+      "loss": 0.8765,
+      "step": 50540
+    },
+    {
+      "epoch": 3.4342301943198805,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.853606128550075e-05,
+      "loss": 0.9169,
+      "step": 50545
+    },
+    {
+      "epoch": 3.434569914390542,
+      "grad_norm": 7.625,
+      "learning_rate": 2.853393803505911e-05,
+      "loss": 0.8854,
+      "step": 50550
+    },
+    {
+      "epoch": 3.434909634461204,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8531814784617476e-05,
+      "loss": 0.9142,
+      "step": 50555
+    },
+    {
+      "epoch": 3.435249354531866,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.852969153417584e-05,
+      "loss": 0.959,
+      "step": 50560
+    },
+    {
+      "epoch": 3.4355890746025275,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.85275682837342e-05,
+      "loss": 0.919,
+      "step": 50565
+    },
+    {
+      "epoch": 3.435928794673189,
+      "grad_norm": 9.0,
+      "learning_rate": 2.852544503329257e-05,
+      "loss": 0.8663,
+      "step": 50570
+    },
+    {
+      "epoch": 3.436268514743851,
+      "grad_norm": 7.5,
+      "learning_rate": 2.8523321782850932e-05,
+      "loss": 0.8831,
+      "step": 50575
+    },
+    {
+      "epoch": 3.436608234814513,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.8521198532409293e-05,
+      "loss": 0.9026,
+      "step": 50580
+    },
+    {
+      "epoch": 3.4369479548851745,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.851907528196766e-05,
+      "loss": 0.8832,
+      "step": 50585
+    },
+    {
+      "epoch": 3.4372876749558365,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8516952031526024e-05,
+      "loss": 0.9232,
+      "step": 50590
+    },
+    {
+      "epoch": 3.437627395026498,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8514828781084392e-05,
+      "loss": 0.9308,
+      "step": 50595
+    },
+    {
+      "epoch": 3.43796711509716,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8512705530642752e-05,
+      "loss": 0.9325,
+      "step": 50600
+    },
+    {
+      "epoch": 3.438306835167822,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.8510582280201113e-05,
+      "loss": 0.9151,
+      "step": 50605
+    },
+    {
+      "epoch": 3.4386465552384835,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8508459029759484e-05,
+      "loss": 0.8802,
+      "step": 50610
+    },
+    {
+      "epoch": 3.438986275309145,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8506335779317844e-05,
+      "loss": 0.8639,
+      "step": 50615
+    },
+    {
+      "epoch": 3.439325995379807,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8504212528876205e-05,
+      "loss": 0.9649,
+      "step": 50620
+    },
+    {
+      "epoch": 3.439665715450469,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.8502089278434572e-05,
+      "loss": 1.0387,
+      "step": 50625
+    },
+    {
+      "epoch": 3.4400054355211305,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8499966027992936e-05,
+      "loss": 0.9015,
+      "step": 50630
+    },
+    {
+      "epoch": 3.4403451555917925,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8497842777551297e-05,
+      "loss": 0.958,
+      "step": 50635
+    },
+    {
+      "epoch": 3.440684875662454,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8495719527109664e-05,
+      "loss": 0.8316,
+      "step": 50640
+    },
+    {
+      "epoch": 3.441024595733116,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.849359627666803e-05,
+      "loss": 0.8866,
+      "step": 50645
+    },
+    {
+      "epoch": 3.441364315803778,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.849147302622639e-05,
+      "loss": 0.9383,
+      "step": 50650
+    },
+    {
+      "epoch": 3.4417040358744395,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8489349775784756e-05,
+      "loss": 0.8622,
+      "step": 50655
+    },
+    {
+      "epoch": 3.442043755945101,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.848722652534312e-05,
+      "loss": 0.9089,
+      "step": 50660
+    },
+    {
+      "epoch": 3.442383476015763,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.848510327490148e-05,
+      "loss": 0.9731,
+      "step": 50665
+    },
+    {
+      "epoch": 3.442723196086425,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.848298002445985e-05,
+      "loss": 1.0353,
+      "step": 50670
+    },
+    {
+      "epoch": 3.4430629161570865,
+      "grad_norm": 7.875,
+      "learning_rate": 2.8480856774018213e-05,
+      "loss": 0.9031,
+      "step": 50675
+    },
+    {
+      "epoch": 3.4434026362277486,
+      "grad_norm": 6.875,
+      "learning_rate": 2.8478733523576573e-05,
+      "loss": 0.9624,
+      "step": 50680
+    },
+    {
+      "epoch": 3.44374235629841,
+      "grad_norm": 8.0,
+      "learning_rate": 2.847661027313494e-05,
+      "loss": 0.9154,
+      "step": 50685
+    },
+    {
+      "epoch": 3.444082076369072,
+      "grad_norm": 8.125,
+      "learning_rate": 2.84744870226933e-05,
+      "loss": 0.9374,
+      "step": 50690
+    },
+    {
+      "epoch": 3.444421796439734,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8472363772251665e-05,
+      "loss": 0.9319,
+      "step": 50695
+    },
+    {
+      "epoch": 3.4447615165103955,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8470240521810033e-05,
+      "loss": 0.8838,
+      "step": 50700
+    },
+    {
+      "epoch": 3.445101236581057,
+      "grad_norm": 8.75,
+      "learning_rate": 2.8468117271368393e-05,
+      "loss": 0.9007,
+      "step": 50705
+    },
+    {
+      "epoch": 3.4454409566517192,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.8465994020926757e-05,
+      "loss": 0.8513,
+      "step": 50710
+    },
+    {
+      "epoch": 3.445780676722381,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8463870770485125e-05,
+      "loss": 0.8827,
+      "step": 50715
+    },
+    {
+      "epoch": 3.4461203967930425,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8461747520043485e-05,
+      "loss": 0.9564,
+      "step": 50720
+    },
+    {
+      "epoch": 3.446460116863704,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.845962426960185e-05,
+      "loss": 1.0269,
+      "step": 50725
+    },
+    {
+      "epoch": 3.446799836934366,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8457501019160217e-05,
+      "loss": 0.9373,
+      "step": 50730
+    },
+    {
+      "epoch": 3.447139557005028,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.8455377768718577e-05,
+      "loss": 0.916,
+      "step": 50735
+    },
+    {
+      "epoch": 3.4474792770756895,
+      "grad_norm": 6.5,
+      "learning_rate": 2.8453254518276938e-05,
+      "loss": 0.9164,
+      "step": 50740
+    },
+    {
+      "epoch": 3.4478189971463515,
+      "grad_norm": 8.0,
+      "learning_rate": 2.845113126783531e-05,
+      "loss": 0.897,
+      "step": 50745
+    },
+    {
+      "epoch": 3.448158717217013,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.844900801739367e-05,
+      "loss": 0.8543,
+      "step": 50750
+    },
+    {
+      "epoch": 3.448498437287675,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.844688476695203e-05,
+      "loss": 0.911,
+      "step": 50755
+    },
+    {
+      "epoch": 3.448838157358337,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.8444761516510397e-05,
+      "loss": 0.8477,
+      "step": 50760
+    },
+    {
+      "epoch": 3.4491778774289985,
+      "grad_norm": 6.0,
+      "learning_rate": 2.844263826606876e-05,
+      "loss": 0.9521,
+      "step": 50765
+    },
+    {
+      "epoch": 3.44951759749966,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8440515015627122e-05,
+      "loss": 0.8634,
+      "step": 50770
+    },
+    {
+      "epoch": 3.449857317570322,
+      "grad_norm": 7.75,
+      "learning_rate": 2.843839176518549e-05,
+      "loss": 0.8782,
+      "step": 50775
+    },
+    {
+      "epoch": 3.450197037640984,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8436268514743853e-05,
+      "loss": 0.9068,
+      "step": 50780
+    },
+    {
+      "epoch": 3.4505367577116455,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8434145264302214e-05,
+      "loss": 0.8935,
+      "step": 50785
+    },
+    {
+      "epoch": 3.4508764777823075,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.843202201386058e-05,
+      "loss": 0.9384,
+      "step": 50790
+    },
+    {
+      "epoch": 3.451216197852969,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8429898763418945e-05,
+      "loss": 0.8533,
+      "step": 50795
+    },
+    {
+      "epoch": 3.451555917923631,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.8427775512977306e-05,
+      "loss": 0.902,
+      "step": 50800
+    },
+    {
+      "epoch": 3.451895637994293,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8425652262535673e-05,
+      "loss": 0.8691,
+      "step": 50805
+    },
+    {
+      "epoch": 3.4522353580649545,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.8423529012094037e-05,
+      "loss": 0.9083,
+      "step": 50810
+    },
+    {
+      "epoch": 3.452575078135616,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.8421405761652398e-05,
+      "loss": 0.931,
+      "step": 50815
+    },
+    {
+      "epoch": 3.452914798206278,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8419282511210765e-05,
+      "loss": 0.8755,
+      "step": 50820
+    },
+    {
+      "epoch": 3.45325451827694,
+      "grad_norm": 6.0625,
+      "learning_rate": 2.8417159260769126e-05,
+      "loss": 1.0174,
+      "step": 50825
+    },
+    {
+      "epoch": 3.4535942383476015,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.841503601032749e-05,
+      "loss": 0.8999,
+      "step": 50830
+    },
+    {
+      "epoch": 3.453933958418263,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8412912759885857e-05,
+      "loss": 0.8682,
+      "step": 50835
+    },
+    {
+      "epoch": 3.454273678488925,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.8410789509444218e-05,
+      "loss": 1.2383,
+      "step": 50840
+    },
+    {
+      "epoch": 3.454613398559587,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.8408666259002582e-05,
+      "loss": 0.8873,
+      "step": 50845
+    },
+    {
+      "epoch": 3.4549531186302485,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.840654300856095e-05,
+      "loss": 0.8889,
+      "step": 50850
+    },
+    {
+      "epoch": 3.4552928387009105,
+      "grad_norm": 7.5,
+      "learning_rate": 2.840441975811931e-05,
+      "loss": 0.8894,
+      "step": 50855
+    },
+    {
+      "epoch": 3.455632558771572,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8402296507677674e-05,
+      "loss": 0.8819,
+      "step": 50860
+    },
+    {
+      "epoch": 3.455972278842234,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.840017325723604e-05,
+      "loss": 0.9624,
+      "step": 50865
+    },
+    {
+      "epoch": 3.456311998912896,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8398050006794402e-05,
+      "loss": 0.8669,
+      "step": 50870
+    },
+    {
+      "epoch": 3.4566517189835575,
+      "grad_norm": 6.5,
+      "learning_rate": 2.8395926756352766e-05,
+      "loss": 0.8941,
+      "step": 50875
+    },
+    {
+      "epoch": 3.456991439054219,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.8393803505911133e-05,
+      "loss": 0.8552,
+      "step": 50880
+    },
+    {
+      "epoch": 3.457331159124881,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8391680255469494e-05,
+      "loss": 0.9185,
+      "step": 50885
+    },
+    {
+      "epoch": 3.457670879195543,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8389557005027855e-05,
+      "loss": 0.8536,
+      "step": 50890
+    },
+    {
+      "epoch": 3.4580105992662045,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.8387433754586222e-05,
+      "loss": 0.8539,
+      "step": 50895
+    },
+    {
+      "epoch": 3.4583503193368665,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8385310504144586e-05,
+      "loss": 0.9251,
+      "step": 50900
+    },
+    {
+      "epoch": 3.458690039407528,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8383187253702947e-05,
+      "loss": 0.9029,
+      "step": 50905
+    },
+    {
+      "epoch": 3.45902975947819,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8381064003261314e-05,
+      "loss": 0.9511,
+      "step": 50910
+    },
+    {
+      "epoch": 3.459369479548852,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.8378940752819678e-05,
+      "loss": 0.8407,
+      "step": 50915
+    },
+    {
+      "epoch": 3.4597091996195135,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.837681750237804e-05,
+      "loss": 0.8823,
+      "step": 50920
+    },
+    {
+      "epoch": 3.460048919690175,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.8374694251936406e-05,
+      "loss": 0.9354,
+      "step": 50925
+    },
+    {
+      "epoch": 3.460388639760837,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.837257100149477e-05,
+      "loss": 0.8052,
+      "step": 50930
+    },
+    {
+      "epoch": 3.460728359831499,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8370447751053137e-05,
+      "loss": 0.9037,
+      "step": 50935
+    },
+    {
+      "epoch": 3.4610680799021605,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.8368324500611498e-05,
+      "loss": 0.8574,
+      "step": 50940
+    },
+    {
+      "epoch": 3.4614077999728226,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.8366201250169862e-05,
+      "loss": 0.9404,
+      "step": 50945
+    },
+    {
+      "epoch": 3.461747520043484,
+      "grad_norm": 7.375,
+      "learning_rate": 2.836407799972823e-05,
+      "loss": 1.0018,
+      "step": 50950
+    },
+    {
+      "epoch": 3.462087240114146,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.836195474928659e-05,
+      "loss": 0.9495,
+      "step": 50955
+    },
+    {
+      "epoch": 3.462426960184808,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.835983149884495e-05,
+      "loss": 0.9782,
+      "step": 50960
+    },
+    {
+      "epoch": 3.4627666802554695,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8357708248403318e-05,
+      "loss": 0.9886,
+      "step": 50965
+    },
+    {
+      "epoch": 3.463106400326131,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.8355584997961682e-05,
+      "loss": 0.8591,
+      "step": 50970
+    },
+    {
+      "epoch": 3.4634461203967932,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8353461747520043e-05,
+      "loss": 0.9516,
+      "step": 50975
+    },
+    {
+      "epoch": 3.463785840467455,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.835133849707841e-05,
+      "loss": 0.9812,
+      "step": 50980
+    },
+    {
+      "epoch": 3.4641255605381165,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.8349215246636774e-05,
+      "loss": 0.8872,
+      "step": 50985
+    },
+    {
+      "epoch": 3.4644652806087786,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.8347091996195135e-05,
+      "loss": 0.876,
+      "step": 50990
+    },
+    {
+      "epoch": 3.46480500067944,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8344968745753502e-05,
+      "loss": 0.9074,
+      "step": 50995
+    },
+    {
+      "epoch": 3.465144720750102,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8342845495311866e-05,
+      "loss": 0.914,
+      "step": 51000
+    },
+    {
+      "epoch": 3.465484440820764,
+      "grad_norm": 7.5,
+      "learning_rate": 2.8340722244870227e-05,
+      "loss": 1.0221,
+      "step": 51005
+    },
+    {
+      "epoch": 3.4658241608914255,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8338598994428594e-05,
+      "loss": 0.9034,
+      "step": 51010
+    },
+    {
+      "epoch": 3.466163880962087,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.8336475743986958e-05,
+      "loss": 0.9291,
+      "step": 51015
+    },
+    {
+      "epoch": 3.4665036010327492,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.833435249354532e-05,
+      "loss": 0.9127,
+      "step": 51020
+    },
+    {
+      "epoch": 3.466843321103411,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.8332229243103686e-05,
+      "loss": 0.9514,
+      "step": 51025
+    },
+    {
+      "epoch": 3.4671830411740725,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8330105992662047e-05,
+      "loss": 1.0046,
+      "step": 51030
+    },
+    {
+      "epoch": 3.4675227612447346,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.832798274222041e-05,
+      "loss": 0.9303,
+      "step": 51035
+    },
+    {
+      "epoch": 3.467862481315396,
+      "grad_norm": 6.875,
+      "learning_rate": 2.8325859491778778e-05,
+      "loss": 0.8151,
+      "step": 51040
+    },
+    {
+      "epoch": 3.468202201386058,
+      "grad_norm": 7.5,
+      "learning_rate": 2.832373624133714e-05,
+      "loss": 0.8182,
+      "step": 51045
+    },
+    {
+      "epoch": 3.46854192145672,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8321612990895503e-05,
+      "loss": 1.0295,
+      "step": 51050
+    },
+    {
+      "epoch": 3.4688816415273815,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.831948974045387e-05,
+      "loss": 0.9377,
+      "step": 51055
+    },
+    {
+      "epoch": 3.469221361598043,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.831736649001223e-05,
+      "loss": 0.8835,
+      "step": 51060
+    },
+    {
+      "epoch": 3.469561081668705,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.8315243239570595e-05,
+      "loss": 0.8389,
+      "step": 51065
+    },
+    {
+      "epoch": 3.469900801739367,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8313119989128962e-05,
+      "loss": 0.9149,
+      "step": 51070
+    },
+    {
+      "epoch": 3.4702405218100285,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8310996738687323e-05,
+      "loss": 0.8989,
+      "step": 51075
+    },
+    {
+      "epoch": 3.47058024188069,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8308873488245687e-05,
+      "loss": 0.8845,
+      "step": 51080
+    },
+    {
+      "epoch": 3.470919961951352,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.8306750237804054e-05,
+      "loss": 0.9093,
+      "step": 51085
+    },
+    {
+      "epoch": 3.471259682022014,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8304626987362415e-05,
+      "loss": 0.967,
+      "step": 51090
+    },
+    {
+      "epoch": 3.4715994020926755,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.8302503736920775e-05,
+      "loss": 0.8771,
+      "step": 51095
+    },
+    {
+      "epoch": 3.4719391221633376,
+      "grad_norm": 8.75,
+      "learning_rate": 2.8300380486479146e-05,
+      "loss": 0.9182,
+      "step": 51100
+    },
+    {
+      "epoch": 3.472278842233999,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8298257236037507e-05,
+      "loss": 0.8285,
+      "step": 51105
+    },
+    {
+      "epoch": 3.472618562304661,
+      "grad_norm": 6.625,
+      "learning_rate": 2.8296133985595867e-05,
+      "loss": 0.8711,
+      "step": 51110
+    },
+    {
+      "epoch": 3.472958282375323,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8294010735154235e-05,
+      "loss": 0.9287,
+      "step": 51115
+    },
+    {
+      "epoch": 3.4732980024459845,
+      "grad_norm": 6.625,
+      "learning_rate": 2.82918874847126e-05,
+      "loss": 0.9356,
+      "step": 51120
+    },
+    {
+      "epoch": 3.473637722516646,
+      "grad_norm": 7.375,
+      "learning_rate": 2.828976423427096e-05,
+      "loss": 0.8704,
+      "step": 51125
+    },
+    {
+      "epoch": 3.4739774425873082,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8287640983829327e-05,
+      "loss": 0.8644,
+      "step": 51130
+    },
+    {
+      "epoch": 3.47431716265797,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.828551773338769e-05,
+      "loss": 0.9313,
+      "step": 51135
+    },
+    {
+      "epoch": 3.4746568827286315,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.828339448294605e-05,
+      "loss": 0.834,
+      "step": 51140
+    },
+    {
+      "epoch": 3.4749966027992936,
+      "grad_norm": 7.125,
+      "learning_rate": 2.828127123250442e-05,
+      "loss": 0.9272,
+      "step": 51145
+    },
+    {
+      "epoch": 3.475336322869955,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.8279147982062783e-05,
+      "loss": 0.9575,
+      "step": 51150
+    },
+    {
+      "epoch": 3.475676042940617,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8277024731621143e-05,
+      "loss": 0.8232,
+      "step": 51155
+    },
+    {
+      "epoch": 3.4760157630112785,
+      "grad_norm": 7.125,
+      "learning_rate": 2.827490148117951e-05,
+      "loss": 0.9926,
+      "step": 51160
+    },
+    {
+      "epoch": 3.4763554830819405,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.827277823073787e-05,
+      "loss": 0.8684,
+      "step": 51165
+    },
+    {
+      "epoch": 3.476695203152602,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8270654980296235e-05,
+      "loss": 0.8186,
+      "step": 51170
+    },
+    {
+      "epoch": 3.477034923223264,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.8268531729854603e-05,
+      "loss": 0.9005,
+      "step": 51175
+    },
+    {
+      "epoch": 3.477374643293926,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8266408479412963e-05,
+      "loss": 0.9332,
+      "step": 51180
+    },
+    {
+      "epoch": 3.4777143633645875,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8264285228971327e-05,
+      "loss": 0.8784,
+      "step": 51185
+    },
+    {
+      "epoch": 3.478054083435249,
+      "grad_norm": 10.5,
+      "learning_rate": 2.8262161978529695e-05,
+      "loss": 0.8571,
+      "step": 51190
+    },
+    {
+      "epoch": 3.478393803505911,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.8260038728088055e-05,
+      "loss": 0.9722,
+      "step": 51195
+    },
+    {
+      "epoch": 3.478733523576573,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.825791547764642e-05,
+      "loss": 0.7966,
+      "step": 51200
+    },
+    {
+      "epoch": 3.4790732436472345,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.8255792227204787e-05,
+      "loss": 0.8615,
+      "step": 51205
+    },
+    {
+      "epoch": 3.4794129637178965,
+      "grad_norm": 7.25,
+      "learning_rate": 2.8253668976763147e-05,
+      "loss": 0.8287,
+      "step": 51210
+    },
+    {
+      "epoch": 3.479752683788558,
+      "grad_norm": 7.5,
+      "learning_rate": 2.825154572632151e-05,
+      "loss": 0.9869,
+      "step": 51215
+    },
+    {
+      "epoch": 3.48009240385922,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.824942247587988e-05,
+      "loss": 0.8876,
+      "step": 51220
+    },
+    {
+      "epoch": 3.480432123929882,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.824729922543824e-05,
+      "loss": 0.9083,
+      "step": 51225
+    },
+    {
+      "epoch": 3.4807718440005435,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.82451759749966e-05,
+      "loss": 0.9053,
+      "step": 51230
+    },
+    {
+      "epoch": 3.481111564071205,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.824305272455497e-05,
+      "loss": 0.8232,
+      "step": 51235
+    },
+    {
+      "epoch": 3.481451284141867,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.824092947411333e-05,
+      "loss": 0.8796,
+      "step": 51240
+    },
+    {
+      "epoch": 3.481791004212529,
+      "grad_norm": 6.375,
+      "learning_rate": 2.8238806223671692e-05,
+      "loss": 0.8195,
+      "step": 51245
+    },
+    {
+      "epoch": 3.4821307242831905,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.823668297323006e-05,
+      "loss": 0.8371,
+      "step": 51250
+    },
+    {
+      "epoch": 3.4824704443538526,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8234559722788423e-05,
+      "loss": 0.9888,
+      "step": 51255
+    },
+    {
+      "epoch": 3.482810164424514,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8232436472346784e-05,
+      "loss": 0.9374,
+      "step": 51260
+    },
+    {
+      "epoch": 3.483149884495176,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.823031322190515e-05,
+      "loss": 0.8594,
+      "step": 51265
+    },
+    {
+      "epoch": 3.483489604565838,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.8228189971463515e-05,
+      "loss": 0.9112,
+      "step": 51270
+    },
+    {
+      "epoch": 3.4838293246364995,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8226066721021883e-05,
+      "loss": 0.9253,
+      "step": 51275
+    },
+    {
+      "epoch": 3.484169044707161,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8223943470580243e-05,
+      "loss": 0.9422,
+      "step": 51280
+    },
+    {
+      "epoch": 3.4845087647778232,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8221820220138607e-05,
+      "loss": 0.8397,
+      "step": 51285
+    },
+    {
+      "epoch": 3.484848484848485,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8219696969696975e-05,
+      "loss": 0.8231,
+      "step": 51290
+    },
+    {
+      "epoch": 3.4851882049191465,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8217573719255335e-05,
+      "loss": 0.9364,
+      "step": 51295
+    },
+    {
+      "epoch": 3.4855279249898086,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.82154504688137e-05,
+      "loss": 1.0015,
+      "step": 51300
+    },
+    {
+      "epoch": 3.48586764506047,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8213327218372067e-05,
+      "loss": 0.9193,
+      "step": 51305
+    },
+    {
+      "epoch": 3.486207365131132,
+      "grad_norm": 7.875,
+      "learning_rate": 2.8211203967930427e-05,
+      "loss": 0.9569,
+      "step": 51310
+    },
+    {
+      "epoch": 3.486547085201794,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8209080717488788e-05,
+      "loss": 0.9602,
+      "step": 51315
+    },
+    {
+      "epoch": 3.4868868052724555,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.8206957467047155e-05,
+      "loss": 0.893,
+      "step": 51320
+    },
+    {
+      "epoch": 3.487226525343117,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.820483421660552e-05,
+      "loss": 0.9007,
+      "step": 51325
+    },
+    {
+      "epoch": 3.4875662454137792,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.820271096616388e-05,
+      "loss": 0.9584,
+      "step": 51330
+    },
+    {
+      "epoch": 3.487905965484441,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8200587715722247e-05,
+      "loss": 0.8958,
+      "step": 51335
+    },
+    {
+      "epoch": 3.4882456855551025,
+      "grad_norm": 8.25,
+      "learning_rate": 2.819846446528061e-05,
+      "loss": 0.8715,
+      "step": 51340
+    },
+    {
+      "epoch": 3.4885854056257646,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8196341214838972e-05,
+      "loss": 0.9119,
+      "step": 51345
+    },
+    {
+      "epoch": 3.488925125696426,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.819421796439734e-05,
+      "loss": 0.873,
+      "step": 51350
+    },
+    {
+      "epoch": 3.489264845767088,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.8192094713955703e-05,
+      "loss": 0.9178,
+      "step": 51355
+    },
+    {
+      "epoch": 3.48960456583775,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.8189971463514064e-05,
+      "loss": 0.9664,
+      "step": 51360
+    },
+    {
+      "epoch": 3.4899442859084115,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.818784821307243e-05,
+      "loss": 0.7838,
+      "step": 51365
+    },
+    {
+      "epoch": 3.490284005979073,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8185724962630795e-05,
+      "loss": 0.9033,
+      "step": 51370
+    },
+    {
+      "epoch": 3.4906237260497353,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.8183601712189156e-05,
+      "loss": 0.9589,
+      "step": 51375
+    },
+    {
+      "epoch": 3.490963446120397,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.8181478461747524e-05,
+      "loss": 0.8953,
+      "step": 51380
+    },
+    {
+      "epoch": 3.4913031661910585,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8179355211305884e-05,
+      "loss": 0.8879,
+      "step": 51385
+    },
+    {
+      "epoch": 3.4916428862617206,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.8177231960864248e-05,
+      "loss": 0.8476,
+      "step": 51390
+    },
+    {
+      "epoch": 3.491982606332382,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8175108710422616e-05,
+      "loss": 0.8603,
+      "step": 51395
+    },
+    {
+      "epoch": 3.492322326403044,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8172985459980976e-05,
+      "loss": 0.937,
+      "step": 51400
+    },
+    {
+      "epoch": 3.4926620464737055,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.817086220953934e-05,
+      "loss": 0.9355,
+      "step": 51405
+    },
+    {
+      "epoch": 3.4930017665443676,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8168738959097708e-05,
+      "loss": 0.9237,
+      "step": 51410
+    },
+    {
+      "epoch": 3.493341486615029,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.8166615708656068e-05,
+      "loss": 0.8787,
+      "step": 51415
+    },
+    {
+      "epoch": 3.493681206685691,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8164492458214432e-05,
+      "loss": 1.0263,
+      "step": 51420
+    },
+    {
+      "epoch": 3.494020926756353,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.81623692077728e-05,
+      "loss": 0.9247,
+      "step": 51425
+    },
+    {
+      "epoch": 3.4943606468270145,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.816024595733116e-05,
+      "loss": 0.9633,
+      "step": 51430
+    },
+    {
+      "epoch": 3.494700366897676,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8158122706889524e-05,
+      "loss": 1.0075,
+      "step": 51435
+    },
+    {
+      "epoch": 3.4950400869683382,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.815599945644789e-05,
+      "loss": 0.9447,
+      "step": 51440
+    },
+    {
+      "epoch": 3.495379807039,
+      "grad_norm": 7.5,
+      "learning_rate": 2.8153876206006252e-05,
+      "loss": 0.8663,
+      "step": 51445
+    },
+    {
+      "epoch": 3.4957195271096615,
+      "grad_norm": 6.375,
+      "learning_rate": 2.8151752955564613e-05,
+      "loss": 0.905,
+      "step": 51450
+    },
+    {
+      "epoch": 3.4960592471803236,
+      "grad_norm": 7.0,
+      "learning_rate": 2.814962970512298e-05,
+      "loss": 0.7801,
+      "step": 51455
+    },
+    {
+      "epoch": 3.496398967250985,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.8147506454681344e-05,
+      "loss": 0.9128,
+      "step": 51460
+    },
+    {
+      "epoch": 3.496738687321647,
+      "grad_norm": 7.875,
+      "learning_rate": 2.8145383204239705e-05,
+      "loss": 0.9664,
+      "step": 51465
+    },
+    {
+      "epoch": 3.497078407392309,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8143259953798072e-05,
+      "loss": 0.9803,
+      "step": 51470
+    },
+    {
+      "epoch": 3.4974181274629705,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.8141136703356436e-05,
+      "loss": 0.9214,
+      "step": 51475
+    },
+    {
+      "epoch": 3.497757847533632,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.8139013452914797e-05,
+      "loss": 0.9797,
+      "step": 51480
+    },
+    {
+      "epoch": 3.4980975676042942,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8136890202473164e-05,
+      "loss": 0.9511,
+      "step": 51485
+    },
+    {
+      "epoch": 3.498437287674956,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.8134766952031528e-05,
+      "loss": 0.9168,
+      "step": 51490
+    },
+    {
+      "epoch": 3.4987770077456175,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.813264370158989e-05,
+      "loss": 0.9658,
+      "step": 51495
+    },
+    {
+      "epoch": 3.499116727816279,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.8130520451148256e-05,
+      "loss": 0.9201,
+      "step": 51500
+    },
+    {
+      "epoch": 3.499456447886941,
+      "grad_norm": 7.5,
+      "learning_rate": 2.812839720070662e-05,
+      "loss": 0.8198,
+      "step": 51505
+    },
+    {
+      "epoch": 3.499796167957603,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.812627395026498e-05,
+      "loss": 0.9351,
+      "step": 51510
+    },
+    {
+      "epoch": 3.5001358880282645,
+      "grad_norm": 8.5,
+      "learning_rate": 2.8124150699823348e-05,
+      "loss": 1.0264,
+      "step": 51515
+    },
+    {
+      "epoch": 3.5004756080989265,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.812202744938171e-05,
+      "loss": 0.8491,
+      "step": 51520
+    },
+    {
+      "epoch": 3.500815328169588,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8119904198940073e-05,
+      "loss": 0.9907,
+      "step": 51525
+    },
+    {
+      "epoch": 3.50115504824025,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.811778094849844e-05,
+      "loss": 0.9775,
+      "step": 51530
+    },
+    {
+      "epoch": 3.501494768310912,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.81156576980568e-05,
+      "loss": 0.9007,
+      "step": 51535
+    },
+    {
+      "epoch": 3.5018344883815735,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.8113534447615165e-05,
+      "loss": 0.9601,
+      "step": 51540
+    },
+    {
+      "epoch": 3.502174208452235,
+      "grad_norm": 6.625,
+      "learning_rate": 2.8111411197173532e-05,
+      "loss": 0.8991,
+      "step": 51545
+    },
+    {
+      "epoch": 3.5025139285228972,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8109287946731893e-05,
+      "loss": 0.9451,
+      "step": 51550
+    },
+    {
+      "epoch": 3.502853648593559,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8107164696290257e-05,
+      "loss": 0.9524,
+      "step": 51555
+    },
+    {
+      "epoch": 3.5031933686642205,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.8105041445848624e-05,
+      "loss": 0.877,
+      "step": 51560
+    },
+    {
+      "epoch": 3.5035330887348826,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8102918195406985e-05,
+      "loss": 0.9039,
+      "step": 51565
+    },
+    {
+      "epoch": 3.503872808805544,
+      "grad_norm": 6.625,
+      "learning_rate": 2.810079494496535e-05,
+      "loss": 0.8668,
+      "step": 51570
+    },
+    {
+      "epoch": 3.504212528876206,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8098671694523716e-05,
+      "loss": 0.8979,
+      "step": 51575
+    },
+    {
+      "epoch": 3.504552248946868,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.8096548444082077e-05,
+      "loss": 0.9518,
+      "step": 51580
+    },
+    {
+      "epoch": 3.5048919690175295,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.8094425193640437e-05,
+      "loss": 0.9055,
+      "step": 51585
+    },
+    {
+      "epoch": 3.505231689088191,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.8092301943198808e-05,
+      "loss": 0.8996,
+      "step": 51590
+    },
+    {
+      "epoch": 3.5055714091588532,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.809017869275717e-05,
+      "loss": 0.807,
+      "step": 51595
+    },
+    {
+      "epoch": 3.505911129229515,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.808805544231553e-05,
+      "loss": 0.8764,
+      "step": 51600
+    },
+    {
+      "epoch": 3.5062508493001765,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8085932191873897e-05,
+      "loss": 0.9447,
+      "step": 51605
+    },
+    {
+      "epoch": 3.5065905693708386,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.808380894143226e-05,
+      "loss": 0.856,
+      "step": 51610
+    },
+    {
+      "epoch": 3.5069302894415,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.808168569099062e-05,
+      "loss": 0.816,
+      "step": 51615
+    },
+    {
+      "epoch": 3.507270009512162,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.807956244054899e-05,
+      "loss": 1.01,
+      "step": 51620
+    },
+    {
+      "epoch": 3.507609729582824,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.8077439190107353e-05,
+      "loss": 0.9013,
+      "step": 51625
+    },
+    {
+      "epoch": 3.5079494496534855,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.807531593966572e-05,
+      "loss": 0.9049,
+      "step": 51630
+    },
+    {
+      "epoch": 3.508289169724147,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.807319268922408e-05,
+      "loss": 0.8124,
+      "step": 51635
+    },
+    {
+      "epoch": 3.5086288897948092,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.8071069438782445e-05,
+      "loss": 0.9616,
+      "step": 51640
+    },
+    {
+      "epoch": 3.508968609865471,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.8068946188340812e-05,
+      "loss": 0.7524,
+      "step": 51645
+    },
+    {
+      "epoch": 3.5093083299361325,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.8066822937899173e-05,
+      "loss": 0.9533,
+      "step": 51650
+    },
+    {
+      "epoch": 3.5096480500067946,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8064699687457534e-05,
+      "loss": 0.9456,
+      "step": 51655
+    },
+    {
+      "epoch": 3.509987770077456,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.8062576437015904e-05,
+      "loss": 0.9072,
+      "step": 51660
+    },
+    {
+      "epoch": 3.510327490148118,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.8060453186574265e-05,
+      "loss": 0.8587,
+      "step": 51665
+    },
+    {
+      "epoch": 3.51066721021878,
+      "grad_norm": 5.8125,
+      "learning_rate": 2.8058329936132626e-05,
+      "loss": 0.9499,
+      "step": 51670
+    },
+    {
+      "epoch": 3.5110069302894416,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.8056206685690993e-05,
+      "loss": 0.9876,
+      "step": 51675
+    },
+    {
+      "epoch": 3.511346650360103,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8054083435249357e-05,
+      "loss": 0.8475,
+      "step": 51680
+    },
+    {
+      "epoch": 3.5116863704307653,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8051960184807718e-05,
+      "loss": 0.8318,
+      "step": 51685
+    },
+    {
+      "epoch": 3.512026090501427,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.8049836934366085e-05,
+      "loss": 0.849,
+      "step": 51690
+    },
+    {
+      "epoch": 3.5123658105720885,
+      "grad_norm": 7.25,
+      "learning_rate": 2.804771368392445e-05,
+      "loss": 1.0044,
+      "step": 51695
+    },
+    {
+      "epoch": 3.5127055306427506,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.804559043348281e-05,
+      "loss": 0.9577,
+      "step": 51700
+    },
+    {
+      "epoch": 3.5130452507134122,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8043467183041177e-05,
+      "loss": 0.8961,
+      "step": 51705
+    },
+    {
+      "epoch": 3.513384970784074,
+      "grad_norm": 7.125,
+      "learning_rate": 2.804134393259954e-05,
+      "loss": 0.9403,
+      "step": 51710
+    },
+    {
+      "epoch": 3.513724690854736,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.80392206821579e-05,
+      "loss": 0.8558,
+      "step": 51715
+    },
+    {
+      "epoch": 3.5140644109253976,
+      "grad_norm": 9.125,
+      "learning_rate": 2.803709743171627e-05,
+      "loss": 0.8546,
+      "step": 51720
+    },
+    {
+      "epoch": 3.514404130996059,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8034974181274633e-05,
+      "loss": 0.9829,
+      "step": 51725
+    },
+    {
+      "epoch": 3.5147438510667213,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.8032850930832994e-05,
+      "loss": 0.8602,
+      "step": 51730
+    },
+    {
+      "epoch": 3.515083571137383,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.803072768039136e-05,
+      "loss": 0.889,
+      "step": 51735
+    },
+    {
+      "epoch": 3.5154232912080445,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.802860442994972e-05,
+      "loss": 0.874,
+      "step": 51740
+    },
+    {
+      "epoch": 3.5157630112787066,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8026481179508086e-05,
+      "loss": 0.9336,
+      "step": 51745
+    },
+    {
+      "epoch": 3.5161027313493682,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.8024357929066453e-05,
+      "loss": 0.887,
+      "step": 51750
+    },
+    {
+      "epoch": 3.51644245142003,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.8022234678624814e-05,
+      "loss": 0.932,
+      "step": 51755
+    },
+    {
+      "epoch": 3.516782171490692,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.8020111428183178e-05,
+      "loss": 0.9479,
+      "step": 51760
+    },
+    {
+      "epoch": 3.5171218915613536,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.8017988177741545e-05,
+      "loss": 0.8331,
+      "step": 51765
+    },
+    {
+      "epoch": 3.517461611632015,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8015864927299906e-05,
+      "loss": 0.9509,
+      "step": 51770
+    },
+    {
+      "epoch": 3.517801331702677,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.801374167685827e-05,
+      "loss": 0.9029,
+      "step": 51775
+    },
+    {
+      "epoch": 3.518141051773339,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.8011618426416637e-05,
+      "loss": 0.8385,
+      "step": 51780
+    },
+    {
+      "epoch": 3.5184807718440005,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.8009495175974998e-05,
+      "loss": 0.9978,
+      "step": 51785
+    },
+    {
+      "epoch": 3.518820491914662,
+      "grad_norm": 8.25,
+      "learning_rate": 2.800737192553336e-05,
+      "loss": 0.9419,
+      "step": 51790
+    },
+    {
+      "epoch": 3.5191602119853242,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.800524867509173e-05,
+      "loss": 0.9001,
+      "step": 51795
+    },
+    {
+      "epoch": 3.519499932055986,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.800312542465009e-05,
+      "loss": 0.9237,
+      "step": 51800
+    },
+    {
+      "epoch": 3.5198396521266475,
+      "grad_norm": 6.0,
+      "learning_rate": 2.800100217420845e-05,
+      "loss": 0.9618,
+      "step": 51805
+    },
+    {
+      "epoch": 3.520179372197309,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7998878923766818e-05,
+      "loss": 0.9506,
+      "step": 51810
+    },
+    {
+      "epoch": 3.520519092267971,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.799675567332518e-05,
+      "loss": 0.9001,
+      "step": 51815
+    },
+    {
+      "epoch": 3.520858812338633,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.7994632422883542e-05,
+      "loss": 0.9018,
+      "step": 51820
+    },
+    {
+      "epoch": 3.5211985324092945,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.799250917244191e-05,
+      "loss": 0.9802,
+      "step": 51825
+    },
+    {
+      "epoch": 3.5215382524799566,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7990385922000274e-05,
+      "loss": 0.974,
+      "step": 51830
+    },
+    {
+      "epoch": 3.521877972550618,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7988262671558634e-05,
+      "loss": 0.8587,
+      "step": 51835
+    },
+    {
+      "epoch": 3.52221769262128,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7986139421117002e-05,
+      "loss": 0.8944,
+      "step": 51840
+    },
+    {
+      "epoch": 3.522557412691942,
+      "grad_norm": 6.375,
+      "learning_rate": 2.7984016170675366e-05,
+      "loss": 0.9163,
+      "step": 51845
+    },
+    {
+      "epoch": 3.5228971327626035,
+      "grad_norm": 8.0,
+      "learning_rate": 2.7981892920233726e-05,
+      "loss": 0.9394,
+      "step": 51850
+    },
+    {
+      "epoch": 3.523236852833265,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7979769669792094e-05,
+      "loss": 0.8637,
+      "step": 51855
+    },
+    {
+      "epoch": 3.5235765729039272,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7977646419350458e-05,
+      "loss": 0.8775,
+      "step": 51860
+    },
+    {
+      "epoch": 3.523916292974589,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.797552316890882e-05,
+      "loss": 0.8631,
+      "step": 51865
+    },
+    {
+      "epoch": 3.5242560130452505,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.7973399918467186e-05,
+      "loss": 0.8904,
+      "step": 51870
+    },
+    {
+      "epoch": 3.5245957331159126,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7971276668025546e-05,
+      "loss": 1.0243,
+      "step": 51875
+    },
+    {
+      "epoch": 3.524935453186574,
+      "grad_norm": 6.875,
+      "learning_rate": 2.796915341758391e-05,
+      "loss": 0.8991,
+      "step": 51880
+    },
+    {
+      "epoch": 3.525275173257236,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.7967030167142278e-05,
+      "loss": 0.8489,
+      "step": 51885
+    },
+    {
+      "epoch": 3.525614893327898,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.796490691670064e-05,
+      "loss": 0.8573,
+      "step": 51890
+    },
+    {
+      "epoch": 3.5259546133985595,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.7962783666259002e-05,
+      "loss": 0.8598,
+      "step": 51895
+    },
+    {
+      "epoch": 3.526294333469221,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.796066041581737e-05,
+      "loss": 1.0068,
+      "step": 51900
+    },
+    {
+      "epoch": 3.5266340535398832,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.795853716537573e-05,
+      "loss": 0.9332,
+      "step": 51905
+    },
+    {
+      "epoch": 3.526973773610545,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7956413914934094e-05,
+      "loss": 0.8455,
+      "step": 51910
+    },
+    {
+      "epoch": 3.5273134936812065,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7954290664492462e-05,
+      "loss": 0.9068,
+      "step": 51915
+    },
+    {
+      "epoch": 3.5276532137518686,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.7952167414050822e-05,
+      "loss": 0.9257,
+      "step": 51920
+    },
+    {
+      "epoch": 3.52799293382253,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.7950044163609186e-05,
+      "loss": 0.9444,
+      "step": 51925
+    },
+    {
+      "epoch": 3.528332653893192,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7947920913167554e-05,
+      "loss": 0.9452,
+      "step": 51930
+    },
+    {
+      "epoch": 3.528672373963854,
+      "grad_norm": 7.125,
+      "learning_rate": 2.7945797662725914e-05,
+      "loss": 0.8175,
+      "step": 51935
+    },
+    {
+      "epoch": 3.5290120940345155,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7943674412284275e-05,
+      "loss": 0.9615,
+      "step": 51940
+    },
+    {
+      "epoch": 3.529351814105177,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7941551161842642e-05,
+      "loss": 0.9808,
+      "step": 51945
+    },
+    {
+      "epoch": 3.5296915341758393,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7939427911401006e-05,
+      "loss": 0.816,
+      "step": 51950
+    },
+    {
+      "epoch": 3.530031254246501,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7937304660959367e-05,
+      "loss": 0.8425,
+      "step": 51955
+    },
+    {
+      "epoch": 3.5303709743171625,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7935181410517734e-05,
+      "loss": 0.8853,
+      "step": 51960
+    },
+    {
+      "epoch": 3.5307106943878246,
+      "grad_norm": 7.375,
+      "learning_rate": 2.79330581600761e-05,
+      "loss": 0.9388,
+      "step": 51965
+    },
+    {
+      "epoch": 3.531050414458486,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.7930934909634466e-05,
+      "loss": 0.9819,
+      "step": 51970
+    },
+    {
+      "epoch": 3.531390134529148,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.7928811659192826e-05,
+      "loss": 0.901,
+      "step": 51975
+    },
+    {
+      "epoch": 3.53172985459981,
+      "grad_norm": 6.09375,
+      "learning_rate": 2.792668840875119e-05,
+      "loss": 0.9037,
+      "step": 51980
+    },
+    {
+      "epoch": 3.5320695746704716,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7924565158309558e-05,
+      "loss": 0.947,
+      "step": 51985
+    },
+    {
+      "epoch": 3.532409294741133,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.792244190786792e-05,
+      "loss": 0.9241,
+      "step": 51990
+    },
+    {
+      "epoch": 3.5327490148117953,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7920318657426282e-05,
+      "loss": 0.8634,
+      "step": 51995
+    },
+    {
+      "epoch": 3.533088734882457,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.791819540698465e-05,
+      "loss": 0.8787,
+      "step": 52000
+    },
+    {
+      "epoch": 3.5334284549531185,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.791607215654301e-05,
+      "loss": 0.9994,
+      "step": 52005
+    },
+    {
+      "epoch": 3.5337681750237806,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.791394890610137e-05,
+      "loss": 0.8648,
+      "step": 52010
+    },
+    {
+      "epoch": 3.5341078950944422,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.7911825655659742e-05,
+      "loss": 0.9472,
+      "step": 52015
+    },
+    {
+      "epoch": 3.534447615165104,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.7909702405218102e-05,
+      "loss": 0.8432,
+      "step": 52020
+    },
+    {
+      "epoch": 3.534787335235766,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7907579154776463e-05,
+      "loss": 0.9912,
+      "step": 52025
+    },
+    {
+      "epoch": 3.5351270553064276,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.790545590433483e-05,
+      "loss": 0.9675,
+      "step": 52030
+    },
+    {
+      "epoch": 3.535466775377089,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.7903332653893194e-05,
+      "loss": 0.9305,
+      "step": 52035
+    },
+    {
+      "epoch": 3.5358064954477513,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.7901209403451555e-05,
+      "loss": 0.8369,
+      "step": 52040
+    },
+    {
+      "epoch": 3.536146215518413,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7899086153009922e-05,
+      "loss": 0.837,
+      "step": 52045
+    },
+    {
+      "epoch": 3.5364859355890745,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7896962902568286e-05,
+      "loss": 0.8226,
+      "step": 52050
+    },
+    {
+      "epoch": 3.5368256556597366,
+      "grad_norm": 7.75,
+      "learning_rate": 2.7894839652126647e-05,
+      "loss": 0.9707,
+      "step": 52055
+    },
+    {
+      "epoch": 3.5371653757303982,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7892716401685014e-05,
+      "loss": 0.9135,
+      "step": 52060
+    },
+    {
+      "epoch": 3.53750509580106,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.789059315124338e-05,
+      "loss": 0.9043,
+      "step": 52065
+    },
+    {
+      "epoch": 3.537844815871722,
+      "grad_norm": 8.125,
+      "learning_rate": 2.788846990080174e-05,
+      "loss": 0.9651,
+      "step": 52070
+    },
+    {
+      "epoch": 3.5381845359423836,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7886346650360106e-05,
+      "loss": 0.9735,
+      "step": 52075
+    },
+    {
+      "epoch": 3.538524256013045,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7884223399918467e-05,
+      "loss": 0.9261,
+      "step": 52080
+    },
+    {
+      "epoch": 3.5388639760837073,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.788210014947683e-05,
+      "loss": 0.8567,
+      "step": 52085
+    },
+    {
+      "epoch": 3.539203696154369,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.78799768990352e-05,
+      "loss": 0.8667,
+      "step": 52090
+    },
+    {
+      "epoch": 3.5395434162250305,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.787785364859356e-05,
+      "loss": 0.9477,
+      "step": 52095
+    },
+    {
+      "epoch": 3.5398831362956926,
+      "grad_norm": 7.0,
+      "learning_rate": 2.7875730398151923e-05,
+      "loss": 0.9322,
+      "step": 52100
+    },
+    {
+      "epoch": 3.5402228563663543,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.787360714771029e-05,
+      "loss": 0.9039,
+      "step": 52105
+    },
+    {
+      "epoch": 3.540562576437016,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.787148389726865e-05,
+      "loss": 0.9047,
+      "step": 52110
+    },
+    {
+      "epoch": 3.5409022965076775,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7869360646827015e-05,
+      "loss": 0.9548,
+      "step": 52115
+    },
+    {
+      "epoch": 3.5412420165783396,
+      "grad_norm": 8.125,
+      "learning_rate": 2.7867237396385383e-05,
+      "loss": 0.9384,
+      "step": 52120
+    },
+    {
+      "epoch": 3.541581736649001,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.7865114145943743e-05,
+      "loss": 0.895,
+      "step": 52125
+    },
+    {
+      "epoch": 3.541921456719663,
+      "grad_norm": 8.0,
+      "learning_rate": 2.7862990895502107e-05,
+      "loss": 0.9569,
+      "step": 52130
+    },
+    {
+      "epoch": 3.542261176790325,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7860867645060475e-05,
+      "loss": 0.9634,
+      "step": 52135
+    },
+    {
+      "epoch": 3.5426008968609866,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.7858744394618835e-05,
+      "loss": 0.9332,
+      "step": 52140
+    },
+    {
+      "epoch": 3.542940616931648,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.7856621144177196e-05,
+      "loss": 0.9112,
+      "step": 52145
+    },
+    {
+      "epoch": 3.54328033700231,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7854497893735567e-05,
+      "loss": 0.939,
+      "step": 52150
+    },
+    {
+      "epoch": 3.543620057072972,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.7852374643293927e-05,
+      "loss": 0.9329,
+      "step": 52155
+    },
+    {
+      "epoch": 3.5439597771436335,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7850251392852288e-05,
+      "loss": 0.8773,
+      "step": 52160
+    },
+    {
+      "epoch": 3.544299497214295,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.7848128142410655e-05,
+      "loss": 0.9291,
+      "step": 52165
+    },
+    {
+      "epoch": 3.5446392172849572,
+      "grad_norm": 7.125,
+      "learning_rate": 2.784600489196902e-05,
+      "loss": 0.8754,
+      "step": 52170
+    },
+    {
+      "epoch": 3.544978937355619,
+      "grad_norm": 7.375,
+      "learning_rate": 2.784388164152738e-05,
+      "loss": 0.8814,
+      "step": 52175
+    },
+    {
+      "epoch": 3.5453186574262805,
+      "grad_norm": 6.75,
+      "learning_rate": 2.7841758391085747e-05,
+      "loss": 0.8636,
+      "step": 52180
+    },
+    {
+      "epoch": 3.5456583774969426,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.783963514064411e-05,
+      "loss": 0.8938,
+      "step": 52185
+    },
+    {
+      "epoch": 3.545998097567604,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7837511890202472e-05,
+      "loss": 0.868,
+      "step": 52190
+    },
+    {
+      "epoch": 3.546337817638266,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.783538863976084e-05,
+      "loss": 0.7927,
+      "step": 52195
+    },
+    {
+      "epoch": 3.546677537708928,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7833265389319203e-05,
+      "loss": 0.8353,
+      "step": 52200
+    },
+    {
+      "epoch": 3.5470172577795895,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7831142138877564e-05,
+      "loss": 0.8757,
+      "step": 52205
+    },
+    {
+      "epoch": 3.547356977850251,
+      "grad_norm": 7.375,
+      "learning_rate": 2.782901888843593e-05,
+      "loss": 0.9559,
+      "step": 52210
+    },
+    {
+      "epoch": 3.5476966979209132,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.7826895637994295e-05,
+      "loss": 0.9258,
+      "step": 52215
+    },
+    {
+      "epoch": 3.548036417991575,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7824772387552656e-05,
+      "loss": 0.9789,
+      "step": 52220
+    },
+    {
+      "epoch": 3.5483761380622365,
+      "grad_norm": 7.75,
+      "learning_rate": 2.7822649137111023e-05,
+      "loss": 0.9378,
+      "step": 52225
+    },
+    {
+      "epoch": 3.5487158581328986,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7820525886669384e-05,
+      "loss": 0.8342,
+      "step": 52230
+    },
+    {
+      "epoch": 3.54905557820356,
+      "grad_norm": 7.125,
+      "learning_rate": 2.7818402636227748e-05,
+      "loss": 0.9043,
+      "step": 52235
+    },
+    {
+      "epoch": 3.549395298274222,
+      "grad_norm": 6.625,
+      "learning_rate": 2.7816279385786115e-05,
+      "loss": 0.9353,
+      "step": 52240
+    },
+    {
+      "epoch": 3.549735018344884,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.7814156135344476e-05,
+      "loss": 0.9567,
+      "step": 52245
+    },
+    {
+      "epoch": 3.5500747384155455,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.781203288490284e-05,
+      "loss": 0.825,
+      "step": 52250
+    },
+    {
+      "epoch": 3.550414458486207,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7809909634461207e-05,
+      "loss": 0.978,
+      "step": 52255
+    },
+    {
+      "epoch": 3.5507541785568693,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.7807786384019568e-05,
+      "loss": 0.9264,
+      "step": 52260
+    },
+    {
+      "epoch": 3.551093898627531,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.7805663133577932e-05,
+      "loss": 0.8738,
+      "step": 52265
+    },
+    {
+      "epoch": 3.5514336186981925,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.78035398831363e-05,
+      "loss": 0.9491,
+      "step": 52270
+    },
+    {
+      "epoch": 3.5517733387688546,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.780141663269466e-05,
+      "loss": 0.9029,
+      "step": 52275
+    },
+    {
+      "epoch": 3.5521130588395162,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.779929338225302e-05,
+      "loss": 0.9653,
+      "step": 52280
+    },
+    {
+      "epoch": 3.552452778910178,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.779717013181139e-05,
+      "loss": 0.948,
+      "step": 52285
+    },
+    {
+      "epoch": 3.55279249898084,
+      "grad_norm": 7.0,
+      "learning_rate": 2.7795046881369752e-05,
+      "loss": 0.9229,
+      "step": 52290
+    },
+    {
+      "epoch": 3.5531322190515016,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7792923630928112e-05,
+      "loss": 0.8941,
+      "step": 52295
+    },
+    {
+      "epoch": 3.553471939122163,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.779080038048648e-05,
+      "loss": 0.9121,
+      "step": 52300
+    },
+    {
+      "epoch": 3.5538116591928253,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7788677130044844e-05,
+      "loss": 0.9493,
+      "step": 52305
+    },
+    {
+      "epoch": 3.554151379263487,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.778655387960321e-05,
+      "loss": 0.9823,
+      "step": 52310
+    },
+    {
+      "epoch": 3.5544910993341485,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.7784430629161572e-05,
+      "loss": 0.989,
+      "step": 52315
+    },
+    {
+      "epoch": 3.5548308194048106,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7782307378719936e-05,
+      "loss": 0.9103,
+      "step": 52320
+    },
+    {
+      "epoch": 3.5551705394754722,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.7780184128278303e-05,
+      "loss": 0.9547,
+      "step": 52325
+    },
+    {
+      "epoch": 3.555510259546134,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7778060877836664e-05,
+      "loss": 0.8805,
+      "step": 52330
+    },
+    {
+      "epoch": 3.555849979616796,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.7775937627395028e-05,
+      "loss": 0.9691,
+      "step": 52335
+    },
+    {
+      "epoch": 3.5561896996874576,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7773814376953395e-05,
+      "loss": 0.8907,
+      "step": 52340
+    },
+    {
+      "epoch": 3.556529419758119,
+      "grad_norm": 5.875,
+      "learning_rate": 2.7771691126511756e-05,
+      "loss": 0.8028,
+      "step": 52345
+    },
+    {
+      "epoch": 3.5568691398287813,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.776956787607012e-05,
+      "loss": 0.8794,
+      "step": 52350
+    },
+    {
+      "epoch": 3.557208859899443,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7767444625628487e-05,
+      "loss": 0.9215,
+      "step": 52355
+    },
+    {
+      "epoch": 3.5575485799701045,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7765321375186848e-05,
+      "loss": 0.8722,
+      "step": 52360
+    },
+    {
+      "epoch": 3.5578883000407666,
+      "grad_norm": 7.5,
+      "learning_rate": 2.776319812474521e-05,
+      "loss": 0.8579,
+      "step": 52365
+    },
+    {
+      "epoch": 3.5582280201114282,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7761074874303576e-05,
+      "loss": 0.8768,
+      "step": 52370
+    },
+    {
+      "epoch": 3.55856774018209,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.775895162386194e-05,
+      "loss": 0.889,
+      "step": 52375
+    },
+    {
+      "epoch": 3.558907460252752,
+      "grad_norm": 8.125,
+      "learning_rate": 2.77568283734203e-05,
+      "loss": 0.9609,
+      "step": 52380
+    },
+    {
+      "epoch": 3.5592471803234136,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7754705122978668e-05,
+      "loss": 0.8761,
+      "step": 52385
+    },
+    {
+      "epoch": 3.559586900394075,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.7752581872537032e-05,
+      "loss": 0.8018,
+      "step": 52390
+    },
+    {
+      "epoch": 3.5599266204647373,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7750458622095393e-05,
+      "loss": 0.9392,
+      "step": 52395
+    },
+    {
+      "epoch": 3.560266340535399,
+      "grad_norm": 6.875,
+      "learning_rate": 2.774833537165376e-05,
+      "loss": 0.8976,
+      "step": 52400
+    },
+    {
+      "epoch": 3.5606060606060606,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7746212121212124e-05,
+      "loss": 0.8685,
+      "step": 52405
+    },
+    {
+      "epoch": 3.5609457806767226,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7744088870770485e-05,
+      "loss": 0.8496,
+      "step": 52410
+    },
+    {
+      "epoch": 3.5612855007473843,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7741965620328852e-05,
+      "loss": 0.9094,
+      "step": 52415
+    },
+    {
+      "epoch": 3.561625220818046,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7739842369887216e-05,
+      "loss": 0.9352,
+      "step": 52420
+    },
+    {
+      "epoch": 3.561964940888708,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7737719119445577e-05,
+      "loss": 0.9701,
+      "step": 52425
+    },
+    {
+      "epoch": 3.5623046609593696,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.7735595869003944e-05,
+      "loss": 0.7818,
+      "step": 52430
+    },
+    {
+      "epoch": 3.5626443810300312,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7733472618562305e-05,
+      "loss": 0.9273,
+      "step": 52435
+    },
+    {
+      "epoch": 3.5629841011006933,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.773134936812067e-05,
+      "loss": 0.8918,
+      "step": 52440
+    },
+    {
+      "epoch": 3.563323821171355,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.7729226117679036e-05,
+      "loss": 0.8508,
+      "step": 52445
+    },
+    {
+      "epoch": 3.5636635412420166,
+      "grad_norm": 5.78125,
+      "learning_rate": 2.7727102867237397e-05,
+      "loss": 0.9862,
+      "step": 52450
+    },
+    {
+      "epoch": 3.564003261312678,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.772497961679576e-05,
+      "loss": 0.9116,
+      "step": 52455
+    },
+    {
+      "epoch": 3.5643429813833403,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.7722856366354128e-05,
+      "loss": 0.9522,
+      "step": 52460
+    },
+    {
+      "epoch": 3.564682701454002,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.772073311591249e-05,
+      "loss": 0.9215,
+      "step": 52465
+    },
+    {
+      "epoch": 3.5650224215246635,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7718609865470853e-05,
+      "loss": 0.9214,
+      "step": 52470
+    },
+    {
+      "epoch": 3.5653621415953256,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.771648661502922e-05,
+      "loss": 0.907,
+      "step": 52475
+    },
+    {
+      "epoch": 3.5657018616659872,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.771436336458758e-05,
+      "loss": 0.9501,
+      "step": 52480
+    },
+    {
+      "epoch": 3.566041581736649,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.7712240114145945e-05,
+      "loss": 0.9514,
+      "step": 52485
+    },
+    {
+      "epoch": 3.5663813018073105,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7710116863704312e-05,
+      "loss": 0.9171,
+      "step": 52490
+    },
+    {
+      "epoch": 3.5667210218779726,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.7707993613262673e-05,
+      "loss": 0.9246,
+      "step": 52495
+    },
+    {
+      "epoch": 3.567060741948634,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7705870362821033e-05,
+      "loss": 0.8121,
+      "step": 52500
+    },
+    {
+      "epoch": 3.567400462019296,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.77037471123794e-05,
+      "loss": 0.9455,
+      "step": 52505
+    },
+    {
+      "epoch": 3.567740182089958,
+      "grad_norm": 8.0,
+      "learning_rate": 2.7701623861937765e-05,
+      "loss": 0.9283,
+      "step": 52510
+    },
+    {
+      "epoch": 3.5680799021606195,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7699500611496125e-05,
+      "loss": 0.9466,
+      "step": 52515
+    },
+    {
+      "epoch": 3.568419622231281,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.7697377361054493e-05,
+      "loss": 0.886,
+      "step": 52520
+    },
+    {
+      "epoch": 3.5687593423019432,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.7695254110612857e-05,
+      "loss": 0.9065,
+      "step": 52525
+    },
+    {
+      "epoch": 3.569099062372605,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7693130860171217e-05,
+      "loss": 0.8734,
+      "step": 52530
+    },
+    {
+      "epoch": 3.5694387824432665,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.7691007609729585e-05,
+      "loss": 0.9462,
+      "step": 52535
+    },
+    {
+      "epoch": 3.5697785025139286,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.768888435928795e-05,
+      "loss": 1.034,
+      "step": 52540
+    },
+    {
+      "epoch": 3.57011822258459,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.768676110884631e-05,
+      "loss": 0.7924,
+      "step": 52545
+    },
+    {
+      "epoch": 3.570457942655252,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.7684637858404677e-05,
+      "loss": 0.9588,
+      "step": 52550
+    },
+    {
+      "epoch": 3.570797662725914,
+      "grad_norm": 7.25,
+      "learning_rate": 2.768251460796304e-05,
+      "loss": 0.8458,
+      "step": 52555
+    },
+    {
+      "epoch": 3.5711373827965756,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.76803913575214e-05,
+      "loss": 0.9337,
+      "step": 52560
+    },
+    {
+      "epoch": 3.571477102867237,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.767826810707977e-05,
+      "loss": 0.9139,
+      "step": 52565
+    },
+    {
+      "epoch": 3.5718168229378993,
+      "grad_norm": 7.875,
+      "learning_rate": 2.767614485663813e-05,
+      "loss": 0.9353,
+      "step": 52570
+    },
+    {
+      "epoch": 3.572156543008561,
+      "grad_norm": 6.625,
+      "learning_rate": 2.7674021606196493e-05,
+      "loss": 0.9098,
+      "step": 52575
+    },
+    {
+      "epoch": 3.5724962630792225,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.767189835575486e-05,
+      "loss": 0.9065,
+      "step": 52580
+    },
+    {
+      "epoch": 3.5728359831498846,
+      "grad_norm": 7.375,
+      "learning_rate": 2.766977510531322e-05,
+      "loss": 0.9307,
+      "step": 52585
+    },
+    {
+      "epoch": 3.5731757032205462,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.7667651854871585e-05,
+      "loss": 0.9662,
+      "step": 52590
+    },
+    {
+      "epoch": 3.573515423291208,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.7665528604429953e-05,
+      "loss": 0.8619,
+      "step": 52595
+    },
+    {
+      "epoch": 3.57385514336187,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.7663405353988313e-05,
+      "loss": 0.9407,
+      "step": 52600
+    },
+    {
+      "epoch": 3.5741948634325316,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.7661282103546677e-05,
+      "loss": 0.9702,
+      "step": 52605
+    },
+    {
+      "epoch": 3.574534583503193,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.7659158853105045e-05,
+      "loss": 0.9541,
+      "step": 52610
+    },
+    {
+      "epoch": 3.5748743035738553,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7657035602663405e-05,
+      "loss": 0.8551,
+      "step": 52615
+    },
+    {
+      "epoch": 3.575214023644517,
+      "grad_norm": 7.5,
+      "learning_rate": 2.765491235222177e-05,
+      "loss": 0.9271,
+      "step": 52620
+    },
+    {
+      "epoch": 3.5755537437151785,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.7652789101780137e-05,
+      "loss": 0.8937,
+      "step": 52625
+    },
+    {
+      "epoch": 3.5758934637858406,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.7650665851338497e-05,
+      "loss": 0.9171,
+      "step": 52630
+    },
+    {
+      "epoch": 3.5762331838565022,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7648542600896858e-05,
+      "loss": 0.8717,
+      "step": 52635
+    },
+    {
+      "epoch": 3.576572903927164,
+      "grad_norm": 8.625,
+      "learning_rate": 2.764641935045523e-05,
+      "loss": 0.9088,
+      "step": 52640
+    },
+    {
+      "epoch": 3.576912623997826,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.764429610001359e-05,
+      "loss": 0.9716,
+      "step": 52645
+    },
+    {
+      "epoch": 3.5772523440684876,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7642172849571957e-05,
+      "loss": 0.8621,
+      "step": 52650
+    },
+    {
+      "epoch": 3.577592064139149,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7640049599130317e-05,
+      "loss": 0.902,
+      "step": 52655
+    },
+    {
+      "epoch": 3.5779317842098113,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.763792634868868e-05,
+      "loss": 0.8156,
+      "step": 52660
+    },
+    {
+      "epoch": 3.578271504280473,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.763580309824705e-05,
+      "loss": 0.9979,
+      "step": 52665
+    },
+    {
+      "epoch": 3.5786112243511345,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.763367984780541e-05,
+      "loss": 0.8311,
+      "step": 52670
+    },
+    {
+      "epoch": 3.5789509444217966,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7631556597363773e-05,
+      "loss": 0.8114,
+      "step": 52675
+    },
+    {
+      "epoch": 3.5792906644924583,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.762943334692214e-05,
+      "loss": 1.0417,
+      "step": 52680
+    },
+    {
+      "epoch": 3.57963038456312,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.76273100964805e-05,
+      "loss": 0.8982,
+      "step": 52685
+    },
+    {
+      "epoch": 3.579970104633782,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7625186846038865e-05,
+      "loss": 0.9676,
+      "step": 52690
+    },
+    {
+      "epoch": 3.5803098247044436,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.7623063595597233e-05,
+      "loss": 0.928,
+      "step": 52695
+    },
+    {
+      "epoch": 3.580649544775105,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.7620940345155593e-05,
+      "loss": 0.9563,
+      "step": 52700
+    },
+    {
+      "epoch": 3.5809892648457673,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.7618817094713957e-05,
+      "loss": 0.8864,
+      "step": 52705
+    },
+    {
+      "epoch": 3.581328984916429,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.7616693844272325e-05,
+      "loss": 1.006,
+      "step": 52710
+    },
+    {
+      "epoch": 3.5816687049870906,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7614570593830685e-05,
+      "loss": 0.9125,
+      "step": 52715
+    },
+    {
+      "epoch": 3.5820084250577526,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7612447343389046e-05,
+      "loss": 0.8977,
+      "step": 52720
+    },
+    {
+      "epoch": 3.5823481451284143,
+      "grad_norm": 7.875,
+      "learning_rate": 2.7610324092947413e-05,
+      "loss": 0.9201,
+      "step": 52725
+    },
+    {
+      "epoch": 3.582687865199076,
+      "grad_norm": 7.0,
+      "learning_rate": 2.7608200842505777e-05,
+      "loss": 0.9206,
+      "step": 52730
+    },
+    {
+      "epoch": 3.583027585269738,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.7606077592064138e-05,
+      "loss": 0.8467,
+      "step": 52735
+    },
+    {
+      "epoch": 3.5833673053403996,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7603954341622505e-05,
+      "loss": 0.8916,
+      "step": 52740
+    },
+    {
+      "epoch": 3.5837070254110612,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.760183109118087e-05,
+      "loss": 0.8353,
+      "step": 52745
+    },
+    {
+      "epoch": 3.5840467454817233,
+      "grad_norm": 7.5,
+      "learning_rate": 2.759970784073923e-05,
+      "loss": 0.9043,
+      "step": 52750
+    },
+    {
+      "epoch": 3.584386465552385,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.7597584590297597e-05,
+      "loss": 0.8135,
+      "step": 52755
+    },
+    {
+      "epoch": 3.5847261856230466,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.759546133985596e-05,
+      "loss": 0.9311,
+      "step": 52760
+    },
+    {
+      "epoch": 3.5850659056937086,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7593338089414322e-05,
+      "loss": 0.9029,
+      "step": 52765
+    },
+    {
+      "epoch": 3.5854056257643703,
+      "grad_norm": 8.375,
+      "learning_rate": 2.759121483897269e-05,
+      "loss": 0.9198,
+      "step": 52770
+    },
+    {
+      "epoch": 3.585745345835032,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.7589091588531053e-05,
+      "loss": 0.8108,
+      "step": 52775
+    },
+    {
+      "epoch": 3.586085065905694,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7586968338089414e-05,
+      "loss": 0.9981,
+      "step": 52780
+    },
+    {
+      "epoch": 3.5864247859763556,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.758484508764778e-05,
+      "loss": 0.8693,
+      "step": 52785
+    },
+    {
+      "epoch": 3.5867645060470172,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.7582721837206142e-05,
+      "loss": 0.9843,
+      "step": 52790
+    },
+    {
+      "epoch": 3.587104226117679,
+      "grad_norm": 6.75,
+      "learning_rate": 2.7580598586764506e-05,
+      "loss": 1.003,
+      "step": 52795
+    },
+    {
+      "epoch": 3.587443946188341,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.7578475336322873e-05,
+      "loss": 0.9543,
+      "step": 52800
+    },
+    {
+      "epoch": 3.5877836662590026,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7576352085881234e-05,
+      "loss": 0.9665,
+      "step": 52805
+    },
+    {
+      "epoch": 3.588123386329664,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.7574228835439598e-05,
+      "loss": 0.9782,
+      "step": 52810
+    },
+    {
+      "epoch": 3.5884631064003263,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.7572105584997966e-05,
+      "loss": 0.9333,
+      "step": 52815
+    },
+    {
+      "epoch": 3.588802826470988,
+      "grad_norm": 8.375,
+      "learning_rate": 2.7569982334556326e-05,
+      "loss": 0.9946,
+      "step": 52820
+    },
+    {
+      "epoch": 3.5891425465416495,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.756785908411469e-05,
+      "loss": 0.9643,
+      "step": 52825
+    },
+    {
+      "epoch": 3.589482266612311,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.7565735833673058e-05,
+      "loss": 0.932,
+      "step": 52830
+    },
+    {
+      "epoch": 3.5898219866829733,
+      "grad_norm": 7.0,
+      "learning_rate": 2.7563612583231418e-05,
+      "loss": 0.8906,
+      "step": 52835
+    },
+    {
+      "epoch": 3.590161706753635,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.7561489332789782e-05,
+      "loss": 0.9908,
+      "step": 52840
+    },
+    {
+      "epoch": 3.5905014268242965,
+      "grad_norm": 8.0,
+      "learning_rate": 2.755936608234815e-05,
+      "loss": 0.9293,
+      "step": 52845
+    },
+    {
+      "epoch": 3.5908411468949586,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.755724283190651e-05,
+      "loss": 0.9907,
+      "step": 52850
+    },
+    {
+      "epoch": 3.59118086696562,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.755511958146487e-05,
+      "loss": 0.8612,
+      "step": 52855
+    },
+    {
+      "epoch": 3.591520587036282,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7552996331023238e-05,
+      "loss": 0.8473,
+      "step": 52860
+    },
+    {
+      "epoch": 3.591860307106944,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7550873080581602e-05,
+      "loss": 0.8533,
+      "step": 52865
+    },
+    {
+      "epoch": 3.5922000271776056,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.7548749830139963e-05,
+      "loss": 0.8774,
+      "step": 52870
+    },
+    {
+      "epoch": 3.592539747248267,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.754662657969833e-05,
+      "loss": 0.9408,
+      "step": 52875
+    },
+    {
+      "epoch": 3.5928794673189293,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.7544503329256694e-05,
+      "loss": 0.9689,
+      "step": 52880
+    },
+    {
+      "epoch": 3.593219187389591,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7542380078815055e-05,
+      "loss": 0.8911,
+      "step": 52885
+    },
+    {
+      "epoch": 3.5935589074602525,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7540256828373422e-05,
+      "loss": 0.9274,
+      "step": 52890
+    },
+    {
+      "epoch": 3.5938986275309146,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7538133577931786e-05,
+      "loss": 0.9574,
+      "step": 52895
+    },
+    {
+      "epoch": 3.5942383476015762,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7536010327490147e-05,
+      "loss": 0.9125,
+      "step": 52900
+    },
+    {
+      "epoch": 3.594578067672238,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7533887077048514e-05,
+      "loss": 0.8801,
+      "step": 52905
+    },
+    {
+      "epoch": 3.5949177877429,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.7531763826606878e-05,
+      "loss": 0.8662,
+      "step": 52910
+    },
+    {
+      "epoch": 3.5952575078135616,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.752964057616524e-05,
+      "loss": 0.9372,
+      "step": 52915
+    },
+    {
+      "epoch": 3.595597227884223,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7527517325723606e-05,
+      "loss": 1.0252,
+      "step": 52920
+    },
+    {
+      "epoch": 3.5959369479548853,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7525394075281967e-05,
+      "loss": 0.9464,
+      "step": 52925
+    },
+    {
+      "epoch": 3.596276668025547,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.752327082484033e-05,
+      "loss": 0.9597,
+      "step": 52930
+    },
+    {
+      "epoch": 3.5966163880962085,
+      "grad_norm": 6.5,
+      "learning_rate": 2.7521147574398698e-05,
+      "loss": 0.9178,
+      "step": 52935
+    },
+    {
+      "epoch": 3.5969561081668706,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.751902432395706e-05,
+      "loss": 0.8962,
+      "step": 52940
+    },
+    {
+      "epoch": 3.5972958282375322,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.7516901073515423e-05,
+      "loss": 0.9155,
+      "step": 52945
+    },
+    {
+      "epoch": 3.597635548308194,
+      "grad_norm": 7.5,
+      "learning_rate": 2.751477782307379e-05,
+      "loss": 1.0166,
+      "step": 52950
+    },
+    {
+      "epoch": 3.597975268378856,
+      "grad_norm": 9.5,
+      "learning_rate": 2.751265457263215e-05,
+      "loss": 0.9052,
+      "step": 52955
+    },
+    {
+      "epoch": 3.5983149884495176,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.7510531322190515e-05,
+      "loss": 0.852,
+      "step": 52960
+    },
+    {
+      "epoch": 3.598654708520179,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.7508408071748882e-05,
+      "loss": 0.9164,
+      "step": 52965
+    },
+    {
+      "epoch": 3.5989944285908413,
+      "grad_norm": 7.75,
+      "learning_rate": 2.7506284821307243e-05,
+      "loss": 0.9954,
+      "step": 52970
+    },
+    {
+      "epoch": 3.599334148661503,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.7504161570865607e-05,
+      "loss": 0.9038,
+      "step": 52975
+    },
+    {
+      "epoch": 3.5996738687321646,
+      "grad_norm": 8.75,
+      "learning_rate": 2.7502038320423974e-05,
+      "loss": 0.7738,
+      "step": 52980
+    },
+    {
+      "epoch": 3.6000135888028266,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.7499915069982335e-05,
+      "loss": 0.8439,
+      "step": 52985
+    },
+    {
+      "epoch": 3.6003533088734883,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7497791819540702e-05,
+      "loss": 0.8924,
+      "step": 52990
+    },
+    {
+      "epoch": 3.60069302894415,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7495668569099063e-05,
+      "loss": 0.9907,
+      "step": 52995
+    },
+    {
+      "epoch": 3.601032749014812,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7493545318657427e-05,
+      "loss": 0.8534,
+      "step": 53000
+    },
+    {
+      "epoch": 3.6013724690854736,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7491422068215794e-05,
+      "loss": 0.8947,
+      "step": 53005
+    },
+    {
+      "epoch": 3.6017121891561352,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7489298817774155e-05,
+      "loss": 0.872,
+      "step": 53010
+    },
+    {
+      "epoch": 3.6020519092267973,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.748717556733252e-05,
+      "loss": 0.8628,
+      "step": 53015
+    },
+    {
+      "epoch": 3.602391629297459,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.7485052316890886e-05,
+      "loss": 0.8911,
+      "step": 53020
+    },
+    {
+      "epoch": 3.6027313493681206,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7482929066449247e-05,
+      "loss": 0.9095,
+      "step": 53025
+    },
+    {
+      "epoch": 3.6030710694387826,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.748080581600761e-05,
+      "loss": 0.9141,
+      "step": 53030
+    },
+    {
+      "epoch": 3.6034107895094443,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.7478682565565978e-05,
+      "loss": 0.8757,
+      "step": 53035
+    },
+    {
+      "epoch": 3.603750509580106,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.747655931512434e-05,
+      "loss": 0.8244,
+      "step": 53040
+    },
+    {
+      "epoch": 3.604090229650768,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7474436064682703e-05,
+      "loss": 0.8076,
+      "step": 53045
+    },
+    {
+      "epoch": 3.6044299497214296,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.747231281424107e-05,
+      "loss": 0.8654,
+      "step": 53050
+    },
+    {
+      "epoch": 3.6047696697920912,
+      "grad_norm": 8.25,
+      "learning_rate": 2.747018956379943e-05,
+      "loss": 0.9506,
+      "step": 53055
+    },
+    {
+      "epoch": 3.6051093898627533,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.746806631335779e-05,
+      "loss": 0.8279,
+      "step": 53060
+    },
+    {
+      "epoch": 3.605449109933415,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.7465943062916162e-05,
+      "loss": 0.8024,
+      "step": 53065
+    },
+    {
+      "epoch": 3.6057888300040766,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7463819812474523e-05,
+      "loss": 0.9185,
+      "step": 53070
+    },
+    {
+      "epoch": 3.6061285500747386,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7461696562032884e-05,
+      "loss": 0.9426,
+      "step": 53075
+    },
+    {
+      "epoch": 3.6064682701454003,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.745957331159125e-05,
+      "loss": 0.9684,
+      "step": 53080
+    },
+    {
+      "epoch": 3.606807990216062,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7457450061149615e-05,
+      "loss": 0.8402,
+      "step": 53085
+    },
+    {
+      "epoch": 3.607147710286724,
+      "grad_norm": 6.875,
+      "learning_rate": 2.7455326810707976e-05,
+      "loss": 0.9488,
+      "step": 53090
+    },
+    {
+      "epoch": 3.6074874303573856,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7453203560266343e-05,
+      "loss": 0.8974,
+      "step": 53095
+    },
+    {
+      "epoch": 3.6078271504280472,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7451080309824707e-05,
+      "loss": 0.9161,
+      "step": 53100
+    },
+    {
+      "epoch": 3.6081668704987093,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.7448957059383068e-05,
+      "loss": 0.8726,
+      "step": 53105
+    },
+    {
+      "epoch": 3.608506590569371,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.7446833808941435e-05,
+      "loss": 0.9119,
+      "step": 53110
+    },
+    {
+      "epoch": 3.6088463106400326,
+      "grad_norm": 7.625,
+      "learning_rate": 2.74447105584998e-05,
+      "loss": 0.9621,
+      "step": 53115
+    },
+    {
+      "epoch": 3.6091860307106947,
+      "grad_norm": 8.125,
+      "learning_rate": 2.744258730805816e-05,
+      "loss": 0.8207,
+      "step": 53120
+    },
+    {
+      "epoch": 3.6095257507813563,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7440464057616527e-05,
+      "loss": 0.8891,
+      "step": 53125
+    },
+    {
+      "epoch": 3.609865470852018,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.743834080717489e-05,
+      "loss": 0.8813,
+      "step": 53130
+    },
+    {
+      "epoch": 3.6102051909226796,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.743621755673325e-05,
+      "loss": 0.999,
+      "step": 53135
+    },
+    {
+      "epoch": 3.6105449109933416,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.743409430629162e-05,
+      "loss": 0.8674,
+      "step": 53140
+    },
+    {
+      "epoch": 3.6108846310640033,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.743197105584998e-05,
+      "loss": 0.8333,
+      "step": 53145
+    },
+    {
+      "epoch": 3.611224351134665,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7429847805408344e-05,
+      "loss": 0.9349,
+      "step": 53150
+    },
+    {
+      "epoch": 3.611564071205327,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.742772455496671e-05,
+      "loss": 0.8685,
+      "step": 53155
+    },
+    {
+      "epoch": 3.6119037912759886,
+      "grad_norm": 7.625,
+      "learning_rate": 2.742560130452507e-05,
+      "loss": 0.8775,
+      "step": 53160
+    },
+    {
+      "epoch": 3.6122435113466502,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7423478054083436e-05,
+      "loss": 0.9592,
+      "step": 53165
+    },
+    {
+      "epoch": 3.612583231417312,
+      "grad_norm": 6.75,
+      "learning_rate": 2.7421354803641803e-05,
+      "loss": 0.9809,
+      "step": 53170
+    },
+    {
+      "epoch": 3.612922951487974,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.7419231553200164e-05,
+      "loss": 0.9444,
+      "step": 53175
+    },
+    {
+      "epoch": 3.6132626715586356,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7417108302758528e-05,
+      "loss": 1.0226,
+      "step": 53180
+    },
+    {
+      "epoch": 3.613602391629297,
+      "grad_norm": 7.75,
+      "learning_rate": 2.7414985052316895e-05,
+      "loss": 0.9442,
+      "step": 53185
+    },
+    {
+      "epoch": 3.6139421116999593,
+      "grad_norm": 6.75,
+      "learning_rate": 2.7412861801875256e-05,
+      "loss": 0.9009,
+      "step": 53190
+    },
+    {
+      "epoch": 3.614281831770621,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7410738551433616e-05,
+      "loss": 0.9063,
+      "step": 53195
+    },
+    {
+      "epoch": 3.6146215518412825,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.7408615300991987e-05,
+      "loss": 0.8689,
+      "step": 53200
+    },
+    {
+      "epoch": 3.6149612719119446,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7406492050550348e-05,
+      "loss": 0.9473,
+      "step": 53205
+    },
+    {
+      "epoch": 3.6153009919826062,
+      "grad_norm": 6.625,
+      "learning_rate": 2.7404368800108708e-05,
+      "loss": 0.9188,
+      "step": 53210
+    },
+    {
+      "epoch": 3.615640712053268,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.7402245549667076e-05,
+      "loss": 0.8663,
+      "step": 53215
+    },
+    {
+      "epoch": 3.61598043212393,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.740012229922544e-05,
+      "loss": 0.8756,
+      "step": 53220
+    },
+    {
+      "epoch": 3.6163201521945916,
+      "grad_norm": 7.375,
+      "learning_rate": 2.73979990487838e-05,
+      "loss": 0.7975,
+      "step": 53225
+    },
+    {
+      "epoch": 3.616659872265253,
+      "grad_norm": 7.125,
+      "learning_rate": 2.7395875798342168e-05,
+      "loss": 0.87,
+      "step": 53230
+    },
+    {
+      "epoch": 3.6169995923359153,
+      "grad_norm": 7.125,
+      "learning_rate": 2.739375254790053e-05,
+      "loss": 0.9223,
+      "step": 53235
+    },
+    {
+      "epoch": 3.617339312406577,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.7391629297458892e-05,
+      "loss": 0.9309,
+      "step": 53240
+    },
+    {
+      "epoch": 3.6176790324772385,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.738950604701726e-05,
+      "loss": 0.8903,
+      "step": 53245
+    },
+    {
+      "epoch": 3.6180187525479006,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.7387382796575624e-05,
+      "loss": 0.8706,
+      "step": 53250
+    },
+    {
+      "epoch": 3.6183584726185622,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.7385259546133984e-05,
+      "loss": 0.8909,
+      "step": 53255
+    },
+    {
+      "epoch": 3.618698192689224,
+      "grad_norm": 7.625,
+      "learning_rate": 2.738313629569235e-05,
+      "loss": 0.8399,
+      "step": 53260
+    },
+    {
+      "epoch": 3.619037912759886,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7381013045250716e-05,
+      "loss": 0.98,
+      "step": 53265
+    },
+    {
+      "epoch": 3.6193776328305476,
+      "grad_norm": 7.875,
+      "learning_rate": 2.7378889794809076e-05,
+      "loss": 0.8221,
+      "step": 53270
+    },
+    {
+      "epoch": 3.619717352901209,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7376766544367444e-05,
+      "loss": 0.858,
+      "step": 53275
+    },
+    {
+      "epoch": 3.6200570729718713,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.7374643293925804e-05,
+      "loss": 0.9407,
+      "step": 53280
+    },
+    {
+      "epoch": 3.620396793042533,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.737252004348417e-05,
+      "loss": 0.8983,
+      "step": 53285
+    },
+    {
+      "epoch": 3.6207365131131946,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.7370396793042536e-05,
+      "loss": 0.8889,
+      "step": 53290
+    },
+    {
+      "epoch": 3.6210762331838566,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.7368273542600896e-05,
+      "loss": 0.9067,
+      "step": 53295
+    },
+    {
+      "epoch": 3.6214159532545183,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.736615029215926e-05,
+      "loss": 0.8869,
+      "step": 53300
+    },
+    {
+      "epoch": 3.62175567332518,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7364027041717628e-05,
+      "loss": 0.903,
+      "step": 53305
+    },
+    {
+      "epoch": 3.622095393395842,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.736190379127599e-05,
+      "loss": 0.9392,
+      "step": 53310
+    },
+    {
+      "epoch": 3.6224351134665036,
+      "grad_norm": 6.625,
+      "learning_rate": 2.7359780540834352e-05,
+      "loss": 0.834,
+      "step": 53315
+    },
+    {
+      "epoch": 3.6227748335371652,
+      "grad_norm": 8.875,
+      "learning_rate": 2.735765729039272e-05,
+      "loss": 0.8328,
+      "step": 53320
+    },
+    {
+      "epoch": 3.6231145536078273,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.735553403995108e-05,
+      "loss": 0.9232,
+      "step": 53325
+    },
+    {
+      "epoch": 3.623454273678489,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.7353410789509448e-05,
+      "loss": 0.8123,
+      "step": 53330
+    },
+    {
+      "epoch": 3.6237939937491506,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7351287539067812e-05,
+      "loss": 0.8241,
+      "step": 53335
+    },
+    {
+      "epoch": 3.6241337138198126,
+      "grad_norm": 6.625,
+      "learning_rate": 2.7349164288626172e-05,
+      "loss": 0.8567,
+      "step": 53340
+    },
+    {
+      "epoch": 3.6244734338904743,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.734704103818454e-05,
+      "loss": 0.9193,
+      "step": 53345
+    },
+    {
+      "epoch": 3.624813153961136,
+      "grad_norm": 6.375,
+      "learning_rate": 2.73449177877429e-05,
+      "loss": 0.8324,
+      "step": 53350
+    },
+    {
+      "epoch": 3.625152874031798,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.7342794537301264e-05,
+      "loss": 0.8736,
+      "step": 53355
+    },
+    {
+      "epoch": 3.6254925941024596,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7340671286859632e-05,
+      "loss": 0.958,
+      "step": 53360
+    },
+    {
+      "epoch": 3.6258323141731212,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7338548036417992e-05,
+      "loss": 0.8822,
+      "step": 53365
+    },
+    {
+      "epoch": 3.6261720342437833,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7336424785976356e-05,
+      "loss": 0.9107,
+      "step": 53370
+    },
+    {
+      "epoch": 3.626511754314445,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.7334301535534724e-05,
+      "loss": 0.8676,
+      "step": 53375
+    },
+    {
+      "epoch": 3.6268514743851066,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.7332178285093084e-05,
+      "loss": 0.9658,
+      "step": 53380
+    },
+    {
+      "epoch": 3.6271911944557687,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.733005503465145e-05,
+      "loss": 0.9799,
+      "step": 53385
+    },
+    {
+      "epoch": 3.6275309145264303,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7327931784209816e-05,
+      "loss": 0.8403,
+      "step": 53390
+    },
+    {
+      "epoch": 3.627870634597092,
+      "grad_norm": 6.09375,
+      "learning_rate": 2.7325808533768176e-05,
+      "loss": 0.8178,
+      "step": 53395
+    },
+    {
+      "epoch": 3.628210354667754,
+      "grad_norm": 6.875,
+      "learning_rate": 2.732368528332654e-05,
+      "loss": 0.8008,
+      "step": 53400
+    },
+    {
+      "epoch": 3.6285500747384156,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7321562032884908e-05,
+      "loss": 0.8495,
+      "step": 53405
+    },
+    {
+      "epoch": 3.6288897948090773,
+      "grad_norm": 6.0,
+      "learning_rate": 2.731943878244327e-05,
+      "loss": 0.8734,
+      "step": 53410
+    },
+    {
+      "epoch": 3.6292295148797393,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.731731553200163e-05,
+      "loss": 0.7761,
+      "step": 53415
+    },
+    {
+      "epoch": 3.629569234950401,
+      "grad_norm": 6.5,
+      "learning_rate": 2.7315192281559996e-05,
+      "loss": 0.8531,
+      "step": 53420
+    },
+    {
+      "epoch": 3.6299089550210626,
+      "grad_norm": 8.25,
+      "learning_rate": 2.731306903111836e-05,
+      "loss": 0.9276,
+      "step": 53425
+    },
+    {
+      "epoch": 3.6302486750917247,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.731094578067672e-05,
+      "loss": 0.8119,
+      "step": 53430
+    },
+    {
+      "epoch": 3.6305883951623863,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.730882253023509e-05,
+      "loss": 0.8384,
+      "step": 53435
+    },
+    {
+      "epoch": 3.630928115233048,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7306699279793452e-05,
+      "loss": 0.8721,
+      "step": 53440
+    },
+    {
+      "epoch": 3.63126783530371,
+      "grad_norm": 7.875,
+      "learning_rate": 2.7304576029351813e-05,
+      "loss": 0.8727,
+      "step": 53445
+    },
+    {
+      "epoch": 3.6316075553743716,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.730245277891018e-05,
+      "loss": 0.9315,
+      "step": 53450
+    },
+    {
+      "epoch": 3.6319472754450333,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.7300329528468544e-05,
+      "loss": 0.9515,
+      "step": 53455
+    },
+    {
+      "epoch": 3.6322869955156953,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.7298206278026905e-05,
+      "loss": 0.8196,
+      "step": 53460
+    },
+    {
+      "epoch": 3.632626715586357,
+      "grad_norm": 6.25,
+      "learning_rate": 2.7296083027585272e-05,
+      "loss": 0.9135,
+      "step": 53465
+    },
+    {
+      "epoch": 3.6329664356570186,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.7293959777143636e-05,
+      "loss": 0.9473,
+      "step": 53470
+    },
+    {
+      "epoch": 3.6333061557276802,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7291836526701997e-05,
+      "loss": 0.9688,
+      "step": 53475
+    },
+    {
+      "epoch": 3.6336458757983423,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.7289713276260364e-05,
+      "loss": 0.8343,
+      "step": 53480
+    },
+    {
+      "epoch": 3.633985595869004,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7287590025818725e-05,
+      "loss": 0.8409,
+      "step": 53485
+    },
+    {
+      "epoch": 3.6343253159396656,
+      "grad_norm": 8.125,
+      "learning_rate": 2.728546677537709e-05,
+      "loss": 0.9461,
+      "step": 53490
+    },
+    {
+      "epoch": 3.6346650360103276,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7283343524935456e-05,
+      "loss": 0.8846,
+      "step": 53495
+    },
+    {
+      "epoch": 3.6350047560809893,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.7281220274493817e-05,
+      "loss": 0.919,
+      "step": 53500
+    },
+    {
+      "epoch": 3.635344476151651,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.727909702405218e-05,
+      "loss": 0.8665,
+      "step": 53505
+    },
+    {
+      "epoch": 3.635684196222313,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.727697377361055e-05,
+      "loss": 0.9812,
+      "step": 53510
+    },
+    {
+      "epoch": 3.6360239162929746,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.727485052316891e-05,
+      "loss": 0.8527,
+      "step": 53515
+    },
+    {
+      "epoch": 3.6363636363636362,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.9176,
+      "step": 53520
+    },
+    {
+      "epoch": 3.636703356434298,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.727060402228564e-05,
+      "loss": 0.8383,
+      "step": 53525
+    },
+    {
+      "epoch": 3.63704307650496,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7268480771844e-05,
+      "loss": 0.957,
+      "step": 53530
+    },
+    {
+      "epoch": 3.6373827965756216,
+      "grad_norm": 7.875,
+      "learning_rate": 2.7266357521402365e-05,
+      "loss": 1.0111,
+      "step": 53535
+    },
+    {
+      "epoch": 3.637722516646283,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.7264234270960733e-05,
+      "loss": 0.9409,
+      "step": 53540
+    },
+    {
+      "epoch": 3.6380622367169453,
+      "grad_norm": 6.625,
+      "learning_rate": 2.7262111020519093e-05,
+      "loss": 0.9375,
+      "step": 53545
+    },
+    {
+      "epoch": 3.638401956787607,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.7259987770077454e-05,
+      "loss": 0.876,
+      "step": 53550
+    },
+    {
+      "epoch": 3.6387416768582685,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7257864519635825e-05,
+      "loss": 0.9817,
+      "step": 53555
+    },
+    {
+      "epoch": 3.6390813969289306,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.7255741269194185e-05,
+      "loss": 0.839,
+      "step": 53560
+    },
+    {
+      "epoch": 3.6394211169995923,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.7253618018752546e-05,
+      "loss": 0.8582,
+      "step": 53565
+    },
+    {
+      "epoch": 3.639760837070254,
+      "grad_norm": 8.125,
+      "learning_rate": 2.7251494768310913e-05,
+      "loss": 0.8958,
+      "step": 53570
+    },
+    {
+      "epoch": 3.640100557140916,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7249371517869277e-05,
+      "loss": 0.9911,
+      "step": 53575
+    },
+    {
+      "epoch": 3.6404402772115776,
+      "grad_norm": 7.125,
+      "learning_rate": 2.7247248267427638e-05,
+      "loss": 0.8691,
+      "step": 53580
+    },
+    {
+      "epoch": 3.6407799972822392,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7245125016986005e-05,
+      "loss": 0.838,
+      "step": 53585
+    },
+    {
+      "epoch": 3.6411197173529013,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.724300176654437e-05,
+      "loss": 0.9758,
+      "step": 53590
+    },
+    {
+      "epoch": 3.641459437423563,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.724087851610273e-05,
+      "loss": 0.9165,
+      "step": 53595
+    },
+    {
+      "epoch": 3.6417991574942246,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.7238755265661097e-05,
+      "loss": 0.9691,
+      "step": 53600
+    },
+    {
+      "epoch": 3.6421388775648866,
+      "grad_norm": 7.375,
+      "learning_rate": 2.723663201521946e-05,
+      "loss": 0.9488,
+      "step": 53605
+    },
+    {
+      "epoch": 3.6424785976355483,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.7234508764777822e-05,
+      "loss": 0.9583,
+      "step": 53610
+    },
+    {
+      "epoch": 3.64281831770621,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.723238551433619e-05,
+      "loss": 0.947,
+      "step": 53615
+    },
+    {
+      "epoch": 3.643158037776872,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.723026226389455e-05,
+      "loss": 0.8616,
+      "step": 53620
+    },
+    {
+      "epoch": 3.6434977578475336,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7228139013452914e-05,
+      "loss": 0.9045,
+      "step": 53625
+    },
+    {
+      "epoch": 3.6438374779181952,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.722601576301128e-05,
+      "loss": 1.0147,
+      "step": 53630
+    },
+    {
+      "epoch": 3.6441771979888573,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.7223892512569642e-05,
+      "loss": 0.8969,
+      "step": 53635
+    },
+    {
+      "epoch": 3.644516918059519,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7221769262128006e-05,
+      "loss": 0.9475,
+      "step": 53640
+    },
+    {
+      "epoch": 3.6448566381301806,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.7219646011686373e-05,
+      "loss": 0.888,
+      "step": 53645
+    },
+    {
+      "epoch": 3.6451963582008426,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.7217522761244734e-05,
+      "loss": 0.8777,
+      "step": 53650
+    },
+    {
+      "epoch": 3.6455360782715043,
+      "grad_norm": 6.75,
+      "learning_rate": 2.7215399510803098e-05,
+      "loss": 0.8846,
+      "step": 53655
+    },
+    {
+      "epoch": 3.645875798342166,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.7213276260361465e-05,
+      "loss": 0.8671,
+      "step": 53660
+    },
+    {
+      "epoch": 3.646215518412828,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.7211153009919826e-05,
+      "loss": 0.9492,
+      "step": 53665
+    },
+    {
+      "epoch": 3.6465552384834896,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7209029759478193e-05,
+      "loss": 0.9087,
+      "step": 53670
+    },
+    {
+      "epoch": 3.6468949585541512,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7206906509036557e-05,
+      "loss": 0.8932,
+      "step": 53675
+    },
+    {
+      "epoch": 3.6472346786248133,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7204783258594918e-05,
+      "loss": 0.8571,
+      "step": 53680
+    },
+    {
+      "epoch": 3.647574398695475,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.7202660008153285e-05,
+      "loss": 0.9914,
+      "step": 53685
+    },
+    {
+      "epoch": 3.6479141187661366,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.720053675771165e-05,
+      "loss": 0.8207,
+      "step": 53690
+    },
+    {
+      "epoch": 3.6482538388367987,
+      "grad_norm": 7.5,
+      "learning_rate": 2.719841350727001e-05,
+      "loss": 0.999,
+      "step": 53695
+    },
+    {
+      "epoch": 3.6485935589074603,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7196290256828377e-05,
+      "loss": 0.9999,
+      "step": 53700
+    },
+    {
+      "epoch": 3.648933278978122,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.7194167006386738e-05,
+      "loss": 0.8201,
+      "step": 53705
+    },
+    {
+      "epoch": 3.649272999048784,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.7192043755945102e-05,
+      "loss": 0.9608,
+      "step": 53710
+    },
+    {
+      "epoch": 3.6496127191194456,
+      "grad_norm": 7.0,
+      "learning_rate": 2.718992050550347e-05,
+      "loss": 0.9156,
+      "step": 53715
+    },
+    {
+      "epoch": 3.6499524391901073,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.718779725506183e-05,
+      "loss": 0.9256,
+      "step": 53720
+    },
+    {
+      "epoch": 3.6502921592607693,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.7185674004620194e-05,
+      "loss": 0.8272,
+      "step": 53725
+    },
+    {
+      "epoch": 3.650631879331431,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.718355075417856e-05,
+      "loss": 0.897,
+      "step": 53730
+    },
+    {
+      "epoch": 3.6509715994020926,
+      "grad_norm": 8.375,
+      "learning_rate": 2.7181427503736922e-05,
+      "loss": 0.8885,
+      "step": 53735
+    },
+    {
+      "epoch": 3.6513113194727547,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7179304253295286e-05,
+      "loss": 0.887,
+      "step": 53740
+    },
+    {
+      "epoch": 3.6516510395434163,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.7177181002853653e-05,
+      "loss": 0.7682,
+      "step": 53745
+    },
+    {
+      "epoch": 3.651990759614078,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.7175057752412014e-05,
+      "loss": 0.8502,
+      "step": 53750
+    },
+    {
+      "epoch": 3.65233047968474,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.7172934501970378e-05,
+      "loss": 0.9107,
+      "step": 53755
+    },
+    {
+      "epoch": 3.6526701997554016,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7170811251528745e-05,
+      "loss": 0.9483,
+      "step": 53760
+    },
+    {
+      "epoch": 3.6530099198260633,
+      "grad_norm": 7.125,
+      "learning_rate": 2.7168688001087106e-05,
+      "loss": 0.9444,
+      "step": 53765
+    },
+    {
+      "epoch": 3.6533496398967253,
+      "grad_norm": 8.375,
+      "learning_rate": 2.7166564750645467e-05,
+      "loss": 0.8704,
+      "step": 53770
+    },
+    {
+      "epoch": 3.653689359967387,
+      "grad_norm": 8.125,
+      "learning_rate": 2.7164441500203834e-05,
+      "loss": 0.912,
+      "step": 53775
+    },
+    {
+      "epoch": 3.6540290800380486,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7162318249762198e-05,
+      "loss": 0.9469,
+      "step": 53780
+    },
+    {
+      "epoch": 3.6543688001087107,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.716019499932056e-05,
+      "loss": 0.9665,
+      "step": 53785
+    },
+    {
+      "epoch": 3.6547085201793723,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7158071748878926e-05,
+      "loss": 0.913,
+      "step": 53790
+    },
+    {
+      "epoch": 3.655048240250034,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.715594849843729e-05,
+      "loss": 0.9283,
+      "step": 53795
+    },
+    {
+      "epoch": 3.655387960320696,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.715382524799565e-05,
+      "loss": 0.8814,
+      "step": 53800
+    },
+    {
+      "epoch": 3.6557276803913576,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.7151701997554018e-05,
+      "loss": 0.8982,
+      "step": 53805
+    },
+    {
+      "epoch": 3.6560674004620193,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.7149578747112382e-05,
+      "loss": 0.8587,
+      "step": 53810
+    },
+    {
+      "epoch": 3.656407120532681,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.7147455496670743e-05,
+      "loss": 0.9218,
+      "step": 53815
+    },
+    {
+      "epoch": 3.656746840603343,
+      "grad_norm": 7.0,
+      "learning_rate": 2.714533224622911e-05,
+      "loss": 0.9088,
+      "step": 53820
+    },
+    {
+      "epoch": 3.6570865606740046,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.7143208995787474e-05,
+      "loss": 0.8353,
+      "step": 53825
+    },
+    {
+      "epoch": 3.6574262807446662,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.7141085745345835e-05,
+      "loss": 0.9219,
+      "step": 53830
+    },
+    {
+      "epoch": 3.6577660008153283,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.7138962494904202e-05,
+      "loss": 0.8202,
+      "step": 53835
+    },
+    {
+      "epoch": 3.65810572088599,
+      "grad_norm": 7.5,
+      "learning_rate": 2.7136839244462563e-05,
+      "loss": 0.8874,
+      "step": 53840
+    },
+    {
+      "epoch": 3.6584454409566516,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.7134715994020927e-05,
+      "loss": 0.931,
+      "step": 53845
+    },
+    {
+      "epoch": 3.6587851610273137,
+      "grad_norm": 8.625,
+      "learning_rate": 2.7132592743579294e-05,
+      "loss": 0.8537,
+      "step": 53850
+    },
+    {
+      "epoch": 3.6591248810979753,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.7130469493137655e-05,
+      "loss": 0.8973,
+      "step": 53855
+    },
+    {
+      "epoch": 3.659464601168637,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.712834624269602e-05,
+      "loss": 0.9057,
+      "step": 53860
+    },
+    {
+      "epoch": 3.6598043212392986,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7126222992254386e-05,
+      "loss": 0.9052,
+      "step": 53865
+    },
+    {
+      "epoch": 3.6601440413099606,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7124099741812747e-05,
+      "loss": 0.86,
+      "step": 53870
+    },
+    {
+      "epoch": 3.6604837613806223,
+      "grad_norm": 7.75,
+      "learning_rate": 2.712197649137111e-05,
+      "loss": 0.9306,
+      "step": 53875
+    },
+    {
+      "epoch": 3.660823481451284,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.7119853240929478e-05,
+      "loss": 0.958,
+      "step": 53880
+    },
+    {
+      "epoch": 3.661163201521946,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.711772999048784e-05,
+      "loss": 0.8332,
+      "step": 53885
+    },
+    {
+      "epoch": 3.6615029215926076,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7115606740046203e-05,
+      "loss": 0.8679,
+      "step": 53890
+    },
+    {
+      "epoch": 3.6618426416632692,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.711348348960457e-05,
+      "loss": 0.9378,
+      "step": 53895
+    },
+    {
+      "epoch": 3.6621823617339313,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.711136023916293e-05,
+      "loss": 0.8845,
+      "step": 53900
+    },
+    {
+      "epoch": 3.662522081804593,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.710923698872129e-05,
+      "loss": 0.9346,
+      "step": 53905
+    },
+    {
+      "epoch": 3.6628618018752546,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.710711373827966e-05,
+      "loss": 0.9156,
+      "step": 53910
+    },
+    {
+      "epoch": 3.6632015219459166,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7104990487838023e-05,
+      "loss": 0.94,
+      "step": 53915
+    },
+    {
+      "epoch": 3.6635412420165783,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7102867237396383e-05,
+      "loss": 0.9815,
+      "step": 53920
+    },
+    {
+      "epoch": 3.66388096208724,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.710074398695475e-05,
+      "loss": 0.9275,
+      "step": 53925
+    },
+    {
+      "epoch": 3.664220682157902,
+      "grad_norm": 8.0,
+      "learning_rate": 2.7098620736513115e-05,
+      "loss": 0.9508,
+      "step": 53930
+    },
+    {
+      "epoch": 3.6645604022285636,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7096497486071475e-05,
+      "loss": 0.9329,
+      "step": 53935
+    },
+    {
+      "epoch": 3.6649001222992252,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.7094374235629843e-05,
+      "loss": 0.857,
+      "step": 53940
+    },
+    {
+      "epoch": 3.6652398423698873,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.7092250985188207e-05,
+      "loss": 0.9111,
+      "step": 53945
+    },
+    {
+      "epoch": 3.665579562440549,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7090127734746567e-05,
+      "loss": 0.8951,
+      "step": 53950
+    },
+    {
+      "epoch": 3.6659192825112106,
+      "grad_norm": 8.0,
+      "learning_rate": 2.7088004484304935e-05,
+      "loss": 0.9453,
+      "step": 53955
+    },
+    {
+      "epoch": 3.6662590025818727,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.70858812338633e-05,
+      "loss": 0.9396,
+      "step": 53960
+    },
+    {
+      "epoch": 3.6665987226525343,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.708375798342166e-05,
+      "loss": 0.8591,
+      "step": 53965
+    },
+    {
+      "epoch": 3.666938442723196,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.7081634732980027e-05,
+      "loss": 0.9504,
+      "step": 53970
+    },
+    {
+      "epoch": 3.667278162793858,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.7079511482538387e-05,
+      "loss": 1.0607,
+      "step": 53975
+    },
+    {
+      "epoch": 3.6676178828645196,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.707738823209675e-05,
+      "loss": 0.9146,
+      "step": 53980
+    },
+    {
+      "epoch": 3.6679576029351812,
+      "grad_norm": 8.0,
+      "learning_rate": 2.707526498165512e-05,
+      "loss": 0.8909,
+      "step": 53985
+    },
+    {
+      "epoch": 3.6682973230058433,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.707314173121348e-05,
+      "loss": 0.9496,
+      "step": 53990
+    },
+    {
+      "epoch": 3.668637043076505,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.7071018480771843e-05,
+      "loss": 0.8691,
+      "step": 53995
+    },
+    {
+      "epoch": 3.6689767631471666,
+      "grad_norm": 7.5,
+      "learning_rate": 2.706889523033021e-05,
+      "loss": 0.902,
+      "step": 54000
+    },
+    {
+      "epoch": 3.6693164832178287,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.706677197988857e-05,
+      "loss": 0.9634,
+      "step": 54005
+    },
+    {
+      "epoch": 3.6696562032884903,
+      "grad_norm": 6.375,
+      "learning_rate": 2.706464872944694e-05,
+      "loss": 0.9292,
+      "step": 54010
+    },
+    {
+      "epoch": 3.669995923359152,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.7062525479005303e-05,
+      "loss": 0.8986,
+      "step": 54015
+    },
+    {
+      "epoch": 3.670335643429814,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7060402228563663e-05,
+      "loss": 0.9469,
+      "step": 54020
+    },
+    {
+      "epoch": 3.6706753635004756,
+      "grad_norm": 7.25,
+      "learning_rate": 2.705827897812203e-05,
+      "loss": 0.9687,
+      "step": 54025
+    },
+    {
+      "epoch": 3.6710150835711373,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7056155727680395e-05,
+      "loss": 0.907,
+      "step": 54030
+    },
+    {
+      "epoch": 3.6713548036417993,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.7054032477238755e-05,
+      "loss": 0.9285,
+      "step": 54035
+    },
+    {
+      "epoch": 3.671694523712461,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.7051909226797123e-05,
+      "loss": 1.0389,
+      "step": 54040
+    },
+    {
+      "epoch": 3.6720342437831226,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.7049785976355487e-05,
+      "loss": 0.9061,
+      "step": 54045
+    },
+    {
+      "epoch": 3.6723739638537847,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7047662725913847e-05,
+      "loss": 0.9287,
+      "step": 54050
+    },
+    {
+      "epoch": 3.6727136839244463,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.7045539475472215e-05,
+      "loss": 0.9333,
+      "step": 54055
+    },
+    {
+      "epoch": 3.673053403995108,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.7043416225030575e-05,
+      "loss": 0.9206,
+      "step": 54060
+    },
+    {
+      "epoch": 3.67339312406577,
+      "grad_norm": 9.0,
+      "learning_rate": 2.704129297458894e-05,
+      "loss": 0.8733,
+      "step": 54065
+    },
+    {
+      "epoch": 3.6737328441364316,
+      "grad_norm": 6.5,
+      "learning_rate": 2.7039169724147307e-05,
+      "loss": 0.8306,
+      "step": 54070
+    },
+    {
+      "epoch": 3.6740725642070933,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7037046473705667e-05,
+      "loss": 0.9515,
+      "step": 54075
+    },
+    {
+      "epoch": 3.6744122842777553,
+      "grad_norm": 6.875,
+      "learning_rate": 2.703492322326403e-05,
+      "loss": 0.936,
+      "step": 54080
+    },
+    {
+      "epoch": 3.674752004348417,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.70327999728224e-05,
+      "loss": 0.8035,
+      "step": 54085
+    },
+    {
+      "epoch": 3.6750917244190786,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.703067672238076e-05,
+      "loss": 0.8997,
+      "step": 54090
+    },
+    {
+      "epoch": 3.6754314444897407,
+      "grad_norm": 7.0,
+      "learning_rate": 2.7028553471939123e-05,
+      "loss": 0.8938,
+      "step": 54095
+    },
+    {
+      "epoch": 3.6757711645604023,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.702643022149749e-05,
+      "loss": 0.9408,
+      "step": 54100
+    },
+    {
+      "epoch": 3.676110884631064,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.702430697105585e-05,
+      "loss": 0.8609,
+      "step": 54105
+    },
+    {
+      "epoch": 3.676450604701726,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7022183720614212e-05,
+      "loss": 0.826,
+      "step": 54110
+    },
+    {
+      "epoch": 3.6767903247723877,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.7020060470172583e-05,
+      "loss": 0.9545,
+      "step": 54115
+    },
+    {
+      "epoch": 3.6771300448430493,
+      "grad_norm": 7.375,
+      "learning_rate": 2.7017937219730943e-05,
+      "loss": 0.9165,
+      "step": 54120
+    },
+    {
+      "epoch": 3.6774697649137114,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7015813969289304e-05,
+      "loss": 0.8539,
+      "step": 54125
+    },
+    {
+      "epoch": 3.677809484984373,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.701369071884767e-05,
+      "loss": 0.889,
+      "step": 54130
+    },
+    {
+      "epoch": 3.6781492050550346,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.7011567468406035e-05,
+      "loss": 1.0299,
+      "step": 54135
+    },
+    {
+      "epoch": 3.6784889251256967,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.7009444217964396e-05,
+      "loss": 0.8842,
+      "step": 54140
+    },
+    {
+      "epoch": 3.6788286451963583,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7007320967522763e-05,
+      "loss": 0.8828,
+      "step": 54145
+    },
+    {
+      "epoch": 3.67916836526702,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.7005197717081127e-05,
+      "loss": 0.9376,
+      "step": 54150
+    },
+    {
+      "epoch": 3.6795080853376816,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7003074466639488e-05,
+      "loss": 0.9398,
+      "step": 54155
+    },
+    {
+      "epoch": 3.6798478054083437,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.7000951216197855e-05,
+      "loss": 0.9611,
+      "step": 54160
+    },
+    {
+      "epoch": 3.6801875254790053,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.699882796575622e-05,
+      "loss": 0.981,
+      "step": 54165
+    },
+    {
+      "epoch": 3.680527245549667,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.699670471531458e-05,
+      "loss": 0.8064,
+      "step": 54170
+    },
+    {
+      "epoch": 3.680866965620329,
+      "grad_norm": 6.5,
+      "learning_rate": 2.6994581464872947e-05,
+      "loss": 0.9212,
+      "step": 54175
+    },
+    {
+      "epoch": 3.6812066856909906,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.699245821443131e-05,
+      "loss": 1.0068,
+      "step": 54180
+    },
+    {
+      "epoch": 3.6815464057616523,
+      "grad_norm": 7.625,
+      "learning_rate": 2.6990334963989672e-05,
+      "loss": 0.9364,
+      "step": 54185
+    },
+    {
+      "epoch": 3.6818861258323143,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.698821171354804e-05,
+      "loss": 0.9409,
+      "step": 54190
+    },
+    {
+      "epoch": 3.682225845902976,
+      "grad_norm": 6.875,
+      "learning_rate": 2.69860884631064e-05,
+      "loss": 0.8922,
+      "step": 54195
+    },
+    {
+      "epoch": 3.6825655659736376,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.6983965212664764e-05,
+      "loss": 0.9109,
+      "step": 54200
+    },
+    {
+      "epoch": 3.6829052860442992,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.698184196222313e-05,
+      "loss": 0.8623,
+      "step": 54205
+    },
+    {
+      "epoch": 3.6832450061149613,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.6979718711781492e-05,
+      "loss": 0.9006,
+      "step": 54210
+    },
+    {
+      "epoch": 3.683584726185623,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6977595461339856e-05,
+      "loss": 0.9616,
+      "step": 54215
+    },
+    {
+      "epoch": 3.6839244462562846,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.6975472210898223e-05,
+      "loss": 0.879,
+      "step": 54220
+    },
+    {
+      "epoch": 3.6842641663269466,
+      "grad_norm": 8.625,
+      "learning_rate": 2.6973348960456584e-05,
+      "loss": 0.9963,
+      "step": 54225
+    },
+    {
+      "epoch": 3.6846038863976083,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6971225710014948e-05,
+      "loss": 1.0113,
+      "step": 54230
+    },
+    {
+      "epoch": 3.68494360646827,
+      "grad_norm": 9.0,
+      "learning_rate": 2.6969102459573315e-05,
+      "loss": 0.9554,
+      "step": 54235
+    },
+    {
+      "epoch": 3.685283326538932,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6966979209131676e-05,
+      "loss": 0.9611,
+      "step": 54240
+    },
+    {
+      "epoch": 3.6856230466095936,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.696485595869004e-05,
+      "loss": 0.8818,
+      "step": 54245
+    },
+    {
+      "epoch": 3.6859627666802552,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.6962732708248408e-05,
+      "loss": 0.8505,
+      "step": 54250
+    },
+    {
+      "epoch": 3.6863024867509173,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.6960609457806768e-05,
+      "loss": 0.9449,
+      "step": 54255
+    },
+    {
+      "epoch": 3.686642206821579,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.695848620736513e-05,
+      "loss": 0.8737,
+      "step": 54260
+    },
+    {
+      "epoch": 3.6869819268922406,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.6956362956923496e-05,
+      "loss": 1.0071,
+      "step": 54265
+    },
+    {
+      "epoch": 3.6873216469629027,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.695423970648186e-05,
+      "loss": 0.8936,
+      "step": 54270
+    },
+    {
+      "epoch": 3.6876613670335643,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.695211645604022e-05,
+      "loss": 0.8727,
+      "step": 54275
+    },
+    {
+      "epoch": 3.688001087104226,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6949993205598588e-05,
+      "loss": 0.8939,
+      "step": 54280
+    },
+    {
+      "epoch": 3.688340807174888,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.6947869955156952e-05,
+      "loss": 0.9272,
+      "step": 54285
+    },
+    {
+      "epoch": 3.6886805272455496,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6945746704715313e-05,
+      "loss": 0.9125,
+      "step": 54290
+    },
+    {
+      "epoch": 3.6890202473162113,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.694362345427368e-05,
+      "loss": 0.9097,
+      "step": 54295
+    },
+    {
+      "epoch": 3.6893599673868733,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6941500203832044e-05,
+      "loss": 0.9106,
+      "step": 54300
+    },
+    {
+      "epoch": 3.689699687457535,
+      "grad_norm": 8.25,
+      "learning_rate": 2.6939376953390405e-05,
+      "loss": 0.8518,
+      "step": 54305
+    },
+    {
+      "epoch": 3.6900394075281966,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6937253702948772e-05,
+      "loss": 0.7962,
+      "step": 54310
+    },
+    {
+      "epoch": 3.6903791275988587,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6935130452507136e-05,
+      "loss": 0.8896,
+      "step": 54315
+    },
+    {
+      "epoch": 3.6907188476695203,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.6933007202065497e-05,
+      "loss": 0.9483,
+      "step": 54320
+    },
+    {
+      "epoch": 3.691058567740182,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6930883951623864e-05,
+      "loss": 0.9253,
+      "step": 54325
+    },
+    {
+      "epoch": 3.691398287810844,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.6928760701182225e-05,
+      "loss": 0.8601,
+      "step": 54330
+    },
+    {
+      "epoch": 3.6917380078815056,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.692663745074059e-05,
+      "loss": 0.9062,
+      "step": 54335
+    },
+    {
+      "epoch": 3.6920777279521673,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6924514200298956e-05,
+      "loss": 0.8612,
+      "step": 54340
+    },
+    {
+      "epoch": 3.6924174480228293,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6922390949857317e-05,
+      "loss": 0.8822,
+      "step": 54345
+    },
+    {
+      "epoch": 3.692757168093491,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.6920267699415684e-05,
+      "loss": 1.0245,
+      "step": 54350
+    },
+    {
+      "epoch": 3.6930968881641526,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6918144448974048e-05,
+      "loss": 0.946,
+      "step": 54355
+    },
+    {
+      "epoch": 3.6934366082348147,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.691602119853241e-05,
+      "loss": 0.8448,
+      "step": 54360
+    },
+    {
+      "epoch": 3.6937763283054763,
+      "grad_norm": 7.75,
+      "learning_rate": 2.6913897948090776e-05,
+      "loss": 0.9696,
+      "step": 54365
+    },
+    {
+      "epoch": 3.694116048376138,
+      "grad_norm": 6.75,
+      "learning_rate": 2.691177469764914e-05,
+      "loss": 0.8796,
+      "step": 54370
+    },
+    {
+      "epoch": 3.6944557684468,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.69096514472075e-05,
+      "loss": 0.8897,
+      "step": 54375
+    },
+    {
+      "epoch": 3.6947954885174616,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6907528196765868e-05,
+      "loss": 0.8289,
+      "step": 54380
+    },
+    {
+      "epoch": 3.6951352085881233,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.6905404946324232e-05,
+      "loss": 0.9399,
+      "step": 54385
+    },
+    {
+      "epoch": 3.6954749286587854,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.6903281695882593e-05,
+      "loss": 0.9334,
+      "step": 54390
+    },
+    {
+      "epoch": 3.695814648729447,
+      "grad_norm": 8.125,
+      "learning_rate": 2.690115844544096e-05,
+      "loss": 0.8334,
+      "step": 54395
+    },
+    {
+      "epoch": 3.6961543688001086,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.689903519499932e-05,
+      "loss": 0.93,
+      "step": 54400
+    },
+    {
+      "epoch": 3.6964940888707707,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.6896911944557685e-05,
+      "loss": 0.9037,
+      "step": 54405
+    },
+    {
+      "epoch": 3.6968338089414323,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.6894788694116052e-05,
+      "loss": 0.9967,
+      "step": 54410
+    },
+    {
+      "epoch": 3.697173529012094,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.6892665443674413e-05,
+      "loss": 0.8303,
+      "step": 54415
+    },
+    {
+      "epoch": 3.697513249082756,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.6890542193232777e-05,
+      "loss": 0.8777,
+      "step": 54420
+    },
+    {
+      "epoch": 3.6978529691534177,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.6888418942791144e-05,
+      "loss": 0.8482,
+      "step": 54425
+    },
+    {
+      "epoch": 3.6981926892240793,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6886295692349505e-05,
+      "loss": 0.878,
+      "step": 54430
+    },
+    {
+      "epoch": 3.6985324092947414,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.688417244190787e-05,
+      "loss": 0.8721,
+      "step": 54435
+    },
+    {
+      "epoch": 3.698872129365403,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.6882049191466236e-05,
+      "loss": 0.8984,
+      "step": 54440
+    },
+    {
+      "epoch": 3.6992118494360646,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6879925941024597e-05,
+      "loss": 0.8357,
+      "step": 54445
+    },
+    {
+      "epoch": 3.6995515695067267,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.687780269058296e-05,
+      "loss": 0.8494,
+      "step": 54450
+    },
+    {
+      "epoch": 3.6998912895773883,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.6875679440141328e-05,
+      "loss": 0.9712,
+      "step": 54455
+    },
+    {
+      "epoch": 3.70023100964805,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.687355618969969e-05,
+      "loss": 0.8993,
+      "step": 54460
+    },
+    {
+      "epoch": 3.700570729718712,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.687143293925805e-05,
+      "loss": 0.8879,
+      "step": 54465
+    },
+    {
+      "epoch": 3.7009104497893737,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.686930968881642e-05,
+      "loss": 0.9958,
+      "step": 54470
+    },
+    {
+      "epoch": 3.7012501698600353,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.686718643837478e-05,
+      "loss": 1.0061,
+      "step": 54475
+    },
+    {
+      "epoch": 3.7015898899306974,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.686506318793314e-05,
+      "loss": 0.8899,
+      "step": 54480
+    },
+    {
+      "epoch": 3.701929610001359,
+      "grad_norm": 6.625,
+      "learning_rate": 2.686293993749151e-05,
+      "loss": 0.8704,
+      "step": 54485
+    },
+    {
+      "epoch": 3.7022693300720206,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6860816687049873e-05,
+      "loss": 0.9403,
+      "step": 54490
+    },
+    {
+      "epoch": 3.7026090501426823,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.6858693436608234e-05,
+      "loss": 0.9801,
+      "step": 54495
+    },
+    {
+      "epoch": 3.7029487702133443,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.68565701861666e-05,
+      "loss": 0.908,
+      "step": 54500
+    },
+    {
+      "epoch": 3.703288490284006,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6854446935724965e-05,
+      "loss": 0.867,
+      "step": 54505
+    },
+    {
+      "epoch": 3.7036282103546676,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6852323685283326e-05,
+      "loss": 0.9983,
+      "step": 54510
+    },
+    {
+      "epoch": 3.7039679304253297,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.6850200434841693e-05,
+      "loss": 0.8557,
+      "step": 54515
+    },
+    {
+      "epoch": 3.7043076504959913,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6848077184400057e-05,
+      "loss": 0.8616,
+      "step": 54520
+    },
+    {
+      "epoch": 3.704647370566653,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.6845953933958418e-05,
+      "loss": 0.8945,
+      "step": 54525
+    },
+    {
+      "epoch": 3.704987090637315,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.6843830683516785e-05,
+      "loss": 0.8981,
+      "step": 54530
+    },
+    {
+      "epoch": 3.7053268107079766,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6841707433075146e-05,
+      "loss": 0.9327,
+      "step": 54535
+    },
+    {
+      "epoch": 3.7056665307786383,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.683958418263351e-05,
+      "loss": 1.0063,
+      "step": 54540
+    },
+    {
+      "epoch": 3.7060062508493,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6837460932191877e-05,
+      "loss": 0.9578,
+      "step": 54545
+    },
+    {
+      "epoch": 3.706345970919962,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.6835337681750238e-05,
+      "loss": 0.9545,
+      "step": 54550
+    },
+    {
+      "epoch": 3.7066856909906236,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.68332144313086e-05,
+      "loss": 0.9204,
+      "step": 54555
+    },
+    {
+      "epoch": 3.7070254110612852,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.683109118086697e-05,
+      "loss": 0.8815,
+      "step": 54560
+    },
+    {
+      "epoch": 3.7073651311319473,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.682896793042533e-05,
+      "loss": 0.9157,
+      "step": 54565
+    },
+    {
+      "epoch": 3.707704851202609,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.6826844679983694e-05,
+      "loss": 0.9159,
+      "step": 54570
+    },
+    {
+      "epoch": 3.7080445712732706,
+      "grad_norm": 7.75,
+      "learning_rate": 2.682472142954206e-05,
+      "loss": 0.9,
+      "step": 54575
+    },
+    {
+      "epoch": 3.7083842913439327,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.682259817910042e-05,
+      "loss": 0.9321,
+      "step": 54580
+    },
+    {
+      "epoch": 3.7087240114145943,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.6820474928658786e-05,
+      "loss": 0.9147,
+      "step": 54585
+    },
+    {
+      "epoch": 3.709063731485256,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6818351678217153e-05,
+      "loss": 0.944,
+      "step": 54590
+    },
+    {
+      "epoch": 3.709403451555918,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6816228427775514e-05,
+      "loss": 0.9057,
+      "step": 54595
+    },
+    {
+      "epoch": 3.7097431716265796,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.6814105177333874e-05,
+      "loss": 0.9487,
+      "step": 54600
+    },
+    {
+      "epoch": 3.7100828916972413,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6811981926892245e-05,
+      "loss": 0.8853,
+      "step": 54605
+    },
+    {
+      "epoch": 3.7104226117679033,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.6809858676450606e-05,
+      "loss": 0.8273,
+      "step": 54610
+    },
+    {
+      "epoch": 3.710762331838565,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.6807735426008966e-05,
+      "loss": 0.9133,
+      "step": 54615
+    },
+    {
+      "epoch": 3.7111020519092266,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.6805612175567334e-05,
+      "loss": 0.8313,
+      "step": 54620
+    },
+    {
+      "epoch": 3.7114417719798887,
+      "grad_norm": 8.125,
+      "learning_rate": 2.6803488925125698e-05,
+      "loss": 0.8464,
+      "step": 54625
+    },
+    {
+      "epoch": 3.7117814920505503,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.6801365674684058e-05,
+      "loss": 0.8815,
+      "step": 54630
+    },
+    {
+      "epoch": 3.712121212121212,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.6799242424242426e-05,
+      "loss": 0.9778,
+      "step": 54635
+    },
+    {
+      "epoch": 3.712460932191874,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.679711917380079e-05,
+      "loss": 0.8482,
+      "step": 54640
+    },
+    {
+      "epoch": 3.7128006522625356,
+      "grad_norm": 9.125,
+      "learning_rate": 2.679499592335915e-05,
+      "loss": 0.9376,
+      "step": 54645
+    },
+    {
+      "epoch": 3.7131403723331973,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6792872672917518e-05,
+      "loss": 0.8578,
+      "step": 54650
+    },
+    {
+      "epoch": 3.7134800924038593,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.679074942247588e-05,
+      "loss": 0.8473,
+      "step": 54655
+    },
+    {
+      "epoch": 3.713819812474521,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6788626172034242e-05,
+      "loss": 0.8404,
+      "step": 54660
+    },
+    {
+      "epoch": 3.7141595325451826,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.678650292159261e-05,
+      "loss": 0.8156,
+      "step": 54665
+    },
+    {
+      "epoch": 3.7144992526158447,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6784379671150974e-05,
+      "loss": 0.9063,
+      "step": 54670
+    },
+    {
+      "epoch": 3.7148389726865063,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6782256420709334e-05,
+      "loss": 0.8568,
+      "step": 54675
+    },
+    {
+      "epoch": 3.715178692757168,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.67801331702677e-05,
+      "loss": 0.8739,
+      "step": 54680
+    },
+    {
+      "epoch": 3.71551841282783,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6778009919826062e-05,
+      "loss": 0.8779,
+      "step": 54685
+    },
+    {
+      "epoch": 3.7158581328984917,
+      "grad_norm": 8.0,
+      "learning_rate": 2.677588666938443e-05,
+      "loss": 0.9724,
+      "step": 54690
+    },
+    {
+      "epoch": 3.7161978529691533,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.6773763418942794e-05,
+      "loss": 0.8714,
+      "step": 54695
+    },
+    {
+      "epoch": 3.7165375730398154,
+      "grad_norm": 7.625,
+      "learning_rate": 2.6771640168501154e-05,
+      "loss": 0.9107,
+      "step": 54700
+    },
+    {
+      "epoch": 3.716877293110477,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.6769516918059522e-05,
+      "loss": 0.8802,
+      "step": 54705
+    },
+    {
+      "epoch": 3.7172170131811386,
+      "grad_norm": 6.125,
+      "learning_rate": 2.6767393667617886e-05,
+      "loss": 0.8733,
+      "step": 54710
+    },
+    {
+      "epoch": 3.7175567332518007,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6765270417176246e-05,
+      "loss": 0.8858,
+      "step": 54715
+    },
+    {
+      "epoch": 3.7178964533224623,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6763147166734614e-05,
+      "loss": 0.9428,
+      "step": 54720
+    },
+    {
+      "epoch": 3.718236173393124,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6761023916292978e-05,
+      "loss": 0.9496,
+      "step": 54725
+    },
+    {
+      "epoch": 3.718575893463786,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.675890066585134e-05,
+      "loss": 0.9816,
+      "step": 54730
+    },
+    {
+      "epoch": 3.7189156135344477,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6756777415409706e-05,
+      "loss": 0.8464,
+      "step": 54735
+    },
+    {
+      "epoch": 3.7192553336051093,
+      "grad_norm": 7.0,
+      "learning_rate": 2.675465416496807e-05,
+      "loss": 0.8971,
+      "step": 54740
+    },
+    {
+      "epoch": 3.7195950536757714,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.675253091452643e-05,
+      "loss": 0.898,
+      "step": 54745
+    },
+    {
+      "epoch": 3.719934773746433,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6750407664084798e-05,
+      "loss": 0.908,
+      "step": 54750
+    },
+    {
+      "epoch": 3.7202744938170946,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.674828441364316e-05,
+      "loss": 0.8848,
+      "step": 54755
+    },
+    {
+      "epoch": 3.7206142138877567,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.6746161163201522e-05,
+      "loss": 0.9695,
+      "step": 54760
+    },
+    {
+      "epoch": 3.7209539339584183,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.674403791275989e-05,
+      "loss": 0.9874,
+      "step": 54765
+    },
+    {
+      "epoch": 3.72129365402908,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.674191466231825e-05,
+      "loss": 0.9512,
+      "step": 54770
+    },
+    {
+      "epoch": 3.721633374099742,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.6739791411876614e-05,
+      "loss": 0.9023,
+      "step": 54775
+    },
+    {
+      "epoch": 3.7219730941704037,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.6737668161434982e-05,
+      "loss": 0.9282,
+      "step": 54780
+    },
+    {
+      "epoch": 3.7223128142410653,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6735544910993342e-05,
+      "loss": 0.913,
+      "step": 54785
+    },
+    {
+      "epoch": 3.7226525343117274,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6733421660551706e-05,
+      "loss": 0.8622,
+      "step": 54790
+    },
+    {
+      "epoch": 3.722992254382389,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.6731298410110074e-05,
+      "loss": 0.882,
+      "step": 54795
+    },
+    {
+      "epoch": 3.7233319744530506,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.6729175159668434e-05,
+      "loss": 0.835,
+      "step": 54800
+    },
+    {
+      "epoch": 3.7236716945237127,
+      "grad_norm": 8.625,
+      "learning_rate": 2.67270519092268e-05,
+      "loss": 0.8633,
+      "step": 54805
+    },
+    {
+      "epoch": 3.7240114145943743,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6724928658785166e-05,
+      "loss": 0.9177,
+      "step": 54810
+    },
+    {
+      "epoch": 3.724351134665036,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6722805408343526e-05,
+      "loss": 0.8944,
+      "step": 54815
+    },
+    {
+      "epoch": 3.724690854735698,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.6720682157901887e-05,
+      "loss": 0.9234,
+      "step": 54820
+    },
+    {
+      "epoch": 3.7250305748063597,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.6718558907460254e-05,
+      "loss": 0.8693,
+      "step": 54825
+    },
+    {
+      "epoch": 3.7253702948770213,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.671643565701862e-05,
+      "loss": 0.8345,
+      "step": 54830
+    },
+    {
+      "epoch": 3.725710014947683,
+      "grad_norm": 7.0,
+      "learning_rate": 2.671431240657698e-05,
+      "loss": 0.8365,
+      "step": 54835
+    },
+    {
+      "epoch": 3.726049735018345,
+      "grad_norm": 7.75,
+      "learning_rate": 2.6712189156135346e-05,
+      "loss": 0.9031,
+      "step": 54840
+    },
+    {
+      "epoch": 3.7263894550890067,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.671006590569371e-05,
+      "loss": 0.8966,
+      "step": 54845
+    },
+    {
+      "epoch": 3.7267291751596683,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.670794265525207e-05,
+      "loss": 0.9087,
+      "step": 54850
+    },
+    {
+      "epoch": 3.7270688952303304,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.670581940481044e-05,
+      "loss": 0.9853,
+      "step": 54855
+    },
+    {
+      "epoch": 3.727408615300992,
+      "grad_norm": 6.875,
+      "learning_rate": 2.6703696154368802e-05,
+      "loss": 0.9198,
+      "step": 54860
+    },
+    {
+      "epoch": 3.7277483353716536,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6701572903927163e-05,
+      "loss": 0.9192,
+      "step": 54865
+    },
+    {
+      "epoch": 3.7280880554423157,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.669944965348553e-05,
+      "loss": 0.9383,
+      "step": 54870
+    },
+    {
+      "epoch": 3.7284277755129773,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6697326403043894e-05,
+      "loss": 1.0353,
+      "step": 54875
+    },
+    {
+      "epoch": 3.728767495583639,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6695203152602255e-05,
+      "loss": 0.9037,
+      "step": 54880
+    },
+    {
+      "epoch": 3.7291072156543006,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.6693079902160622e-05,
+      "loss": 0.8947,
+      "step": 54885
+    },
+    {
+      "epoch": 3.7294469357249627,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.6690956651718983e-05,
+      "loss": 0.9214,
+      "step": 54890
+    },
+    {
+      "epoch": 3.7297866557956243,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6688833401277347e-05,
+      "loss": 0.8546,
+      "step": 54895
+    },
+    {
+      "epoch": 3.730126375866286,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.6686710150835714e-05,
+      "loss": 0.8496,
+      "step": 54900
+    },
+    {
+      "epoch": 3.730466095936948,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6684586900394075e-05,
+      "loss": 0.8798,
+      "step": 54905
+    },
+    {
+      "epoch": 3.7308058160076096,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.668246364995244e-05,
+      "loss": 0.9244,
+      "step": 54910
+    },
+    {
+      "epoch": 3.7311455360782713,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6680340399510806e-05,
+      "loss": 0.8769,
+      "step": 54915
+    },
+    {
+      "epoch": 3.7314852561489333,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.6678217149069167e-05,
+      "loss": 0.8948,
+      "step": 54920
+    },
+    {
+      "epoch": 3.731824976219595,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.667609389862753e-05,
+      "loss": 0.9904,
+      "step": 54925
+    },
+    {
+      "epoch": 3.7321646962902566,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.66739706481859e-05,
+      "loss": 0.8642,
+      "step": 54930
+    },
+    {
+      "epoch": 3.7325044163609187,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.667184739774426e-05,
+      "loss": 0.9875,
+      "step": 54935
+    },
+    {
+      "epoch": 3.7328441364315803,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.6669724147302623e-05,
+      "loss": 0.8213,
+      "step": 54940
+    },
+    {
+      "epoch": 3.733183856502242,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.666760089686099e-05,
+      "loss": 0.8275,
+      "step": 54945
+    },
+    {
+      "epoch": 3.733523576572904,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.666547764641935e-05,
+      "loss": 0.9412,
+      "step": 54950
+    },
+    {
+      "epoch": 3.7338632966435656,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6663354395977712e-05,
+      "loss": 0.8451,
+      "step": 54955
+    },
+    {
+      "epoch": 3.7342030167142273,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6661231145536083e-05,
+      "loss": 0.9488,
+      "step": 54960
+    },
+    {
+      "epoch": 3.7345427367848893,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6659107895094443e-05,
+      "loss": 0.8563,
+      "step": 54965
+    },
+    {
+      "epoch": 3.734882456855551,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6656984644652804e-05,
+      "loss": 0.9842,
+      "step": 54970
+    },
+    {
+      "epoch": 3.7352221769262126,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.665486139421117e-05,
+      "loss": 0.8471,
+      "step": 54975
+    },
+    {
+      "epoch": 3.7355618969968747,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6652738143769535e-05,
+      "loss": 0.8548,
+      "step": 54980
+    },
+    {
+      "epoch": 3.7359016170675363,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.6650614893327896e-05,
+      "loss": 0.9035,
+      "step": 54985
+    },
+    {
+      "epoch": 3.736241337138198,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6648491642886263e-05,
+      "loss": 0.967,
+      "step": 54990
+    },
+    {
+      "epoch": 3.73658105720886,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.6646368392444627e-05,
+      "loss": 0.889,
+      "step": 54995
+    },
+    {
+      "epoch": 3.7369207772795217,
+      "grad_norm": 10.6875,
+      "learning_rate": 2.6644245142002988e-05,
+      "loss": 0.8387,
+      "step": 55000
+    },
+    {
+      "epoch": 3.7372604973501833,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6642121891561355e-05,
+      "loss": 0.8988,
+      "step": 55005
+    },
+    {
+      "epoch": 3.7376002174208454,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.663999864111972e-05,
+      "loss": 0.9077,
+      "step": 55010
+    },
+    {
+      "epoch": 3.737939937491507,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.663787539067808e-05,
+      "loss": 0.9349,
+      "step": 55015
+    },
+    {
+      "epoch": 3.7382796575621686,
+      "grad_norm": 7.75,
+      "learning_rate": 2.6635752140236447e-05,
+      "loss": 0.8976,
+      "step": 55020
+    },
+    {
+      "epoch": 3.7386193776328307,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.6633628889794808e-05,
+      "loss": 0.9028,
+      "step": 55025
+    },
+    {
+      "epoch": 3.7389590977034923,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.663150563935318e-05,
+      "loss": 0.9525,
+      "step": 55030
+    },
+    {
+      "epoch": 3.739298817774154,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.662938238891154e-05,
+      "loss": 0.861,
+      "step": 55035
+    },
+    {
+      "epoch": 3.739638537844816,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.66272591384699e-05,
+      "loss": 0.9108,
+      "step": 55040
+    },
+    {
+      "epoch": 3.7399782579154777,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6625135888028267e-05,
+      "loss": 0.8405,
+      "step": 55045
+    },
+    {
+      "epoch": 3.7403179779861393,
+      "grad_norm": 6.875,
+      "learning_rate": 2.662301263758663e-05,
+      "loss": 0.839,
+      "step": 55050
+    },
+    {
+      "epoch": 3.7406576980568014,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.6620889387144992e-05,
+      "loss": 0.9241,
+      "step": 55055
+    },
+    {
+      "epoch": 3.740997418127463,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.661876613670336e-05,
+      "loss": 0.9271,
+      "step": 55060
+    },
+    {
+      "epoch": 3.7413371381981246,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.6616642886261723e-05,
+      "loss": 0.9623,
+      "step": 55065
+    },
+    {
+      "epoch": 3.7416768582687867,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.6614519635820084e-05,
+      "loss": 0.9066,
+      "step": 55070
+    },
+    {
+      "epoch": 3.7420165783394483,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.661239638537845e-05,
+      "loss": 0.995,
+      "step": 55075
+    },
+    {
+      "epoch": 3.74235629841011,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6610273134936815e-05,
+      "loss": 0.9197,
+      "step": 55080
+    },
+    {
+      "epoch": 3.742696018480772,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6608149884495176e-05,
+      "loss": 0.9762,
+      "step": 55085
+    },
+    {
+      "epoch": 3.7430357385514337,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6606026634053543e-05,
+      "loss": 0.8687,
+      "step": 55090
+    },
+    {
+      "epoch": 3.7433754586220953,
+      "grad_norm": 6.5,
+      "learning_rate": 2.6603903383611907e-05,
+      "loss": 0.8705,
+      "step": 55095
+    },
+    {
+      "epoch": 3.7437151786927574,
+      "grad_norm": 8.5,
+      "learning_rate": 2.6601780133170268e-05,
+      "loss": 0.8349,
+      "step": 55100
+    },
+    {
+      "epoch": 3.744054898763419,
+      "grad_norm": 7.625,
+      "learning_rate": 2.6599656882728635e-05,
+      "loss": 0.8121,
+      "step": 55105
+    },
+    {
+      "epoch": 3.7443946188340806,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.6597533632286996e-05,
+      "loss": 0.9399,
+      "step": 55110
+    },
+    {
+      "epoch": 3.7447343389047427,
+      "grad_norm": 7.125,
+      "learning_rate": 2.659541038184536e-05,
+      "loss": 0.9233,
+      "step": 55115
+    },
+    {
+      "epoch": 3.7450740589754044,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6593287131403727e-05,
+      "loss": 0.8578,
+      "step": 55120
+    },
+    {
+      "epoch": 3.745413779046066,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.6591163880962088e-05,
+      "loss": 0.9249,
+      "step": 55125
+    },
+    {
+      "epoch": 3.745753499116728,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.6589040630520452e-05,
+      "loss": 0.8571,
+      "step": 55130
+    },
+    {
+      "epoch": 3.7460932191873897,
+      "grad_norm": 7.25,
+      "learning_rate": 2.658691738007882e-05,
+      "loss": 0.9447,
+      "step": 55135
+    },
+    {
+      "epoch": 3.7464329392580513,
+      "grad_norm": 7.625,
+      "learning_rate": 2.658479412963718e-05,
+      "loss": 0.9309,
+      "step": 55140
+    },
+    {
+      "epoch": 3.7467726593287134,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.6582670879195544e-05,
+      "loss": 0.9013,
+      "step": 55145
+    },
+    {
+      "epoch": 3.747112379399375,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.658054762875391e-05,
+      "loss": 0.938,
+      "step": 55150
+    },
+    {
+      "epoch": 3.7474520994700367,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6578424378312272e-05,
+      "loss": 0.9411,
+      "step": 55155
+    },
+    {
+      "epoch": 3.7477918195406987,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.6576301127870636e-05,
+      "loss": 0.8834,
+      "step": 55160
+    },
+    {
+      "epoch": 3.7481315396113604,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6574177877429003e-05,
+      "loss": 0.906,
+      "step": 55165
+    },
+    {
+      "epoch": 3.748471259682022,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.6572054626987364e-05,
+      "loss": 0.8243,
+      "step": 55170
+    },
+    {
+      "epoch": 3.7488109797526836,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6569931376545725e-05,
+      "loss": 0.9938,
+      "step": 55175
+    },
+    {
+      "epoch": 3.7491506998233457,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6567808126104092e-05,
+      "loss": 0.9642,
+      "step": 55180
+    },
+    {
+      "epoch": 3.7494904198940073,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6565684875662456e-05,
+      "loss": 0.8762,
+      "step": 55185
+    },
+    {
+      "epoch": 3.749830139964669,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.6563561625220817e-05,
+      "loss": 0.9681,
+      "step": 55190
+    },
+    {
+      "epoch": 3.750169860035331,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6561438374779184e-05,
+      "loss": 0.8413,
+      "step": 55195
+    },
+    {
+      "epoch": 3.7505095801059927,
+      "grad_norm": 8.125,
+      "learning_rate": 2.6559315124337548e-05,
+      "loss": 0.8722,
+      "step": 55200
+    },
+    {
+      "epoch": 3.7508493001766543,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.655719187389591e-05,
+      "loss": 0.9105,
+      "step": 55205
+    },
+    {
+      "epoch": 3.7511890202473164,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.6555068623454276e-05,
+      "loss": 0.8722,
+      "step": 55210
+    },
+    {
+      "epoch": 3.751528740317978,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.655294537301264e-05,
+      "loss": 0.8491,
+      "step": 55215
+    },
+    {
+      "epoch": 3.7518684603886396,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.6550822122571e-05,
+      "loss": 0.9442,
+      "step": 55220
+    },
+    {
+      "epoch": 3.7522081804593013,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6548698872129368e-05,
+      "loss": 0.9627,
+      "step": 55225
+    },
+    {
+      "epoch": 3.7525479005299633,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.6546575621687732e-05,
+      "loss": 0.9791,
+      "step": 55230
+    },
+    {
+      "epoch": 3.752887620600625,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6544452371246093e-05,
+      "loss": 0.9049,
+      "step": 55235
+    },
+    {
+      "epoch": 3.7532273406712866,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.654232912080446e-05,
+      "loss": 0.8478,
+      "step": 55240
+    },
+    {
+      "epoch": 3.7535670607419487,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.654020587036282e-05,
+      "loss": 0.9865,
+      "step": 55245
+    },
+    {
+      "epoch": 3.7539067808126103,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6538082619921185e-05,
+      "loss": 0.8869,
+      "step": 55250
+    },
+    {
+      "epoch": 3.754246500883272,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6535959369479552e-05,
+      "loss": 0.9008,
+      "step": 55255
+    },
+    {
+      "epoch": 3.754586220953934,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6533836119037913e-05,
+      "loss": 0.9507,
+      "step": 55260
+    },
+    {
+      "epoch": 3.7549259410245956,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6531712868596277e-05,
+      "loss": 0.8908,
+      "step": 55265
+    },
+    {
+      "epoch": 3.7552656610952573,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.6529589618154644e-05,
+      "loss": 0.8415,
+      "step": 55270
+    },
+    {
+      "epoch": 3.7556053811659194,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.6527466367713005e-05,
+      "loss": 0.9389,
+      "step": 55275
+    },
+    {
+      "epoch": 3.755945101236581,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.652534311727137e-05,
+      "loss": 0.9226,
+      "step": 55280
+    },
+    {
+      "epoch": 3.7562848213072426,
+      "grad_norm": 7.625,
+      "learning_rate": 2.6523219866829736e-05,
+      "loss": 0.8538,
+      "step": 55285
+    },
+    {
+      "epoch": 3.7566245413779047,
+      "grad_norm": 8.125,
+      "learning_rate": 2.6521096616388097e-05,
+      "loss": 0.9121,
+      "step": 55290
+    },
+    {
+      "epoch": 3.7569642614485663,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.651897336594646e-05,
+      "loss": 0.8946,
+      "step": 55295
+    },
+    {
+      "epoch": 3.757303981519228,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6516850115504828e-05,
+      "loss": 0.9118,
+      "step": 55300
+    },
+    {
+      "epoch": 3.75764370158989,
+      "grad_norm": 7.375,
+      "learning_rate": 2.651472686506319e-05,
+      "loss": 0.8799,
+      "step": 55305
+    },
+    {
+      "epoch": 3.7579834216605517,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.651260361462155e-05,
+      "loss": 0.8817,
+      "step": 55310
+    },
+    {
+      "epoch": 3.7583231417312133,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.6510480364179917e-05,
+      "loss": 0.9525,
+      "step": 55315
+    },
+    {
+      "epoch": 3.7586628618018754,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.650835711373828e-05,
+      "loss": 0.7895,
+      "step": 55320
+    },
+    {
+      "epoch": 3.759002581872537,
+      "grad_norm": 8.5,
+      "learning_rate": 2.650623386329664e-05,
+      "loss": 0.9236,
+      "step": 55325
+    },
+    {
+      "epoch": 3.7593423019431986,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.650411061285501e-05,
+      "loss": 0.952,
+      "step": 55330
+    },
+    {
+      "epoch": 3.7596820220138607,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6501987362413373e-05,
+      "loss": 0.9258,
+      "step": 55335
+    },
+    {
+      "epoch": 3.7600217420845223,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.6499864111971733e-05,
+      "loss": 0.9642,
+      "step": 55340
+    },
+    {
+      "epoch": 3.760361462155184,
+      "grad_norm": 7.25,
+      "learning_rate": 2.64977408615301e-05,
+      "loss": 0.9465,
+      "step": 55345
+    },
+    {
+      "epoch": 3.760701182225846,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6495617611088465e-05,
+      "loss": 0.8289,
+      "step": 55350
+    },
+    {
+      "epoch": 3.7610409022965077,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.6493494360646825e-05,
+      "loss": 1.0676,
+      "step": 55355
+    },
+    {
+      "epoch": 3.7613806223671693,
+      "grad_norm": 6.5,
+      "learning_rate": 2.6491371110205193e-05,
+      "loss": 0.9163,
+      "step": 55360
+    },
+    {
+      "epoch": 3.7617203424378314,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.6489247859763557e-05,
+      "loss": 0.8631,
+      "step": 55365
+    },
+    {
+      "epoch": 3.762060062508493,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6487124609321924e-05,
+      "loss": 0.7744,
+      "step": 55370
+    },
+    {
+      "epoch": 3.7623997825791546,
+      "grad_norm": 9.6875,
+      "learning_rate": 2.6485001358880285e-05,
+      "loss": 0.9,
+      "step": 55375
+    },
+    {
+      "epoch": 3.7627395026498167,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6482878108438645e-05,
+      "loss": 1.0487,
+      "step": 55380
+    },
+    {
+      "epoch": 3.7630792227204783,
+      "grad_norm": 6.625,
+      "learning_rate": 2.6480754857997016e-05,
+      "loss": 0.8807,
+      "step": 55385
+    },
+    {
+      "epoch": 3.76341894279114,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.6478631607555377e-05,
+      "loss": 0.8728,
+      "step": 55390
+    },
+    {
+      "epoch": 3.763758662861802,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6476508357113737e-05,
+      "loss": 1.0186,
+      "step": 55395
+    },
+    {
+      "epoch": 3.7640983829324637,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6474385106672105e-05,
+      "loss": 0.8515,
+      "step": 55400
+    },
+    {
+      "epoch": 3.7644381030031253,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.647226185623047e-05,
+      "loss": 0.8901,
+      "step": 55405
+    },
+    {
+      "epoch": 3.7647778230737874,
+      "grad_norm": 8.0,
+      "learning_rate": 2.647013860578883e-05,
+      "loss": 0.8901,
+      "step": 55410
+    },
+    {
+      "epoch": 3.765117543144449,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6468015355347197e-05,
+      "loss": 0.952,
+      "step": 55415
+    },
+    {
+      "epoch": 3.7654572632151107,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.646589210490556e-05,
+      "loss": 0.9033,
+      "step": 55420
+    },
+    {
+      "epoch": 3.7657969832857727,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.646376885446392e-05,
+      "loss": 0.8539,
+      "step": 55425
+    },
+    {
+      "epoch": 3.7661367033564344,
+      "grad_norm": 7.625,
+      "learning_rate": 2.646164560402229e-05,
+      "loss": 0.8756,
+      "step": 55430
+    },
+    {
+      "epoch": 3.766476423427096,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6459522353580653e-05,
+      "loss": 0.8892,
+      "step": 55435
+    },
+    {
+      "epoch": 3.766816143497758,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6457399103139013e-05,
+      "loss": 0.9774,
+      "step": 55440
+    },
+    {
+      "epoch": 3.7671558635684197,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.645527585269738e-05,
+      "loss": 0.7492,
+      "step": 55445
+    },
+    {
+      "epoch": 3.7674955836390813,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.645315260225574e-05,
+      "loss": 0.9275,
+      "step": 55450
+    },
+    {
+      "epoch": 3.7678353037097434,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.6451029351814105e-05,
+      "loss": 0.8844,
+      "step": 55455
+    },
+    {
+      "epoch": 3.768175023780405,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.6448906101372473e-05,
+      "loss": 0.8219,
+      "step": 55460
+    },
+    {
+      "epoch": 3.7685147438510667,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.6446782850930833e-05,
+      "loss": 0.8785,
+      "step": 55465
+    },
+    {
+      "epoch": 3.7688544639217287,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6444659600489197e-05,
+      "loss": 0.8438,
+      "step": 55470
+    },
+    {
+      "epoch": 3.7691941839923904,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.6442536350047565e-05,
+      "loss": 0.8824,
+      "step": 55475
+    },
+    {
+      "epoch": 3.769533904063052,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6440413099605925e-05,
+      "loss": 0.8722,
+      "step": 55480
+    },
+    {
+      "epoch": 3.769873624133714,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.643828984916429e-05,
+      "loss": 0.9519,
+      "step": 55485
+    },
+    {
+      "epoch": 3.7702133442043757,
+      "grad_norm": 6.5,
+      "learning_rate": 2.6436166598722657e-05,
+      "loss": 0.8856,
+      "step": 55490
+    },
+    {
+      "epoch": 3.7705530642750373,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.6434043348281017e-05,
+      "loss": 1.0389,
+      "step": 55495
+    },
+    {
+      "epoch": 3.7708927843456994,
+      "grad_norm": 7.5,
+      "learning_rate": 2.643192009783938e-05,
+      "loss": 0.9429,
+      "step": 55500
+    },
+    {
+      "epoch": 3.771232504416361,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.642979684739775e-05,
+      "loss": 0.848,
+      "step": 55505
+    },
+    {
+      "epoch": 3.7715722244870227,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.642767359695611e-05,
+      "loss": 0.8987,
+      "step": 55510
+    },
+    {
+      "epoch": 3.7719119445576843,
+      "grad_norm": 7.75,
+      "learning_rate": 2.642555034651447e-05,
+      "loss": 0.8842,
+      "step": 55515
+    },
+    {
+      "epoch": 3.7722516646283464,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.642342709607284e-05,
+      "loss": 0.9764,
+      "step": 55520
+    },
+    {
+      "epoch": 3.772591384699008,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.64213038456312e-05,
+      "loss": 0.8703,
+      "step": 55525
+    },
+    {
+      "epoch": 3.7729311047696696,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6419180595189562e-05,
+      "loss": 0.9984,
+      "step": 55530
+    },
+    {
+      "epoch": 3.7732708248403317,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.641705734474793e-05,
+      "loss": 0.8675,
+      "step": 55535
+    },
+    {
+      "epoch": 3.7736105449109933,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6414934094306293e-05,
+      "loss": 0.9705,
+      "step": 55540
+    },
+    {
+      "epoch": 3.773950264981655,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6412810843864654e-05,
+      "loss": 0.9781,
+      "step": 55545
+    },
+    {
+      "epoch": 3.774289985052317,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.641068759342302e-05,
+      "loss": 0.9386,
+      "step": 55550
+    },
+    {
+      "epoch": 3.7746297051229787,
+      "grad_norm": 7.625,
+      "learning_rate": 2.6408564342981385e-05,
+      "loss": 1.0925,
+      "step": 55555
+    },
+    {
+      "epoch": 3.7749694251936403,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.6406441092539746e-05,
+      "loss": 1.0035,
+      "step": 55560
+    },
+    {
+      "epoch": 3.775309145264302,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.6404317842098113e-05,
+      "loss": 0.9149,
+      "step": 55565
+    },
+    {
+      "epoch": 3.775648865334964,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.6402194591656477e-05,
+      "loss": 0.8869,
+      "step": 55570
+    },
+    {
+      "epoch": 3.7759885854056257,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6400071341214838e-05,
+      "loss": 0.9692,
+      "step": 55575
+    },
+    {
+      "epoch": 3.7763283054762873,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6397948090773205e-05,
+      "loss": 0.9053,
+      "step": 55580
+    },
+    {
+      "epoch": 3.7766680255469494,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.639582484033157e-05,
+      "loss": 0.9606,
+      "step": 55585
+    },
+    {
+      "epoch": 3.777007745617611,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.639370158988993e-05,
+      "loss": 0.9042,
+      "step": 55590
+    },
+    {
+      "epoch": 3.7773474656882726,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6391578339448297e-05,
+      "loss": 0.9596,
+      "step": 55595
+    },
+    {
+      "epoch": 3.7776871857589347,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.6389455089006658e-05,
+      "loss": 0.8887,
+      "step": 55600
+    },
+    {
+      "epoch": 3.7780269058295963,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6387331838565022e-05,
+      "loss": 0.8051,
+      "step": 55605
+    },
+    {
+      "epoch": 3.778366625900258,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.638520858812339e-05,
+      "loss": 0.9499,
+      "step": 55610
+    },
+    {
+      "epoch": 3.77870634597092,
+      "grad_norm": 6.875,
+      "learning_rate": 2.638308533768175e-05,
+      "loss": 0.9136,
+      "step": 55615
+    },
+    {
+      "epoch": 3.7790460660415817,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6380962087240114e-05,
+      "loss": 0.851,
+      "step": 55620
+    },
+    {
+      "epoch": 3.7793857861122433,
+      "grad_norm": 7.375,
+      "learning_rate": 2.637883883679848e-05,
+      "loss": 0.9404,
+      "step": 55625
+    },
+    {
+      "epoch": 3.7797255061829054,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.6376715586356842e-05,
+      "loss": 0.8962,
+      "step": 55630
+    },
+    {
+      "epoch": 3.780065226253567,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.6374592335915206e-05,
+      "loss": 0.8978,
+      "step": 55635
+    },
+    {
+      "epoch": 3.7804049463242286,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6372469085473573e-05,
+      "loss": 0.8509,
+      "step": 55640
+    },
+    {
+      "epoch": 3.7807446663948907,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6370345835031934e-05,
+      "loss": 0.916,
+      "step": 55645
+    },
+    {
+      "epoch": 3.7810843864655523,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.6368222584590295e-05,
+      "loss": 0.9125,
+      "step": 55650
+    },
+    {
+      "epoch": 3.781424106536214,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.6366099334148665e-05,
+      "loss": 0.9242,
+      "step": 55655
+    },
+    {
+      "epoch": 3.781763826606876,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6363976083707026e-05,
+      "loss": 0.788,
+      "step": 55660
+    },
+    {
+      "epoch": 3.7821035466775377,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.6361852833265387e-05,
+      "loss": 0.9275,
+      "step": 55665
+    },
+    {
+      "epoch": 3.7824432667481993,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.6359729582823754e-05,
+      "loss": 0.9929,
+      "step": 55670
+    },
+    {
+      "epoch": 3.7827829868188614,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6357606332382118e-05,
+      "loss": 0.8917,
+      "step": 55675
+    },
+    {
+      "epoch": 3.783122706889523,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.635548308194048e-05,
+      "loss": 0.8878,
+      "step": 55680
+    },
+    {
+      "epoch": 3.7834624269601846,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6353359831498846e-05,
+      "loss": 0.8627,
+      "step": 55685
+    },
+    {
+      "epoch": 3.7838021470308467,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.635123658105721e-05,
+      "loss": 0.8917,
+      "step": 55690
+    },
+    {
+      "epoch": 3.7841418671015083,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.634911333061557e-05,
+      "loss": 0.9441,
+      "step": 55695
+    },
+    {
+      "epoch": 3.78448158717217,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6346990080173938e-05,
+      "loss": 0.9327,
+      "step": 55700
+    },
+    {
+      "epoch": 3.784821307242832,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6344866829732302e-05,
+      "loss": 0.9227,
+      "step": 55705
+    },
+    {
+      "epoch": 3.7851610273134937,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.634274357929067e-05,
+      "loss": 0.9038,
+      "step": 55710
+    },
+    {
+      "epoch": 3.7855007473841553,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.634062032884903e-05,
+      "loss": 0.8629,
+      "step": 55715
+    },
+    {
+      "epoch": 3.7858404674548174,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6338497078407394e-05,
+      "loss": 0.8738,
+      "step": 55720
+    },
+    {
+      "epoch": 3.786180187525479,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.633637382796576e-05,
+      "loss": 0.9835,
+      "step": 55725
+    },
+    {
+      "epoch": 3.7865199075961407,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6334250577524122e-05,
+      "loss": 0.8726,
+      "step": 55730
+    },
+    {
+      "epoch": 3.7868596276668027,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6332127327082483e-05,
+      "loss": 0.8742,
+      "step": 55735
+    },
+    {
+      "epoch": 3.7871993477374644,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.633000407664085e-05,
+      "loss": 0.9312,
+      "step": 55740
+    },
+    {
+      "epoch": 3.787539067808126,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6327880826199214e-05,
+      "loss": 1.0008,
+      "step": 55745
+    },
+    {
+      "epoch": 3.787878787878788,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6325757575757575e-05,
+      "loss": 0.9237,
+      "step": 55750
+    },
+    {
+      "epoch": 3.7882185079494497,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.6323634325315942e-05,
+      "loss": 0.937,
+      "step": 55755
+    },
+    {
+      "epoch": 3.7885582280201113,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6321511074874306e-05,
+      "loss": 0.8886,
+      "step": 55760
+    },
+    {
+      "epoch": 3.7888979480907734,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6319387824432667e-05,
+      "loss": 0.8381,
+      "step": 55765
+    },
+    {
+      "epoch": 3.789237668161435,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6317264573991034e-05,
+      "loss": 0.9502,
+      "step": 55770
+    },
+    {
+      "epoch": 3.7895773882320967,
+      "grad_norm": 6.75,
+      "learning_rate": 2.6315141323549398e-05,
+      "loss": 0.933,
+      "step": 55775
+    },
+    {
+      "epoch": 3.7899171083027587,
+      "grad_norm": 8.25,
+      "learning_rate": 2.631301807310776e-05,
+      "loss": 0.9168,
+      "step": 55780
+    },
+    {
+      "epoch": 3.7902568283734204,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6310894822666126e-05,
+      "loss": 0.8795,
+      "step": 55785
+    },
+    {
+      "epoch": 3.790596548444082,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.630877157222449e-05,
+      "loss": 0.9337,
+      "step": 55790
+    },
+    {
+      "epoch": 3.790936268514744,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.630664832178285e-05,
+      "loss": 0.8116,
+      "step": 55795
+    },
+    {
+      "epoch": 3.7912759885854057,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.6304525071341218e-05,
+      "loss": 0.8886,
+      "step": 55800
+    },
+    {
+      "epoch": 3.7916157086560673,
+      "grad_norm": 7.75,
+      "learning_rate": 2.630240182089958e-05,
+      "loss": 0.924,
+      "step": 55805
+    },
+    {
+      "epoch": 3.7919554287267294,
+      "grad_norm": 7.75,
+      "learning_rate": 2.6300278570457943e-05,
+      "loss": 0.9395,
+      "step": 55810
+    },
+    {
+      "epoch": 3.792295148797391,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.629815532001631e-05,
+      "loss": 0.8535,
+      "step": 55815
+    },
+    {
+      "epoch": 3.7926348688680527,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.629603206957467e-05,
+      "loss": 1.1154,
+      "step": 55820
+    },
+    {
+      "epoch": 3.7929745889387148,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6293908819133035e-05,
+      "loss": 0.8753,
+      "step": 55825
+    },
+    {
+      "epoch": 3.7933143090093764,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6291785568691402e-05,
+      "loss": 1.0203,
+      "step": 55830
+    },
+    {
+      "epoch": 3.793654029080038,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.6289662318249763e-05,
+      "loss": 0.8371,
+      "step": 55835
+    },
+    {
+      "epoch": 3.7939937491507,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6287539067808127e-05,
+      "loss": 0.939,
+      "step": 55840
+    },
+    {
+      "epoch": 3.7943334692213617,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.6285415817366494e-05,
+      "loss": 0.9075,
+      "step": 55845
+    },
+    {
+      "epoch": 3.7946731892920234,
+      "grad_norm": 7.125,
+      "learning_rate": 2.6283292566924855e-05,
+      "loss": 0.8074,
+      "step": 55850
+    },
+    {
+      "epoch": 3.795012909362685,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.628116931648322e-05,
+      "loss": 0.8826,
+      "step": 55855
+    },
+    {
+      "epoch": 3.795352629433347,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.6279046066041586e-05,
+      "loss": 0.9889,
+      "step": 55860
+    },
+    {
+      "epoch": 3.7956923495040087,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6276922815599947e-05,
+      "loss": 0.8239,
+      "step": 55865
+    },
+    {
+      "epoch": 3.7960320695746703,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.6274799565158307e-05,
+      "loss": 0.909,
+      "step": 55870
+    },
+    {
+      "epoch": 3.7963717896453324,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.6272676314716675e-05,
+      "loss": 0.9098,
+      "step": 55875
+    },
+    {
+      "epoch": 3.796711509715994,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.627055306427504e-05,
+      "loss": 0.8243,
+      "step": 55880
+    },
+    {
+      "epoch": 3.7970512297866557,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.62684298138334e-05,
+      "loss": 0.9485,
+      "step": 55885
+    },
+    {
+      "epoch": 3.7973909498573177,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.6266306563391767e-05,
+      "loss": 0.9016,
+      "step": 55890
+    },
+    {
+      "epoch": 3.7977306699279794,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.626418331295013e-05,
+      "loss": 0.8366,
+      "step": 55895
+    },
+    {
+      "epoch": 3.798070389998641,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.626206006250849e-05,
+      "loss": 0.9163,
+      "step": 55900
+    },
+    {
+      "epoch": 3.7984101100693026,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.625993681206686e-05,
+      "loss": 0.863,
+      "step": 55905
+    },
+    {
+      "epoch": 3.7987498301399647,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.6257813561625223e-05,
+      "loss": 0.8074,
+      "step": 55910
+    },
+    {
+      "epoch": 3.7990895502106263,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6255690311183584e-05,
+      "loss": 0.8871,
+      "step": 55915
+    },
+    {
+      "epoch": 3.799429270281288,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.625356706074195e-05,
+      "loss": 0.89,
+      "step": 55920
+    },
+    {
+      "epoch": 3.79976899035195,
+      "grad_norm": 5.78125,
+      "learning_rate": 2.6251443810300315e-05,
+      "loss": 0.9226,
+      "step": 55925
+    },
+    {
+      "epoch": 3.8001087104226117,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6249320559858676e-05,
+      "loss": 1.1811,
+      "step": 55930
+    },
+    {
+      "epoch": 3.8004484304932733,
+      "grad_norm": 7.75,
+      "learning_rate": 2.6247197309417043e-05,
+      "loss": 0.8818,
+      "step": 55935
+    },
+    {
+      "epoch": 3.8007881505639354,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.6245074058975404e-05,
+      "loss": 0.8225,
+      "step": 55940
+    },
+    {
+      "epoch": 3.801127870634597,
+      "grad_norm": 5.875,
+      "learning_rate": 2.6242950808533768e-05,
+      "loss": 0.9482,
+      "step": 55945
+    },
+    {
+      "epoch": 3.8014675907052586,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6240827558092135e-05,
+      "loss": 0.935,
+      "step": 55950
+    },
+    {
+      "epoch": 3.8018073107759207,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6238704307650496e-05,
+      "loss": 0.7941,
+      "step": 55955
+    },
+    {
+      "epoch": 3.8021470308465823,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.623658105720886e-05,
+      "loss": 0.9186,
+      "step": 55960
+    },
+    {
+      "epoch": 3.802486750917244,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6234457806767227e-05,
+      "loss": 0.9657,
+      "step": 55965
+    },
+    {
+      "epoch": 3.802826470987906,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.6232334556325588e-05,
+      "loss": 0.9212,
+      "step": 55970
+    },
+    {
+      "epoch": 3.8031661910585677,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.623021130588395e-05,
+      "loss": 0.864,
+      "step": 55975
+    },
+    {
+      "epoch": 3.8035059111292293,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.622808805544232e-05,
+      "loss": 0.933,
+      "step": 55980
+    },
+    {
+      "epoch": 3.8038456311998914,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.622596480500068e-05,
+      "loss": 0.8978,
+      "step": 55985
+    },
+    {
+      "epoch": 3.804185351270553,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6223841554559044e-05,
+      "loss": 0.9342,
+      "step": 55990
+    },
+    {
+      "epoch": 3.8045250713412146,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.622171830411741e-05,
+      "loss": 0.9523,
+      "step": 55995
+    },
+    {
+      "epoch": 3.8048647914118767,
+      "grad_norm": 7.125,
+      "learning_rate": 2.621959505367577e-05,
+      "loss": 0.8975,
+      "step": 56000
+    },
+    {
+      "epoch": 3.8052045114825384,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.6217471803234132e-05,
+      "loss": 0.8541,
+      "step": 56005
+    },
+    {
+      "epoch": 3.8055442315532,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6215348552792503e-05,
+      "loss": 0.8508,
+      "step": 56010
+    },
+    {
+      "epoch": 3.805883951623862,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6213225302350864e-05,
+      "loss": 0.8486,
+      "step": 56015
+    },
+    {
+      "epoch": 3.8062236716945237,
+      "grad_norm": 7.375,
+      "learning_rate": 2.6211102051909224e-05,
+      "loss": 0.9273,
+      "step": 56020
+    },
+    {
+      "epoch": 3.8065633917651853,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.620897880146759e-05,
+      "loss": 0.8516,
+      "step": 56025
+    },
+    {
+      "epoch": 3.8069031118358474,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6206855551025956e-05,
+      "loss": 0.894,
+      "step": 56030
+    },
+    {
+      "epoch": 3.807242831906509,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6204732300584316e-05,
+      "loss": 0.8103,
+      "step": 56035
+    },
+    {
+      "epoch": 3.8075825519771707,
+      "grad_norm": 8.125,
+      "learning_rate": 2.6202609050142684e-05,
+      "loss": 0.9798,
+      "step": 56040
+    },
+    {
+      "epoch": 3.8079222720478327,
+      "grad_norm": 8.75,
+      "learning_rate": 2.6200485799701048e-05,
+      "loss": 0.8908,
+      "step": 56045
+    },
+    {
+      "epoch": 3.8082619921184944,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.6198362549259415e-05,
+      "loss": 0.9252,
+      "step": 56050
+    },
+    {
+      "epoch": 3.808601712189156,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.6196239298817776e-05,
+      "loss": 0.8852,
+      "step": 56055
+    },
+    {
+      "epoch": 3.808941432259818,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.619411604837614e-05,
+      "loss": 0.952,
+      "step": 56060
+    },
+    {
+      "epoch": 3.8092811523304797,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6191992797934507e-05,
+      "loss": 0.8967,
+      "step": 56065
+    },
+    {
+      "epoch": 3.8096208724011413,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.6189869547492868e-05,
+      "loss": 0.9303,
+      "step": 56070
+    },
+    {
+      "epoch": 3.8099605924718034,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.6187746297051228e-05,
+      "loss": 0.814,
+      "step": 56075
+    },
+    {
+      "epoch": 3.810300312542465,
+      "grad_norm": 6.75,
+      "learning_rate": 2.61856230466096e-05,
+      "loss": 0.9242,
+      "step": 56080
+    },
+    {
+      "epoch": 3.8106400326131267,
+      "grad_norm": 7.625,
+      "learning_rate": 2.618349979616796e-05,
+      "loss": 0.8843,
+      "step": 56085
+    },
+    {
+      "epoch": 3.8109797526837887,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.618137654572632e-05,
+      "loss": 0.8878,
+      "step": 56090
+    },
+    {
+      "epoch": 3.8113194727544504,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.6179253295284688e-05,
+      "loss": 0.8943,
+      "step": 56095
+    },
+    {
+      "epoch": 3.811659192825112,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.617713004484305e-05,
+      "loss": 0.9846,
+      "step": 56100
+    },
+    {
+      "epoch": 3.811998912895774,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6175006794401412e-05,
+      "loss": 0.7819,
+      "step": 56105
+    },
+    {
+      "epoch": 3.8123386329664357,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.617288354395978e-05,
+      "loss": 0.8285,
+      "step": 56110
+    },
+    {
+      "epoch": 3.8126783530370973,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6170760293518144e-05,
+      "loss": 0.9294,
+      "step": 56115
+    },
+    {
+      "epoch": 3.8130180731077594,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6168637043076504e-05,
+      "loss": 0.9278,
+      "step": 56120
+    },
+    {
+      "epoch": 3.813357793178421,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.616651379263487e-05,
+      "loss": 0.8875,
+      "step": 56125
+    },
+    {
+      "epoch": 3.8136975132490827,
+      "grad_norm": 8.25,
+      "learning_rate": 2.6164390542193236e-05,
+      "loss": 0.8959,
+      "step": 56130
+    },
+    {
+      "epoch": 3.8140372333197448,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.6162267291751596e-05,
+      "loss": 0.9269,
+      "step": 56135
+    },
+    {
+      "epoch": 3.8143769533904064,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6160144041309964e-05,
+      "loss": 0.9026,
+      "step": 56140
+    },
+    {
+      "epoch": 3.814716673461068,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6158020790868328e-05,
+      "loss": 0.8543,
+      "step": 56145
+    },
+    {
+      "epoch": 3.81505639353173,
+      "grad_norm": 7.125,
+      "learning_rate": 2.615589754042669e-05,
+      "loss": 0.9497,
+      "step": 56150
+    },
+    {
+      "epoch": 3.8153961136023917,
+      "grad_norm": 8.25,
+      "learning_rate": 2.6153774289985056e-05,
+      "loss": 0.9623,
+      "step": 56155
+    },
+    {
+      "epoch": 3.8157358336730534,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6151651039543416e-05,
+      "loss": 0.8985,
+      "step": 56160
+    },
+    {
+      "epoch": 3.8160755537437154,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.614952778910178e-05,
+      "loss": 0.919,
+      "step": 56165
+    },
+    {
+      "epoch": 3.816415273814377,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6147404538660148e-05,
+      "loss": 0.8228,
+      "step": 56170
+    },
+    {
+      "epoch": 3.8167549938850387,
+      "grad_norm": 8.25,
+      "learning_rate": 2.614528128821851e-05,
+      "loss": 1.0244,
+      "step": 56175
+    },
+    {
+      "epoch": 3.8170947139557008,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.6143158037776872e-05,
+      "loss": 0.8249,
+      "step": 56180
+    },
+    {
+      "epoch": 3.8174344340263624,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.614103478733524e-05,
+      "loss": 0.9086,
+      "step": 56185
+    },
+    {
+      "epoch": 3.817774154097024,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.61389115368936e-05,
+      "loss": 0.9639,
+      "step": 56190
+    },
+    {
+      "epoch": 3.8181138741676857,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.6136788286451964e-05,
+      "loss": 0.9277,
+      "step": 56195
+    },
+    {
+      "epoch": 3.8184535942383477,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.6134665036010332e-05,
+      "loss": 0.9338,
+      "step": 56200
+    },
+    {
+      "epoch": 3.8187933143090094,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.6132541785568692e-05,
+      "loss": 0.8583,
+      "step": 56205
+    },
+    {
+      "epoch": 3.819133034379671,
+      "grad_norm": 6.5,
+      "learning_rate": 2.6130418535127056e-05,
+      "loss": 0.9523,
+      "step": 56210
+    },
+    {
+      "epoch": 3.819472754450333,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6128295284685424e-05,
+      "loss": 0.901,
+      "step": 56215
+    },
+    {
+      "epoch": 3.8198124745209947,
+      "grad_norm": 6.75,
+      "learning_rate": 2.6126172034243784e-05,
+      "loss": 0.9411,
+      "step": 56220
+    },
+    {
+      "epoch": 3.8201521945916563,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6124048783802145e-05,
+      "loss": 0.8543,
+      "step": 56225
+    },
+    {
+      "epoch": 3.8204919146623184,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.6121925533360512e-05,
+      "loss": 0.926,
+      "step": 56230
+    },
+    {
+      "epoch": 3.82083163473298,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.6119802282918876e-05,
+      "loss": 0.959,
+      "step": 56235
+    },
+    {
+      "epoch": 3.8211713548036417,
+      "grad_norm": 7.0,
+      "learning_rate": 2.6117679032477237e-05,
+      "loss": 0.8746,
+      "step": 56240
+    },
+    {
+      "epoch": 3.8215110748743033,
+      "grad_norm": 6.875,
+      "learning_rate": 2.6115555782035604e-05,
+      "loss": 0.8697,
+      "step": 56245
+    },
+    {
+      "epoch": 3.8218507949449654,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.611343253159397e-05,
+      "loss": 0.9051,
+      "step": 56250
+    },
+    {
+      "epoch": 3.822190515015627,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.611130928115233e-05,
+      "loss": 0.929,
+      "step": 56255
+    },
+    {
+      "epoch": 3.8225302350862886,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.6109186030710696e-05,
+      "loss": 0.9205,
+      "step": 56260
+    },
+    {
+      "epoch": 3.8228699551569507,
+      "grad_norm": 7.875,
+      "learning_rate": 2.610706278026906e-05,
+      "loss": 0.877,
+      "step": 56265
+    },
+    {
+      "epoch": 3.8232096752276123,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.610493952982742e-05,
+      "loss": 0.8352,
+      "step": 56270
+    },
+    {
+      "epoch": 3.823549395298274,
+      "grad_norm": 6.875,
+      "learning_rate": 2.610281627938579e-05,
+      "loss": 0.8806,
+      "step": 56275
+    },
+    {
+      "epoch": 3.823889115368936,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.6100693028944152e-05,
+      "loss": 0.9181,
+      "step": 56280
+    },
+    {
+      "epoch": 3.8242288354395977,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6098569778502513e-05,
+      "loss": 0.9357,
+      "step": 56285
+    },
+    {
+      "epoch": 3.8245685555102593,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.609644652806088e-05,
+      "loss": 0.8383,
+      "step": 56290
+    },
+    {
+      "epoch": 3.8249082755809214,
+      "grad_norm": 8.25,
+      "learning_rate": 2.609432327761924e-05,
+      "loss": 0.9125,
+      "step": 56295
+    },
+    {
+      "epoch": 3.825247995651583,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.6092200027177605e-05,
+      "loss": 0.9451,
+      "step": 56300
+    },
+    {
+      "epoch": 3.8255877157222447,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6090076776735972e-05,
+      "loss": 0.9697,
+      "step": 56305
+    },
+    {
+      "epoch": 3.8259274357929067,
+      "grad_norm": 7.25,
+      "learning_rate": 2.6087953526294333e-05,
+      "loss": 0.9858,
+      "step": 56310
+    },
+    {
+      "epoch": 3.8262671558635684,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6085830275852697e-05,
+      "loss": 0.945,
+      "step": 56315
+    },
+    {
+      "epoch": 3.82660687593423,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.6083707025411064e-05,
+      "loss": 0.8361,
+      "step": 56320
+    },
+    {
+      "epoch": 3.826946596004892,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6081583774969425e-05,
+      "loss": 0.9496,
+      "step": 56325
+    },
+    {
+      "epoch": 3.8272863160755537,
+      "grad_norm": 7.375,
+      "learning_rate": 2.607946052452779e-05,
+      "loss": 0.8976,
+      "step": 56330
+    },
+    {
+      "epoch": 3.8276260361462153,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6077337274086156e-05,
+      "loss": 0.9996,
+      "step": 56335
+    },
+    {
+      "epoch": 3.8279657562168774,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.6075214023644517e-05,
+      "loss": 1.0098,
+      "step": 56340
+    },
+    {
+      "epoch": 3.828305476287539,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.607309077320288e-05,
+      "loss": 0.9167,
+      "step": 56345
+    },
+    {
+      "epoch": 3.8286451963582007,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.607096752276125e-05,
+      "loss": 1.0297,
+      "step": 56350
+    },
+    {
+      "epoch": 3.8289849164288627,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.606884427231961e-05,
+      "loss": 0.8448,
+      "step": 56355
+    },
+    {
+      "epoch": 3.8293246364995244,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.606672102187797e-05,
+      "loss": 0.8825,
+      "step": 56360
+    },
+    {
+      "epoch": 3.829664356570186,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.6064597771436337e-05,
+      "loss": 0.9184,
+      "step": 56365
+    },
+    {
+      "epoch": 3.830004076640848,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.60624745209947e-05,
+      "loss": 0.892,
+      "step": 56370
+    },
+    {
+      "epoch": 3.8303437967115097,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.6060351270553062e-05,
+      "loss": 0.9196,
+      "step": 56375
+    },
+    {
+      "epoch": 3.8306835167821713,
+      "grad_norm": 8.375,
+      "learning_rate": 2.605822802011143e-05,
+      "loss": 0.9453,
+      "step": 56380
+    },
+    {
+      "epoch": 3.8310232368528334,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.6056104769669793e-05,
+      "loss": 0.8976,
+      "step": 56385
+    },
+    {
+      "epoch": 3.831362956923495,
+      "grad_norm": 7.375,
+      "learning_rate": 2.605398151922816e-05,
+      "loss": 0.9218,
+      "step": 56390
+    },
+    {
+      "epoch": 3.8317026769941567,
+      "grad_norm": 8.0,
+      "learning_rate": 2.605185826878652e-05,
+      "loss": 0.8767,
+      "step": 56395
+    },
+    {
+      "epoch": 3.8320423970648188,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.6049735018344885e-05,
+      "loss": 0.857,
+      "step": 56400
+    },
+    {
+      "epoch": 3.8323821171354804,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.6047611767903253e-05,
+      "loss": 0.9766,
+      "step": 56405
+    },
+    {
+      "epoch": 3.832721837206142,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.6045488517461613e-05,
+      "loss": 0.8723,
+      "step": 56410
+    },
+    {
+      "epoch": 3.833061557276804,
+      "grad_norm": 7.5,
+      "learning_rate": 2.6043365267019977e-05,
+      "loss": 0.8677,
+      "step": 56415
+    },
+    {
+      "epoch": 3.8334012773474657,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.6041242016578345e-05,
+      "loss": 0.8564,
+      "step": 56420
+    },
+    {
+      "epoch": 3.8337409974181273,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6039118766136705e-05,
+      "loss": 0.8794,
+      "step": 56425
+    },
+    {
+      "epoch": 3.8340807174887894,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.6036995515695066e-05,
+      "loss": 0.9248,
+      "step": 56430
+    },
+    {
+      "epoch": 3.834420437559451,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.6034872265253437e-05,
+      "loss": 0.8936,
+      "step": 56435
+    },
+    {
+      "epoch": 3.8347601576301127,
+      "grad_norm": 8.375,
+      "learning_rate": 2.6032749014811797e-05,
+      "loss": 0.9766,
+      "step": 56440
+    },
+    {
+      "epoch": 3.8350998777007748,
+      "grad_norm": 8.5,
+      "learning_rate": 2.6030625764370158e-05,
+      "loss": 0.8913,
+      "step": 56445
+    },
+    {
+      "epoch": 3.8354395977714364,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.6028502513928525e-05,
+      "loss": 1.0181,
+      "step": 56450
+    },
+    {
+      "epoch": 3.835779317842098,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.602637926348689e-05,
+      "loss": 0.8862,
+      "step": 56455
+    },
+    {
+      "epoch": 3.83611903791276,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.602425601304525e-05,
+      "loss": 0.9457,
+      "step": 56460
+    },
+    {
+      "epoch": 3.8364587579834217,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6022132762603617e-05,
+      "loss": 0.902,
+      "step": 56465
+    },
+    {
+      "epoch": 3.8367984780540834,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.602000951216198e-05,
+      "loss": 0.8271,
+      "step": 56470
+    },
+    {
+      "epoch": 3.8371381981247454,
+      "grad_norm": 8.375,
+      "learning_rate": 2.6017886261720342e-05,
+      "loss": 0.9049,
+      "step": 56475
+    },
+    {
+      "epoch": 3.837477918195407,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.601576301127871e-05,
+      "loss": 0.8655,
+      "step": 56480
+    },
+    {
+      "epoch": 3.8378176382660687,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.6013639760837073e-05,
+      "loss": 0.958,
+      "step": 56485
+    },
+    {
+      "epoch": 3.8381573583367308,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.6011516510395434e-05,
+      "loss": 0.9519,
+      "step": 56490
+    },
+    {
+      "epoch": 3.8384970784073924,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.60093932599538e-05,
+      "loss": 0.9649,
+      "step": 56495
+    },
+    {
+      "epoch": 3.838836798478054,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.6007270009512165e-05,
+      "loss": 0.8385,
+      "step": 56500
+    },
+    {
+      "epoch": 3.839176518548716,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.6005146759070526e-05,
+      "loss": 0.8973,
+      "step": 56505
+    },
+    {
+      "epoch": 3.8395162386193777,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.6003023508628893e-05,
+      "loss": 0.9302,
+      "step": 56510
+    },
+    {
+      "epoch": 3.8398559586900394,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.6000900258187254e-05,
+      "loss": 0.9538,
+      "step": 56515
+    },
+    {
+      "epoch": 3.8401956787607014,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.5998777007745618e-05,
+      "loss": 0.9881,
+      "step": 56520
+    },
+    {
+      "epoch": 3.840535398831363,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.5996653757303985e-05,
+      "loss": 0.8775,
+      "step": 56525
+    },
+    {
+      "epoch": 3.8408751189020247,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5994530506862346e-05,
+      "loss": 0.9613,
+      "step": 56530
+    },
+    {
+      "epoch": 3.8412148389726863,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.599240725642071e-05,
+      "loss": 0.8689,
+      "step": 56535
+    },
+    {
+      "epoch": 3.8415545590433484,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5990284005979077e-05,
+      "loss": 0.8972,
+      "step": 56540
+    },
+    {
+      "epoch": 3.84189427911401,
+      "grad_norm": 8.5,
+      "learning_rate": 2.5988160755537438e-05,
+      "loss": 0.8787,
+      "step": 56545
+    },
+    {
+      "epoch": 3.8422339991846717,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5986037505095802e-05,
+      "loss": 0.9288,
+      "step": 56550
+    },
+    {
+      "epoch": 3.8425737192553338,
+      "grad_norm": 8.5,
+      "learning_rate": 2.598391425465417e-05,
+      "loss": 0.8759,
+      "step": 56555
+    },
+    {
+      "epoch": 3.8429134393259954,
+      "grad_norm": 8.375,
+      "learning_rate": 2.598179100421253e-05,
+      "loss": 0.9112,
+      "step": 56560
+    },
+    {
+      "epoch": 3.843253159396657,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.597966775377089e-05,
+      "loss": 0.8869,
+      "step": 56565
+    },
+    {
+      "epoch": 3.843592879467319,
+      "grad_norm": 7.625,
+      "learning_rate": 2.597754450332926e-05,
+      "loss": 0.9961,
+      "step": 56570
+    },
+    {
+      "epoch": 3.8439325995379807,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.5975421252887622e-05,
+      "loss": 0.9699,
+      "step": 56575
+    },
+    {
+      "epoch": 3.8442723196086424,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.5973298002445982e-05,
+      "loss": 1.0022,
+      "step": 56580
+    },
+    {
+      "epoch": 3.844612039679304,
+      "grad_norm": 7.125,
+      "learning_rate": 2.597117475200435e-05,
+      "loss": 0.9425,
+      "step": 56585
+    },
+    {
+      "epoch": 3.844951759749966,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.5969051501562714e-05,
+      "loss": 0.9059,
+      "step": 56590
+    },
+    {
+      "epoch": 3.8452914798206277,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5966928251121074e-05,
+      "loss": 0.9575,
+      "step": 56595
+    },
+    {
+      "epoch": 3.8456311998912893,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5964805000679442e-05,
+      "loss": 1.0735,
+      "step": 56600
+    },
+    {
+      "epoch": 3.8459709199619514,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5962681750237806e-05,
+      "loss": 0.9106,
+      "step": 56605
+    },
+    {
+      "epoch": 3.846310640032613,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5960558499796167e-05,
+      "loss": 0.9511,
+      "step": 56610
+    },
+    {
+      "epoch": 3.8466503601032747,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.5958435249354534e-05,
+      "loss": 0.9038,
+      "step": 56615
+    },
+    {
+      "epoch": 3.8469900801739367,
+      "grad_norm": 8.5,
+      "learning_rate": 2.5956311998912898e-05,
+      "loss": 0.9758,
+      "step": 56620
+    },
+    {
+      "epoch": 3.8473298002445984,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.595418874847126e-05,
+      "loss": 0.9454,
+      "step": 56625
+    },
+    {
+      "epoch": 3.84766952031526,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5952065498029626e-05,
+      "loss": 0.9845,
+      "step": 56630
+    },
+    {
+      "epoch": 3.848009240385922,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.594994224758799e-05,
+      "loss": 1.0394,
+      "step": 56635
+    },
+    {
+      "epoch": 3.8483489604565837,
+      "grad_norm": 6.75,
+      "learning_rate": 2.594781899714635e-05,
+      "loss": 0.9134,
+      "step": 56640
+    },
+    {
+      "epoch": 3.8486886805272453,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5945695746704718e-05,
+      "loss": 0.9451,
+      "step": 56645
+    },
+    {
+      "epoch": 3.8490284005979074,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.594357249626308e-05,
+      "loss": 0.8823,
+      "step": 56650
+    },
+    {
+      "epoch": 3.849368120668569,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5941449245821443e-05,
+      "loss": 0.9306,
+      "step": 56655
+    },
+    {
+      "epoch": 3.8497078407392307,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.593932599537981e-05,
+      "loss": 0.9407,
+      "step": 56660
+    },
+    {
+      "epoch": 3.8500475608098927,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.593720274493817e-05,
+      "loss": 0.9289,
+      "step": 56665
+    },
+    {
+      "epoch": 3.8503872808805544,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5935079494496535e-05,
+      "loss": 0.9734,
+      "step": 56670
+    },
+    {
+      "epoch": 3.850727000951216,
+      "grad_norm": 7.875,
+      "learning_rate": 2.5932956244054902e-05,
+      "loss": 0.8782,
+      "step": 56675
+    },
+    {
+      "epoch": 3.851066721021878,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.5930832993613263e-05,
+      "loss": 0.8481,
+      "step": 56680
+    },
+    {
+      "epoch": 3.8514064410925397,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.5928709743171627e-05,
+      "loss": 0.8389,
+      "step": 56685
+    },
+    {
+      "epoch": 3.8517461611632013,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.5926586492729994e-05,
+      "loss": 0.8538,
+      "step": 56690
+    },
+    {
+      "epoch": 3.8520858812338634,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5924463242288355e-05,
+      "loss": 0.9669,
+      "step": 56695
+    },
+    {
+      "epoch": 3.852425601304525,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.592233999184672e-05,
+      "loss": 0.8889,
+      "step": 56700
+    },
+    {
+      "epoch": 3.8527653213751867,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5920216741405086e-05,
+      "loss": 0.8504,
+      "step": 56705
+    },
+    {
+      "epoch": 3.8531050414458488,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5918093490963447e-05,
+      "loss": 0.9246,
+      "step": 56710
+    },
+    {
+      "epoch": 3.8534447615165104,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.5915970240521807e-05,
+      "loss": 0.8697,
+      "step": 56715
+    },
+    {
+      "epoch": 3.853784481587172,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5913846990080175e-05,
+      "loss": 0.9057,
+      "step": 56720
+    },
+    {
+      "epoch": 3.854124201657834,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.591172373963854e-05,
+      "loss": 0.9643,
+      "step": 56725
+    },
+    {
+      "epoch": 3.8544639217284957,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5909600489196906e-05,
+      "loss": 0.9334,
+      "step": 56730
+    },
+    {
+      "epoch": 3.8548036417991574,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5907477238755267e-05,
+      "loss": 0.882,
+      "step": 56735
+    },
+    {
+      "epoch": 3.8551433618698194,
+      "grad_norm": 8.375,
+      "learning_rate": 2.590535398831363e-05,
+      "loss": 0.9074,
+      "step": 56740
+    },
+    {
+      "epoch": 3.855483081940481,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.5903230737871998e-05,
+      "loss": 0.8744,
+      "step": 56745
+    },
+    {
+      "epoch": 3.8558228020111427,
+      "grad_norm": 7.125,
+      "learning_rate": 2.590110748743036e-05,
+      "loss": 0.9144,
+      "step": 56750
+    },
+    {
+      "epoch": 3.8561625220818048,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5898984236988723e-05,
+      "loss": 0.8713,
+      "step": 56755
+    },
+    {
+      "epoch": 3.8565022421524664,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.589686098654709e-05,
+      "loss": 0.8835,
+      "step": 56760
+    },
+    {
+      "epoch": 3.856841962223128,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.589473773610545e-05,
+      "loss": 0.9306,
+      "step": 56765
+    },
+    {
+      "epoch": 3.85718168229379,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.5892614485663815e-05,
+      "loss": 0.8969,
+      "step": 56770
+    },
+    {
+      "epoch": 3.8575214023644517,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5890491235222182e-05,
+      "loss": 0.9461,
+      "step": 56775
+    },
+    {
+      "epoch": 3.8578611224351134,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.5888367984780543e-05,
+      "loss": 0.8978,
+      "step": 56780
+    },
+    {
+      "epoch": 3.8582008425057754,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5886244734338903e-05,
+      "loss": 0.9299,
+      "step": 56785
+    },
+    {
+      "epoch": 3.858540562576437,
+      "grad_norm": 7.0,
+      "learning_rate": 2.588412148389727e-05,
+      "loss": 0.8468,
+      "step": 56790
+    },
+    {
+      "epoch": 3.8588802826470987,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.5881998233455635e-05,
+      "loss": 0.8942,
+      "step": 56795
+    },
+    {
+      "epoch": 3.8592200027177608,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.5879874983013995e-05,
+      "loss": 0.8176,
+      "step": 56800
+    },
+    {
+      "epoch": 3.8595597227884224,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.5877751732572363e-05,
+      "loss": 0.9577,
+      "step": 56805
+    },
+    {
+      "epoch": 3.859899442859084,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.5875628482130727e-05,
+      "loss": 0.8664,
+      "step": 56810
+    },
+    {
+      "epoch": 3.860239162929746,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.5873505231689087e-05,
+      "loss": 1.1526,
+      "step": 56815
+    },
+    {
+      "epoch": 3.8605788830004077,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.5871381981247455e-05,
+      "loss": 1.0302,
+      "step": 56820
+    },
+    {
+      "epoch": 3.8609186030710694,
+      "grad_norm": 7.625,
+      "learning_rate": 2.586925873080582e-05,
+      "loss": 0.966,
+      "step": 56825
+    },
+    {
+      "epoch": 3.8612583231417315,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.586713548036418e-05,
+      "loss": 0.9226,
+      "step": 56830
+    },
+    {
+      "epoch": 3.861598043212393,
+      "grad_norm": 6.5,
+      "learning_rate": 2.5865012229922547e-05,
+      "loss": 0.9357,
+      "step": 56835
+    },
+    {
+      "epoch": 3.8619377632830547,
+      "grad_norm": 5.125,
+      "learning_rate": 2.586288897948091e-05,
+      "loss": 0.9386,
+      "step": 56840
+    },
+    {
+      "epoch": 3.862277483353717,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.586076572903927e-05,
+      "loss": 0.891,
+      "step": 56845
+    },
+    {
+      "epoch": 3.8626172034243784,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.585864247859764e-05,
+      "loss": 0.9405,
+      "step": 56850
+    },
+    {
+      "epoch": 3.86295692349504,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.5856519228156e-05,
+      "loss": 0.8503,
+      "step": 56855
+    },
+    {
+      "epoch": 3.863296643565702,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5854395977714363e-05,
+      "loss": 0.8759,
+      "step": 56860
+    },
+    {
+      "epoch": 3.8636363636363638,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.585227272727273e-05,
+      "loss": 0.9145,
+      "step": 56865
+    },
+    {
+      "epoch": 3.8639760837070254,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.585014947683109e-05,
+      "loss": 0.9862,
+      "step": 56870
+    },
+    {
+      "epoch": 3.864315803777687,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5848026226389455e-05,
+      "loss": 0.8719,
+      "step": 56875
+    },
+    {
+      "epoch": 3.864655523848349,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.5845902975947823e-05,
+      "loss": 0.9527,
+      "step": 56880
+    },
+    {
+      "epoch": 3.8649952439190107,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5843779725506183e-05,
+      "loss": 0.8858,
+      "step": 56885
+    },
+    {
+      "epoch": 3.8653349639896724,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.5841656475064547e-05,
+      "loss": 0.8831,
+      "step": 56890
+    },
+    {
+      "epoch": 3.8656746840603344,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5839533224622915e-05,
+      "loss": 0.9144,
+      "step": 56895
+    },
+    {
+      "epoch": 3.866014404130996,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5837409974181275e-05,
+      "loss": 0.8188,
+      "step": 56900
+    },
+    {
+      "epoch": 3.8663541242016577,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.583528672373964e-05,
+      "loss": 0.9451,
+      "step": 56905
+    },
+    {
+      "epoch": 3.8666938442723198,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5833163473298007e-05,
+      "loss": 0.9882,
+      "step": 56910
+    },
+    {
+      "epoch": 3.8670335643429814,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5831040222856367e-05,
+      "loss": 0.8963,
+      "step": 56915
+    },
+    {
+      "epoch": 3.867373284413643,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5828916972414728e-05,
+      "loss": 0.967,
+      "step": 56920
+    },
+    {
+      "epoch": 3.8677130044843047,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.58267937219731e-05,
+      "loss": 0.9964,
+      "step": 56925
+    },
+    {
+      "epoch": 3.8680527245549667,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.582467047153146e-05,
+      "loss": 0.8963,
+      "step": 56930
+    },
+    {
+      "epoch": 3.8683924446256284,
+      "grad_norm": 7.875,
+      "learning_rate": 2.582254722108982e-05,
+      "loss": 0.8536,
+      "step": 56935
+    },
+    {
+      "epoch": 3.86873216469629,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5820423970648187e-05,
+      "loss": 0.8654,
+      "step": 56940
+    },
+    {
+      "epoch": 3.869071884766952,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.581830072020655e-05,
+      "loss": 0.949,
+      "step": 56945
+    },
+    {
+      "epoch": 3.8694116048376137,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.5816177469764912e-05,
+      "loss": 0.9212,
+      "step": 56950
+    },
+    {
+      "epoch": 3.8697513249082753,
+      "grad_norm": 7.25,
+      "learning_rate": 2.581405421932328e-05,
+      "loss": 0.8545,
+      "step": 56955
+    },
+    {
+      "epoch": 3.8700910449789374,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.5811930968881643e-05,
+      "loss": 1.0045,
+      "step": 56960
+    },
+    {
+      "epoch": 3.870430765049599,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5809807718440004e-05,
+      "loss": 1.0228,
+      "step": 56965
+    },
+    {
+      "epoch": 3.8707704851202607,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.580768446799837e-05,
+      "loss": 0.8584,
+      "step": 56970
+    },
+    {
+      "epoch": 3.8711102051909227,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5805561217556735e-05,
+      "loss": 0.8497,
+      "step": 56975
+    },
+    {
+      "epoch": 3.8714499252615844,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5803437967115096e-05,
+      "loss": 1.0047,
+      "step": 56980
+    },
+    {
+      "epoch": 3.871789645332246,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5801314716673463e-05,
+      "loss": 0.77,
+      "step": 56985
+    },
+    {
+      "epoch": 3.872129365402908,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.5799191466231824e-05,
+      "loss": 0.8884,
+      "step": 56990
+    },
+    {
+      "epoch": 3.8724690854735697,
+      "grad_norm": 6.875,
+      "learning_rate": 2.5797068215790188e-05,
+      "loss": 0.8922,
+      "step": 56995
+    },
+    {
+      "epoch": 3.8728088055442313,
+      "grad_norm": 8.25,
+      "learning_rate": 2.5794944965348555e-05,
+      "loss": 0.9028,
+      "step": 57000
+    },
+    {
+      "epoch": 3.8731485256148934,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5792821714906916e-05,
+      "loss": 0.8813,
+      "step": 57005
+    },
+    {
+      "epoch": 3.873488245685555,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.579069846446528e-05,
+      "loss": 0.8971,
+      "step": 57010
+    },
+    {
+      "epoch": 3.8738279657562167,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.5788575214023647e-05,
+      "loss": 0.9509,
+      "step": 57015
+    },
+    {
+      "epoch": 3.8741676858268788,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5786451963582008e-05,
+      "loss": 0.9654,
+      "step": 57020
+    },
+    {
+      "epoch": 3.8745074058975404,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5784328713140372e-05,
+      "loss": 0.9158,
+      "step": 57025
+    },
+    {
+      "epoch": 3.874847125968202,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.578220546269874e-05,
+      "loss": 0.8774,
+      "step": 57030
+    },
+    {
+      "epoch": 3.875186846038864,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.57800822122571e-05,
+      "loss": 0.8381,
+      "step": 57035
+    },
+    {
+      "epoch": 3.8755265661095257,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5777958961815464e-05,
+      "loss": 0.9318,
+      "step": 57040
+    },
+    {
+      "epoch": 3.8758662861801874,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.577583571137383e-05,
+      "loss": 0.9154,
+      "step": 57045
+    },
+    {
+      "epoch": 3.8762060062508494,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5773712460932192e-05,
+      "loss": 0.8386,
+      "step": 57050
+    },
+    {
+      "epoch": 3.876545726321511,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5771589210490553e-05,
+      "loss": 0.8823,
+      "step": 57055
+    },
+    {
+      "epoch": 3.8768854463921727,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.5769465960048923e-05,
+      "loss": 0.9069,
+      "step": 57060
+    },
+    {
+      "epoch": 3.8772251664628348,
+      "grad_norm": 6.625,
+      "learning_rate": 2.5767342709607284e-05,
+      "loss": 0.8718,
+      "step": 57065
+    },
+    {
+      "epoch": 3.8775648865334964,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.576521945916565e-05,
+      "loss": 0.971,
+      "step": 57070
+    },
+    {
+      "epoch": 3.877904606604158,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5763096208724012e-05,
+      "loss": 1.0076,
+      "step": 57075
+    },
+    {
+      "epoch": 3.87824432667482,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.5760972958282376e-05,
+      "loss": 0.8621,
+      "step": 57080
+    },
+    {
+      "epoch": 3.8785840467454817,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5758849707840743e-05,
+      "loss": 1.0356,
+      "step": 57085
+    },
+    {
+      "epoch": 3.8789237668161434,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5756726457399104e-05,
+      "loss": 1.0135,
+      "step": 57090
+    },
+    {
+      "epoch": 3.8792634868868054,
+      "grad_norm": 6.25,
+      "learning_rate": 2.5754603206957468e-05,
+      "loss": 0.9268,
+      "step": 57095
+    },
+    {
+      "epoch": 3.879603206957467,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5752479956515835e-05,
+      "loss": 0.9488,
+      "step": 57100
+    },
+    {
+      "epoch": 3.8799429270281287,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5750356706074196e-05,
+      "loss": 0.8769,
+      "step": 57105
+    },
+    {
+      "epoch": 3.880282647098791,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.574823345563256e-05,
+      "loss": 0.9049,
+      "step": 57110
+    },
+    {
+      "epoch": 3.8806223671694524,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.5746110205190928e-05,
+      "loss": 0.9159,
+      "step": 57115
+    },
+    {
+      "epoch": 3.880962087240114,
+      "grad_norm": 6.5,
+      "learning_rate": 2.5743986954749288e-05,
+      "loss": 0.9199,
+      "step": 57120
+    },
+    {
+      "epoch": 3.881301807310776,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.5741863704307652e-05,
+      "loss": 0.8877,
+      "step": 57125
+    },
+    {
+      "epoch": 3.8816415273814378,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.573974045386602e-05,
+      "loss": 0.821,
+      "step": 57130
+    },
+    {
+      "epoch": 3.8819812474520994,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.573761720342438e-05,
+      "loss": 0.8781,
+      "step": 57135
+    },
+    {
+      "epoch": 3.8823209675227615,
+      "grad_norm": 6.875,
+      "learning_rate": 2.573549395298274e-05,
+      "loss": 0.869,
+      "step": 57140
+    },
+    {
+      "epoch": 3.882660687593423,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.5733370702541108e-05,
+      "loss": 0.9729,
+      "step": 57145
+    },
+    {
+      "epoch": 3.8830004076640847,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5731247452099472e-05,
+      "loss": 0.8328,
+      "step": 57150
+    },
+    {
+      "epoch": 3.883340127734747,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.5729124201657833e-05,
+      "loss": 0.9646,
+      "step": 57155
+    },
+    {
+      "epoch": 3.8836798478054084,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.57270009512162e-05,
+      "loss": 1.0179,
+      "step": 57160
+    },
+    {
+      "epoch": 3.88401956787607,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5724877700774564e-05,
+      "loss": 0.9132,
+      "step": 57165
+    },
+    {
+      "epoch": 3.884359287946732,
+      "grad_norm": 7.0,
+      "learning_rate": 2.5722754450332925e-05,
+      "loss": 1.035,
+      "step": 57170
+    },
+    {
+      "epoch": 3.8846990080173938,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.5720631199891292e-05,
+      "loss": 0.8641,
+      "step": 57175
+    },
+    {
+      "epoch": 3.8850387280880554,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5718507949449656e-05,
+      "loss": 0.9405,
+      "step": 57180
+    },
+    {
+      "epoch": 3.8853784481587175,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5716384699008017e-05,
+      "loss": 0.8977,
+      "step": 57185
+    },
+    {
+      "epoch": 3.885718168229379,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.5714261448566384e-05,
+      "loss": 0.9438,
+      "step": 57190
+    },
+    {
+      "epoch": 3.8860578883000407,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5712138198124748e-05,
+      "loss": 0.9689,
+      "step": 57195
+    },
+    {
+      "epoch": 3.886397608370703,
+      "grad_norm": 8.0,
+      "learning_rate": 2.571001494768311e-05,
+      "loss": 0.9832,
+      "step": 57200
+    },
+    {
+      "epoch": 3.8867373284413644,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.5707891697241476e-05,
+      "loss": 0.9836,
+      "step": 57205
+    },
+    {
+      "epoch": 3.887077048512026,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.5705768446799837e-05,
+      "loss": 1.0553,
+      "step": 57210
+    },
+    {
+      "epoch": 3.887416768582688,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.57036451963582e-05,
+      "loss": 0.8711,
+      "step": 57215
+    },
+    {
+      "epoch": 3.8877564886533498,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.5701521945916568e-05,
+      "loss": 0.8789,
+      "step": 57220
+    },
+    {
+      "epoch": 3.8880962087240114,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.569939869547493e-05,
+      "loss": 0.9124,
+      "step": 57225
+    },
+    {
+      "epoch": 3.888435928794673,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.5697275445033293e-05,
+      "loss": 0.8921,
+      "step": 57230
+    },
+    {
+      "epoch": 3.888775648865335,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.569515219459166e-05,
+      "loss": 0.9601,
+      "step": 57235
+    },
+    {
+      "epoch": 3.8891153689359967,
+      "grad_norm": 7.0,
+      "learning_rate": 2.569302894415002e-05,
+      "loss": 0.9013,
+      "step": 57240
+    },
+    {
+      "epoch": 3.8894550890066584,
+      "grad_norm": 6.875,
+      "learning_rate": 2.5690905693708385e-05,
+      "loss": 0.8996,
+      "step": 57245
+    },
+    {
+      "epoch": 3.8897948090773204,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5688782443266752e-05,
+      "loss": 0.9273,
+      "step": 57250
+    },
+    {
+      "epoch": 3.890134529147982,
+      "grad_norm": 8.375,
+      "learning_rate": 2.5686659192825113e-05,
+      "loss": 0.8845,
+      "step": 57255
+    },
+    {
+      "epoch": 3.8904742492186437,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.5684535942383477e-05,
+      "loss": 0.8814,
+      "step": 57260
+    },
+    {
+      "epoch": 3.8908139692893053,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.5682412691941844e-05,
+      "loss": 0.8917,
+      "step": 57265
+    },
+    {
+      "epoch": 3.8911536893599674,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5680289441500205e-05,
+      "loss": 0.8224,
+      "step": 57270
+    },
+    {
+      "epoch": 3.891493409430629,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5678166191058565e-05,
+      "loss": 0.8849,
+      "step": 57275
+    },
+    {
+      "epoch": 3.8918331295012907,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5676042940616933e-05,
+      "loss": 0.8749,
+      "step": 57280
+    },
+    {
+      "epoch": 3.8921728495719528,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.5673919690175297e-05,
+      "loss": 0.9364,
+      "step": 57285
+    },
+    {
+      "epoch": 3.8925125696426144,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5671796439733657e-05,
+      "loss": 0.8922,
+      "step": 57290
+    },
+    {
+      "epoch": 3.892852289713276,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.5669673189292025e-05,
+      "loss": 0.9479,
+      "step": 57295
+    },
+    {
+      "epoch": 3.893192009783938,
+      "grad_norm": 7.375,
+      "learning_rate": 2.566754993885039e-05,
+      "loss": 0.809,
+      "step": 57300
+    },
+    {
+      "epoch": 3.8935317298545997,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.566542668840875e-05,
+      "loss": 0.8347,
+      "step": 57305
+    },
+    {
+      "epoch": 3.8938714499252614,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.5663303437967117e-05,
+      "loss": 0.864,
+      "step": 57310
+    },
+    {
+      "epoch": 3.8942111699959234,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.566118018752548e-05,
+      "loss": 0.9057,
+      "step": 57315
+    },
+    {
+      "epoch": 3.894550890066585,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.565905693708384e-05,
+      "loss": 0.8266,
+      "step": 57320
+    },
+    {
+      "epoch": 3.8948906101372467,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.565693368664221e-05,
+      "loss": 0.9809,
+      "step": 57325
+    },
+    {
+      "epoch": 3.8952303302079088,
+      "grad_norm": 6.0625,
+      "learning_rate": 2.5654810436200573e-05,
+      "loss": 1.0356,
+      "step": 57330
+    },
+    {
+      "epoch": 3.8955700502785704,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.5652687185758934e-05,
+      "loss": 0.9475,
+      "step": 57335
+    },
+    {
+      "epoch": 3.895909770349232,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.56505639353173e-05,
+      "loss": 0.9371,
+      "step": 57340
+    },
+    {
+      "epoch": 3.896249490419894,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.564844068487566e-05,
+      "loss": 0.9287,
+      "step": 57345
+    },
+    {
+      "epoch": 3.8965892104905557,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.5646317434434026e-05,
+      "loss": 1.0574,
+      "step": 57350
+    },
+    {
+      "epoch": 3.8969289305612174,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.5644194183992393e-05,
+      "loss": 0.9529,
+      "step": 57355
+    },
+    {
+      "epoch": 3.8972686506318794,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.5642070933550754e-05,
+      "loss": 0.8916,
+      "step": 57360
+    },
+    {
+      "epoch": 3.897608370702541,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.5639947683109118e-05,
+      "loss": 1.0478,
+      "step": 57365
+    },
+    {
+      "epoch": 3.8979480907732027,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.5637824432667485e-05,
+      "loss": 0.984,
+      "step": 57370
+    },
+    {
+      "epoch": 3.8982878108438648,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.5635701182225846e-05,
+      "loss": 1.018,
+      "step": 57375
+    },
+    {
+      "epoch": 3.8986275309145264,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.563357793178421e-05,
+      "loss": 0.9195,
+      "step": 57380
+    },
+    {
+      "epoch": 3.898967250985188,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5631454681342577e-05,
+      "loss": 0.9505,
+      "step": 57385
+    },
+    {
+      "epoch": 3.89930697105585,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.5629331430900938e-05,
+      "loss": 0.9253,
+      "step": 57390
+    },
+    {
+      "epoch": 3.8996466911265117,
+      "grad_norm": 7.375,
+      "learning_rate": 2.56272081804593e-05,
+      "loss": 0.9075,
+      "step": 57395
+    },
+    {
+      "epoch": 3.8999864111971734,
+      "grad_norm": 8.0,
+      "learning_rate": 2.562508493001767e-05,
+      "loss": 0.8521,
+      "step": 57400
+    },
+    {
+      "epoch": 3.9003261312678354,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.562296167957603e-05,
+      "loss": 0.8614,
+      "step": 57405
+    },
+    {
+      "epoch": 3.900665851338497,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5620838429134397e-05,
+      "loss": 0.8845,
+      "step": 57410
+    },
+    {
+      "epoch": 3.9010055714091587,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.561871517869276e-05,
+      "loss": 0.7948,
+      "step": 57415
+    },
+    {
+      "epoch": 3.901345291479821,
+      "grad_norm": 8.625,
+      "learning_rate": 2.561659192825112e-05,
+      "loss": 0.9267,
+      "step": 57420
+    },
+    {
+      "epoch": 3.9016850115504824,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.561446867780949e-05,
+      "loss": 0.9537,
+      "step": 57425
+    },
+    {
+      "epoch": 3.902024731621144,
+      "grad_norm": 7.625,
+      "learning_rate": 2.561234542736785e-05,
+      "loss": 0.9906,
+      "step": 57430
+    },
+    {
+      "epoch": 3.902364451691806,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5610222176926214e-05,
+      "loss": 0.94,
+      "step": 57435
+    },
+    {
+      "epoch": 3.9027041717624678,
+      "grad_norm": 6.75,
+      "learning_rate": 2.560809892648458e-05,
+      "loss": 0.8587,
+      "step": 57440
+    },
+    {
+      "epoch": 3.9030438918331294,
+      "grad_norm": 7.5,
+      "learning_rate": 2.560597567604294e-05,
+      "loss": 0.7174,
+      "step": 57445
+    },
+    {
+      "epoch": 3.9033836119037915,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5603852425601306e-05,
+      "loss": 0.9355,
+      "step": 57450
+    },
+    {
+      "epoch": 3.903723331974453,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5601729175159673e-05,
+      "loss": 1.0156,
+      "step": 57455
+    },
+    {
+      "epoch": 3.9040630520451147,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.5599605924718034e-05,
+      "loss": 0.8321,
+      "step": 57460
+    },
+    {
+      "epoch": 3.904402772115777,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.5597482674276398e-05,
+      "loss": 0.9855,
+      "step": 57465
+    },
+    {
+      "epoch": 3.9047424921864384,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5595359423834765e-05,
+      "loss": 0.9121,
+      "step": 57470
+    },
+    {
+      "epoch": 3.9050822122571,
+      "grad_norm": 6.0625,
+      "learning_rate": 2.5593236173393126e-05,
+      "loss": 0.9611,
+      "step": 57475
+    },
+    {
+      "epoch": 3.905421932327762,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5591112922951486e-05,
+      "loss": 0.871,
+      "step": 57480
+    },
+    {
+      "epoch": 3.9057616523984238,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.5588989672509857e-05,
+      "loss": 0.8707,
+      "step": 57485
+    },
+    {
+      "epoch": 3.9061013724690854,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.5586866422068218e-05,
+      "loss": 0.8075,
+      "step": 57490
+    },
+    {
+      "epoch": 3.9064410925397475,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5584743171626578e-05,
+      "loss": 0.9104,
+      "step": 57495
+    },
+    {
+      "epoch": 3.906780812610409,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5582619921184946e-05,
+      "loss": 0.9826,
+      "step": 57500
+    },
+    {
+      "epoch": 3.9071205326810707,
+      "grad_norm": 8.25,
+      "learning_rate": 2.558049667074331e-05,
+      "loss": 0.9524,
+      "step": 57505
+    },
+    {
+      "epoch": 3.907460252751733,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.557837342030167e-05,
+      "loss": 0.8621,
+      "step": 57510
+    },
+    {
+      "epoch": 3.9077999728223944,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5576250169860038e-05,
+      "loss": 0.9427,
+      "step": 57515
+    },
+    {
+      "epoch": 3.908139692893056,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.55741269194184e-05,
+      "loss": 0.9738,
+      "step": 57520
+    },
+    {
+      "epoch": 3.908479412963718,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.5572003668976762e-05,
+      "loss": 0.8019,
+      "step": 57525
+    },
+    {
+      "epoch": 3.90881913303438,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.556988041853513e-05,
+      "loss": 0.9059,
+      "step": 57530
+    },
+    {
+      "epoch": 3.9091588531050414,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.5567757168093494e-05,
+      "loss": 0.8781,
+      "step": 57535
+    },
+    {
+      "epoch": 3.9094985731757035,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5565633917651854e-05,
+      "loss": 0.787,
+      "step": 57540
+    },
+    {
+      "epoch": 3.909838293246365,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.556351066721022e-05,
+      "loss": 0.9519,
+      "step": 57545
+    },
+    {
+      "epoch": 3.9101780133170267,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5561387416768586e-05,
+      "loss": 0.7924,
+      "step": 57550
+    },
+    {
+      "epoch": 3.910517733387689,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5559264166326946e-05,
+      "loss": 0.9352,
+      "step": 57555
+    },
+    {
+      "epoch": 3.9108574534583505,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.5557140915885314e-05,
+      "loss": 0.9451,
+      "step": 57560
+    },
+    {
+      "epoch": 3.911197173529012,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.5555017665443674e-05,
+      "loss": 0.8889,
+      "step": 57565
+    },
+    {
+      "epoch": 3.9115368935996737,
+      "grad_norm": 7.25,
+      "learning_rate": 2.555289441500204e-05,
+      "loss": 0.9537,
+      "step": 57570
+    },
+    {
+      "epoch": 3.911876613670336,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.5550771164560406e-05,
+      "loss": 0.8838,
+      "step": 57575
+    },
+    {
+      "epoch": 3.9122163337409974,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5548647914118766e-05,
+      "loss": 0.8908,
+      "step": 57580
+    },
+    {
+      "epoch": 3.912556053811659,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.554652466367713e-05,
+      "loss": 0.918,
+      "step": 57585
+    },
+    {
+      "epoch": 3.912895773882321,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.5544401413235498e-05,
+      "loss": 0.9533,
+      "step": 57590
+    },
+    {
+      "epoch": 3.9132354939529828,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.554227816279386e-05,
+      "loss": 0.8258,
+      "step": 57595
+    },
+    {
+      "epoch": 3.9135752140236444,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.5540154912352222e-05,
+      "loss": 1.0169,
+      "step": 57600
+    },
+    {
+      "epoch": 3.913914934094306,
+      "grad_norm": 10.875,
+      "learning_rate": 2.553803166191059e-05,
+      "loss": 0.9599,
+      "step": 57605
+    },
+    {
+      "epoch": 3.914254654164968,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.553590841146895e-05,
+      "loss": 0.888,
+      "step": 57610
+    },
+    {
+      "epoch": 3.9145943742356297,
+      "grad_norm": 6.625,
+      "learning_rate": 2.5533785161027314e-05,
+      "loss": 0.8317,
+      "step": 57615
+    },
+    {
+      "epoch": 3.9149340943062914,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5531661910585682e-05,
+      "loss": 0.8908,
+      "step": 57620
+    },
+    {
+      "epoch": 3.9152738143769534,
+      "grad_norm": 6.625,
+      "learning_rate": 2.5529538660144042e-05,
+      "loss": 1.0912,
+      "step": 57625
+    },
+    {
+      "epoch": 3.915613534447615,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5527415409702403e-05,
+      "loss": 0.8802,
+      "step": 57630
+    },
+    {
+      "epoch": 3.9159532545182767,
+      "grad_norm": 7.75,
+      "learning_rate": 2.552529215926077e-05,
+      "loss": 1.005,
+      "step": 57635
+    },
+    {
+      "epoch": 3.9162929745889388,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5523168908819134e-05,
+      "loss": 0.9263,
+      "step": 57640
+    },
+    {
+      "epoch": 3.9166326946596004,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.5521045658377495e-05,
+      "loss": 0.9918,
+      "step": 57645
+    },
+    {
+      "epoch": 3.916972414730262,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5518922407935862e-05,
+      "loss": 0.9309,
+      "step": 57650
+    },
+    {
+      "epoch": 3.917312134800924,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.5516799157494226e-05,
+      "loss": 0.9445,
+      "step": 57655
+    },
+    {
+      "epoch": 3.9176518548715857,
+      "grad_norm": 8.375,
+      "learning_rate": 2.5514675907052587e-05,
+      "loss": 0.858,
+      "step": 57660
+    },
+    {
+      "epoch": 3.9179915749422474,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5512552656610954e-05,
+      "loss": 0.9971,
+      "step": 57665
+    },
+    {
+      "epoch": 3.9183312950129094,
+      "grad_norm": 7.0,
+      "learning_rate": 2.551042940616932e-05,
+      "loss": 0.9265,
+      "step": 57670
+    },
+    {
+      "epoch": 3.918671015083571,
+      "grad_norm": 8.625,
+      "learning_rate": 2.550830615572768e-05,
+      "loss": 0.9434,
+      "step": 57675
+    },
+    {
+      "epoch": 3.9190107351542327,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.5506182905286046e-05,
+      "loss": 0.9257,
+      "step": 57680
+    },
+    {
+      "epoch": 3.919350455224895,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.550405965484441e-05,
+      "loss": 0.8848,
+      "step": 57685
+    },
+    {
+      "epoch": 3.9196901752955564,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.550193640440277e-05,
+      "loss": 0.9441,
+      "step": 57690
+    },
+    {
+      "epoch": 3.920029895366218,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.549981315396114e-05,
+      "loss": 0.9488,
+      "step": 57695
+    },
+    {
+      "epoch": 3.92036961543688,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.54976899035195e-05,
+      "loss": 0.91,
+      "step": 57700
+    },
+    {
+      "epoch": 3.9207093355075417,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.5495566653077863e-05,
+      "loss": 0.9317,
+      "step": 57705
+    },
+    {
+      "epoch": 3.9210490555782034,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.549344340263623e-05,
+      "loss": 0.9976,
+      "step": 57710
+    },
+    {
+      "epoch": 3.9213887756488655,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.549132015219459e-05,
+      "loss": 0.9331,
+      "step": 57715
+    },
+    {
+      "epoch": 3.921728495719527,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.5489196901752955e-05,
+      "loss": 0.8799,
+      "step": 57720
+    },
+    {
+      "epoch": 3.9220682157901887,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.5487073651311322e-05,
+      "loss": 0.8784,
+      "step": 57725
+    },
+    {
+      "epoch": 3.922407935860851,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5484950400869683e-05,
+      "loss": 0.9187,
+      "step": 57730
+    },
+    {
+      "epoch": 3.9227476559315124,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5482827150428047e-05,
+      "loss": 0.8941,
+      "step": 57735
+    },
+    {
+      "epoch": 3.923087376002174,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5480703899986414e-05,
+      "loss": 0.8936,
+      "step": 57740
+    },
+    {
+      "epoch": 3.923427096072836,
+      "grad_norm": 6.375,
+      "learning_rate": 2.5478580649544775e-05,
+      "loss": 0.8976,
+      "step": 57745
+    },
+    {
+      "epoch": 3.9237668161434978,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5476457399103142e-05,
+      "loss": 0.8844,
+      "step": 57750
+    },
+    {
+      "epoch": 3.9241065362141594,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.5474334148661506e-05,
+      "loss": 1.027,
+      "step": 57755
+    },
+    {
+      "epoch": 3.9244462562848215,
+      "grad_norm": 11.6875,
+      "learning_rate": 2.5472210898219867e-05,
+      "loss": 0.8703,
+      "step": 57760
+    },
+    {
+      "epoch": 3.924785976355483,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5470087647778234e-05,
+      "loss": 0.911,
+      "step": 57765
+    },
+    {
+      "epoch": 3.9251256964261447,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5467964397336595e-05,
+      "loss": 0.8949,
+      "step": 57770
+    },
+    {
+      "epoch": 3.925465416496807,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.546584114689496e-05,
+      "loss": 0.9684,
+      "step": 57775
+    },
+    {
+      "epoch": 3.9258051365674684,
+      "grad_norm": 8.0,
+      "learning_rate": 2.5463717896453326e-05,
+      "loss": 0.9865,
+      "step": 57780
+    },
+    {
+      "epoch": 3.92614485663813,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.5461594646011687e-05,
+      "loss": 0.8772,
+      "step": 57785
+    },
+    {
+      "epoch": 3.926484576708792,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.545947139557005e-05,
+      "loss": 0.8778,
+      "step": 57790
+    },
+    {
+      "epoch": 3.9268242967794538,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.545734814512842e-05,
+      "loss": 1.0109,
+      "step": 57795
+    },
+    {
+      "epoch": 3.9271640168501154,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.545522489468678e-05,
+      "loss": 0.8994,
+      "step": 57800
+    },
+    {
+      "epoch": 3.9275037369207775,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.5453101644245143e-05,
+      "loss": 0.9813,
+      "step": 57805
+    },
+    {
+      "epoch": 3.927843456991439,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.545097839380351e-05,
+      "loss": 0.9226,
+      "step": 57810
+    },
+    {
+      "epoch": 3.9281831770621007,
+      "grad_norm": 7.75,
+      "learning_rate": 2.544885514336187e-05,
+      "loss": 0.9848,
+      "step": 57815
+    },
+    {
+      "epoch": 3.928522897132763,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.5446731892920235e-05,
+      "loss": 0.8475,
+      "step": 57820
+    },
+    {
+      "epoch": 3.9288626172034244,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5444608642478603e-05,
+      "loss": 0.9004,
+      "step": 57825
+    },
+    {
+      "epoch": 3.929202337274086,
+      "grad_norm": 8.0,
+      "learning_rate": 2.5442485392036963e-05,
+      "loss": 0.925,
+      "step": 57830
+    },
+    {
+      "epoch": 3.929542057344748,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5440362141595324e-05,
+      "loss": 0.7925,
+      "step": 57835
+    },
+    {
+      "epoch": 3.92988177741541,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.5438238891153695e-05,
+      "loss": 0.9013,
+      "step": 57840
+    },
+    {
+      "epoch": 3.9302214974860714,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5436115640712055e-05,
+      "loss": 0.9129,
+      "step": 57845
+    },
+    {
+      "epoch": 3.9305612175567335,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.5433992390270416e-05,
+      "loss": 0.9207,
+      "step": 57850
+    },
+    {
+      "epoch": 3.930900937627395,
+      "grad_norm": 8.5,
+      "learning_rate": 2.5431869139828783e-05,
+      "loss": 1.022,
+      "step": 57855
+    },
+    {
+      "epoch": 3.9312406576980568,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5429745889387147e-05,
+      "loss": 0.9575,
+      "step": 57860
+    },
+    {
+      "epoch": 3.931580377768719,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5427622638945508e-05,
+      "loss": 0.8609,
+      "step": 57865
+    },
+    {
+      "epoch": 3.9319200978393805,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.5425499388503875e-05,
+      "loss": 0.9635,
+      "step": 57870
+    },
+    {
+      "epoch": 3.932259817910042,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.542337613806224e-05,
+      "loss": 0.9042,
+      "step": 57875
+    },
+    {
+      "epoch": 3.932599537980704,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.54212528876206e-05,
+      "loss": 0.953,
+      "step": 57880
+    },
+    {
+      "epoch": 3.932939258051366,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.5419129637178967e-05,
+      "loss": 0.8776,
+      "step": 57885
+    },
+    {
+      "epoch": 3.9332789781220274,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.541700638673733e-05,
+      "loss": 0.8709,
+      "step": 57890
+    },
+    {
+      "epoch": 3.9336186981926895,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5414883136295692e-05,
+      "loss": 0.9039,
+      "step": 57895
+    },
+    {
+      "epoch": 3.933958418263351,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.541275988585406e-05,
+      "loss": 0.8994,
+      "step": 57900
+    },
+    {
+      "epoch": 3.9342981383340128,
+      "grad_norm": 7.125,
+      "learning_rate": 2.541063663541242e-05,
+      "loss": 0.9513,
+      "step": 57905
+    },
+    {
+      "epoch": 3.9346378584046744,
+      "grad_norm": 7.875,
+      "learning_rate": 2.5408513384970784e-05,
+      "loss": 0.8827,
+      "step": 57910
+    },
+    {
+      "epoch": 3.9349775784753365,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.540639013452915e-05,
+      "loss": 0.8537,
+      "step": 57915
+    },
+    {
+      "epoch": 3.935317298545998,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5404266884087512e-05,
+      "loss": 0.8814,
+      "step": 57920
+    },
+    {
+      "epoch": 3.9356570186166597,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5402143633645876e-05,
+      "loss": 0.8982,
+      "step": 57925
+    },
+    {
+      "epoch": 3.935996738687322,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5400020383204243e-05,
+      "loss": 0.9033,
+      "step": 57930
+    },
+    {
+      "epoch": 3.9363364587579834,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5397897132762604e-05,
+      "loss": 0.9314,
+      "step": 57935
+    },
+    {
+      "epoch": 3.936676178828645,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5395773882320968e-05,
+      "loss": 0.9486,
+      "step": 57940
+    },
+    {
+      "epoch": 3.9370158988993067,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5393650631879335e-05,
+      "loss": 0.7469,
+      "step": 57945
+    },
+    {
+      "epoch": 3.9373556189699688,
+      "grad_norm": 7.0,
+      "learning_rate": 2.5391527381437696e-05,
+      "loss": 0.9357,
+      "step": 57950
+    },
+    {
+      "epoch": 3.9376953390406304,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.538940413099606e-05,
+      "loss": 1.0178,
+      "step": 57955
+    },
+    {
+      "epoch": 3.938035059111292,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5387280880554427e-05,
+      "loss": 0.892,
+      "step": 57960
+    },
+    {
+      "epoch": 3.938374779181954,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.5385157630112788e-05,
+      "loss": 0.8073,
+      "step": 57965
+    },
+    {
+      "epoch": 3.9387144992526157,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.538303437967115e-05,
+      "loss": 0.9114,
+      "step": 57970
+    },
+    {
+      "epoch": 3.9390542193232774,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.538091112922952e-05,
+      "loss": 0.998,
+      "step": 57975
+    },
+    {
+      "epoch": 3.9393939393939394,
+      "grad_norm": 7.625,
+      "learning_rate": 2.537878787878788e-05,
+      "loss": 0.856,
+      "step": 57980
+    },
+    {
+      "epoch": 3.939733659464601,
+      "grad_norm": 7.25,
+      "learning_rate": 2.537666462834624e-05,
+      "loss": 1.026,
+      "step": 57985
+    },
+    {
+      "epoch": 3.9400733795352627,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5374541377904608e-05,
+      "loss": 0.9658,
+      "step": 57990
+    },
+    {
+      "epoch": 3.940413099605925,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5372418127462972e-05,
+      "loss": 1.0155,
+      "step": 57995
+    },
+    {
+      "epoch": 3.9407528196765864,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.5370294877021332e-05,
+      "loss": 0.9032,
+      "step": 58000
+    },
+    {
+      "epoch": 3.941092539747248,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.53681716265797e-05,
+      "loss": 0.9155,
+      "step": 58005
+    },
+    {
+      "epoch": 3.94143225981791,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.5366048376138064e-05,
+      "loss": 0.9539,
+      "step": 58010
+    },
+    {
+      "epoch": 3.9417719798885718,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.5363925125696424e-05,
+      "loss": 0.831,
+      "step": 58015
+    },
+    {
+      "epoch": 3.9421116999592334,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.5361801875254792e-05,
+      "loss": 0.8156,
+      "step": 58020
+    },
+    {
+      "epoch": 3.9424514200298955,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5359678624813156e-05,
+      "loss": 0.8825,
+      "step": 58025
+    },
+    {
+      "epoch": 3.942791140100557,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5357555374371517e-05,
+      "loss": 0.9283,
+      "step": 58030
+    },
+    {
+      "epoch": 3.9431308601712187,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.5355432123929884e-05,
+      "loss": 0.8895,
+      "step": 58035
+    },
+    {
+      "epoch": 3.943470580241881,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.5353308873488248e-05,
+      "loss": 0.9018,
+      "step": 58040
+    },
+    {
+      "epoch": 3.9438103003125424,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.535118562304661e-05,
+      "loss": 0.8928,
+      "step": 58045
+    },
+    {
+      "epoch": 3.944150020383204,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.5349062372604976e-05,
+      "loss": 0.9064,
+      "step": 58050
+    },
+    {
+      "epoch": 3.944489740453866,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5346939122163337e-05,
+      "loss": 0.9854,
+      "step": 58055
+    },
+    {
+      "epoch": 3.9448294605245278,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.53448158717217e-05,
+      "loss": 0.8084,
+      "step": 58060
+    },
+    {
+      "epoch": 3.9451691805951894,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.5342692621280068e-05,
+      "loss": 0.9673,
+      "step": 58065
+    },
+    {
+      "epoch": 3.9455089006658515,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.534056937083843e-05,
+      "loss": 0.9825,
+      "step": 58070
+    },
+    {
+      "epoch": 3.945848620736513,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.5338446120396793e-05,
+      "loss": 0.9347,
+      "step": 58075
+    },
+    {
+      "epoch": 3.9461883408071747,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.533632286995516e-05,
+      "loss": 0.8969,
+      "step": 58080
+    },
+    {
+      "epoch": 3.946528060877837,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.533419961951352e-05,
+      "loss": 0.9096,
+      "step": 58085
+    },
+    {
+      "epoch": 3.9468677809484984,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5332076369071888e-05,
+      "loss": 0.8913,
+      "step": 58090
+    },
+    {
+      "epoch": 3.94720750101916,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5329953118630252e-05,
+      "loss": 0.8404,
+      "step": 58095
+    },
+    {
+      "epoch": 3.947547221089822,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5327829868188613e-05,
+      "loss": 0.8916,
+      "step": 58100
+    },
+    {
+      "epoch": 3.9478869411604838,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.532570661774698e-05,
+      "loss": 0.8983,
+      "step": 58105
+    },
+    {
+      "epoch": 3.9482266612311454,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.5323583367305344e-05,
+      "loss": 0.9502,
+      "step": 58110
+    },
+    {
+      "epoch": 3.9485663813018075,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5321460116863705e-05,
+      "loss": 0.941,
+      "step": 58115
+    },
+    {
+      "epoch": 3.948906101372469,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5319336866422072e-05,
+      "loss": 0.941,
+      "step": 58120
+    },
+    {
+      "epoch": 3.9492458214431307,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5317213615980433e-05,
+      "loss": 0.9017,
+      "step": 58125
+    },
+    {
+      "epoch": 3.949585541513793,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5315090365538797e-05,
+      "loss": 0.9695,
+      "step": 58130
+    },
+    {
+      "epoch": 3.9499252615844545,
+      "grad_norm": 9.0,
+      "learning_rate": 2.5312967115097164e-05,
+      "loss": 0.9513,
+      "step": 58135
+    },
+    {
+      "epoch": 3.950264981655116,
+      "grad_norm": 7.875,
+      "learning_rate": 2.5310843864655525e-05,
+      "loss": 0.8883,
+      "step": 58140
+    },
+    {
+      "epoch": 3.950604701725778,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.530872061421389e-05,
+      "loss": 1.0503,
+      "step": 58145
+    },
+    {
+      "epoch": 3.95094442179644,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.5306597363772256e-05,
+      "loss": 0.9188,
+      "step": 58150
+    },
+    {
+      "epoch": 3.9512841418671014,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.5304474113330617e-05,
+      "loss": 0.9569,
+      "step": 58155
+    },
+    {
+      "epoch": 3.9516238619377635,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.530235086288898e-05,
+      "loss": 0.9468,
+      "step": 58160
+    },
+    {
+      "epoch": 3.951963582008425,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5300227612447348e-05,
+      "loss": 0.9955,
+      "step": 58165
+    },
+    {
+      "epoch": 3.9523033020790868,
+      "grad_norm": 8.125,
+      "learning_rate": 2.529810436200571e-05,
+      "loss": 0.939,
+      "step": 58170
+    },
+    {
+      "epoch": 3.952643022149749,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.5295981111564073e-05,
+      "loss": 0.8674,
+      "step": 58175
+    },
+    {
+      "epoch": 3.9529827422204105,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.529385786112244e-05,
+      "loss": 0.9158,
+      "step": 58180
+    },
+    {
+      "epoch": 3.953322462291072,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.52917346106808e-05,
+      "loss": 0.9233,
+      "step": 58185
+    },
+    {
+      "epoch": 3.953662182361734,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.528961136023916e-05,
+      "loss": 0.8652,
+      "step": 58190
+    },
+    {
+      "epoch": 3.954001902432396,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.528748810979753e-05,
+      "loss": 0.9235,
+      "step": 58195
+    },
+    {
+      "epoch": 3.9543416225030574,
+      "grad_norm": 8.875,
+      "learning_rate": 2.5285364859355893e-05,
+      "loss": 0.8956,
+      "step": 58200
+    },
+    {
+      "epoch": 3.9546813425737195,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.5283241608914253e-05,
+      "loss": 0.9755,
+      "step": 58205
+    },
+    {
+      "epoch": 3.955021062644381,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.528111835847262e-05,
+      "loss": 0.9069,
+      "step": 58210
+    },
+    {
+      "epoch": 3.9553607827150428,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5278995108030985e-05,
+      "loss": 0.8844,
+      "step": 58215
+    },
+    {
+      "epoch": 3.955700502785705,
+      "grad_norm": 8.25,
+      "learning_rate": 2.5276871857589345e-05,
+      "loss": 0.881,
+      "step": 58220
+    },
+    {
+      "epoch": 3.9560402228563665,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5274748607147713e-05,
+      "loss": 0.8532,
+      "step": 58225
+    },
+    {
+      "epoch": 3.956379942927028,
+      "grad_norm": 6.0,
+      "learning_rate": 2.5272625356706077e-05,
+      "loss": 0.9224,
+      "step": 58230
+    },
+    {
+      "epoch": 3.95671966299769,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5270502106264437e-05,
+      "loss": 0.8954,
+      "step": 58235
+    },
+    {
+      "epoch": 3.957059383068352,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.5268378855822805e-05,
+      "loss": 0.9472,
+      "step": 58240
+    },
+    {
+      "epoch": 3.9573991031390134,
+      "grad_norm": 8.5,
+      "learning_rate": 2.526625560538117e-05,
+      "loss": 0.8919,
+      "step": 58245
+    },
+    {
+      "epoch": 3.957738823209675,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.526413235493953e-05,
+      "loss": 0.8607,
+      "step": 58250
+    },
+    {
+      "epoch": 3.958078543280337,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5262009104497897e-05,
+      "loss": 0.8044,
+      "step": 58255
+    },
+    {
+      "epoch": 3.958418263350999,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5259885854056257e-05,
+      "loss": 0.8859,
+      "step": 58260
+    },
+    {
+      "epoch": 3.9587579834216604,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.525776260361462e-05,
+      "loss": 0.9225,
+      "step": 58265
+    },
+    {
+      "epoch": 3.9590977034923225,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.525563935317299e-05,
+      "loss": 0.9247,
+      "step": 58270
+    },
+    {
+      "epoch": 3.959437423562984,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.525351610273135e-05,
+      "loss": 0.8909,
+      "step": 58275
+    },
+    {
+      "epoch": 3.9597771436336457,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5251392852289713e-05,
+      "loss": 0.9733,
+      "step": 58280
+    },
+    {
+      "epoch": 3.9601168637043074,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.524926960184808e-05,
+      "loss": 0.8215,
+      "step": 58285
+    },
+    {
+      "epoch": 3.9604565837749695,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.524714635140644e-05,
+      "loss": 0.9607,
+      "step": 58290
+    },
+    {
+      "epoch": 3.960796303845631,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5245023100964805e-05,
+      "loss": 0.9122,
+      "step": 58295
+    },
+    {
+      "epoch": 3.9611360239162927,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.5242899850523173e-05,
+      "loss": 0.9283,
+      "step": 58300
+    },
+    {
+      "epoch": 3.961475743986955,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.5240776600081533e-05,
+      "loss": 0.8988,
+      "step": 58305
+    },
+    {
+      "epoch": 3.9618154640576164,
+      "grad_norm": 6.875,
+      "learning_rate": 2.5238653349639897e-05,
+      "loss": 0.9016,
+      "step": 58310
+    },
+    {
+      "epoch": 3.962155184128278,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.5236530099198265e-05,
+      "loss": 0.942,
+      "step": 58315
+    },
+    {
+      "epoch": 3.96249490419894,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5234406848756625e-05,
+      "loss": 0.8639,
+      "step": 58320
+    },
+    {
+      "epoch": 3.9628346242696018,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5232283598314986e-05,
+      "loss": 0.9435,
+      "step": 58325
+    },
+    {
+      "epoch": 3.9631743443402634,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.5230160347873353e-05,
+      "loss": 0.8731,
+      "step": 58330
+    },
+    {
+      "epoch": 3.9635140644109255,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.5228037097431717e-05,
+      "loss": 0.9273,
+      "step": 58335
+    },
+    {
+      "epoch": 3.963853784481587,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5225913846990078e-05,
+      "loss": 0.9073,
+      "step": 58340
+    },
+    {
+      "epoch": 3.9641935045522487,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.5223790596548445e-05,
+      "loss": 0.8558,
+      "step": 58345
+    },
+    {
+      "epoch": 3.964533224622911,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.522166734610681e-05,
+      "loss": 0.9954,
+      "step": 58350
+    },
+    {
+      "epoch": 3.9648729446935724,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.521954409566517e-05,
+      "loss": 0.8778,
+      "step": 58355
+    },
+    {
+      "epoch": 3.965212664764234,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.5217420845223537e-05,
+      "loss": 0.8841,
+      "step": 58360
+    },
+    {
+      "epoch": 3.965552384834896,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.52152975947819e-05,
+      "loss": 0.9444,
+      "step": 58365
+    },
+    {
+      "epoch": 3.9658921049055578,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.5213174344340262e-05,
+      "loss": 0.9145,
+      "step": 58370
+    },
+    {
+      "epoch": 3.9662318249762194,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.521105109389863e-05,
+      "loss": 0.8732,
+      "step": 58375
+    },
+    {
+      "epoch": 3.9665715450468815,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5208927843456993e-05,
+      "loss": 0.9538,
+      "step": 58380
+    },
+    {
+      "epoch": 3.966911265117543,
+      "grad_norm": 8.5,
+      "learning_rate": 2.5206804593015354e-05,
+      "loss": 0.8618,
+      "step": 58385
+    },
+    {
+      "epoch": 3.9672509851882047,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.520468134257372e-05,
+      "loss": 0.9578,
+      "step": 58390
+    },
+    {
+      "epoch": 3.967590705258867,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5202558092132082e-05,
+      "loss": 0.9625,
+      "step": 58395
+    },
+    {
+      "epoch": 3.9679304253295284,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.5200434841690446e-05,
+      "loss": 0.8927,
+      "step": 58400
+    },
+    {
+      "epoch": 3.96827014540019,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.5198311591248813e-05,
+      "loss": 0.881,
+      "step": 58405
+    },
+    {
+      "epoch": 3.968609865470852,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5196188340807174e-05,
+      "loss": 0.8982,
+      "step": 58410
+    },
+    {
+      "epoch": 3.968949585541514,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.5194065090365538e-05,
+      "loss": 0.8817,
+      "step": 58415
+    },
+    {
+      "epoch": 3.9692893056121754,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5191941839923905e-05,
+      "loss": 0.883,
+      "step": 58420
+    },
+    {
+      "epoch": 3.9696290256828375,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5189818589482266e-05,
+      "loss": 0.9387,
+      "step": 58425
+    },
+    {
+      "epoch": 3.969968745753499,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5187695339040633e-05,
+      "loss": 0.8039,
+      "step": 58430
+    },
+    {
+      "epoch": 3.9703084658241607,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5185572088598997e-05,
+      "loss": 0.8764,
+      "step": 58435
+    },
+    {
+      "epoch": 3.970648185894823,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.5183448838157358e-05,
+      "loss": 0.85,
+      "step": 58440
+    },
+    {
+      "epoch": 3.9709879059654845,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5181325587715725e-05,
+      "loss": 0.953,
+      "step": 58445
+    },
+    {
+      "epoch": 3.971327626036146,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.517920233727409e-05,
+      "loss": 0.7702,
+      "step": 58450
+    },
+    {
+      "epoch": 3.971667346106808,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.517707908683245e-05,
+      "loss": 0.9604,
+      "step": 58455
+    },
+    {
+      "epoch": 3.97200706617747,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5174955836390817e-05,
+      "loss": 0.9693,
+      "step": 58460
+    },
+    {
+      "epoch": 3.9723467862481314,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.517283258594918e-05,
+      "loss": 0.7644,
+      "step": 58465
+    },
+    {
+      "epoch": 3.9726865063187935,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.5170709335507542e-05,
+      "loss": 0.807,
+      "step": 58470
+    },
+    {
+      "epoch": 3.973026226389455,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.516858608506591e-05,
+      "loss": 0.844,
+      "step": 58475
+    },
+    {
+      "epoch": 3.9733659464601168,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.516646283462427e-05,
+      "loss": 0.9422,
+      "step": 58480
+    },
+    {
+      "epoch": 3.973705666530779,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5164339584182634e-05,
+      "loss": 0.8169,
+      "step": 58485
+    },
+    {
+      "epoch": 3.9740453866014405,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.5162216333741e-05,
+      "loss": 0.98,
+      "step": 58490
+    },
+    {
+      "epoch": 3.974385106672102,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.5160093083299362e-05,
+      "loss": 0.9168,
+      "step": 58495
+    },
+    {
+      "epoch": 3.974724826742764,
+      "grad_norm": 7.375,
+      "learning_rate": 2.5157969832857726e-05,
+      "loss": 0.9285,
+      "step": 58500
+    },
+    {
+      "epoch": 3.975064546813426,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.5155846582416093e-05,
+      "loss": 0.9767,
+      "step": 58505
+    },
+    {
+      "epoch": 3.9754042668840874,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5153723331974454e-05,
+      "loss": 0.963,
+      "step": 58510
+    },
+    {
+      "epoch": 3.9757439869547495,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.5151600081532818e-05,
+      "loss": 0.9717,
+      "step": 58515
+    },
+    {
+      "epoch": 3.976083707025411,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5149476831091185e-05,
+      "loss": 0.9535,
+      "step": 58520
+    },
+    {
+      "epoch": 3.9764234270960728,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5147353580649546e-05,
+      "loss": 0.8955,
+      "step": 58525
+    },
+    {
+      "epoch": 3.976763147166735,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.5145230330207907e-05,
+      "loss": 0.7789,
+      "step": 58530
+    },
+    {
+      "epoch": 3.9771028672373965,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5143107079766278e-05,
+      "loss": 0.7268,
+      "step": 58535
+    },
+    {
+      "epoch": 3.977442587308058,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.5140983829324638e-05,
+      "loss": 0.9254,
+      "step": 58540
+    },
+    {
+      "epoch": 3.97778230737872,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5138860578883e-05,
+      "loss": 0.8312,
+      "step": 58545
+    },
+    {
+      "epoch": 3.978122027449382,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5136737328441366e-05,
+      "loss": 0.9728,
+      "step": 58550
+    },
+    {
+      "epoch": 3.9784617475200434,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.513461407799973e-05,
+      "loss": 0.9293,
+      "step": 58555
+    },
+    {
+      "epoch": 3.9788014675907055,
+      "grad_norm": 7.0,
+      "learning_rate": 2.513249082755809e-05,
+      "loss": 0.942,
+      "step": 58560
+    },
+    {
+      "epoch": 3.979141187661367,
+      "grad_norm": 7.0,
+      "learning_rate": 2.5130367577116458e-05,
+      "loss": 0.9103,
+      "step": 58565
+    },
+    {
+      "epoch": 3.979480907732029,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5128244326674822e-05,
+      "loss": 0.9583,
+      "step": 58570
+    },
+    {
+      "epoch": 3.979820627802691,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.5126121076233183e-05,
+      "loss": 0.9752,
+      "step": 58575
+    },
+    {
+      "epoch": 3.9801603478733525,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.512399782579155e-05,
+      "loss": 0.9753,
+      "step": 58580
+    },
+    {
+      "epoch": 3.980500067944014,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.5121874575349914e-05,
+      "loss": 0.825,
+      "step": 58585
+    },
+    {
+      "epoch": 3.9808397880146758,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5119751324908275e-05,
+      "loss": 0.8793,
+      "step": 58590
+    },
+    {
+      "epoch": 3.981179508085338,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5117628074466642e-05,
+      "loss": 0.8466,
+      "step": 58595
+    },
+    {
+      "epoch": 3.9815192281559995,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5115504824025006e-05,
+      "loss": 0.8535,
+      "step": 58600
+    },
+    {
+      "epoch": 3.981858948226661,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.5113381573583367e-05,
+      "loss": 0.914,
+      "step": 58605
+    },
+    {
+      "epoch": 3.982198668297323,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.5111258323141734e-05,
+      "loss": 0.864,
+      "step": 58610
+    },
+    {
+      "epoch": 3.982538388367985,
+      "grad_norm": 8.125,
+      "learning_rate": 2.5109135072700095e-05,
+      "loss": 0.9105,
+      "step": 58615
+    },
+    {
+      "epoch": 3.9828781084386464,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.510701182225846e-05,
+      "loss": 0.9973,
+      "step": 58620
+    },
+    {
+      "epoch": 3.983217828509308,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.5104888571816826e-05,
+      "loss": 0.8427,
+      "step": 58625
+    },
+    {
+      "epoch": 3.98355754857997,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5102765321375187e-05,
+      "loss": 0.9732,
+      "step": 58630
+    },
+    {
+      "epoch": 3.9838972686506318,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.510064207093355e-05,
+      "loss": 0.9264,
+      "step": 58635
+    },
+    {
+      "epoch": 3.9842369887212934,
+      "grad_norm": 6.625,
+      "learning_rate": 2.5098518820491918e-05,
+      "loss": 0.9942,
+      "step": 58640
+    },
+    {
+      "epoch": 3.9845767087919555,
+      "grad_norm": 7.875,
+      "learning_rate": 2.509639557005028e-05,
+      "loss": 0.9439,
+      "step": 58645
+    },
+    {
+      "epoch": 3.984916428862617,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.5094272319608643e-05,
+      "loss": 0.9853,
+      "step": 58650
+    },
+    {
+      "epoch": 3.9852561489332787,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.509214906916701e-05,
+      "loss": 0.9119,
+      "step": 58655
+    },
+    {
+      "epoch": 3.985595869003941,
+      "grad_norm": 7.0,
+      "learning_rate": 2.509002581872537e-05,
+      "loss": 0.8804,
+      "step": 58660
+    },
+    {
+      "epoch": 3.9859355890746024,
+      "grad_norm": 7.875,
+      "learning_rate": 2.5087902568283735e-05,
+      "loss": 0.9508,
+      "step": 58665
+    },
+    {
+      "epoch": 3.986275309145264,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.5085779317842102e-05,
+      "loss": 0.8335,
+      "step": 58670
+    },
+    {
+      "epoch": 3.986615029215926,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.5083656067400463e-05,
+      "loss": 0.8915,
+      "step": 58675
+    },
+    {
+      "epoch": 3.9869547492865878,
+      "grad_norm": 7.0,
+      "learning_rate": 2.5081532816958823e-05,
+      "loss": 0.9595,
+      "step": 58680
+    },
+    {
+      "epoch": 3.9872944693572494,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.507940956651719e-05,
+      "loss": 0.8359,
+      "step": 58685
+    },
+    {
+      "epoch": 3.9876341894279115,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.5077286316075555e-05,
+      "loss": 0.847,
+      "step": 58690
+    },
+    {
+      "epoch": 3.987973909498573,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5075163065633915e-05,
+      "loss": 0.9938,
+      "step": 58695
+    },
+    {
+      "epoch": 3.9883136295692347,
+      "grad_norm": 7.875,
+      "learning_rate": 2.5073039815192283e-05,
+      "loss": 0.9655,
+      "step": 58700
+    },
+    {
+      "epoch": 3.988653349639897,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5070916564750647e-05,
+      "loss": 0.9487,
+      "step": 58705
+    },
+    {
+      "epoch": 3.9889930697105584,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.5068793314309007e-05,
+      "loss": 0.9148,
+      "step": 58710
+    },
+    {
+      "epoch": 3.98933278978122,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5066670063867375e-05,
+      "loss": 0.8738,
+      "step": 58715
+    },
+    {
+      "epoch": 3.989672509851882,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.506454681342574e-05,
+      "loss": 0.905,
+      "step": 58720
+    },
+    {
+      "epoch": 3.990012229922544,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.50624235629841e-05,
+      "loss": 0.8297,
+      "step": 58725
+    },
+    {
+      "epoch": 3.9903519499932054,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5060300312542467e-05,
+      "loss": 0.9424,
+      "step": 58730
+    },
+    {
+      "epoch": 3.9906916700638675,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.505817706210083e-05,
+      "loss": 0.8008,
+      "step": 58735
+    },
+    {
+      "epoch": 3.991031390134529,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.505605381165919e-05,
+      "loss": 0.909,
+      "step": 58740
+    },
+    {
+      "epoch": 3.9913711102051908,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.505393056121756e-05,
+      "loss": 0.9675,
+      "step": 58745
+    },
+    {
+      "epoch": 3.991710830275853,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.505180731077592e-05,
+      "loss": 0.8694,
+      "step": 58750
+    },
+    {
+      "epoch": 3.9920505503465145,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.5049684060334284e-05,
+      "loss": 0.8957,
+      "step": 58755
+    },
+    {
+      "epoch": 3.992390270417176,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.504756080989265e-05,
+      "loss": 0.8959,
+      "step": 58760
+    },
+    {
+      "epoch": 3.992729990487838,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.504543755945101e-05,
+      "loss": 0.9108,
+      "step": 58765
+    },
+    {
+      "epoch": 3.9930697105585,
+      "grad_norm": 7.125,
+      "learning_rate": 2.504331430900938e-05,
+      "loss": 0.851,
+      "step": 58770
+    },
+    {
+      "epoch": 3.9934094306291614,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.5041191058567743e-05,
+      "loss": 0.8031,
+      "step": 58775
+    },
+    {
+      "epoch": 3.9937491506998235,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5039067808126104e-05,
+      "loss": 0.8297,
+      "step": 58780
+    },
+    {
+      "epoch": 3.994088870770485,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.503694455768447e-05,
+      "loss": 0.9718,
+      "step": 58785
+    },
+    {
+      "epoch": 3.9944285908411468,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5034821307242835e-05,
+      "loss": 0.7892,
+      "step": 58790
+    },
+    {
+      "epoch": 3.994768310911809,
+      "grad_norm": 7.625,
+      "learning_rate": 2.5032698056801196e-05,
+      "loss": 0.8772,
+      "step": 58795
+    },
+    {
+      "epoch": 3.9951080309824705,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.5030574806359563e-05,
+      "loss": 0.9699,
+      "step": 58800
+    },
+    {
+      "epoch": 3.995447751053132,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.5028451555917927e-05,
+      "loss": 1.0885,
+      "step": 58805
+    },
+    {
+      "epoch": 3.995787471123794,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.5026328305476288e-05,
+      "loss": 0.926,
+      "step": 58810
+    },
+    {
+      "epoch": 3.996127191194456,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5024205055034655e-05,
+      "loss": 0.7955,
+      "step": 58815
+    },
+    {
+      "epoch": 3.9964669112651174,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.5022081804593016e-05,
+      "loss": 0.9037,
+      "step": 58820
+    },
+    {
+      "epoch": 3.9968066313357795,
+      "grad_norm": 6.375,
+      "learning_rate": 2.501995855415138e-05,
+      "loss": 0.9339,
+      "step": 58825
+    },
+    {
+      "epoch": 3.997146351406441,
+      "grad_norm": 7.125,
+      "learning_rate": 2.5017835303709747e-05,
+      "loss": 0.9644,
+      "step": 58830
+    },
+    {
+      "epoch": 3.9974860714771028,
+      "grad_norm": 8.0,
+      "learning_rate": 2.5015712053268108e-05,
+      "loss": 0.937,
+      "step": 58835
+    },
+    {
+      "epoch": 3.997825791547765,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.501358880282647e-05,
+      "loss": 0.943,
+      "step": 58840
+    },
+    {
+      "epoch": 3.9981655116184265,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.501146555238484e-05,
+      "loss": 0.9474,
+      "step": 58845
+    },
+    {
+      "epoch": 3.998505231689088,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.50093423019432e-05,
+      "loss": 0.8662,
+      "step": 58850
+    },
+    {
+      "epoch": 3.99884495175975,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5007219051501564e-05,
+      "loss": 0.9375,
+      "step": 58855
+    },
+    {
+      "epoch": 3.999184671830412,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.500509580105993e-05,
+      "loss": 0.9034,
+      "step": 58860
+    },
+    {
+      "epoch": 3.9995243919010735,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.500297255061829e-05,
+      "loss": 0.9168,
+      "step": 58865
+    },
+    {
+      "epoch": 3.9998641119717355,
+      "grad_norm": 7.5,
+      "learning_rate": 2.5000849300176656e-05,
+      "loss": 0.875,
+      "step": 58870
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 1.7055662870407104,
+      "eval_runtime": 146.8008,
+      "eval_samples_per_second": 70.293,
+      "eval_steps_per_second": 8.787,
+      "step": 58872
+    },
+    {
+      "epoch": 4.000203832042397,
+      "grad_norm": 6.25,
+      "learning_rate": 2.499872604973502e-05,
+      "loss": 0.9218,
+      "step": 58875
+    },
+    {
+      "epoch": 4.000543552113059,
+      "grad_norm": 6.75,
+      "learning_rate": 2.4996602799293384e-05,
+      "loss": 0.8901,
+      "step": 58880
+    },
+    {
+      "epoch": 4.000883272183721,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.4994479548851748e-05,
+      "loss": 0.8735,
+      "step": 58885
+    },
+    {
+      "epoch": 4.001222992254382,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.499235629841011e-05,
+      "loss": 0.9163,
+      "step": 58890
+    },
+    {
+      "epoch": 4.001562712325044,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.4990233047968476e-05,
+      "loss": 0.9171,
+      "step": 58895
+    },
+    {
+      "epoch": 4.001902432395706,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.498810979752684e-05,
+      "loss": 0.8234,
+      "step": 58900
+    },
+    {
+      "epoch": 4.002242152466367,
+      "grad_norm": 7.875,
+      "learning_rate": 2.4985986547085204e-05,
+      "loss": 0.904,
+      "step": 58905
+    },
+    {
+      "epoch": 4.0025818725370295,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4983863296643568e-05,
+      "loss": 0.9079,
+      "step": 58910
+    },
+    {
+      "epoch": 4.0029215926076915,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.498174004620193e-05,
+      "loss": 0.8381,
+      "step": 58915
+    },
+    {
+      "epoch": 4.003261312678353,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.4979616795760292e-05,
+      "loss": 0.8858,
+      "step": 58920
+    },
+    {
+      "epoch": 4.003601032749015,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.497749354531866e-05,
+      "loss": 0.8739,
+      "step": 58925
+    },
+    {
+      "epoch": 4.003940752819677,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4975370294877024e-05,
+      "loss": 0.9668,
+      "step": 58930
+    },
+    {
+      "epoch": 4.004280472890338,
+      "grad_norm": 7.125,
+      "learning_rate": 2.4973247044435384e-05,
+      "loss": 0.8245,
+      "step": 58935
+    },
+    {
+      "epoch": 4.004620192961,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.497112379399375e-05,
+      "loss": 0.8307,
+      "step": 58940
+    },
+    {
+      "epoch": 4.004959913031662,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.4969000543552116e-05,
+      "loss": 0.8516,
+      "step": 58945
+    },
+    {
+      "epoch": 4.005299633102323,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.4966877293110476e-05,
+      "loss": 0.8493,
+      "step": 58950
+    },
+    {
+      "epoch": 4.0056393531729855,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.4964754042668844e-05,
+      "loss": 0.9435,
+      "step": 58955
+    },
+    {
+      "epoch": 4.0059790732436475,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4962630792227208e-05,
+      "loss": 0.8471,
+      "step": 58960
+    },
+    {
+      "epoch": 4.006318793314309,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.4960507541785568e-05,
+      "loss": 0.774,
+      "step": 58965
+    },
+    {
+      "epoch": 4.006658513384971,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.4958384291343932e-05,
+      "loss": 0.8322,
+      "step": 58970
+    },
+    {
+      "epoch": 4.006998233455633,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.49562610409023e-05,
+      "loss": 0.9007,
+      "step": 58975
+    },
+    {
+      "epoch": 4.007337953526294,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.495413779046066e-05,
+      "loss": 0.8344,
+      "step": 58980
+    },
+    {
+      "epoch": 4.007677673596956,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4952014540019024e-05,
+      "loss": 0.8809,
+      "step": 58985
+    },
+    {
+      "epoch": 4.008017393667618,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.494989128957739e-05,
+      "loss": 0.8541,
+      "step": 58990
+    },
+    {
+      "epoch": 4.008357113738279,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.4947768039135752e-05,
+      "loss": 0.9382,
+      "step": 58995
+    },
+    {
+      "epoch": 4.0086968338089415,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.4945644788694116e-05,
+      "loss": 0.8485,
+      "step": 59000
+    },
+    {
+      "epoch": 4.009036553879604,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.494352153825248e-05,
+      "loss": 0.9038,
+      "step": 59005
+    },
+    {
+      "epoch": 4.009376273950265,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4941398287810844e-05,
+      "loss": 0.9014,
+      "step": 59010
+    },
+    {
+      "epoch": 4.009715994020927,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.493927503736921e-05,
+      "loss": 0.7956,
+      "step": 59015
+    },
+    {
+      "epoch": 4.010055714091589,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.4937151786927572e-05,
+      "loss": 0.9194,
+      "step": 59020
+    },
+    {
+      "epoch": 4.01039543416225,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.4935028536485936e-05,
+      "loss": 0.8959,
+      "step": 59025
+    },
+    {
+      "epoch": 4.010735154232912,
+      "grad_norm": 7.5,
+      "learning_rate": 2.49329052860443e-05,
+      "loss": 0.883,
+      "step": 59030
+    },
+    {
+      "epoch": 4.011074874303574,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.4930782035602664e-05,
+      "loss": 0.9271,
+      "step": 59035
+    },
+    {
+      "epoch": 4.011414594374235,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.492865878516103e-05,
+      "loss": 0.9175,
+      "step": 59040
+    },
+    {
+      "epoch": 4.0117543144448975,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.4926535534719392e-05,
+      "loss": 0.9027,
+      "step": 59045
+    },
+    {
+      "epoch": 4.01209403451556,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4924412284277756e-05,
+      "loss": 0.7921,
+      "step": 59050
+    },
+    {
+      "epoch": 4.012433754586221,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.492228903383612e-05,
+      "loss": 0.9577,
+      "step": 59055
+    },
+    {
+      "epoch": 4.012773474656883,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4920165783394484e-05,
+      "loss": 0.898,
+      "step": 59060
+    },
+    {
+      "epoch": 4.013113194727545,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.491804253295285e-05,
+      "loss": 0.8756,
+      "step": 59065
+    },
+    {
+      "epoch": 4.013452914798206,
+      "grad_norm": 8.0,
+      "learning_rate": 2.4915919282511212e-05,
+      "loss": 0.925,
+      "step": 59070
+    },
+    {
+      "epoch": 4.013792634868868,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4913796032069576e-05,
+      "loss": 0.7433,
+      "step": 59075
+    },
+    {
+      "epoch": 4.01413235493953,
+      "grad_norm": 7.875,
+      "learning_rate": 2.491167278162794e-05,
+      "loss": 0.8941,
+      "step": 59080
+    },
+    {
+      "epoch": 4.014472075010191,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.4909549531186304e-05,
+      "loss": 0.8524,
+      "step": 59085
+    },
+    {
+      "epoch": 4.0148117950808535,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.490742628074467e-05,
+      "loss": 0.9202,
+      "step": 59090
+    },
+    {
+      "epoch": 4.015151515151516,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.4905303030303032e-05,
+      "loss": 0.8639,
+      "step": 59095
+    },
+    {
+      "epoch": 4.015491235222177,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4903179779861396e-05,
+      "loss": 0.8924,
+      "step": 59100
+    },
+    {
+      "epoch": 4.015830955292839,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4901056529419757e-05,
+      "loss": 0.7467,
+      "step": 59105
+    },
+    {
+      "epoch": 4.016170675363501,
+      "grad_norm": 8.0,
+      "learning_rate": 2.4898933278978124e-05,
+      "loss": 0.8422,
+      "step": 59110
+    },
+    {
+      "epoch": 4.016510395434162,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.489681002853649e-05,
+      "loss": 0.8181,
+      "step": 59115
+    },
+    {
+      "epoch": 4.016850115504824,
+      "grad_norm": 8.25,
+      "learning_rate": 2.489468677809485e-05,
+      "loss": 0.8094,
+      "step": 59120
+    },
+    {
+      "epoch": 4.017189835575485,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.4892563527653216e-05,
+      "loss": 0.7714,
+      "step": 59125
+    },
+    {
+      "epoch": 4.017529555646147,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.489044027721158e-05,
+      "loss": 0.9313,
+      "step": 59130
+    },
+    {
+      "epoch": 4.0178692757168095,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.488831702676994e-05,
+      "loss": 0.7933,
+      "step": 59135
+    },
+    {
+      "epoch": 4.018208995787471,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4886193776328305e-05,
+      "loss": 0.8262,
+      "step": 59140
+    },
+    {
+      "epoch": 4.018548715858133,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.4884070525886672e-05,
+      "loss": 0.8931,
+      "step": 59145
+    },
+    {
+      "epoch": 4.018888435928795,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4881947275445033e-05,
+      "loss": 0.8322,
+      "step": 59150
+    },
+    {
+      "epoch": 4.019228155999456,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4879824025003397e-05,
+      "loss": 0.9819,
+      "step": 59155
+    },
+    {
+      "epoch": 4.019567876070118,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.4877700774561764e-05,
+      "loss": 0.7972,
+      "step": 59160
+    },
+    {
+      "epoch": 4.01990759614078,
+      "grad_norm": 7.75,
+      "learning_rate": 2.4875577524120125e-05,
+      "loss": 0.7987,
+      "step": 59165
+    },
+    {
+      "epoch": 4.020247316211441,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.487345427367849e-05,
+      "loss": 0.8249,
+      "step": 59170
+    },
+    {
+      "epoch": 4.0205870362821035,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.4871331023236853e-05,
+      "loss": 0.9223,
+      "step": 59175
+    },
+    {
+      "epoch": 4.0209267563527655,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4869207772795217e-05,
+      "loss": 0.8602,
+      "step": 59180
+    },
+    {
+      "epoch": 4.021266476423427,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.486708452235358e-05,
+      "loss": 0.9298,
+      "step": 59185
+    },
+    {
+      "epoch": 4.021606196494089,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4864961271911945e-05,
+      "loss": 0.882,
+      "step": 59190
+    },
+    {
+      "epoch": 4.021945916564751,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.486283802147031e-05,
+      "loss": 0.8584,
+      "step": 59195
+    },
+    {
+      "epoch": 4.022285636635412,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4860714771028673e-05,
+      "loss": 0.8039,
+      "step": 59200
+    },
+    {
+      "epoch": 4.022625356706074,
+      "grad_norm": 7.0,
+      "learning_rate": 2.4858591520587037e-05,
+      "loss": 0.8106,
+      "step": 59205
+    },
+    {
+      "epoch": 4.022965076776736,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.48564682701454e-05,
+      "loss": 0.8775,
+      "step": 59210
+    },
+    {
+      "epoch": 4.023304796847397,
+      "grad_norm": 8.75,
+      "learning_rate": 2.4854345019703765e-05,
+      "loss": 0.939,
+      "step": 59215
+    },
+    {
+      "epoch": 4.0236445169180595,
+      "grad_norm": 7.125,
+      "learning_rate": 2.485222176926213e-05,
+      "loss": 0.8662,
+      "step": 59220
+    },
+    {
+      "epoch": 4.0239842369887215,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.4850098518820493e-05,
+      "loss": 0.9204,
+      "step": 59225
+    },
+    {
+      "epoch": 4.024323957059383,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4847975268378857e-05,
+      "loss": 0.9834,
+      "step": 59230
+    },
+    {
+      "epoch": 4.024663677130045,
+      "grad_norm": 7.5,
+      "learning_rate": 2.484585201793722e-05,
+      "loss": 0.9807,
+      "step": 59235
+    },
+    {
+      "epoch": 4.025003397200707,
+      "grad_norm": 7.0,
+      "learning_rate": 2.4843728767495585e-05,
+      "loss": 0.9018,
+      "step": 59240
+    },
+    {
+      "epoch": 4.025343117271368,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.484160551705395e-05,
+      "loss": 0.8442,
+      "step": 59245
+    },
+    {
+      "epoch": 4.02568283734203,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4839482266612313e-05,
+      "loss": 0.9297,
+      "step": 59250
+    },
+    {
+      "epoch": 4.026022557412692,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4837359016170677e-05,
+      "loss": 0.9165,
+      "step": 59255
+    },
+    {
+      "epoch": 4.026362277483353,
+      "grad_norm": 8.375,
+      "learning_rate": 2.483523576572904e-05,
+      "loss": 0.9669,
+      "step": 59260
+    },
+    {
+      "epoch": 4.0267019975540155,
+      "grad_norm": 6.75,
+      "learning_rate": 2.4833112515287405e-05,
+      "loss": 0.8612,
+      "step": 59265
+    },
+    {
+      "epoch": 4.0270417176246776,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.483098926484577e-05,
+      "loss": 0.8361,
+      "step": 59270
+    },
+    {
+      "epoch": 4.027381437695339,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.482886601440413e-05,
+      "loss": 0.8721,
+      "step": 59275
+    },
+    {
+      "epoch": 4.027721157766001,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4826742763962497e-05,
+      "loss": 0.8733,
+      "step": 59280
+    },
+    {
+      "epoch": 4.028060877836663,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.482461951352086e-05,
+      "loss": 0.7809,
+      "step": 59285
+    },
+    {
+      "epoch": 4.028400597907324,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.4822496263079222e-05,
+      "loss": 0.8499,
+      "step": 59290
+    },
+    {
+      "epoch": 4.028740317977986,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.482037301263759e-05,
+      "loss": 0.8229,
+      "step": 59295
+    },
+    {
+      "epoch": 4.029080038048648,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4818249762195953e-05,
+      "loss": 0.8263,
+      "step": 59300
+    },
+    {
+      "epoch": 4.029419758119309,
+      "grad_norm": 6.875,
+      "learning_rate": 2.4816126511754314e-05,
+      "loss": 0.8472,
+      "step": 59305
+    },
+    {
+      "epoch": 4.0297594781899715,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.4814003261312678e-05,
+      "loss": 0.8672,
+      "step": 59310
+    },
+    {
+      "epoch": 4.030099198260634,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4811880010871045e-05,
+      "loss": 0.8519,
+      "step": 59315
+    },
+    {
+      "epoch": 4.030438918331295,
+      "grad_norm": 7.75,
+      "learning_rate": 2.4809756760429406e-05,
+      "loss": 0.9545,
+      "step": 59320
+    },
+    {
+      "epoch": 4.030778638401957,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.480763350998777e-05,
+      "loss": 0.9082,
+      "step": 59325
+    },
+    {
+      "epoch": 4.031118358472619,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4805510259546137e-05,
+      "loss": 0.8859,
+      "step": 59330
+    },
+    {
+      "epoch": 4.03145807854328,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4803387009104498e-05,
+      "loss": 0.9172,
+      "step": 59335
+    },
+    {
+      "epoch": 4.031797798613942,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4801263758662862e-05,
+      "loss": 0.8778,
+      "step": 59340
+    },
+    {
+      "epoch": 4.032137518684604,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.4799140508221226e-05,
+      "loss": 0.798,
+      "step": 59345
+    },
+    {
+      "epoch": 4.032477238755265,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.479701725777959e-05,
+      "loss": 0.8542,
+      "step": 59350
+    },
+    {
+      "epoch": 4.0328169588259275,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.4794894007337954e-05,
+      "loss": 0.8819,
+      "step": 59355
+    },
+    {
+      "epoch": 4.03315667889659,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.4792770756896318e-05,
+      "loss": 0.8404,
+      "step": 59360
+    },
+    {
+      "epoch": 4.033496398967251,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.4790647506454682e-05,
+      "loss": 0.8345,
+      "step": 59365
+    },
+    {
+      "epoch": 4.033836119037913,
+      "grad_norm": 7.875,
+      "learning_rate": 2.4788524256013046e-05,
+      "loss": 0.9629,
+      "step": 59370
+    },
+    {
+      "epoch": 4.034175839108575,
+      "grad_norm": 5.9375,
+      "learning_rate": 2.478640100557141e-05,
+      "loss": 0.8654,
+      "step": 59375
+    },
+    {
+      "epoch": 4.034515559179236,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.4784277755129777e-05,
+      "loss": 0.856,
+      "step": 59380
+    },
+    {
+      "epoch": 4.034855279249898,
+      "grad_norm": 5.625,
+      "learning_rate": 2.4782154504688138e-05,
+      "loss": 0.8443,
+      "step": 59385
+    },
+    {
+      "epoch": 4.03519499932056,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.4780031254246502e-05,
+      "loss": 0.8571,
+      "step": 59390
+    },
+    {
+      "epoch": 4.035534719391221,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.4777908003804866e-05,
+      "loss": 0.7775,
+      "step": 59395
+    },
+    {
+      "epoch": 4.0358744394618835,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.477578475336323e-05,
+      "loss": 0.6776,
+      "step": 59400
+    },
+    {
+      "epoch": 4.036214159532546,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4773661502921594e-05,
+      "loss": 0.8402,
+      "step": 59405
+    },
+    {
+      "epoch": 4.036553879603207,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4771538252479958e-05,
+      "loss": 0.8138,
+      "step": 59410
+    },
+    {
+      "epoch": 4.036893599673869,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4769415002038322e-05,
+      "loss": 0.7987,
+      "step": 59415
+    },
+    {
+      "epoch": 4.037233319744531,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.4767291751596686e-05,
+      "loss": 0.8127,
+      "step": 59420
+    },
+    {
+      "epoch": 4.037573039815192,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.476516850115505e-05,
+      "loss": 0.8708,
+      "step": 59425
+    },
+    {
+      "epoch": 4.037912759885854,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.4763045250713414e-05,
+      "loss": 0.7788,
+      "step": 59430
+    },
+    {
+      "epoch": 4.038252479956516,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4760922000271778e-05,
+      "loss": 0.9123,
+      "step": 59435
+    },
+    {
+      "epoch": 4.0385922000271774,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4758798749830142e-05,
+      "loss": 0.8789,
+      "step": 59440
+    },
+    {
+      "epoch": 4.0389319200978395,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.4756675499388502e-05,
+      "loss": 0.8979,
+      "step": 59445
+    },
+    {
+      "epoch": 4.039271640168501,
+      "grad_norm": 8.375,
+      "learning_rate": 2.475455224894687e-05,
+      "loss": 0.8173,
+      "step": 59450
+    },
+    {
+      "epoch": 4.039611360239163,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4752428998505234e-05,
+      "loss": 0.9001,
+      "step": 59455
+    },
+    {
+      "epoch": 4.039951080309825,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.4750305748063594e-05,
+      "loss": 0.9089,
+      "step": 59460
+    },
+    {
+      "epoch": 4.040290800380486,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.4748182497621962e-05,
+      "loss": 0.7913,
+      "step": 59465
+    },
+    {
+      "epoch": 4.040630520451148,
+      "grad_norm": 8.75,
+      "learning_rate": 2.4746059247180326e-05,
+      "loss": 0.9317,
+      "step": 59470
+    },
+    {
+      "epoch": 4.04097024052181,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4743935996738687e-05,
+      "loss": 0.862,
+      "step": 59475
+    },
+    {
+      "epoch": 4.041309960592471,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.4741812746297054e-05,
+      "loss": 0.8472,
+      "step": 59480
+    },
+    {
+      "epoch": 4.0416496806631335,
+      "grad_norm": 7.875,
+      "learning_rate": 2.4739689495855418e-05,
+      "loss": 0.9107,
+      "step": 59485
+    },
+    {
+      "epoch": 4.0419894007337955,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.473756624541378e-05,
+      "loss": 0.9446,
+      "step": 59490
+    },
+    {
+      "epoch": 4.042329120804457,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.4735442994972143e-05,
+      "loss": 0.9456,
+      "step": 59495
+    },
+    {
+      "epoch": 4.042668840875119,
+      "grad_norm": 8.0,
+      "learning_rate": 2.473331974453051e-05,
+      "loss": 0.9428,
+      "step": 59500
+    },
+    {
+      "epoch": 4.043008560945781,
+      "grad_norm": 7.0,
+      "learning_rate": 2.473119649408887e-05,
+      "loss": 0.8673,
+      "step": 59505
+    },
+    {
+      "epoch": 4.043348281016442,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.4729073243647235e-05,
+      "loss": 0.8006,
+      "step": 59510
+    },
+    {
+      "epoch": 4.043688001087104,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.4726949993205602e-05,
+      "loss": 0.8256,
+      "step": 59515
+    },
+    {
+      "epoch": 4.044027721157766,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4724826742763963e-05,
+      "loss": 0.8713,
+      "step": 59520
+    },
+    {
+      "epoch": 4.044367441228427,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.4722703492322327e-05,
+      "loss": 0.9033,
+      "step": 59525
+    },
+    {
+      "epoch": 4.0447071612990895,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.472058024188069e-05,
+      "loss": 0.852,
+      "step": 59530
+    },
+    {
+      "epoch": 4.0450468813697515,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4718456991439055e-05,
+      "loss": 0.8631,
+      "step": 59535
+    },
+    {
+      "epoch": 4.045386601440413,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.471633374099742e-05,
+      "loss": 0.9475,
+      "step": 59540
+    },
+    {
+      "epoch": 4.045726321511075,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4714210490555783e-05,
+      "loss": 0.8589,
+      "step": 59545
+    },
+    {
+      "epoch": 4.046066041581737,
+      "grad_norm": 8.0,
+      "learning_rate": 2.471208724011415e-05,
+      "loss": 0.8248,
+      "step": 59550
+    },
+    {
+      "epoch": 4.046405761652398,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.470996398967251e-05,
+      "loss": 0.8196,
+      "step": 59555
+    },
+    {
+      "epoch": 4.04674548172306,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.4707840739230875e-05,
+      "loss": 0.9891,
+      "step": 59560
+    },
+    {
+      "epoch": 4.047085201793722,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.470571748878924e-05,
+      "loss": 0.7879,
+      "step": 59565
+    },
+    {
+      "epoch": 4.047424921864383,
+      "grad_norm": 8.875,
+      "learning_rate": 2.4703594238347603e-05,
+      "loss": 0.8221,
+      "step": 59570
+    },
+    {
+      "epoch": 4.0477646419350455,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.4701470987905967e-05,
+      "loss": 0.8578,
+      "step": 59575
+    },
+    {
+      "epoch": 4.048104362005708,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.469934773746433e-05,
+      "loss": 0.835,
+      "step": 59580
+    },
+    {
+      "epoch": 4.048444082076369,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4697224487022695e-05,
+      "loss": 0.8729,
+      "step": 59585
+    },
+    {
+      "epoch": 4.048783802147031,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.469510123658106e-05,
+      "loss": 0.8545,
+      "step": 59590
+    },
+    {
+      "epoch": 4.049123522217693,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.4692977986139423e-05,
+      "loss": 0.8635,
+      "step": 59595
+    },
+    {
+      "epoch": 4.049463242288354,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.4690854735697787e-05,
+      "loss": 0.9041,
+      "step": 59600
+    },
+    {
+      "epoch": 4.049802962359016,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.468873148525615e-05,
+      "loss": 0.7878,
+      "step": 59605
+    },
+    {
+      "epoch": 4.050142682429678,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4686608234814515e-05,
+      "loss": 0.8055,
+      "step": 59610
+    },
+    {
+      "epoch": 4.050482402500339,
+      "grad_norm": 7.125,
+      "learning_rate": 2.468448498437288e-05,
+      "loss": 0.9243,
+      "step": 59615
+    },
+    {
+      "epoch": 4.0508221225710015,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4682361733931243e-05,
+      "loss": 0.8806,
+      "step": 59620
+    },
+    {
+      "epoch": 4.051161842641664,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4680238483489607e-05,
+      "loss": 0.8416,
+      "step": 59625
+    },
+    {
+      "epoch": 4.051501562712325,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4678115233047967e-05,
+      "loss": 0.8856,
+      "step": 59630
+    },
+    {
+      "epoch": 4.051841282782987,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.4675991982606335e-05,
+      "loss": 0.91,
+      "step": 59635
+    },
+    {
+      "epoch": 4.052181002853649,
+      "grad_norm": 7.25,
+      "learning_rate": 2.46738687321647e-05,
+      "loss": 0.9115,
+      "step": 59640
+    },
+    {
+      "epoch": 4.05252072292431,
+      "grad_norm": 6.75,
+      "learning_rate": 2.467174548172306e-05,
+      "loss": 0.8238,
+      "step": 59645
+    },
+    {
+      "epoch": 4.052860442994972,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4669622231281427e-05,
+      "loss": 0.9049,
+      "step": 59650
+    },
+    {
+      "epoch": 4.053200163065634,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.466749898083979e-05,
+      "loss": 0.8417,
+      "step": 59655
+    },
+    {
+      "epoch": 4.053539883136295,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.466537573039815e-05,
+      "loss": 0.8661,
+      "step": 59660
+    },
+    {
+      "epoch": 4.0538796032069575,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4663252479956515e-05,
+      "loss": 0.8706,
+      "step": 59665
+    },
+    {
+      "epoch": 4.05421932327762,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.4661129229514883e-05,
+      "loss": 0.8823,
+      "step": 59670
+    },
+    {
+      "epoch": 4.054559043348281,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4659005979073243e-05,
+      "loss": 0.8041,
+      "step": 59675
+    },
+    {
+      "epoch": 4.054898763418943,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4656882728631607e-05,
+      "loss": 0.8953,
+      "step": 59680
+    },
+    {
+      "epoch": 4.055238483489605,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4654759478189975e-05,
+      "loss": 0.8193,
+      "step": 59685
+    },
+    {
+      "epoch": 4.055578203560266,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.4652636227748335e-05,
+      "loss": 0.8438,
+      "step": 59690
+    },
+    {
+      "epoch": 4.055917923630928,
+      "grad_norm": 8.0,
+      "learning_rate": 2.46505129773067e-05,
+      "loss": 0.8482,
+      "step": 59695
+    },
+    {
+      "epoch": 4.05625764370159,
+      "grad_norm": 8.75,
+      "learning_rate": 2.4648389726865063e-05,
+      "loss": 0.8121,
+      "step": 59700
+    },
+    {
+      "epoch": 4.056597363772251,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.4646266476423427e-05,
+      "loss": 0.7734,
+      "step": 59705
+    },
+    {
+      "epoch": 4.0569370838429135,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.464414322598179e-05,
+      "loss": 0.8076,
+      "step": 59710
+    },
+    {
+      "epoch": 4.057276803913576,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.4642019975540155e-05,
+      "loss": 0.8528,
+      "step": 59715
+    },
+    {
+      "epoch": 4.057616523984237,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.4639896725098523e-05,
+      "loss": 0.8702,
+      "step": 59720
+    },
+    {
+      "epoch": 4.057956244054899,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.4637773474656883e-05,
+      "loss": 0.8458,
+      "step": 59725
+    },
+    {
+      "epoch": 4.058295964125561,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.4635650224215247e-05,
+      "loss": 0.8883,
+      "step": 59730
+    },
+    {
+      "epoch": 4.058635684196222,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.463352697377361e-05,
+      "loss": 0.9251,
+      "step": 59735
+    },
+    {
+      "epoch": 4.058975404266884,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.4631403723331975e-05,
+      "loss": 0.9214,
+      "step": 59740
+    },
+    {
+      "epoch": 4.059315124337546,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.462928047289034e-05,
+      "loss": 0.9798,
+      "step": 59745
+    },
+    {
+      "epoch": 4.0596548444082075,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4627157222448703e-05,
+      "loss": 0.9013,
+      "step": 59750
+    },
+    {
+      "epoch": 4.0599945644788695,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4625033972007067e-05,
+      "loss": 0.8304,
+      "step": 59755
+    },
+    {
+      "epoch": 4.060334284549532,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.462291072156543e-05,
+      "loss": 0.8997,
+      "step": 59760
+    },
+    {
+      "epoch": 4.060674004620193,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4620787471123795e-05,
+      "loss": 0.8293,
+      "step": 59765
+    },
+    {
+      "epoch": 4.061013724690855,
+      "grad_norm": 7.625,
+      "learning_rate": 2.461866422068216e-05,
+      "loss": 0.9105,
+      "step": 59770
+    },
+    {
+      "epoch": 4.061353444761517,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.4616540970240523e-05,
+      "loss": 0.8849,
+      "step": 59775
+    },
+    {
+      "epoch": 4.061693164832178,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4614417719798887e-05,
+      "loss": 0.8153,
+      "step": 59780
+    },
+    {
+      "epoch": 4.06203288490284,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.461229446935725e-05,
+      "loss": 0.8499,
+      "step": 59785
+    },
+    {
+      "epoch": 4.062372604973502,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4610171218915615e-05,
+      "loss": 0.9076,
+      "step": 59790
+    },
+    {
+      "epoch": 4.0627123250441635,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.460804796847398e-05,
+      "loss": 0.9207,
+      "step": 59795
+    },
+    {
+      "epoch": 4.0630520451148255,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.460592471803234e-05,
+      "loss": 0.8952,
+      "step": 59800
+    },
+    {
+      "epoch": 4.063391765185487,
+      "grad_norm": 7.125,
+      "learning_rate": 2.4603801467590707e-05,
+      "loss": 0.8769,
+      "step": 59805
+    },
+    {
+      "epoch": 4.063731485256149,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.460167821714907e-05,
+      "loss": 0.7852,
+      "step": 59810
+    },
+    {
+      "epoch": 4.064071205326811,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.4599554966707432e-05,
+      "loss": 0.7897,
+      "step": 59815
+    },
+    {
+      "epoch": 4.064410925397472,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.45974317162658e-05,
+      "loss": 0.8461,
+      "step": 59820
+    },
+    {
+      "epoch": 4.064750645468134,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4595308465824163e-05,
+      "loss": 0.8438,
+      "step": 59825
+    },
+    {
+      "epoch": 4.065090365538796,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.4593185215382524e-05,
+      "loss": 0.7754,
+      "step": 59830
+    },
+    {
+      "epoch": 4.065430085609457,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4591061964940888e-05,
+      "loss": 0.8561,
+      "step": 59835
+    },
+    {
+      "epoch": 4.0657698056801195,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4588938714499255e-05,
+      "loss": 0.9353,
+      "step": 59840
+    },
+    {
+      "epoch": 4.0661095257507816,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.4586815464057616e-05,
+      "loss": 0.9325,
+      "step": 59845
+    },
+    {
+      "epoch": 4.066449245821443,
+      "grad_norm": 7.625,
+      "learning_rate": 2.458469221361598e-05,
+      "loss": 0.799,
+      "step": 59850
+    },
+    {
+      "epoch": 4.066788965892105,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4582568963174347e-05,
+      "loss": 0.8339,
+      "step": 59855
+    },
+    {
+      "epoch": 4.067128685962767,
+      "grad_norm": 7.25,
+      "learning_rate": 2.4580445712732708e-05,
+      "loss": 0.8309,
+      "step": 59860
+    },
+    {
+      "epoch": 4.067468406033428,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.4578322462291072e-05,
+      "loss": 0.8808,
+      "step": 59865
+    },
+    {
+      "epoch": 4.06780812610409,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.457619921184944e-05,
+      "loss": 0.9376,
+      "step": 59870
+    },
+    {
+      "epoch": 4.068147846174752,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.45740759614078e-05,
+      "loss": 0.8993,
+      "step": 59875
+    },
+    {
+      "epoch": 4.068487566245413,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4571952710966164e-05,
+      "loss": 0.8288,
+      "step": 59880
+    },
+    {
+      "epoch": 4.0688272863160755,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4569829460524528e-05,
+      "loss": 0.8356,
+      "step": 59885
+    },
+    {
+      "epoch": 4.069167006386738,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4567706210082895e-05,
+      "loss": 0.8538,
+      "step": 59890
+    },
+    {
+      "epoch": 4.069506726457399,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4565582959641256e-05,
+      "loss": 0.878,
+      "step": 59895
+    },
+    {
+      "epoch": 4.069846446528061,
+      "grad_norm": 7.75,
+      "learning_rate": 2.456345970919962e-05,
+      "loss": 0.9569,
+      "step": 59900
+    },
+    {
+      "epoch": 4.070186166598723,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4561336458757987e-05,
+      "loss": 0.8276,
+      "step": 59905
+    },
+    {
+      "epoch": 4.070525886669384,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4559213208316348e-05,
+      "loss": 0.8237,
+      "step": 59910
+    },
+    {
+      "epoch": 4.070865606740046,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4557089957874712e-05,
+      "loss": 0.9727,
+      "step": 59915
+    },
+    {
+      "epoch": 4.071205326810708,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.4554966707433076e-05,
+      "loss": 0.883,
+      "step": 59920
+    },
+    {
+      "epoch": 4.071545046881369,
+      "grad_norm": 7.375,
+      "learning_rate": 2.455284345699144e-05,
+      "loss": 0.846,
+      "step": 59925
+    },
+    {
+      "epoch": 4.0718847669520315,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4550720206549804e-05,
+      "loss": 0.9596,
+      "step": 59930
+    },
+    {
+      "epoch": 4.072224487022694,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.4548596956108168e-05,
+      "loss": 0.8959,
+      "step": 59935
+    },
+    {
+      "epoch": 4.072564207093355,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.4546473705666532e-05,
+      "loss": 0.9724,
+      "step": 59940
+    },
+    {
+      "epoch": 4.072903927164017,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4544350455224896e-05,
+      "loss": 0.8758,
+      "step": 59945
+    },
+    {
+      "epoch": 4.073243647234679,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.454222720478326e-05,
+      "loss": 0.8411,
+      "step": 59950
+    },
+    {
+      "epoch": 4.07358336730534,
+      "grad_norm": 7.125,
+      "learning_rate": 2.4540103954341624e-05,
+      "loss": 0.8831,
+      "step": 59955
+    },
+    {
+      "epoch": 4.073923087376002,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.4537980703899988e-05,
+      "loss": 0.9097,
+      "step": 59960
+    },
+    {
+      "epoch": 4.074262807446664,
+      "grad_norm": 9.8125,
+      "learning_rate": 2.4535857453458352e-05,
+      "loss": 0.7328,
+      "step": 59965
+    },
+    {
+      "epoch": 4.074602527517325,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.4533734203016716e-05,
+      "loss": 1.1448,
+      "step": 59970
+    },
+    {
+      "epoch": 4.0749422475879875,
+      "grad_norm": 7.75,
+      "learning_rate": 2.453161095257508e-05,
+      "loss": 0.8796,
+      "step": 59975
+    },
+    {
+      "epoch": 4.07528196765865,
+      "grad_norm": 6.875,
+      "learning_rate": 2.4529487702133444e-05,
+      "loss": 0.8638,
+      "step": 59980
+    },
+    {
+      "epoch": 4.075621687729311,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.4527364451691805e-05,
+      "loss": 0.8527,
+      "step": 59985
+    },
+    {
+      "epoch": 4.075961407799973,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4525241201250172e-05,
+      "loss": 0.879,
+      "step": 59990
+    },
+    {
+      "epoch": 4.076301127870635,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.4523117950808536e-05,
+      "loss": 0.8404,
+      "step": 59995
+    },
+    {
+      "epoch": 4.076640847941296,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4520994700366897e-05,
+      "loss": 0.9151,
+      "step": 60000
+    },
+    {
+      "epoch": 4.076980568011958,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4518871449925264e-05,
+      "loss": 0.8792,
+      "step": 60005
+    },
+    {
+      "epoch": 4.07732028808262,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.4516748199483628e-05,
+      "loss": 0.8891,
+      "step": 60010
+    },
+    {
+      "epoch": 4.0776600081532814,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.451462494904199e-05,
+      "loss": 0.8011,
+      "step": 60015
+    },
+    {
+      "epoch": 4.0779997282239435,
+      "grad_norm": 8.0,
+      "learning_rate": 2.4512501698600353e-05,
+      "loss": 0.723,
+      "step": 60020
+    },
+    {
+      "epoch": 4.078339448294606,
+      "grad_norm": 8.125,
+      "learning_rate": 2.451037844815872e-05,
+      "loss": 0.8588,
+      "step": 60025
+    },
+    {
+      "epoch": 4.078679168365267,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.450825519771708e-05,
+      "loss": 0.8598,
+      "step": 60030
+    },
+    {
+      "epoch": 4.079018888435929,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.4506131947275445e-05,
+      "loss": 0.7916,
+      "step": 60035
+    },
+    {
+      "epoch": 4.079358608506591,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.4504008696833812e-05,
+      "loss": 0.8968,
+      "step": 60040
+    },
+    {
+      "epoch": 4.079698328577252,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.4501885446392173e-05,
+      "loss": 0.8877,
+      "step": 60045
+    },
+    {
+      "epoch": 4.080038048647914,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.4499762195950537e-05,
+      "loss": 0.8491,
+      "step": 60050
+    },
+    {
+      "epoch": 4.080377768718576,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.44976389455089e-05,
+      "loss": 0.9167,
+      "step": 60055
+    },
+    {
+      "epoch": 4.0807174887892375,
+      "grad_norm": 7.125,
+      "learning_rate": 2.4495515695067268e-05,
+      "loss": 0.9158,
+      "step": 60060
+    },
+    {
+      "epoch": 4.0810572088598995,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.449339244462563e-05,
+      "loss": 0.8757,
+      "step": 60065
+    },
+    {
+      "epoch": 4.081396928930562,
+      "grad_norm": 7.25,
+      "learning_rate": 2.4491269194183993e-05,
+      "loss": 0.8621,
+      "step": 60070
+    },
+    {
+      "epoch": 4.081736649001223,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.448914594374236e-05,
+      "loss": 0.8894,
+      "step": 60075
+    },
+    {
+      "epoch": 4.082076369071885,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.448702269330072e-05,
+      "loss": 0.7737,
+      "step": 60080
+    },
+    {
+      "epoch": 4.082416089142547,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4484899442859085e-05,
+      "loss": 0.902,
+      "step": 60085
+    },
+    {
+      "epoch": 4.082755809213208,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.448277619241745e-05,
+      "loss": 0.8829,
+      "step": 60090
+    },
+    {
+      "epoch": 4.08309552928387,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4480652941975813e-05,
+      "loss": 0.9115,
+      "step": 60095
+    },
+    {
+      "epoch": 4.083435249354532,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4478529691534177e-05,
+      "loss": 0.8705,
+      "step": 60100
+    },
+    {
+      "epoch": 4.0837749694251935,
+      "grad_norm": 8.25,
+      "learning_rate": 2.447640644109254e-05,
+      "loss": 0.8451,
+      "step": 60105
+    },
+    {
+      "epoch": 4.0841146894958555,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4474283190650905e-05,
+      "loss": 0.8934,
+      "step": 60110
+    },
+    {
+      "epoch": 4.084454409566518,
+      "grad_norm": 7.75,
+      "learning_rate": 2.447215994020927e-05,
+      "loss": 0.7911,
+      "step": 60115
+    },
+    {
+      "epoch": 4.084794129637179,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4470036689767633e-05,
+      "loss": 0.8908,
+      "step": 60120
+    },
+    {
+      "epoch": 4.085133849707841,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4467913439325997e-05,
+      "loss": 0.8156,
+      "step": 60125
+    },
+    {
+      "epoch": 4.085473569778502,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.446579018888436e-05,
+      "loss": 0.9386,
+      "step": 60130
+    },
+    {
+      "epoch": 4.085813289849164,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4463666938442725e-05,
+      "loss": 0.9407,
+      "step": 60135
+    },
+    {
+      "epoch": 4.086153009919826,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.446154368800109e-05,
+      "loss": 0.9539,
+      "step": 60140
+    },
+    {
+      "epoch": 4.086492729990487,
+      "grad_norm": 9.6875,
+      "learning_rate": 2.4459420437559453e-05,
+      "loss": 0.9325,
+      "step": 60145
+    },
+    {
+      "epoch": 4.0868324500611495,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.4457297187117817e-05,
+      "loss": 0.85,
+      "step": 60150
+    },
+    {
+      "epoch": 4.0871721701318116,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.4455173936676177e-05,
+      "loss": 0.9834,
+      "step": 60155
+    },
+    {
+      "epoch": 4.087511890202473,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4453050686234545e-05,
+      "loss": 0.8575,
+      "step": 60160
+    },
+    {
+      "epoch": 4.087851610273135,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.445092743579291e-05,
+      "loss": 0.885,
+      "step": 60165
+    },
+    {
+      "epoch": 4.088191330343797,
+      "grad_norm": 6.75,
+      "learning_rate": 2.444880418535127e-05,
+      "loss": 0.832,
+      "step": 60170
+    },
+    {
+      "epoch": 4.088531050414458,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4446680934909637e-05,
+      "loss": 0.853,
+      "step": 60175
+    },
+    {
+      "epoch": 4.08887077048512,
+      "grad_norm": 8.0,
+      "learning_rate": 2.4444557684468e-05,
+      "loss": 0.8825,
+      "step": 60180
+    },
+    {
+      "epoch": 4.089210490555782,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.444243443402636e-05,
+      "loss": 0.8159,
+      "step": 60185
+    },
+    {
+      "epoch": 4.089550210626443,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.4440311183584726e-05,
+      "loss": 0.8518,
+      "step": 60190
+    },
+    {
+      "epoch": 4.0898899306971055,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4438187933143093e-05,
+      "loss": 0.8071,
+      "step": 60195
+    },
+    {
+      "epoch": 4.090229650767768,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4436064682701454e-05,
+      "loss": 0.9249,
+      "step": 60200
+    },
+    {
+      "epoch": 4.090569370838429,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4433941432259818e-05,
+      "loss": 0.8508,
+      "step": 60205
+    },
+    {
+      "epoch": 4.090909090909091,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4431818181818185e-05,
+      "loss": 0.8826,
+      "step": 60210
+    },
+    {
+      "epoch": 4.091248810979753,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4429694931376546e-05,
+      "loss": 0.9402,
+      "step": 60215
+    },
+    {
+      "epoch": 4.091588531050414,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.442757168093491e-05,
+      "loss": 0.7972,
+      "step": 60220
+    },
+    {
+      "epoch": 4.091928251121076,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.4425448430493274e-05,
+      "loss": 0.8919,
+      "step": 60225
+    },
+    {
+      "epoch": 4.092267971191738,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.442332518005164e-05,
+      "loss": 0.8174,
+      "step": 60230
+    },
+    {
+      "epoch": 4.092607691262399,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.442120192961e-05,
+      "loss": 0.8619,
+      "step": 60235
+    },
+    {
+      "epoch": 4.0929474113330615,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.4419078679168366e-05,
+      "loss": 0.8619,
+      "step": 60240
+    },
+    {
+      "epoch": 4.093287131403724,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.4416955428726733e-05,
+      "loss": 0.8258,
+      "step": 60245
+    },
+    {
+      "epoch": 4.093626851474385,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.4414832178285094e-05,
+      "loss": 0.7186,
+      "step": 60250
+    },
+    {
+      "epoch": 4.093966571545047,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.4412708927843458e-05,
+      "loss": 0.8725,
+      "step": 60255
+    },
+    {
+      "epoch": 4.094306291615709,
+      "grad_norm": 8.5,
+      "learning_rate": 2.441058567740182e-05,
+      "loss": 0.8772,
+      "step": 60260
+    },
+    {
+      "epoch": 4.09464601168637,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4408462426960186e-05,
+      "loss": 0.8576,
+      "step": 60265
+    },
+    {
+      "epoch": 4.094985731757032,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.440633917651855e-05,
+      "loss": 0.8647,
+      "step": 60270
+    },
+    {
+      "epoch": 4.095325451827694,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4404215926076914e-05,
+      "loss": 0.8235,
+      "step": 60275
+    },
+    {
+      "epoch": 4.095665171898355,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.4402092675635278e-05,
+      "loss": 0.808,
+      "step": 60280
+    },
+    {
+      "epoch": 4.0960048919690175,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.439996942519364e-05,
+      "loss": 0.8079,
+      "step": 60285
+    },
+    {
+      "epoch": 4.09634461203968,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4397846174752006e-05,
+      "loss": 0.8606,
+      "step": 60290
+    },
+    {
+      "epoch": 4.096684332110341,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.439572292431037e-05,
+      "loss": 0.8714,
+      "step": 60295
+    },
+    {
+      "epoch": 4.097024052181003,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4393599673868734e-05,
+      "loss": 0.9364,
+      "step": 60300
+    },
+    {
+      "epoch": 4.097363772251665,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4391476423427098e-05,
+      "loss": 0.9406,
+      "step": 60305
+    },
+    {
+      "epoch": 4.097703492322326,
+      "grad_norm": 8.0,
+      "learning_rate": 2.438935317298546e-05,
+      "loss": 0.9099,
+      "step": 60310
+    },
+    {
+      "epoch": 4.098043212392988,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.4387229922543826e-05,
+      "loss": 0.8631,
+      "step": 60315
+    },
+    {
+      "epoch": 4.09838293246365,
+      "grad_norm": 6.75,
+      "learning_rate": 2.438510667210219e-05,
+      "loss": 0.8258,
+      "step": 60320
+    },
+    {
+      "epoch": 4.0987226525343115,
+      "grad_norm": 7.125,
+      "learning_rate": 2.438298342166055e-05,
+      "loss": 0.8677,
+      "step": 60325
+    },
+    {
+      "epoch": 4.0990623726049735,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.4380860171218918e-05,
+      "loss": 0.8836,
+      "step": 60330
+    },
+    {
+      "epoch": 4.099402092675636,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.437873692077728e-05,
+      "loss": 0.8156,
+      "step": 60335
+    },
+    {
+      "epoch": 4.099741812746297,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4376613670335642e-05,
+      "loss": 0.8582,
+      "step": 60340
+    },
+    {
+      "epoch": 4.100081532816959,
+      "grad_norm": 8.375,
+      "learning_rate": 2.437449041989401e-05,
+      "loss": 0.8569,
+      "step": 60345
+    },
+    {
+      "epoch": 4.100421252887621,
+      "grad_norm": 8.5,
+      "learning_rate": 2.4372367169452374e-05,
+      "loss": 0.9193,
+      "step": 60350
+    },
+    {
+      "epoch": 4.100760972958282,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4370243919010734e-05,
+      "loss": 0.8562,
+      "step": 60355
+    },
+    {
+      "epoch": 4.101100693028944,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4368120668569098e-05,
+      "loss": 0.9054,
+      "step": 60360
+    },
+    {
+      "epoch": 4.101440413099606,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.4365997418127466e-05,
+      "loss": 0.8576,
+      "step": 60365
+    },
+    {
+      "epoch": 4.1017801331702675,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.4363874167685826e-05,
+      "loss": 0.8188,
+      "step": 60370
+    },
+    {
+      "epoch": 4.1021198532409295,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.436175091724419e-05,
+      "loss": 0.8187,
+      "step": 60375
+    },
+    {
+      "epoch": 4.102459573311592,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4359627666802558e-05,
+      "loss": 0.8138,
+      "step": 60380
+    },
+    {
+      "epoch": 4.102799293382253,
+      "grad_norm": 6.625,
+      "learning_rate": 2.4357504416360918e-05,
+      "loss": 0.8955,
+      "step": 60385
+    },
+    {
+      "epoch": 4.103139013452915,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4355381165919282e-05,
+      "loss": 0.838,
+      "step": 60390
+    },
+    {
+      "epoch": 4.103478733523577,
+      "grad_norm": 7.375,
+      "learning_rate": 2.435325791547765e-05,
+      "loss": 0.8323,
+      "step": 60395
+    },
+    {
+      "epoch": 4.103818453594238,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4351134665036014e-05,
+      "loss": 0.8373,
+      "step": 60400
+    },
+    {
+      "epoch": 4.1041581736649,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.4349011414594374e-05,
+      "loss": 0.9399,
+      "step": 60405
+    },
+    {
+      "epoch": 4.104497893735562,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4346888164152738e-05,
+      "loss": 0.8828,
+      "step": 60410
+    },
+    {
+      "epoch": 4.1048376138062235,
+      "grad_norm": 7.25,
+      "learning_rate": 2.4344764913711106e-05,
+      "loss": 0.9124,
+      "step": 60415
+    },
+    {
+      "epoch": 4.1051773338768855,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4342641663269466e-05,
+      "loss": 0.8269,
+      "step": 60420
+    },
+    {
+      "epoch": 4.105517053947548,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.434051841282783e-05,
+      "loss": 0.8773,
+      "step": 60425
+    },
+    {
+      "epoch": 4.105856774018209,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4338395162386198e-05,
+      "loss": 0.8894,
+      "step": 60430
+    },
+    {
+      "epoch": 4.106196494088871,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.433627191194456e-05,
+      "loss": 0.8203,
+      "step": 60435
+    },
+    {
+      "epoch": 4.106536214159533,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4334148661502922e-05,
+      "loss": 0.903,
+      "step": 60440
+    },
+    {
+      "epoch": 4.106875934230194,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4332025411061286e-05,
+      "loss": 0.9484,
+      "step": 60445
+    },
+    {
+      "epoch": 4.107215654300856,
+      "grad_norm": 8.375,
+      "learning_rate": 2.432990216061965e-05,
+      "loss": 0.9437,
+      "step": 60450
+    },
+    {
+      "epoch": 4.107555374371518,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4327778910178014e-05,
+      "loss": 0.7932,
+      "step": 60455
+    },
+    {
+      "epoch": 4.1078950944421795,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.432565565973638e-05,
+      "loss": 0.8919,
+      "step": 60460
+    },
+    {
+      "epoch": 4.108234814512842,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.4323532409294742e-05,
+      "loss": 0.8777,
+      "step": 60465
+    },
+    {
+      "epoch": 4.108574534583504,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4321409158853106e-05,
+      "loss": 0.8253,
+      "step": 60470
+    },
+    {
+      "epoch": 4.108914254654165,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.431928590841147e-05,
+      "loss": 0.9285,
+      "step": 60475
+    },
+    {
+      "epoch": 4.109253974724827,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.4317162657969834e-05,
+      "loss": 0.8648,
+      "step": 60480
+    },
+    {
+      "epoch": 4.109593694795488,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.43150394075282e-05,
+      "loss": 0.8721,
+      "step": 60485
+    },
+    {
+      "epoch": 4.10993341486615,
+      "grad_norm": 7.25,
+      "learning_rate": 2.4312916157086562e-05,
+      "loss": 0.8916,
+      "step": 60490
+    },
+    {
+      "epoch": 4.110273134936812,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.4310792906644926e-05,
+      "loss": 0.8814,
+      "step": 60495
+    },
+    {
+      "epoch": 4.110612855007473,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.430866965620329e-05,
+      "loss": 0.7887,
+      "step": 60500
+    },
+    {
+      "epoch": 4.1109525750781355,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4306546405761654e-05,
+      "loss": 0.8426,
+      "step": 60505
+    },
+    {
+      "epoch": 4.111292295148798,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.4304423155320015e-05,
+      "loss": 0.9014,
+      "step": 60510
+    },
+    {
+      "epoch": 4.111632015219459,
+      "grad_norm": 8.0,
+      "learning_rate": 2.4302299904878382e-05,
+      "loss": 0.9355,
+      "step": 60515
+    },
+    {
+      "epoch": 4.111971735290121,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4300176654436746e-05,
+      "loss": 0.9102,
+      "step": 60520
+    },
+    {
+      "epoch": 4.112311455360783,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.4298053403995107e-05,
+      "loss": 0.887,
+      "step": 60525
+    },
+    {
+      "epoch": 4.112651175431444,
+      "grad_norm": 8.5,
+      "learning_rate": 2.4295930153553474e-05,
+      "loss": 0.8272,
+      "step": 60530
+    },
+    {
+      "epoch": 4.112990895502106,
+      "grad_norm": 8.75,
+      "learning_rate": 2.429380690311184e-05,
+      "loss": 0.806,
+      "step": 60535
+    },
+    {
+      "epoch": 4.113330615572768,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.42916836526702e-05,
+      "loss": 0.8549,
+      "step": 60540
+    },
+    {
+      "epoch": 4.113670335643429,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.4289560402228563e-05,
+      "loss": 0.9157,
+      "step": 60545
+    },
+    {
+      "epoch": 4.1140100557140915,
+      "grad_norm": 7.5,
+      "learning_rate": 2.428743715178693e-05,
+      "loss": 0.7709,
+      "step": 60550
+    },
+    {
+      "epoch": 4.114349775784754,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.428531390134529e-05,
+      "loss": 0.7898,
+      "step": 60555
+    },
+    {
+      "epoch": 4.114689495855415,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4283190650903655e-05,
+      "loss": 0.7574,
+      "step": 60560
+    },
+    {
+      "epoch": 4.115029215926077,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4281067400462022e-05,
+      "loss": 0.827,
+      "step": 60565
+    },
+    {
+      "epoch": 4.115368935996739,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.4278944150020386e-05,
+      "loss": 0.8217,
+      "step": 60570
+    },
+    {
+      "epoch": 4.1157086560674,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4276820899578747e-05,
+      "loss": 0.7976,
+      "step": 60575
+    },
+    {
+      "epoch": 4.116048376138062,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.427469764913711e-05,
+      "loss": 0.8551,
+      "step": 60580
+    },
+    {
+      "epoch": 4.116388096208724,
+      "grad_norm": 8.125,
+      "learning_rate": 2.427257439869548e-05,
+      "loss": 0.89,
+      "step": 60585
+    },
+    {
+      "epoch": 4.116727816279385,
+      "grad_norm": 9.0,
+      "learning_rate": 2.427045114825384e-05,
+      "loss": 0.9019,
+      "step": 60590
+    },
+    {
+      "epoch": 4.1170675363500475,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.4268327897812203e-05,
+      "loss": 0.9488,
+      "step": 60595
+    },
+    {
+      "epoch": 4.11740725642071,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.426620464737057e-05,
+      "loss": 0.8998,
+      "step": 60600
+    },
+    {
+      "epoch": 4.117746976491371,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.426408139692893e-05,
+      "loss": 0.9494,
+      "step": 60605
+    },
+    {
+      "epoch": 4.118086696562033,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.4261958146487295e-05,
+      "loss": 0.8011,
+      "step": 60610
+    },
+    {
+      "epoch": 4.118426416632695,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.425983489604566e-05,
+      "loss": 0.8025,
+      "step": 60615
+    },
+    {
+      "epoch": 4.118766136703356,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.4257711645604023e-05,
+      "loss": 0.8615,
+      "step": 60620
+    },
+    {
+      "epoch": 4.119105856774018,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4255588395162387e-05,
+      "loss": 0.91,
+      "step": 60625
+    },
+    {
+      "epoch": 4.11944557684468,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.425346514472075e-05,
+      "loss": 0.9194,
+      "step": 60630
+    },
+    {
+      "epoch": 4.1197852969153415,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4251341894279115e-05,
+      "loss": 0.8501,
+      "step": 60635
+    },
+    {
+      "epoch": 4.1201250169860035,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.424921864383748e-05,
+      "loss": 0.857,
+      "step": 60640
+    },
+    {
+      "epoch": 4.120464737056666,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.4247095393395843e-05,
+      "loss": 0.8374,
+      "step": 60645
+    },
+    {
+      "epoch": 4.120804457127327,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.4244972142954207e-05,
+      "loss": 1.1849,
+      "step": 60650
+    },
+    {
+      "epoch": 4.121144177197989,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.424284889251257e-05,
+      "loss": 0.8979,
+      "step": 60655
+    },
+    {
+      "epoch": 4.121483897268651,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4240725642070935e-05,
+      "loss": 0.841,
+      "step": 60660
+    },
+    {
+      "epoch": 4.121823617339312,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.42386023916293e-05,
+      "loss": 0.8102,
+      "step": 60665
+    },
+    {
+      "epoch": 4.122163337409974,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4236479141187663e-05,
+      "loss": 0.8262,
+      "step": 60670
+    },
+    {
+      "epoch": 4.122503057480636,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4234355890746027e-05,
+      "loss": 0.8391,
+      "step": 60675
+    },
+    {
+      "epoch": 4.1228427775512975,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.4232232640304388e-05,
+      "loss": 0.8119,
+      "step": 60680
+    },
+    {
+      "epoch": 4.1231824976219595,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4230109389862755e-05,
+      "loss": 0.7946,
+      "step": 60685
+    },
+    {
+      "epoch": 4.123522217692622,
+      "grad_norm": 7.25,
+      "learning_rate": 2.422798613942112e-05,
+      "loss": 1.0137,
+      "step": 60690
+    },
+    {
+      "epoch": 4.123861937763283,
+      "grad_norm": 8.5,
+      "learning_rate": 2.422586288897948e-05,
+      "loss": 1.1287,
+      "step": 60695
+    },
+    {
+      "epoch": 4.124201657833945,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4223739638537847e-05,
+      "loss": 0.861,
+      "step": 60700
+    },
+    {
+      "epoch": 4.124541377904607,
+      "grad_norm": 7.625,
+      "learning_rate": 2.422161638809621e-05,
+      "loss": 0.9049,
+      "step": 60705
+    },
+    {
+      "epoch": 4.124881097975268,
+      "grad_norm": 7.75,
+      "learning_rate": 2.4219493137654572e-05,
+      "loss": 0.8789,
+      "step": 60710
+    },
+    {
+      "epoch": 4.12522081804593,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4217369887212936e-05,
+      "loss": 0.942,
+      "step": 60715
+    },
+    {
+      "epoch": 4.125560538116592,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4215246636771303e-05,
+      "loss": 0.7905,
+      "step": 60720
+    },
+    {
+      "epoch": 4.1259002581872535,
+      "grad_norm": 8.5,
+      "learning_rate": 2.4213123386329664e-05,
+      "loss": 0.9462,
+      "step": 60725
+    },
+    {
+      "epoch": 4.1262399782579156,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4211000135888028e-05,
+      "loss": 0.9062,
+      "step": 60730
+    },
+    {
+      "epoch": 4.126579698328578,
+      "grad_norm": 7.875,
+      "learning_rate": 2.4208876885446395e-05,
+      "loss": 0.7667,
+      "step": 60735
+    },
+    {
+      "epoch": 4.126919418399239,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.420675363500476e-05,
+      "loss": 0.8209,
+      "step": 60740
+    },
+    {
+      "epoch": 4.127259138469901,
+      "grad_norm": 7.375,
+      "learning_rate": 2.420463038456312e-05,
+      "loss": 0.7579,
+      "step": 60745
+    },
+    {
+      "epoch": 4.127598858540563,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.4202507134121484e-05,
+      "loss": 0.8869,
+      "step": 60750
+    },
+    {
+      "epoch": 4.127938578611224,
+      "grad_norm": 7.375,
+      "learning_rate": 2.420038388367985e-05,
+      "loss": 0.8341,
+      "step": 60755
+    },
+    {
+      "epoch": 4.128278298681886,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.4198260633238212e-05,
+      "loss": 0.9118,
+      "step": 60760
+    },
+    {
+      "epoch": 4.128618018752548,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.4196137382796576e-05,
+      "loss": 0.8605,
+      "step": 60765
+    },
+    {
+      "epoch": 4.1289577388232095,
+      "grad_norm": 7.125,
+      "learning_rate": 2.4194014132354943e-05,
+      "loss": 0.8387,
+      "step": 60770
+    },
+    {
+      "epoch": 4.129297458893872,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.4191890881913304e-05,
+      "loss": 0.7648,
+      "step": 60775
+    },
+    {
+      "epoch": 4.129637178964534,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.4189767631471668e-05,
+      "loss": 0.8302,
+      "step": 60780
+    },
+    {
+      "epoch": 4.129976899035195,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.4187644381030032e-05,
+      "loss": 0.9699,
+      "step": 60785
+    },
+    {
+      "epoch": 4.130316619105857,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4185521130588396e-05,
+      "loss": 0.8487,
+      "step": 60790
+    },
+    {
+      "epoch": 4.130656339176519,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.418339788014676e-05,
+      "loss": 0.904,
+      "step": 60795
+    },
+    {
+      "epoch": 4.13099605924718,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.4181274629705124e-05,
+      "loss": 0.8208,
+      "step": 60800
+    },
+    {
+      "epoch": 4.131335779317842,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.4179151379263488e-05,
+      "loss": 0.8683,
+      "step": 60805
+    },
+    {
+      "epoch": 4.131675499388503,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.4177028128821852e-05,
+      "loss": 0.8305,
+      "step": 60810
+    },
+    {
+      "epoch": 4.1320152194591655,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4174904878380216e-05,
+      "loss": 0.8914,
+      "step": 60815
+    },
+    {
+      "epoch": 4.132354939529828,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.417278162793858e-05,
+      "loss": 0.855,
+      "step": 60820
+    },
+    {
+      "epoch": 4.132694659600489,
+      "grad_norm": 7.75,
+      "learning_rate": 2.4170658377496944e-05,
+      "loss": 0.8628,
+      "step": 60825
+    },
+    {
+      "epoch": 4.133034379671151,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.4168535127055308e-05,
+      "loss": 0.8355,
+      "step": 60830
+    },
+    {
+      "epoch": 4.133374099741813,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4166411876613672e-05,
+      "loss": 0.8188,
+      "step": 60835
+    },
+    {
+      "epoch": 4.133713819812474,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4164288626172036e-05,
+      "loss": 0.9213,
+      "step": 60840
+    },
+    {
+      "epoch": 4.134053539883136,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.41621653757304e-05,
+      "loss": 0.8791,
+      "step": 60845
+    },
+    {
+      "epoch": 4.134393259953798,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.416004212528876e-05,
+      "loss": 0.8311,
+      "step": 60850
+    },
+    {
+      "epoch": 4.134732980024459,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.4157918874847128e-05,
+      "loss": 0.8447,
+      "step": 60855
+    },
+    {
+      "epoch": 4.1350727000951215,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4155795624405492e-05,
+      "loss": 0.7752,
+      "step": 60860
+    },
+    {
+      "epoch": 4.135412420165784,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.4153672373963852e-05,
+      "loss": 1.0179,
+      "step": 60865
+    },
+    {
+      "epoch": 4.135752140236445,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.415154912352222e-05,
+      "loss": 0.8147,
+      "step": 60870
+    },
+    {
+      "epoch": 4.136091860307107,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4149425873080584e-05,
+      "loss": 0.8693,
+      "step": 60875
+    },
+    {
+      "epoch": 4.136431580377769,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4147302622638944e-05,
+      "loss": 0.881,
+      "step": 60880
+    },
+    {
+      "epoch": 4.13677130044843,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.414517937219731e-05,
+      "loss": 0.7572,
+      "step": 60885
+    },
+    {
+      "epoch": 4.137111020519092,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4143056121755676e-05,
+      "loss": 0.9374,
+      "step": 60890
+    },
+    {
+      "epoch": 4.137450740589754,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4140932871314037e-05,
+      "loss": 0.9102,
+      "step": 60895
+    },
+    {
+      "epoch": 4.1377904606604154,
+      "grad_norm": 9.5,
+      "learning_rate": 2.41388096208724e-05,
+      "loss": 0.8606,
+      "step": 60900
+    },
+    {
+      "epoch": 4.1381301807310775,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4136686370430768e-05,
+      "loss": 0.9806,
+      "step": 60905
+    },
+    {
+      "epoch": 4.13846990080174,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.4134563119989132e-05,
+      "loss": 0.8929,
+      "step": 60910
+    },
+    {
+      "epoch": 4.138809620872401,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4132439869547493e-05,
+      "loss": 0.8858,
+      "step": 60915
+    },
+    {
+      "epoch": 4.139149340943063,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.413031661910586e-05,
+      "loss": 0.8161,
+      "step": 60920
+    },
+    {
+      "epoch": 4.139489061013725,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4128193368664224e-05,
+      "loss": 0.9165,
+      "step": 60925
+    },
+    {
+      "epoch": 4.139828781084386,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.4126070118222585e-05,
+      "loss": 0.8888,
+      "step": 60930
+    },
+    {
+      "epoch": 4.140168501155048,
+      "grad_norm": 7.25,
+      "learning_rate": 2.412394686778095e-05,
+      "loss": 0.8199,
+      "step": 60935
+    },
+    {
+      "epoch": 4.14050822122571,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.4121823617339316e-05,
+      "loss": 0.8587,
+      "step": 60940
+    },
+    {
+      "epoch": 4.1408479412963715,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.4119700366897677e-05,
+      "loss": 0.7716,
+      "step": 60945
+    },
+    {
+      "epoch": 4.1411876613670335,
+      "grad_norm": 8.25,
+      "learning_rate": 2.411757711645604e-05,
+      "loss": 0.8855,
+      "step": 60950
+    },
+    {
+      "epoch": 4.141527381437696,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4115453866014408e-05,
+      "loss": 0.9281,
+      "step": 60955
+    },
+    {
+      "epoch": 4.141867101508357,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.411333061557277e-05,
+      "loss": 0.7779,
+      "step": 60960
+    },
+    {
+      "epoch": 4.142206821579019,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.4111207365131133e-05,
+      "loss": 0.8354,
+      "step": 60965
+    },
+    {
+      "epoch": 4.142546541649681,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.4109084114689497e-05,
+      "loss": 0.8967,
+      "step": 60970
+    },
+    {
+      "epoch": 4.142886261720342,
+      "grad_norm": 7.875,
+      "learning_rate": 2.410696086424786e-05,
+      "loss": 0.7471,
+      "step": 60975
+    },
+    {
+      "epoch": 4.143225981791004,
+      "grad_norm": 8.25,
+      "learning_rate": 2.4104837613806225e-05,
+      "loss": 0.8375,
+      "step": 60980
+    },
+    {
+      "epoch": 4.143565701861666,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.410271436336459e-05,
+      "loss": 0.8656,
+      "step": 60985
+    },
+    {
+      "epoch": 4.1439054219323275,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4100591112922953e-05,
+      "loss": 0.8596,
+      "step": 60990
+    },
+    {
+      "epoch": 4.1442451420029895,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4098467862481317e-05,
+      "loss": 0.9046,
+      "step": 60995
+    },
+    {
+      "epoch": 4.144584862073652,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.409634461203968e-05,
+      "loss": 0.8529,
+      "step": 61000
+    },
+    {
+      "epoch": 4.144924582144313,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.4094221361598045e-05,
+      "loss": 0.9752,
+      "step": 61005
+    },
+    {
+      "epoch": 4.145264302214975,
+      "grad_norm": 8.125,
+      "learning_rate": 2.409209811115641e-05,
+      "loss": 0.8349,
+      "step": 61010
+    },
+    {
+      "epoch": 4.145604022285637,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.4089974860714773e-05,
+      "loss": 0.8555,
+      "step": 61015
+    },
+    {
+      "epoch": 4.145943742356298,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4087851610273137e-05,
+      "loss": 0.8346,
+      "step": 61020
+    },
+    {
+      "epoch": 4.14628346242696,
+      "grad_norm": 8.125,
+      "learning_rate": 2.40857283598315e-05,
+      "loss": 0.9158,
+      "step": 61025
+    },
+    {
+      "epoch": 4.146623182497622,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.4083605109389865e-05,
+      "loss": 0.9372,
+      "step": 61030
+    },
+    {
+      "epoch": 4.1469629025682835,
+      "grad_norm": 7.625,
+      "learning_rate": 2.4081481858948225e-05,
+      "loss": 0.8765,
+      "step": 61035
+    },
+    {
+      "epoch": 4.147302622638946,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.4079358608506593e-05,
+      "loss": 0.9541,
+      "step": 61040
+    },
+    {
+      "epoch": 4.147642342709608,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4077235358064957e-05,
+      "loss": 0.8978,
+      "step": 61045
+    },
+    {
+      "epoch": 4.147982062780269,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4075112107623317e-05,
+      "loss": 0.8171,
+      "step": 61050
+    },
+    {
+      "epoch": 4.148321782850931,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.4072988857181685e-05,
+      "loss": 0.9514,
+      "step": 61055
+    },
+    {
+      "epoch": 4.148661502921593,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.407086560674005e-05,
+      "loss": 1.0136,
+      "step": 61060
+    },
+    {
+      "epoch": 4.149001222992254,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.406874235629841e-05,
+      "loss": 0.8707,
+      "step": 61065
+    },
+    {
+      "epoch": 4.149340943062916,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.4066619105856773e-05,
+      "loss": 0.9393,
+      "step": 61070
+    },
+    {
+      "epoch": 4.149680663133578,
+      "grad_norm": 8.375,
+      "learning_rate": 2.406449585541514e-05,
+      "loss": 0.8798,
+      "step": 61075
+    },
+    {
+      "epoch": 4.1500203832042395,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.4062372604973505e-05,
+      "loss": 0.8591,
+      "step": 61080
+    },
+    {
+      "epoch": 4.150360103274902,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4060249354531865e-05,
+      "loss": 0.8691,
+      "step": 61085
+    },
+    {
+      "epoch": 4.150699823345564,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.4058126104090233e-05,
+      "loss": 0.8644,
+      "step": 61090
+    },
+    {
+      "epoch": 4.151039543416225,
+      "grad_norm": 7.125,
+      "learning_rate": 2.4056002853648597e-05,
+      "loss": 0.8843,
+      "step": 61095
+    },
+    {
+      "epoch": 4.151379263486887,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.4053879603206957e-05,
+      "loss": 0.8134,
+      "step": 61100
+    },
+    {
+      "epoch": 4.151718983557549,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.405175635276532e-05,
+      "loss": 0.9446,
+      "step": 61105
+    },
+    {
+      "epoch": 4.15205870362821,
+      "grad_norm": 7.75,
+      "learning_rate": 2.404963310232369e-05,
+      "loss": 0.9687,
+      "step": 61110
+    },
+    {
+      "epoch": 4.152398423698872,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.404750985188205e-05,
+      "loss": 0.8418,
+      "step": 61115
+    },
+    {
+      "epoch": 4.152738143769534,
+      "grad_norm": 7.5,
+      "learning_rate": 2.4045386601440413e-05,
+      "loss": 0.8922,
+      "step": 61120
+    },
+    {
+      "epoch": 4.1530778638401955,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.404326335099878e-05,
+      "loss": 0.9758,
+      "step": 61125
+    },
+    {
+      "epoch": 4.153417583910858,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.404114010055714e-05,
+      "loss": 0.8138,
+      "step": 61130
+    },
+    {
+      "epoch": 4.15375730398152,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4039016850115505e-05,
+      "loss": 0.8463,
+      "step": 61135
+    },
+    {
+      "epoch": 4.154097024052181,
+      "grad_norm": 7.375,
+      "learning_rate": 2.403689359967387e-05,
+      "loss": 0.8387,
+      "step": 61140
+    },
+    {
+      "epoch": 4.154436744122843,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4034770349232233e-05,
+      "loss": 0.8275,
+      "step": 61145
+    },
+    {
+      "epoch": 4.154776464193505,
+      "grad_norm": 7.75,
+      "learning_rate": 2.4032647098790597e-05,
+      "loss": 0.8799,
+      "step": 61150
+    },
+    {
+      "epoch": 4.155116184264166,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.403052384834896e-05,
+      "loss": 0.8768,
+      "step": 61155
+    },
+    {
+      "epoch": 4.155455904334828,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.4028400597907325e-05,
+      "loss": 0.8478,
+      "step": 61160
+    },
+    {
+      "epoch": 4.15579562440549,
+      "grad_norm": 7.625,
+      "learning_rate": 2.402627734746569e-05,
+      "loss": 0.8615,
+      "step": 61165
+    },
+    {
+      "epoch": 4.1561353444761515,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.4024154097024053e-05,
+      "loss": 1.0006,
+      "step": 61170
+    },
+    {
+      "epoch": 4.156475064546814,
+      "grad_norm": 13.3125,
+      "learning_rate": 2.4022030846582417e-05,
+      "loss": 0.8026,
+      "step": 61175
+    },
+    {
+      "epoch": 4.156814784617475,
+      "grad_norm": 7.625,
+      "learning_rate": 2.401990759614078e-05,
+      "loss": 0.8092,
+      "step": 61180
+    },
+    {
+      "epoch": 4.157154504688137,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.4017784345699145e-05,
+      "loss": 0.8079,
+      "step": 61185
+    },
+    {
+      "epoch": 4.157494224758799,
+      "grad_norm": 8.625,
+      "learning_rate": 2.401566109525751e-05,
+      "loss": 0.9628,
+      "step": 61190
+    },
+    {
+      "epoch": 4.15783394482946,
+      "grad_norm": 8.125,
+      "learning_rate": 2.4013537844815873e-05,
+      "loss": 0.9393,
+      "step": 61195
+    },
+    {
+      "epoch": 4.158173664900122,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.4011414594374237e-05,
+      "loss": 0.8632,
+      "step": 61200
+    },
+    {
+      "epoch": 4.158513384970784,
+      "grad_norm": 6.0625,
+      "learning_rate": 2.4009291343932598e-05,
+      "loss": 0.7577,
+      "step": 61205
+    },
+    {
+      "epoch": 4.1588531050414455,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4007168093490965e-05,
+      "loss": 0.8943,
+      "step": 61210
+    },
+    {
+      "epoch": 4.1591928251121075,
+      "grad_norm": 7.375,
+      "learning_rate": 2.400504484304933e-05,
+      "loss": 0.8135,
+      "step": 61215
+    },
+    {
+      "epoch": 4.15953254518277,
+      "grad_norm": 5.5,
+      "learning_rate": 2.400292159260769e-05,
+      "loss": 0.8076,
+      "step": 61220
+    },
+    {
+      "epoch": 4.159872265253431,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.4000798342166057e-05,
+      "loss": 0.8851,
+      "step": 61225
+    },
+    {
+      "epoch": 4.160211985324093,
+      "grad_norm": 8.375,
+      "learning_rate": 2.399867509172442e-05,
+      "loss": 0.8968,
+      "step": 61230
+    },
+    {
+      "epoch": 4.160551705394755,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.3996551841282782e-05,
+      "loss": 0.8351,
+      "step": 61235
+    },
+    {
+      "epoch": 4.160891425465416,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3994428590841146e-05,
+      "loss": 0.8587,
+      "step": 61240
+    },
+    {
+      "epoch": 4.161231145536078,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3992305340399513e-05,
+      "loss": 0.921,
+      "step": 61245
+    },
+    {
+      "epoch": 4.16157086560674,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.3990182089957877e-05,
+      "loss": 0.8515,
+      "step": 61250
+    },
+    {
+      "epoch": 4.1619105856774015,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3988058839516238e-05,
+      "loss": 0.8625,
+      "step": 61255
+    },
+    {
+      "epoch": 4.1622503057480635,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3985935589074605e-05,
+      "loss": 0.9581,
+      "step": 61260
+    },
+    {
+      "epoch": 4.162590025818726,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.398381233863297e-05,
+      "loss": 0.8778,
+      "step": 61265
+    },
+    {
+      "epoch": 4.162929745889387,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.398168908819133e-05,
+      "loss": 0.8935,
+      "step": 61270
+    },
+    {
+      "epoch": 4.163269465960049,
+      "grad_norm": 7.375,
+      "learning_rate": 2.3979565837749694e-05,
+      "loss": 0.9309,
+      "step": 61275
+    },
+    {
+      "epoch": 4.163609186030711,
+      "grad_norm": 8.5,
+      "learning_rate": 2.397744258730806e-05,
+      "loss": 0.8021,
+      "step": 61280
+    },
+    {
+      "epoch": 4.163948906101372,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3975319336866422e-05,
+      "loss": 0.8805,
+      "step": 61285
+    },
+    {
+      "epoch": 4.164288626172034,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.3973196086424786e-05,
+      "loss": 0.771,
+      "step": 61290
+    },
+    {
+      "epoch": 4.164628346242696,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3971072835983153e-05,
+      "loss": 0.9118,
+      "step": 61295
+    },
+    {
+      "epoch": 4.1649680663133575,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3968949585541514e-05,
+      "loss": 0.8619,
+      "step": 61300
+    },
+    {
+      "epoch": 4.1653077863840196,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3966826335099878e-05,
+      "loss": 0.8846,
+      "step": 61305
+    },
+    {
+      "epoch": 4.165647506454682,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3964703084658245e-05,
+      "loss": 0.8,
+      "step": 61310
+    },
+    {
+      "epoch": 4.165987226525343,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3962579834216606e-05,
+      "loss": 0.9179,
+      "step": 61315
+    },
+    {
+      "epoch": 4.166326946596005,
+      "grad_norm": 7.875,
+      "learning_rate": 2.396045658377497e-05,
+      "loss": 0.8442,
+      "step": 61320
+    },
+    {
+      "epoch": 4.166666666666667,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.3958333333333334e-05,
+      "loss": 0.8262,
+      "step": 61325
+    },
+    {
+      "epoch": 4.167006386737328,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3956210082891698e-05,
+      "loss": 0.9345,
+      "step": 61330
+    },
+    {
+      "epoch": 4.16734610680799,
+      "grad_norm": 8.5,
+      "learning_rate": 2.3954086832450062e-05,
+      "loss": 0.88,
+      "step": 61335
+    },
+    {
+      "epoch": 4.167685826878652,
+      "grad_norm": 7.125,
+      "learning_rate": 2.3951963582008426e-05,
+      "loss": 0.8828,
+      "step": 61340
+    },
+    {
+      "epoch": 4.1680255469493135,
+      "grad_norm": 7.0,
+      "learning_rate": 2.394984033156679e-05,
+      "loss": 0.8219,
+      "step": 61345
+    },
+    {
+      "epoch": 4.168365267019976,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3947717081125154e-05,
+      "loss": 0.8646,
+      "step": 61350
+    },
+    {
+      "epoch": 4.168704987090638,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3945593830683518e-05,
+      "loss": 0.9061,
+      "step": 61355
+    },
+    {
+      "epoch": 4.169044707161299,
+      "grad_norm": 7.25,
+      "learning_rate": 2.3943470580241882e-05,
+      "loss": 0.9178,
+      "step": 61360
+    },
+    {
+      "epoch": 4.169384427231961,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.3941347329800246e-05,
+      "loss": 0.9349,
+      "step": 61365
+    },
+    {
+      "epoch": 4.169724147302623,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.393922407935861e-05,
+      "loss": 0.8056,
+      "step": 61370
+    },
+    {
+      "epoch": 4.170063867373284,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.393710082891697e-05,
+      "loss": 0.7389,
+      "step": 61375
+    },
+    {
+      "epoch": 4.170403587443946,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3934977578475338e-05,
+      "loss": 0.9124,
+      "step": 61380
+    },
+    {
+      "epoch": 4.170743307514608,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3932854328033702e-05,
+      "loss": 0.9545,
+      "step": 61385
+    },
+    {
+      "epoch": 4.1710830275852695,
+      "grad_norm": 6.75,
+      "learning_rate": 2.3930731077592063e-05,
+      "loss": 0.788,
+      "step": 61390
+    },
+    {
+      "epoch": 4.171422747655932,
+      "grad_norm": 7.875,
+      "learning_rate": 2.392860782715043e-05,
+      "loss": 0.9067,
+      "step": 61395
+    },
+    {
+      "epoch": 4.171762467726594,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3926484576708794e-05,
+      "loss": 0.9758,
+      "step": 61400
+    },
+    {
+      "epoch": 4.172102187797255,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.3924361326267155e-05,
+      "loss": 0.8185,
+      "step": 61405
+    },
+    {
+      "epoch": 4.172441907867917,
+      "grad_norm": 9.125,
+      "learning_rate": 2.3922238075825522e-05,
+      "loss": 0.8486,
+      "step": 61410
+    },
+    {
+      "epoch": 4.172781627938579,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.3920114825383886e-05,
+      "loss": 0.8531,
+      "step": 61415
+    },
+    {
+      "epoch": 4.17312134800924,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.391799157494225e-05,
+      "loss": 0.8514,
+      "step": 61420
+    },
+    {
+      "epoch": 4.173461068079902,
+      "grad_norm": 7.25,
+      "learning_rate": 2.391586832450061e-05,
+      "loss": 0.8455,
+      "step": 61425
+    },
+    {
+      "epoch": 4.173800788150564,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3913745074058978e-05,
+      "loss": 0.9313,
+      "step": 61430
+    },
+    {
+      "epoch": 4.1741405082212255,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3911621823617342e-05,
+      "loss": 0.8701,
+      "step": 61435
+    },
+    {
+      "epoch": 4.174480228291888,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3909498573175703e-05,
+      "loss": 0.881,
+      "step": 61440
+    },
+    {
+      "epoch": 4.17481994836255,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.390737532273407e-05,
+      "loss": 0.8906,
+      "step": 61445
+    },
+    {
+      "epoch": 4.175159668433211,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.3905252072292434e-05,
+      "loss": 0.9077,
+      "step": 61450
+    },
+    {
+      "epoch": 4.175499388503873,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3903128821850795e-05,
+      "loss": 0.9095,
+      "step": 61455
+    },
+    {
+      "epoch": 4.175839108574535,
+      "grad_norm": 7.75,
+      "learning_rate": 2.390100557140916e-05,
+      "loss": 0.8298,
+      "step": 61460
+    },
+    {
+      "epoch": 4.176178828645196,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.3898882320967526e-05,
+      "loss": 0.8384,
+      "step": 61465
+    },
+    {
+      "epoch": 4.176518548715858,
+      "grad_norm": 7.75,
+      "learning_rate": 2.3896759070525887e-05,
+      "loss": 0.8611,
+      "step": 61470
+    },
+    {
+      "epoch": 4.17685826878652,
+      "grad_norm": 7.375,
+      "learning_rate": 2.389463582008425e-05,
+      "loss": 0.8611,
+      "step": 61475
+    },
+    {
+      "epoch": 4.1771979888571815,
+      "grad_norm": 8.625,
+      "learning_rate": 2.3892512569642618e-05,
+      "loss": 0.8923,
+      "step": 61480
+    },
+    {
+      "epoch": 4.177537708927844,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.389038931920098e-05,
+      "loss": 0.9162,
+      "step": 61485
+    },
+    {
+      "epoch": 4.177877428998505,
+      "grad_norm": 8.75,
+      "learning_rate": 2.3888266068759343e-05,
+      "loss": 0.8629,
+      "step": 61490
+    },
+    {
+      "epoch": 4.178217149069167,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3886142818317707e-05,
+      "loss": 0.7766,
+      "step": 61495
+    },
+    {
+      "epoch": 4.178556869139829,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.388401956787607e-05,
+      "loss": 0.8562,
+      "step": 61500
+    },
+    {
+      "epoch": 4.17889658921049,
+      "grad_norm": 9.25,
+      "learning_rate": 2.3881896317434435e-05,
+      "loss": 0.8446,
+      "step": 61505
+    },
+    {
+      "epoch": 4.179236309281152,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.38797730669928e-05,
+      "loss": 0.8119,
+      "step": 61510
+    },
+    {
+      "epoch": 4.179576029351814,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3877649816551163e-05,
+      "loss": 0.8174,
+      "step": 61515
+    },
+    {
+      "epoch": 4.1799157494224755,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.3875526566109527e-05,
+      "loss": 0.8256,
+      "step": 61520
+    },
+    {
+      "epoch": 4.1802554694931375,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.387340331566789e-05,
+      "loss": 0.7863,
+      "step": 61525
+    },
+    {
+      "epoch": 4.1805951895638,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3871280065226255e-05,
+      "loss": 0.9603,
+      "step": 61530
+    },
+    {
+      "epoch": 4.180934909634461,
+      "grad_norm": 7.25,
+      "learning_rate": 2.386915681478462e-05,
+      "loss": 0.8331,
+      "step": 61535
+    },
+    {
+      "epoch": 4.181274629705123,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.3867033564342983e-05,
+      "loss": 0.9112,
+      "step": 61540
+    },
+    {
+      "epoch": 4.181614349775785,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3864910313901347e-05,
+      "loss": 0.8241,
+      "step": 61545
+    },
+    {
+      "epoch": 4.181954069846446,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.386278706345971e-05,
+      "loss": 0.7949,
+      "step": 61550
+    },
+    {
+      "epoch": 4.182293789917108,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3860663813018075e-05,
+      "loss": 0.9351,
+      "step": 61555
+    },
+    {
+      "epoch": 4.18263350998777,
+      "grad_norm": 8.875,
+      "learning_rate": 2.3858540562576435e-05,
+      "loss": 0.8233,
+      "step": 61560
+    },
+    {
+      "epoch": 4.1829732300584315,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3856417312134803e-05,
+      "loss": 0.8982,
+      "step": 61565
+    },
+    {
+      "epoch": 4.1833129501290935,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3854294061693167e-05,
+      "loss": 0.8509,
+      "step": 61570
+    },
+    {
+      "epoch": 4.183652670199756,
+      "grad_norm": 8.75,
+      "learning_rate": 2.3852170811251527e-05,
+      "loss": 0.9779,
+      "step": 61575
+    },
+    {
+      "epoch": 4.183992390270417,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.3850047560809895e-05,
+      "loss": 0.8364,
+      "step": 61580
+    },
+    {
+      "epoch": 4.184332110341079,
+      "grad_norm": 7.875,
+      "learning_rate": 2.384792431036826e-05,
+      "loss": 0.8759,
+      "step": 61585
+    },
+    {
+      "epoch": 4.184671830411741,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3845801059926623e-05,
+      "loss": 0.8823,
+      "step": 61590
+    },
+    {
+      "epoch": 4.185011550482402,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3843677809484983e-05,
+      "loss": 0.9618,
+      "step": 61595
+    },
+    {
+      "epoch": 4.185351270553064,
+      "grad_norm": 8.375,
+      "learning_rate": 2.384155455904335e-05,
+      "loss": 0.869,
+      "step": 61600
+    },
+    {
+      "epoch": 4.185690990623726,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3839431308601715e-05,
+      "loss": 0.9671,
+      "step": 61605
+    },
+    {
+      "epoch": 4.1860307106943875,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3837308058160076e-05,
+      "loss": 0.8774,
+      "step": 61610
+    },
+    {
+      "epoch": 4.1863704307650496,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3835184807718443e-05,
+      "loss": 0.9218,
+      "step": 61615
+    },
+    {
+      "epoch": 4.186710150835712,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3833061557276807e-05,
+      "loss": 0.8956,
+      "step": 61620
+    },
+    {
+      "epoch": 4.187049870906373,
+      "grad_norm": 8.375,
+      "learning_rate": 2.3830938306835168e-05,
+      "loss": 0.9035,
+      "step": 61625
+    },
+    {
+      "epoch": 4.187389590977035,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.382881505639353e-05,
+      "loss": 0.8164,
+      "step": 61630
+    },
+    {
+      "epoch": 4.187729311047697,
+      "grad_norm": 8.0,
+      "learning_rate": 2.38266918059519e-05,
+      "loss": 0.8397,
+      "step": 61635
+    },
+    {
+      "epoch": 4.188069031118358,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.382456855551026e-05,
+      "loss": 0.8596,
+      "step": 61640
+    },
+    {
+      "epoch": 4.18840875118902,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3822445305068624e-05,
+      "loss": 0.7931,
+      "step": 61645
+    },
+    {
+      "epoch": 4.188748471259682,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.382032205462699e-05,
+      "loss": 0.8013,
+      "step": 61650
+    },
+    {
+      "epoch": 4.1890881913303435,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.381819880418535e-05,
+      "loss": 0.9215,
+      "step": 61655
+    },
+    {
+      "epoch": 4.189427911401006,
+      "grad_norm": 8.375,
+      "learning_rate": 2.3816075553743716e-05,
+      "loss": 0.7424,
+      "step": 61660
+    },
+    {
+      "epoch": 4.189767631471668,
+      "grad_norm": 8.125,
+      "learning_rate": 2.381395230330208e-05,
+      "loss": 0.9505,
+      "step": 61665
+    },
+    {
+      "epoch": 4.190107351542329,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.3811829052860444e-05,
+      "loss": 0.8053,
+      "step": 61670
+    },
+    {
+      "epoch": 4.190447071612991,
+      "grad_norm": 9.125,
+      "learning_rate": 2.3809705802418808e-05,
+      "loss": 0.8132,
+      "step": 61675
+    },
+    {
+      "epoch": 4.190786791683653,
+      "grad_norm": 8.25,
+      "learning_rate": 2.380758255197717e-05,
+      "loss": 0.8726,
+      "step": 61680
+    },
+    {
+      "epoch": 4.191126511754314,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3805459301535536e-05,
+      "loss": 0.8191,
+      "step": 61685
+    },
+    {
+      "epoch": 4.191466231824976,
+      "grad_norm": 7.5,
+      "learning_rate": 2.38033360510939e-05,
+      "loss": 0.9344,
+      "step": 61690
+    },
+    {
+      "epoch": 4.191805951895638,
+      "grad_norm": 9.375,
+      "learning_rate": 2.3801212800652264e-05,
+      "loss": 0.9416,
+      "step": 61695
+    },
+    {
+      "epoch": 4.1921456719662995,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3799089550210628e-05,
+      "loss": 0.9102,
+      "step": 61700
+    },
+    {
+      "epoch": 4.192485392036962,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.379696629976899e-05,
+      "loss": 0.8354,
+      "step": 61705
+    },
+    {
+      "epoch": 4.192825112107624,
+      "grad_norm": 8.375,
+      "learning_rate": 2.3794843049327356e-05,
+      "loss": 0.8324,
+      "step": 61710
+    },
+    {
+      "epoch": 4.193164832178285,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.379271979888572e-05,
+      "loss": 0.8379,
+      "step": 61715
+    },
+    {
+      "epoch": 4.193504552248947,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.3790596548444084e-05,
+      "loss": 0.8624,
+      "step": 61720
+    },
+    {
+      "epoch": 4.193844272319609,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3788473298002448e-05,
+      "loss": 0.9015,
+      "step": 61725
+    },
+    {
+      "epoch": 4.19418399239027,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3786350047560808e-05,
+      "loss": 0.8554,
+      "step": 61730
+    },
+    {
+      "epoch": 4.194523712460932,
+      "grad_norm": 9.3125,
+      "learning_rate": 2.3784226797119176e-05,
+      "loss": 0.8897,
+      "step": 61735
+    },
+    {
+      "epoch": 4.194863432531594,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.378210354667754e-05,
+      "loss": 0.8536,
+      "step": 61740
+    },
+    {
+      "epoch": 4.1952031526022555,
+      "grad_norm": 6.375,
+      "learning_rate": 2.37799802962359e-05,
+      "loss": 0.8957,
+      "step": 61745
+    },
+    {
+      "epoch": 4.195542872672918,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3777857045794268e-05,
+      "loss": 0.8702,
+      "step": 61750
+    },
+    {
+      "epoch": 4.19588259274358,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.377573379535263e-05,
+      "loss": 0.8694,
+      "step": 61755
+    },
+    {
+      "epoch": 4.196222312814241,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3773610544910996e-05,
+      "loss": 0.8812,
+      "step": 61760
+    },
+    {
+      "epoch": 4.196562032884903,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.3771487294469356e-05,
+      "loss": 0.8466,
+      "step": 61765
+    },
+    {
+      "epoch": 4.196901752955565,
+      "grad_norm": 8.75,
+      "learning_rate": 2.3769364044027724e-05,
+      "loss": 0.9513,
+      "step": 61770
+    },
+    {
+      "epoch": 4.197241473026226,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3767240793586088e-05,
+      "loss": 0.8591,
+      "step": 61775
+    },
+    {
+      "epoch": 4.197581193096888,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.3765117543144448e-05,
+      "loss": 0.8282,
+      "step": 61780
+    },
+    {
+      "epoch": 4.19792091316755,
+      "grad_norm": 9.625,
+      "learning_rate": 2.3762994292702816e-05,
+      "loss": 0.8898,
+      "step": 61785
+    },
+    {
+      "epoch": 4.1982606332382115,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.376087104226118e-05,
+      "loss": 0.9069,
+      "step": 61790
+    },
+    {
+      "epoch": 4.198600353308874,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.375874779181954e-05,
+      "loss": 0.8611,
+      "step": 61795
+    },
+    {
+      "epoch": 4.198940073379536,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.3756624541377904e-05,
+      "loss": 0.9235,
+      "step": 61800
+    },
+    {
+      "epoch": 4.199279793450197,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.375450129093627e-05,
+      "loss": 0.8635,
+      "step": 61805
+    },
+    {
+      "epoch": 4.199619513520859,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3752378040494632e-05,
+      "loss": 0.9386,
+      "step": 61810
+    },
+    {
+      "epoch": 4.199959233591521,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3750254790052996e-05,
+      "loss": 0.803,
+      "step": 61815
+    },
+    {
+      "epoch": 4.200298953662182,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3748131539611364e-05,
+      "loss": 0.7231,
+      "step": 61820
+    },
+    {
+      "epoch": 4.200638673732844,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.3746008289169724e-05,
+      "loss": 0.8909,
+      "step": 61825
+    },
+    {
+      "epoch": 4.200978393803506,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3743885038728088e-05,
+      "loss": 0.9041,
+      "step": 61830
+    },
+    {
+      "epoch": 4.2013181138741675,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3741761788286456e-05,
+      "loss": 0.8763,
+      "step": 61835
+    },
+    {
+      "epoch": 4.20165783394483,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3739638537844816e-05,
+      "loss": 0.8403,
+      "step": 61840
+    },
+    {
+      "epoch": 4.201997554015492,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.373751528740318e-05,
+      "loss": 0.8516,
+      "step": 61845
+    },
+    {
+      "epoch": 4.202337274086153,
+      "grad_norm": 7.75,
+      "learning_rate": 2.3735392036961544e-05,
+      "loss": 0.8589,
+      "step": 61850
+    },
+    {
+      "epoch": 4.202676994156815,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3733268786519908e-05,
+      "loss": 0.9102,
+      "step": 61855
+    },
+    {
+      "epoch": 4.203016714227476,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3731145536078272e-05,
+      "loss": 0.868,
+      "step": 61860
+    },
+    {
+      "epoch": 4.203356434298138,
+      "grad_norm": 9.375,
+      "learning_rate": 2.3729022285636636e-05,
+      "loss": 0.8025,
+      "step": 61865
+    },
+    {
+      "epoch": 4.2036961543688,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.3726899035195e-05,
+      "loss": 0.7788,
+      "step": 61870
+    },
+    {
+      "epoch": 4.2040358744394615,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3724775784753364e-05,
+      "loss": 0.819,
+      "step": 61875
+    },
+    {
+      "epoch": 4.2043755945101235,
+      "grad_norm": 7.75,
+      "learning_rate": 2.372265253431173e-05,
+      "loss": 0.9383,
+      "step": 61880
+    },
+    {
+      "epoch": 4.204715314580786,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3720529283870092e-05,
+      "loss": 0.9661,
+      "step": 61885
+    },
+    {
+      "epoch": 4.205055034651447,
+      "grad_norm": 7.125,
+      "learning_rate": 2.3718406033428456e-05,
+      "loss": 0.8691,
+      "step": 61890
+    },
+    {
+      "epoch": 4.205394754722109,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.371628278298682e-05,
+      "loss": 0.8309,
+      "step": 61895
+    },
+    {
+      "epoch": 4.205734474792771,
+      "grad_norm": 7.875,
+      "learning_rate": 2.371415953254518e-05,
+      "loss": 0.8512,
+      "step": 61900
+    },
+    {
+      "epoch": 4.206074194863432,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.371203628210355e-05,
+      "loss": 0.8635,
+      "step": 61905
+    },
+    {
+      "epoch": 4.206413914934094,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3709913031661912e-05,
+      "loss": 0.8331,
+      "step": 61910
+    },
+    {
+      "epoch": 4.206753635004756,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3707789781220273e-05,
+      "loss": 0.9606,
+      "step": 61915
+    },
+    {
+      "epoch": 4.2070933550754175,
+      "grad_norm": 6.625,
+      "learning_rate": 2.370566653077864e-05,
+      "loss": 0.7646,
+      "step": 61920
+    },
+    {
+      "epoch": 4.20743307514608,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3703543280337004e-05,
+      "loss": 0.8517,
+      "step": 61925
+    },
+    {
+      "epoch": 4.207772795216742,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.370142002989537e-05,
+      "loss": 0.9351,
+      "step": 61930
+    },
+    {
+      "epoch": 4.208112515287403,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3699296779453732e-05,
+      "loss": 0.9877,
+      "step": 61935
+    },
+    {
+      "epoch": 4.208452235358065,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.3697173529012096e-05,
+      "loss": 0.8334,
+      "step": 61940
+    },
+    {
+      "epoch": 4.208791955428727,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.369505027857046e-05,
+      "loss": 0.7984,
+      "step": 61945
+    },
+    {
+      "epoch": 4.209131675499388,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.369292702812882e-05,
+      "loss": 0.8749,
+      "step": 61950
+    },
+    {
+      "epoch": 4.20947139557005,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.369080377768719e-05,
+      "loss": 0.8464,
+      "step": 61955
+    },
+    {
+      "epoch": 4.209811115640712,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.3688680527245552e-05,
+      "loss": 0.8121,
+      "step": 61960
+    },
+    {
+      "epoch": 4.2101508357113735,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3686557276803913e-05,
+      "loss": 0.8719,
+      "step": 61965
+    },
+    {
+      "epoch": 4.210490555782036,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.368443402636228e-05,
+      "loss": 0.8441,
+      "step": 61970
+    },
+    {
+      "epoch": 4.210830275852698,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3682310775920644e-05,
+      "loss": 0.9174,
+      "step": 61975
+    },
+    {
+      "epoch": 4.211169995923359,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.3680187525479005e-05,
+      "loss": 0.8769,
+      "step": 61980
+    },
+    {
+      "epoch": 4.211509715994021,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.367806427503737e-05,
+      "loss": 0.9138,
+      "step": 61985
+    },
+    {
+      "epoch": 4.211849436064683,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3675941024595736e-05,
+      "loss": 0.81,
+      "step": 61990
+    },
+    {
+      "epoch": 4.212189156135344,
+      "grad_norm": 7.0,
+      "learning_rate": 2.3673817774154097e-05,
+      "loss": 0.9904,
+      "step": 61995
+    },
+    {
+      "epoch": 4.212528876206006,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.367169452371246e-05,
+      "loss": 0.9533,
+      "step": 62000
+    },
+    {
+      "epoch": 4.212868596276668,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.366957127327083e-05,
+      "loss": 0.8403,
+      "step": 62005
+    },
+    {
+      "epoch": 4.2132083163473295,
+      "grad_norm": 7.75,
+      "learning_rate": 2.366744802282919e-05,
+      "loss": 0.7764,
+      "step": 62010
+    },
+    {
+      "epoch": 4.213548036417992,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3665324772387553e-05,
+      "loss": 0.8219,
+      "step": 62015
+    },
+    {
+      "epoch": 4.213887756488654,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3663201521945917e-05,
+      "loss": 0.8977,
+      "step": 62020
+    },
+    {
+      "epoch": 4.214227476559315,
+      "grad_norm": 8.0,
+      "learning_rate": 2.366107827150428e-05,
+      "loss": 0.8191,
+      "step": 62025
+    },
+    {
+      "epoch": 4.214567196629977,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3658955021062645e-05,
+      "loss": 0.7876,
+      "step": 62030
+    },
+    {
+      "epoch": 4.214906916700639,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.365683177062101e-05,
+      "loss": 0.8715,
+      "step": 62035
+    },
+    {
+      "epoch": 4.2152466367713,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.3654708520179373e-05,
+      "loss": 0.8464,
+      "step": 62040
+    },
+    {
+      "epoch": 4.215586356841962,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3652585269737737e-05,
+      "loss": 0.907,
+      "step": 62045
+    },
+    {
+      "epoch": 4.215926076912624,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.36504620192961e-05,
+      "loss": 0.8713,
+      "step": 62050
+    },
+    {
+      "epoch": 4.2162657969832855,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3648338768854465e-05,
+      "loss": 0.7589,
+      "step": 62055
+    },
+    {
+      "epoch": 4.216605517053948,
+      "grad_norm": 7.75,
+      "learning_rate": 2.364621551841283e-05,
+      "loss": 0.8683,
+      "step": 62060
+    },
+    {
+      "epoch": 4.21694523712461,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3644092267971193e-05,
+      "loss": 0.9678,
+      "step": 62065
+    },
+    {
+      "epoch": 4.217284957195271,
+      "grad_norm": 7.25,
+      "learning_rate": 2.3641969017529557e-05,
+      "loss": 0.9268,
+      "step": 62070
+    },
+    {
+      "epoch": 4.217624677265933,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.363984576708792e-05,
+      "loss": 0.8477,
+      "step": 62075
+    },
+    {
+      "epoch": 4.217964397336595,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3637722516646285e-05,
+      "loss": 0.7707,
+      "step": 62080
+    },
+    {
+      "epoch": 4.218304117407256,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3635599266204646e-05,
+      "loss": 0.9437,
+      "step": 62085
+    },
+    {
+      "epoch": 4.218643837477918,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.3633476015763013e-05,
+      "loss": 0.8649,
+      "step": 62090
+    },
+    {
+      "epoch": 4.21898355754858,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3631352765321377e-05,
+      "loss": 0.9322,
+      "step": 62095
+    },
+    {
+      "epoch": 4.2193232776192415,
+      "grad_norm": 9.0,
+      "learning_rate": 2.362922951487974e-05,
+      "loss": 0.8359,
+      "step": 62100
+    },
+    {
+      "epoch": 4.219662997689904,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3627106264438105e-05,
+      "loss": 0.936,
+      "step": 62105
+    },
+    {
+      "epoch": 4.220002717760566,
+      "grad_norm": 8.625,
+      "learning_rate": 2.362498301399647e-05,
+      "loss": 0.8466,
+      "step": 62110
+    },
+    {
+      "epoch": 4.220342437831227,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3622859763554833e-05,
+      "loss": 0.7595,
+      "step": 62115
+    },
+    {
+      "epoch": 4.220682157901889,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3620736513113194e-05,
+      "loss": 0.9308,
+      "step": 62120
+    },
+    {
+      "epoch": 4.221021877972551,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.361861326267156e-05,
+      "loss": 0.8786,
+      "step": 62125
+    },
+    {
+      "epoch": 4.221361598043212,
+      "grad_norm": 9.3125,
+      "learning_rate": 2.3616490012229925e-05,
+      "loss": 0.8798,
+      "step": 62130
+    },
+    {
+      "epoch": 4.221701318113874,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.3614366761788286e-05,
+      "loss": 0.8348,
+      "step": 62135
+    },
+    {
+      "epoch": 4.222041038184536,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3612243511346653e-05,
+      "loss": 0.8653,
+      "step": 62140
+    },
+    {
+      "epoch": 4.2223807582551975,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.3610120260905017e-05,
+      "loss": 0.8694,
+      "step": 62145
+    },
+    {
+      "epoch": 4.22272047832586,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3607997010463378e-05,
+      "loss": 0.8577,
+      "step": 62150
+    },
+    {
+      "epoch": 4.223060198396522,
+      "grad_norm": 8.5,
+      "learning_rate": 2.3605873760021742e-05,
+      "loss": 0.9431,
+      "step": 62155
+    },
+    {
+      "epoch": 4.223399918467183,
+      "grad_norm": 10.3125,
+      "learning_rate": 2.360375050958011e-05,
+      "loss": 0.8774,
+      "step": 62160
+    },
+    {
+      "epoch": 4.223739638537845,
+      "grad_norm": 7.875,
+      "learning_rate": 2.360162725913847e-05,
+      "loss": 0.8318,
+      "step": 62165
+    },
+    {
+      "epoch": 4.224079358608506,
+      "grad_norm": 10.5625,
+      "learning_rate": 2.3599504008696834e-05,
+      "loss": 0.8893,
+      "step": 62170
+    },
+    {
+      "epoch": 4.224419078679168,
+      "grad_norm": 8.375,
+      "learning_rate": 2.35973807582552e-05,
+      "loss": 0.8751,
+      "step": 62175
+    },
+    {
+      "epoch": 4.22475879874983,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.3595257507813562e-05,
+      "loss": 0.7783,
+      "step": 62180
+    },
+    {
+      "epoch": 4.2250985188204915,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3593134257371926e-05,
+      "loss": 0.8203,
+      "step": 62185
+    },
+    {
+      "epoch": 4.2254382388911536,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.359101100693029e-05,
+      "loss": 0.842,
+      "step": 62190
+    },
+    {
+      "epoch": 4.225777958961816,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3588887756488654e-05,
+      "loss": 0.8788,
+      "step": 62195
+    },
+    {
+      "epoch": 4.226117679032477,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3586764506047018e-05,
+      "loss": 0.8461,
+      "step": 62200
+    },
+    {
+      "epoch": 4.226457399103139,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.3584641255605382e-05,
+      "loss": 0.9448,
+      "step": 62205
+    },
+    {
+      "epoch": 4.226797119173801,
+      "grad_norm": 8.625,
+      "learning_rate": 2.3582518005163746e-05,
+      "loss": 0.8569,
+      "step": 62210
+    },
+    {
+      "epoch": 4.227136839244462,
+      "grad_norm": 8.625,
+      "learning_rate": 2.358039475472211e-05,
+      "loss": 0.9141,
+      "step": 62215
+    },
+    {
+      "epoch": 4.227476559315124,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3578271504280474e-05,
+      "loss": 0.8845,
+      "step": 62220
+    },
+    {
+      "epoch": 4.227816279385786,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3576148253838838e-05,
+      "loss": 0.8953,
+      "step": 62225
+    },
+    {
+      "epoch": 4.2281559994564475,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3574025003397202e-05,
+      "loss": 0.8213,
+      "step": 62230
+    },
+    {
+      "epoch": 4.22849571952711,
+      "grad_norm": 7.0,
+      "learning_rate": 2.3571901752955566e-05,
+      "loss": 0.7292,
+      "step": 62235
+    },
+    {
+      "epoch": 4.228835439597772,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.356977850251393e-05,
+      "loss": 0.7971,
+      "step": 62240
+    },
+    {
+      "epoch": 4.229175159668433,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3567655252072294e-05,
+      "loss": 0.7781,
+      "step": 62245
+    },
+    {
+      "epoch": 4.229514879739095,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3565532001630658e-05,
+      "loss": 0.8627,
+      "step": 62250
+    },
+    {
+      "epoch": 4.229854599809757,
+      "grad_norm": 8.125,
+      "learning_rate": 2.356340875118902e-05,
+      "loss": 0.8842,
+      "step": 62255
+    },
+    {
+      "epoch": 4.230194319880418,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3561285500747386e-05,
+      "loss": 0.9191,
+      "step": 62260
+    },
+    {
+      "epoch": 4.23053403995108,
+      "grad_norm": 7.375,
+      "learning_rate": 2.355916225030575e-05,
+      "loss": 0.8434,
+      "step": 62265
+    },
+    {
+      "epoch": 4.230873760021742,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3557038999864114e-05,
+      "loss": 0.9537,
+      "step": 62270
+    },
+    {
+      "epoch": 4.2312134800924035,
+      "grad_norm": 7.75,
+      "learning_rate": 2.3554915749422478e-05,
+      "loss": 0.9086,
+      "step": 62275
+    },
+    {
+      "epoch": 4.231553200163066,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3552792498980842e-05,
+      "loss": 0.8381,
+      "step": 62280
+    },
+    {
+      "epoch": 4.231892920233728,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3550669248539206e-05,
+      "loss": 0.7831,
+      "step": 62285
+    },
+    {
+      "epoch": 4.232232640304389,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3548545998097566e-05,
+      "loss": 0.9506,
+      "step": 62290
+    },
+    {
+      "epoch": 4.232572360375051,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3546422747655934e-05,
+      "loss": 0.8303,
+      "step": 62295
+    },
+    {
+      "epoch": 4.232912080445713,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3544299497214298e-05,
+      "loss": 0.8714,
+      "step": 62300
+    },
+    {
+      "epoch": 4.233251800516374,
+      "grad_norm": 7.375,
+      "learning_rate": 2.354217624677266e-05,
+      "loss": 0.8537,
+      "step": 62305
+    },
+    {
+      "epoch": 4.233591520587036,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3540052996331026e-05,
+      "loss": 0.8094,
+      "step": 62310
+    },
+    {
+      "epoch": 4.233931240657698,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.353792974588939e-05,
+      "loss": 0.8733,
+      "step": 62315
+    },
+    {
+      "epoch": 4.2342709607283595,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.353580649544775e-05,
+      "loss": 0.7764,
+      "step": 62320
+    },
+    {
+      "epoch": 4.234610680799022,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.3533683245006118e-05,
+      "loss": 0.8193,
+      "step": 62325
+    },
+    {
+      "epoch": 4.234950400869684,
+      "grad_norm": 9.25,
+      "learning_rate": 2.3531559994564482e-05,
+      "loss": 0.7986,
+      "step": 62330
+    },
+    {
+      "epoch": 4.235290120940345,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3529436744122843e-05,
+      "loss": 0.9509,
+      "step": 62335
+    },
+    {
+      "epoch": 4.235629841011007,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.3527313493681207e-05,
+      "loss": 0.8316,
+      "step": 62340
+    },
+    {
+      "epoch": 4.235969561081669,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3525190243239574e-05,
+      "loss": 0.8443,
+      "step": 62345
+    },
+    {
+      "epoch": 4.23630928115233,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3523066992797935e-05,
+      "loss": 0.9352,
+      "step": 62350
+    },
+    {
+      "epoch": 4.236649001222992,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.35209437423563e-05,
+      "loss": 0.9082,
+      "step": 62355
+    },
+    {
+      "epoch": 4.236988721293654,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3518820491914666e-05,
+      "loss": 0.8919,
+      "step": 62360
+    },
+    {
+      "epoch": 4.2373284413643155,
+      "grad_norm": 6.875,
+      "learning_rate": 2.3516697241473027e-05,
+      "loss": 0.8534,
+      "step": 62365
+    },
+    {
+      "epoch": 4.237668161434978,
+      "grad_norm": 7.375,
+      "learning_rate": 2.351457399103139e-05,
+      "loss": 0.8353,
+      "step": 62370
+    },
+    {
+      "epoch": 4.23800788150564,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3512450740589755e-05,
+      "loss": 0.8073,
+      "step": 62375
+    },
+    {
+      "epoch": 4.238347601576301,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.351032749014812e-05,
+      "loss": 0.8097,
+      "step": 62380
+    },
+    {
+      "epoch": 4.238687321646963,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3508204239706483e-05,
+      "loss": 0.9504,
+      "step": 62385
+    },
+    {
+      "epoch": 4.239027041717625,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3506080989264847e-05,
+      "loss": 0.8653,
+      "step": 62390
+    },
+    {
+      "epoch": 4.239366761788286,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.350395773882321e-05,
+      "loss": 0.8336,
+      "step": 62395
+    },
+    {
+      "epoch": 4.239706481858948,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3501834488381575e-05,
+      "loss": 0.9033,
+      "step": 62400
+    },
+    {
+      "epoch": 4.24004620192961,
+      "grad_norm": 8.25,
+      "learning_rate": 2.349971123793994e-05,
+      "loss": 0.8482,
+      "step": 62405
+    },
+    {
+      "epoch": 4.2403859220002715,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.3497587987498303e-05,
+      "loss": 0.9837,
+      "step": 62410
+    },
+    {
+      "epoch": 4.240725642070934,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3495464737056667e-05,
+      "loss": 0.949,
+      "step": 62415
+    },
+    {
+      "epoch": 4.241065362141596,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.349334148661503e-05,
+      "loss": 0.9155,
+      "step": 62420
+    },
+    {
+      "epoch": 4.241405082212257,
+      "grad_norm": 8.625,
+      "learning_rate": 2.3491218236173395e-05,
+      "loss": 0.9201,
+      "step": 62425
+    },
+    {
+      "epoch": 4.241744802282919,
+      "grad_norm": 8.625,
+      "learning_rate": 2.348909498573176e-05,
+      "loss": 0.9124,
+      "step": 62430
+    },
+    {
+      "epoch": 4.242084522353581,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3486971735290123e-05,
+      "loss": 0.7797,
+      "step": 62435
+    },
+    {
+      "epoch": 4.242424242424242,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3484848484848487e-05,
+      "loss": 0.9026,
+      "step": 62440
+    },
+    {
+      "epoch": 4.242763962494904,
+      "grad_norm": 6.75,
+      "learning_rate": 2.348272523440685e-05,
+      "loss": 0.8368,
+      "step": 62445
+    },
+    {
+      "epoch": 4.243103682565566,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3480601983965215e-05,
+      "loss": 0.8862,
+      "step": 62450
+    },
+    {
+      "epoch": 4.2434434026362275,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.347847873352358e-05,
+      "loss": 0.8583,
+      "step": 62455
+    },
+    {
+      "epoch": 4.24378312270689,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3476355483081943e-05,
+      "loss": 1.0189,
+      "step": 62460
+    },
+    {
+      "epoch": 4.244122842777552,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3474232232640307e-05,
+      "loss": 0.8438,
+      "step": 62465
+    },
+    {
+      "epoch": 4.244462562848213,
+      "grad_norm": 8.125,
+      "learning_rate": 2.347210898219867e-05,
+      "loss": 0.8143,
+      "step": 62470
+    },
+    {
+      "epoch": 4.244802282918875,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.346998573175703e-05,
+      "loss": 0.9162,
+      "step": 62475
+    },
+    {
+      "epoch": 4.245142002989537,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.34678624813154e-05,
+      "loss": 0.8868,
+      "step": 62480
+    },
+    {
+      "epoch": 4.245481723060198,
+      "grad_norm": 7.75,
+      "learning_rate": 2.3465739230873763e-05,
+      "loss": 0.8179,
+      "step": 62485
+    },
+    {
+      "epoch": 4.24582144313086,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3463615980432123e-05,
+      "loss": 0.7846,
+      "step": 62490
+    },
+    {
+      "epoch": 4.246161163201522,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.346149272999049e-05,
+      "loss": 0.8279,
+      "step": 62495
+    },
+    {
+      "epoch": 4.246500883272184,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.3459369479548855e-05,
+      "loss": 0.939,
+      "step": 62500
+    },
+    {
+      "epoch": 4.246840603342846,
+      "grad_norm": 7.25,
+      "learning_rate": 2.3457246229107215e-05,
+      "loss": 0.9857,
+      "step": 62505
+    },
+    {
+      "epoch": 4.247180323413508,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.345512297866558e-05,
+      "loss": 0.9039,
+      "step": 62510
+    },
+    {
+      "epoch": 4.247520043484169,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.3452999728223947e-05,
+      "loss": 0.974,
+      "step": 62515
+    },
+    {
+      "epoch": 4.247859763554831,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3450876477782307e-05,
+      "loss": 0.8814,
+      "step": 62520
+    },
+    {
+      "epoch": 4.248199483625493,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.344875322734067e-05,
+      "loss": 0.8411,
+      "step": 62525
+    },
+    {
+      "epoch": 4.248539203696154,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.344662997689904e-05,
+      "loss": 0.8727,
+      "step": 62530
+    },
+    {
+      "epoch": 4.248878923766816,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.34445067264574e-05,
+      "loss": 0.9195,
+      "step": 62535
+    },
+    {
+      "epoch": 4.2492186438374775,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.3442383476015763e-05,
+      "loss": 0.8403,
+      "step": 62540
+    },
+    {
+      "epoch": 4.24955836390814,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3440260225574127e-05,
+      "loss": 0.8307,
+      "step": 62545
+    },
+    {
+      "epoch": 4.249898083978802,
+      "grad_norm": 8.375,
+      "learning_rate": 2.343813697513249e-05,
+      "loss": 0.8465,
+      "step": 62550
+    },
+    {
+      "epoch": 4.250237804049463,
+      "grad_norm": 6.75,
+      "learning_rate": 2.3436013724690855e-05,
+      "loss": 0.9091,
+      "step": 62555
+    },
+    {
+      "epoch": 4.250577524120125,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.343389047424922e-05,
+      "loss": 0.883,
+      "step": 62560
+    },
+    {
+      "epoch": 4.250917244190787,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3431767223807583e-05,
+      "loss": 0.9072,
+      "step": 62565
+    },
+    {
+      "epoch": 4.251256964261448,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.3429643973365947e-05,
+      "loss": 0.8713,
+      "step": 62570
+    },
+    {
+      "epoch": 4.25159668433211,
+      "grad_norm": 8.5,
+      "learning_rate": 2.342752072292431e-05,
+      "loss": 0.7635,
+      "step": 62575
+    },
+    {
+      "epoch": 4.251936404402772,
+      "grad_norm": 7.75,
+      "learning_rate": 2.3425397472482675e-05,
+      "loss": 0.8431,
+      "step": 62580
+    },
+    {
+      "epoch": 4.2522761244734335,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.342327422204104e-05,
+      "loss": 0.8148,
+      "step": 62585
+    },
+    {
+      "epoch": 4.252615844544096,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3421150971599403e-05,
+      "loss": 0.8468,
+      "step": 62590
+    },
+    {
+      "epoch": 4.252955564614758,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3419027721157767e-05,
+      "loss": 0.9407,
+      "step": 62595
+    },
+    {
+      "epoch": 4.253295284685419,
+      "grad_norm": 7.125,
+      "learning_rate": 2.341690447071613e-05,
+      "loss": 0.8807,
+      "step": 62600
+    },
+    {
+      "epoch": 4.253635004756081,
+      "grad_norm": 5.5625,
+      "learning_rate": 2.3414781220274495e-05,
+      "loss": 0.827,
+      "step": 62605
+    },
+    {
+      "epoch": 4.253974724826743,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.341265796983286e-05,
+      "loss": 0.7816,
+      "step": 62610
+    },
+    {
+      "epoch": 4.254314444897404,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3410534719391223e-05,
+      "loss": 0.9045,
+      "step": 62615
+    },
+    {
+      "epoch": 4.254654164968066,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.3408411468949587e-05,
+      "loss": 0.8889,
+      "step": 62620
+    },
+    {
+      "epoch": 4.254993885038728,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.340628821850795e-05,
+      "loss": 0.912,
+      "step": 62625
+    },
+    {
+      "epoch": 4.2553336051093895,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.3404164968066315e-05,
+      "loss": 0.8631,
+      "step": 62630
+    },
+    {
+      "epoch": 4.255673325180052,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.340204171762468e-05,
+      "loss": 0.8517,
+      "step": 62635
+    },
+    {
+      "epoch": 4.256013045250714,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.3399918467183043e-05,
+      "loss": 0.7827,
+      "step": 62640
+    },
+    {
+      "epoch": 4.256352765321375,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3397795216741404e-05,
+      "loss": 0.8511,
+      "step": 62645
+    },
+    {
+      "epoch": 4.256692485392037,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.339567196629977e-05,
+      "loss": 0.8301,
+      "step": 62650
+    },
+    {
+      "epoch": 4.257032205462699,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3393548715858135e-05,
+      "loss": 0.885,
+      "step": 62655
+    },
+    {
+      "epoch": 4.25737192553336,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.3391425465416496e-05,
+      "loss": 0.9043,
+      "step": 62660
+    },
+    {
+      "epoch": 4.257711645604022,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3389302214974863e-05,
+      "loss": 0.9694,
+      "step": 62665
+    },
+    {
+      "epoch": 4.258051365674684,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.3387178964533227e-05,
+      "loss": 0.9263,
+      "step": 62670
+    },
+    {
+      "epoch": 4.2583910857453455,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3385055714091588e-05,
+      "loss": 0.9404,
+      "step": 62675
+    },
+    {
+      "epoch": 4.258730805816008,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3382932463649952e-05,
+      "loss": 0.8566,
+      "step": 62680
+    },
+    {
+      "epoch": 4.25907052588667,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.338080921320832e-05,
+      "loss": 0.8917,
+      "step": 62685
+    },
+    {
+      "epoch": 4.259410245957331,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.337868596276668e-05,
+      "loss": 0.7409,
+      "step": 62690
+    },
+    {
+      "epoch": 4.259749966027993,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3376562712325044e-05,
+      "loss": 0.8746,
+      "step": 62695
+    },
+    {
+      "epoch": 4.260089686098655,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.337443946188341e-05,
+      "loss": 0.8233,
+      "step": 62700
+    },
+    {
+      "epoch": 4.260429406169316,
+      "grad_norm": 8.5,
+      "learning_rate": 2.3372316211441772e-05,
+      "loss": 0.8007,
+      "step": 62705
+    },
+    {
+      "epoch": 4.260769126239978,
+      "grad_norm": 8.875,
+      "learning_rate": 2.3370192961000136e-05,
+      "loss": 0.8883,
+      "step": 62710
+    },
+    {
+      "epoch": 4.26110884631064,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.33680697105585e-05,
+      "loss": 0.8871,
+      "step": 62715
+    },
+    {
+      "epoch": 4.2614485663813015,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3365946460116864e-05,
+      "loss": 0.8547,
+      "step": 62720
+    },
+    {
+      "epoch": 4.261788286451964,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.3363823209675228e-05,
+      "loss": 0.8713,
+      "step": 62725
+    },
+    {
+      "epoch": 4.262128006522626,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3361699959233592e-05,
+      "loss": 0.9123,
+      "step": 62730
+    },
+    {
+      "epoch": 4.262467726593287,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.3359576708791956e-05,
+      "loss": 0.8841,
+      "step": 62735
+    },
+    {
+      "epoch": 4.262807446663949,
+      "grad_norm": 8.0,
+      "learning_rate": 2.335745345835032e-05,
+      "loss": 0.862,
+      "step": 62740
+    },
+    {
+      "epoch": 4.263147166734611,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.3355330207908684e-05,
+      "loss": 1.0163,
+      "step": 62745
+    },
+    {
+      "epoch": 4.263486886805272,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3353206957467048e-05,
+      "loss": 0.823,
+      "step": 62750
+    },
+    {
+      "epoch": 4.263826606875934,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.3351083707025412e-05,
+      "loss": 0.8261,
+      "step": 62755
+    },
+    {
+      "epoch": 4.264166326946596,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3348960456583776e-05,
+      "loss": 0.9207,
+      "step": 62760
+    },
+    {
+      "epoch": 4.2645060470172576,
+      "grad_norm": 8.0,
+      "learning_rate": 2.334683720614214e-05,
+      "loss": 0.8664,
+      "step": 62765
+    },
+    {
+      "epoch": 4.26484576708792,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3344713955700504e-05,
+      "loss": 0.8167,
+      "step": 62770
+    },
+    {
+      "epoch": 4.265185487158582,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3342590705258868e-05,
+      "loss": 0.8615,
+      "step": 62775
+    },
+    {
+      "epoch": 4.265525207229243,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3340467454817232e-05,
+      "loss": 0.8185,
+      "step": 62780
+    },
+    {
+      "epoch": 4.265864927299905,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3338344204375596e-05,
+      "loss": 0.8921,
+      "step": 62785
+    },
+    {
+      "epoch": 4.266204647370567,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.333622095393396e-05,
+      "loss": 0.8938,
+      "step": 62790
+    },
+    {
+      "epoch": 4.266544367441228,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3334097703492324e-05,
+      "loss": 0.8303,
+      "step": 62795
+    },
+    {
+      "epoch": 4.26688408751189,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3331974453050688e-05,
+      "loss": 0.8308,
+      "step": 62800
+    },
+    {
+      "epoch": 4.267223807582552,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.3329851202609052e-05,
+      "loss": 0.8042,
+      "step": 62805
+    },
+    {
+      "epoch": 4.267563527653214,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3327727952167416e-05,
+      "loss": 0.9052,
+      "step": 62810
+    },
+    {
+      "epoch": 4.267903247723876,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3325604701725777e-05,
+      "loss": 0.8644,
+      "step": 62815
+    },
+    {
+      "epoch": 4.268242967794538,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3323481451284144e-05,
+      "loss": 0.7812,
+      "step": 62820
+    },
+    {
+      "epoch": 4.268582687865199,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3321358200842508e-05,
+      "loss": 0.897,
+      "step": 62825
+    },
+    {
+      "epoch": 4.268922407935861,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.331923495040087e-05,
+      "loss": 0.8654,
+      "step": 62830
+    },
+    {
+      "epoch": 4.269262128006522,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.3317111699959236e-05,
+      "loss": 0.8958,
+      "step": 62835
+    },
+    {
+      "epoch": 4.269601848077184,
+      "grad_norm": 8.125,
+      "learning_rate": 2.33149884495176e-05,
+      "loss": 0.8666,
+      "step": 62840
+    },
+    {
+      "epoch": 4.269941568147846,
+      "grad_norm": 7.625,
+      "learning_rate": 2.331286519907596e-05,
+      "loss": 0.8781,
+      "step": 62845
+    },
+    {
+      "epoch": 4.2702812882185075,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3310741948634328e-05,
+      "loss": 0.9515,
+      "step": 62850
+    },
+    {
+      "epoch": 4.27062100828917,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.3308618698192692e-05,
+      "loss": 0.9082,
+      "step": 62855
+    },
+    {
+      "epoch": 4.270960728359832,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3306495447751053e-05,
+      "loss": 0.8392,
+      "step": 62860
+    },
+    {
+      "epoch": 4.271300448430493,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3304372197309417e-05,
+      "loss": 0.9305,
+      "step": 62865
+    },
+    {
+      "epoch": 4.271640168501155,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.3302248946867784e-05,
+      "loss": 0.9053,
+      "step": 62870
+    },
+    {
+      "epoch": 4.271979888571817,
+      "grad_norm": 7.25,
+      "learning_rate": 2.3300125696426145e-05,
+      "loss": 1.016,
+      "step": 62875
+    },
+    {
+      "epoch": 4.272319608642478,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.329800244598451e-05,
+      "loss": 0.8917,
+      "step": 62880
+    },
+    {
+      "epoch": 4.27265932871314,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.3295879195542876e-05,
+      "loss": 0.8449,
+      "step": 62885
+    },
+    {
+      "epoch": 4.272999048783802,
+      "grad_norm": 8.375,
+      "learning_rate": 2.3293755945101237e-05,
+      "loss": 0.9472,
+      "step": 62890
+    },
+    {
+      "epoch": 4.2733387688544635,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.32916326946596e-05,
+      "loss": 0.855,
+      "step": 62895
+    },
+    {
+      "epoch": 4.273678488925126,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.3289509444217965e-05,
+      "loss": 0.7365,
+      "step": 62900
+    },
+    {
+      "epoch": 4.274018208995788,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.328738619377633e-05,
+      "loss": 0.8736,
+      "step": 62905
+    },
+    {
+      "epoch": 4.274357929066449,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.3285262943334693e-05,
+      "loss": 0.8321,
+      "step": 62910
+    },
+    {
+      "epoch": 4.274697649137111,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3283139692893057e-05,
+      "loss": 0.8899,
+      "step": 62915
+    },
+    {
+      "epoch": 4.275037369207773,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.328101644245142e-05,
+      "loss": 0.8806,
+      "step": 62920
+    },
+    {
+      "epoch": 4.275377089278434,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3278893192009785e-05,
+      "loss": 0.8637,
+      "step": 62925
+    },
+    {
+      "epoch": 4.275716809349096,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.327676994156815e-05,
+      "loss": 0.822,
+      "step": 62930
+    },
+    {
+      "epoch": 4.276056529419758,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3274646691126513e-05,
+      "loss": 0.8742,
+      "step": 62935
+    },
+    {
+      "epoch": 4.2763962494904195,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.3272523440684877e-05,
+      "loss": 0.9596,
+      "step": 62940
+    },
+    {
+      "epoch": 4.276735969561082,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.327040019024324e-05,
+      "loss": 0.8791,
+      "step": 62945
+    },
+    {
+      "epoch": 4.277075689631744,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3268276939801605e-05,
+      "loss": 0.8832,
+      "step": 62950
+    },
+    {
+      "epoch": 4.277415409702405,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.326615368935997e-05,
+      "loss": 0.8756,
+      "step": 62955
+    },
+    {
+      "epoch": 4.277755129773067,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.3264030438918333e-05,
+      "loss": 0.9408,
+      "step": 62960
+    },
+    {
+      "epoch": 4.278094849843729,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.3261907188476697e-05,
+      "loss": 0.9457,
+      "step": 62965
+    },
+    {
+      "epoch": 4.27843456991439,
+      "grad_norm": 7.25,
+      "learning_rate": 2.325978393803506e-05,
+      "loss": 0.7705,
+      "step": 62970
+    },
+    {
+      "epoch": 4.278774289985052,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.3257660687593425e-05,
+      "loss": 0.9126,
+      "step": 62975
+    },
+    {
+      "epoch": 4.279114010055714,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.325553743715179e-05,
+      "loss": 0.9176,
+      "step": 62980
+    },
+    {
+      "epoch": 4.2794537301263755,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3253414186710153e-05,
+      "loss": 0.8211,
+      "step": 62985
+    },
+    {
+      "epoch": 4.279793450197038,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.3251290936268517e-05,
+      "loss": 0.9203,
+      "step": 62990
+    },
+    {
+      "epoch": 4.2801331702677,
+      "grad_norm": 8.625,
+      "learning_rate": 2.324916768582688e-05,
+      "loss": 0.8477,
+      "step": 62995
+    },
+    {
+      "epoch": 4.280472890338361,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.324704443538524e-05,
+      "loss": 0.8263,
+      "step": 63000
+    },
+    {
+      "epoch": 4.280812610409023,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.324492118494361e-05,
+      "loss": 0.8567,
+      "step": 63005
+    },
+    {
+      "epoch": 4.281152330479685,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3242797934501973e-05,
+      "loss": 0.9158,
+      "step": 63010
+    },
+    {
+      "epoch": 4.281492050550346,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3240674684060333e-05,
+      "loss": 0.839,
+      "step": 63015
+    },
+    {
+      "epoch": 4.281831770621008,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.32385514336187e-05,
+      "loss": 0.9367,
+      "step": 63020
+    },
+    {
+      "epoch": 4.28217149069167,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.3236428183177065e-05,
+      "loss": 0.826,
+      "step": 63025
+    },
+    {
+      "epoch": 4.2825112107623315,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3234304932735425e-05,
+      "loss": 0.9249,
+      "step": 63030
+    },
+    {
+      "epoch": 4.282850930832994,
+      "grad_norm": 6.875,
+      "learning_rate": 2.323218168229379e-05,
+      "loss": 0.9292,
+      "step": 63035
+    },
+    {
+      "epoch": 4.283190650903656,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3230058431852157e-05,
+      "loss": 0.92,
+      "step": 63040
+    },
+    {
+      "epoch": 4.283530370974317,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3227935181410518e-05,
+      "loss": 0.8849,
+      "step": 63045
+    },
+    {
+      "epoch": 4.283870091044979,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.322581193096888e-05,
+      "loss": 0.8187,
+      "step": 63050
+    },
+    {
+      "epoch": 4.284209811115641,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.322368868052725e-05,
+      "loss": 0.9423,
+      "step": 63055
+    },
+    {
+      "epoch": 4.284549531186302,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.322156543008561e-05,
+      "loss": 0.8507,
+      "step": 63060
+    },
+    {
+      "epoch": 4.284889251256964,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3219442179643974e-05,
+      "loss": 0.8772,
+      "step": 63065
+    },
+    {
+      "epoch": 4.285228971327626,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3217318929202338e-05,
+      "loss": 0.8875,
+      "step": 63070
+    },
+    {
+      "epoch": 4.2855686913982876,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.32151956787607e-05,
+      "loss": 0.8994,
+      "step": 63075
+    },
+    {
+      "epoch": 4.28590841146895,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3213072428319066e-05,
+      "loss": 0.829,
+      "step": 63080
+    },
+    {
+      "epoch": 4.286248131539612,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.321094917787743e-05,
+      "loss": 0.7444,
+      "step": 63085
+    },
+    {
+      "epoch": 4.286587851610273,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3208825927435794e-05,
+      "loss": 0.8572,
+      "step": 63090
+    },
+    {
+      "epoch": 4.286927571680935,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3206702676994158e-05,
+      "loss": 0.8075,
+      "step": 63095
+    },
+    {
+      "epoch": 4.287267291751597,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.320457942655252e-05,
+      "loss": 0.8857,
+      "step": 63100
+    },
+    {
+      "epoch": 4.287607011822258,
+      "grad_norm": 7.75,
+      "learning_rate": 2.3202456176110886e-05,
+      "loss": 0.8935,
+      "step": 63105
+    },
+    {
+      "epoch": 4.28794673189292,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.320033292566925e-05,
+      "loss": 0.8991,
+      "step": 63110
+    },
+    {
+      "epoch": 4.288286451963582,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.3198209675227614e-05,
+      "loss": 0.9511,
+      "step": 63115
+    },
+    {
+      "epoch": 4.288626172034244,
+      "grad_norm": 7.125,
+      "learning_rate": 2.3196086424785978e-05,
+      "loss": 0.8528,
+      "step": 63120
+    },
+    {
+      "epoch": 4.288965892104906,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.319396317434434e-05,
+      "loss": 0.961,
+      "step": 63125
+    },
+    {
+      "epoch": 4.289305612175568,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3191839923902706e-05,
+      "loss": 0.8129,
+      "step": 63130
+    },
+    {
+      "epoch": 4.289645332246229,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.318971667346107e-05,
+      "loss": 0.8963,
+      "step": 63135
+    },
+    {
+      "epoch": 4.289985052316891,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3187593423019434e-05,
+      "loss": 0.8736,
+      "step": 63140
+    },
+    {
+      "epoch": 4.290324772387553,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.3185470172577798e-05,
+      "loss": 0.8824,
+      "step": 63145
+    },
+    {
+      "epoch": 4.290664492458214,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.318334692213616e-05,
+      "loss": 0.8742,
+      "step": 63150
+    },
+    {
+      "epoch": 4.291004212528876,
+      "grad_norm": 8.375,
+      "learning_rate": 2.3181223671694526e-05,
+      "loss": 0.8459,
+      "step": 63155
+    },
+    {
+      "epoch": 4.291343932599538,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.317910042125289e-05,
+      "loss": 0.9229,
+      "step": 63160
+    },
+    {
+      "epoch": 4.2916836526702,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3176977170811254e-05,
+      "loss": 0.7611,
+      "step": 63165
+    },
+    {
+      "epoch": 4.292023372740862,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3174853920369614e-05,
+      "loss": 0.865,
+      "step": 63170
+    },
+    {
+      "epoch": 4.292363092811524,
+      "grad_norm": 7.375,
+      "learning_rate": 2.317273066992798e-05,
+      "loss": 0.8369,
+      "step": 63175
+    },
+    {
+      "epoch": 4.292702812882185,
+      "grad_norm": 9.5,
+      "learning_rate": 2.3170607419486346e-05,
+      "loss": 0.8949,
+      "step": 63180
+    },
+    {
+      "epoch": 4.293042532952847,
+      "grad_norm": 7.125,
+      "learning_rate": 2.3168484169044706e-05,
+      "loss": 0.8541,
+      "step": 63185
+    },
+    {
+      "epoch": 4.293382253023509,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3166360918603074e-05,
+      "loss": 0.8184,
+      "step": 63190
+    },
+    {
+      "epoch": 4.29372197309417,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3164237668161438e-05,
+      "loss": 0.9221,
+      "step": 63195
+    },
+    {
+      "epoch": 4.294061693164832,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3162114417719798e-05,
+      "loss": 0.8416,
+      "step": 63200
+    },
+    {
+      "epoch": 4.294401413235494,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.3159991167278162e-05,
+      "loss": 0.8986,
+      "step": 63205
+    },
+    {
+      "epoch": 4.294741133306156,
+      "grad_norm": 8.125,
+      "learning_rate": 2.315786791683653e-05,
+      "loss": 0.8537,
+      "step": 63210
+    },
+    {
+      "epoch": 4.295080853376818,
+      "grad_norm": 7.875,
+      "learning_rate": 2.315574466639489e-05,
+      "loss": 0.7841,
+      "step": 63215
+    },
+    {
+      "epoch": 4.29542057344748,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3153621415953254e-05,
+      "loss": 0.904,
+      "step": 63220
+    },
+    {
+      "epoch": 4.295760293518141,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.315149816551162e-05,
+      "loss": 0.8011,
+      "step": 63225
+    },
+    {
+      "epoch": 4.296100013588803,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.3149374915069982e-05,
+      "loss": 0.9086,
+      "step": 63230
+    },
+    {
+      "epoch": 4.296439733659464,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3147251664628346e-05,
+      "loss": 0.8542,
+      "step": 63235
+    },
+    {
+      "epoch": 4.296779453730126,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3145128414186714e-05,
+      "loss": 0.8418,
+      "step": 63240
+    },
+    {
+      "epoch": 4.297119173800788,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3143005163745074e-05,
+      "loss": 0.8885,
+      "step": 63245
+    },
+    {
+      "epoch": 4.2974588938714495,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.3140881913303438e-05,
+      "loss": 0.9345,
+      "step": 63250
+    },
+    {
+      "epoch": 4.297798613942112,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3138758662861802e-05,
+      "loss": 0.9415,
+      "step": 63255
+    },
+    {
+      "epoch": 4.298138334012774,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.3136635412420166e-05,
+      "loss": 0.889,
+      "step": 63260
+    },
+    {
+      "epoch": 4.298478054083435,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.313451216197853e-05,
+      "loss": 0.9041,
+      "step": 63265
+    },
+    {
+      "epoch": 4.298817774154097,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.3132388911536894e-05,
+      "loss": 0.907,
+      "step": 63270
+    },
+    {
+      "epoch": 4.299157494224759,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.3130265661095258e-05,
+      "loss": 0.8402,
+      "step": 63275
+    },
+    {
+      "epoch": 4.29949721429542,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3128142410653622e-05,
+      "loss": 0.8,
+      "step": 63280
+    },
+    {
+      "epoch": 4.299836934366082,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.3126019160211986e-05,
+      "loss": 0.9003,
+      "step": 63285
+    },
+    {
+      "epoch": 4.300176654436744,
+      "grad_norm": 9.0,
+      "learning_rate": 2.312389590977035e-05,
+      "loss": 0.8918,
+      "step": 63290
+    },
+    {
+      "epoch": 4.3005163745074055,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3121772659328714e-05,
+      "loss": 0.9415,
+      "step": 63295
+    },
+    {
+      "epoch": 4.300856094578068,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.311964940888708e-05,
+      "loss": 0.7676,
+      "step": 63300
+    },
+    {
+      "epoch": 4.30119581464873,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.3117526158445442e-05,
+      "loss": 0.8876,
+      "step": 63305
+    },
+    {
+      "epoch": 4.301535534719391,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.3115402908003806e-05,
+      "loss": 0.8792,
+      "step": 63310
+    },
+    {
+      "epoch": 4.301875254790053,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.311327965756217e-05,
+      "loss": 0.8094,
+      "step": 63315
+    },
+    {
+      "epoch": 4.302214974860715,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3111156407120534e-05,
+      "loss": 1.0076,
+      "step": 63320
+    },
+    {
+      "epoch": 4.302554694931376,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.31090331566789e-05,
+      "loss": 0.8958,
+      "step": 63325
+    },
+    {
+      "epoch": 4.302894415002038,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.3106909906237262e-05,
+      "loss": 0.8659,
+      "step": 63330
+    },
+    {
+      "epoch": 4.3032341350727,
+      "grad_norm": 8.0,
+      "learning_rate": 2.3104786655795626e-05,
+      "loss": 0.8131,
+      "step": 63335
+    },
+    {
+      "epoch": 4.3035738551433615,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.310266340535399e-05,
+      "loss": 0.8208,
+      "step": 63340
+    },
+    {
+      "epoch": 4.303913575214024,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3100540154912354e-05,
+      "loss": 0.8704,
+      "step": 63345
+    },
+    {
+      "epoch": 4.304253295284686,
+      "grad_norm": 7.75,
+      "learning_rate": 2.309841690447072e-05,
+      "loss": 0.8768,
+      "step": 63350
+    },
+    {
+      "epoch": 4.304593015355347,
+      "grad_norm": 8.375,
+      "learning_rate": 2.309629365402908e-05,
+      "loss": 0.8739,
+      "step": 63355
+    },
+    {
+      "epoch": 4.304932735426009,
+      "grad_norm": 7.375,
+      "learning_rate": 2.3094170403587446e-05,
+      "loss": 0.8744,
+      "step": 63360
+    },
+    {
+      "epoch": 4.305272455496671,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.309204715314581e-05,
+      "loss": 0.9276,
+      "step": 63365
+    },
+    {
+      "epoch": 4.305612175567332,
+      "grad_norm": 8.375,
+      "learning_rate": 2.308992390270417e-05,
+      "loss": 0.7833,
+      "step": 63370
+    },
+    {
+      "epoch": 4.305951895637994,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.308780065226254e-05,
+      "loss": 0.9177,
+      "step": 63375
+    },
+    {
+      "epoch": 4.306291615708656,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.3085677401820902e-05,
+      "loss": 0.8279,
+      "step": 63380
+    },
+    {
+      "epoch": 4.306631335779318,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.3083554151379263e-05,
+      "loss": 0.8359,
+      "step": 63385
+    },
+    {
+      "epoch": 4.30697105584998,
+      "grad_norm": 8.875,
+      "learning_rate": 2.3081430900937627e-05,
+      "loss": 0.8352,
+      "step": 63390
+    },
+    {
+      "epoch": 4.307310775920642,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.3079307650495994e-05,
+      "loss": 0.9263,
+      "step": 63395
+    },
+    {
+      "epoch": 4.307650495991303,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.3077184400054355e-05,
+      "loss": 0.9737,
+      "step": 63400
+    },
+    {
+      "epoch": 4.307990216061965,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.307506114961272e-05,
+      "loss": 0.865,
+      "step": 63405
+    },
+    {
+      "epoch": 4.308329936132627,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.3072937899171086e-05,
+      "loss": 0.9319,
+      "step": 63410
+    },
+    {
+      "epoch": 4.308669656203288,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.3070814648729447e-05,
+      "loss": 0.9796,
+      "step": 63415
+    },
+    {
+      "epoch": 4.30900937627395,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.306869139828781e-05,
+      "loss": 0.9059,
+      "step": 63420
+    },
+    {
+      "epoch": 4.309349096344612,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.3066568147846175e-05,
+      "loss": 0.9302,
+      "step": 63425
+    },
+    {
+      "epoch": 4.309688816415274,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.306444489740454e-05,
+      "loss": 0.9241,
+      "step": 63430
+    },
+    {
+      "epoch": 4.310028536485936,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.3062321646962903e-05,
+      "loss": 0.8789,
+      "step": 63435
+    },
+    {
+      "epoch": 4.310368256556598,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.3060198396521267e-05,
+      "loss": 0.8639,
+      "step": 63440
+    },
+    {
+      "epoch": 4.310707976627259,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.305807514607963e-05,
+      "loss": 0.8054,
+      "step": 63445
+    },
+    {
+      "epoch": 4.311047696697921,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3055951895637995e-05,
+      "loss": 0.9565,
+      "step": 63450
+    },
+    {
+      "epoch": 4.311387416768583,
+      "grad_norm": 8.5,
+      "learning_rate": 2.305382864519636e-05,
+      "loss": 0.9406,
+      "step": 63455
+    },
+    {
+      "epoch": 4.311727136839244,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3051705394754723e-05,
+      "loss": 0.8738,
+      "step": 63460
+    },
+    {
+      "epoch": 4.312066856909906,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.3049582144313087e-05,
+      "loss": 0.8051,
+      "step": 63465
+    },
+    {
+      "epoch": 4.312406576980568,
+      "grad_norm": 7.5,
+      "learning_rate": 2.304745889387145e-05,
+      "loss": 0.7896,
+      "step": 63470
+    },
+    {
+      "epoch": 4.31274629705123,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3045335643429815e-05,
+      "loss": 0.9223,
+      "step": 63475
+    },
+    {
+      "epoch": 4.313086017121892,
+      "grad_norm": 8.25,
+      "learning_rate": 2.304321239298818e-05,
+      "loss": 0.8619,
+      "step": 63480
+    },
+    {
+      "epoch": 4.313425737192554,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.3041089142546543e-05,
+      "loss": 0.821,
+      "step": 63485
+    },
+    {
+      "epoch": 4.313765457263215,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.3038965892104907e-05,
+      "loss": 0.8896,
+      "step": 63490
+    },
+    {
+      "epoch": 4.314105177333877,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.303684264166327e-05,
+      "loss": 0.8735,
+      "step": 63495
+    },
+    {
+      "epoch": 4.314444897404539,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3034719391221635e-05,
+      "loss": 0.9926,
+      "step": 63500
+    },
+    {
+      "epoch": 4.3147846174752,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.303259614078e-05,
+      "loss": 0.8736,
+      "step": 63505
+    },
+    {
+      "epoch": 4.315124337545862,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.3030472890338363e-05,
+      "loss": 0.8523,
+      "step": 63510
+    },
+    {
+      "epoch": 4.3154640576165235,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3028349639896727e-05,
+      "loss": 0.9372,
+      "step": 63515
+    },
+    {
+      "epoch": 4.315803777687186,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.302622638945509e-05,
+      "loss": 0.8913,
+      "step": 63520
+    },
+    {
+      "epoch": 4.316143497757848,
+      "grad_norm": 8.5,
+      "learning_rate": 2.3024103139013452e-05,
+      "loss": 0.9178,
+      "step": 63525
+    },
+    {
+      "epoch": 4.316483217828509,
+      "grad_norm": 5.65625,
+      "learning_rate": 2.302197988857182e-05,
+      "loss": 0.8483,
+      "step": 63530
+    },
+    {
+      "epoch": 4.316822937899171,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.3019856638130183e-05,
+      "loss": 0.8966,
+      "step": 63535
+    },
+    {
+      "epoch": 4.317162657969833,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.3017733387688544e-05,
+      "loss": 0.8338,
+      "step": 63540
+    },
+    {
+      "epoch": 4.317502378040494,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.301561013724691e-05,
+      "loss": 0.8406,
+      "step": 63545
+    },
+    {
+      "epoch": 4.317842098111156,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.3013486886805275e-05,
+      "loss": 0.6896,
+      "step": 63550
+    },
+    {
+      "epoch": 4.318181818181818,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.3011363636363636e-05,
+      "loss": 0.8788,
+      "step": 63555
+    },
+    {
+      "epoch": 4.3185215382524795,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.3009240385922e-05,
+      "loss": 0.9156,
+      "step": 63560
+    },
+    {
+      "epoch": 4.318861258323142,
+      "grad_norm": 8.375,
+      "learning_rate": 2.3007117135480367e-05,
+      "loss": 0.8349,
+      "step": 63565
+    },
+    {
+      "epoch": 4.319200978393804,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.3004993885038728e-05,
+      "loss": 0.8599,
+      "step": 63570
+    },
+    {
+      "epoch": 4.319540698464465,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3002870634597092e-05,
+      "loss": 0.8775,
+      "step": 63575
+    },
+    {
+      "epoch": 4.319880418535127,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.300074738415546e-05,
+      "loss": 0.9195,
+      "step": 63580
+    },
+    {
+      "epoch": 4.320220138605789,
+      "grad_norm": 8.125,
+      "learning_rate": 2.299862413371382e-05,
+      "loss": 0.941,
+      "step": 63585
+    },
+    {
+      "epoch": 4.32055985867645,
+      "grad_norm": 8.375,
+      "learning_rate": 2.2996500883272184e-05,
+      "loss": 0.8711,
+      "step": 63590
+    },
+    {
+      "epoch": 4.320899578747112,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.2994377632830548e-05,
+      "loss": 0.853,
+      "step": 63595
+    },
+    {
+      "epoch": 4.321239298817774,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2992254382388912e-05,
+      "loss": 0.8956,
+      "step": 63600
+    },
+    {
+      "epoch": 4.3215790188884355,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.2990131131947276e-05,
+      "loss": 0.859,
+      "step": 63605
+    },
+    {
+      "epoch": 4.321918738959098,
+      "grad_norm": 7.5,
+      "learning_rate": 2.298800788150564e-05,
+      "loss": 0.8586,
+      "step": 63610
+    },
+    {
+      "epoch": 4.32225845902976,
+      "grad_norm": 7.5,
+      "learning_rate": 2.2985884631064004e-05,
+      "loss": 0.8978,
+      "step": 63615
+    },
+    {
+      "epoch": 4.322598179100421,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2983761380622368e-05,
+      "loss": 0.8363,
+      "step": 63620
+    },
+    {
+      "epoch": 4.322937899171083,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.2981638130180732e-05,
+      "loss": 0.8336,
+      "step": 63625
+    },
+    {
+      "epoch": 4.323277619241745,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2979514879739096e-05,
+      "loss": 0.8954,
+      "step": 63630
+    },
+    {
+      "epoch": 4.323617339312406,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.297739162929746e-05,
+      "loss": 0.8275,
+      "step": 63635
+    },
+    {
+      "epoch": 4.323957059383068,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.2975268378855824e-05,
+      "loss": 0.9235,
+      "step": 63640
+    },
+    {
+      "epoch": 4.32429677945373,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2973145128414188e-05,
+      "loss": 0.8111,
+      "step": 63645
+    },
+    {
+      "epoch": 4.3246364995243916,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.2971021877972552e-05,
+      "loss": 0.8561,
+      "step": 63650
+    },
+    {
+      "epoch": 4.324976219595054,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2968898627530916e-05,
+      "loss": 0.8959,
+      "step": 63655
+    },
+    {
+      "epoch": 4.325315939665716,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.296677537708928e-05,
+      "loss": 0.8019,
+      "step": 63660
+    },
+    {
+      "epoch": 4.325655659736377,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2964652126647644e-05,
+      "loss": 0.8856,
+      "step": 63665
+    },
+    {
+      "epoch": 4.325995379807039,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2962528876206008e-05,
+      "loss": 0.8424,
+      "step": 63670
+    },
+    {
+      "epoch": 4.326335099877701,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2960405625764372e-05,
+      "loss": 0.9175,
+      "step": 63675
+    },
+    {
+      "epoch": 4.326674819948362,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2958282375322736e-05,
+      "loss": 0.9113,
+      "step": 63680
+    },
+    {
+      "epoch": 4.327014540019024,
+      "grad_norm": 7.5,
+      "learning_rate": 2.29561591248811e-05,
+      "loss": 0.8343,
+      "step": 63685
+    },
+    {
+      "epoch": 4.327354260089686,
+      "grad_norm": 8.75,
+      "learning_rate": 2.2954035874439464e-05,
+      "loss": 0.9065,
+      "step": 63690
+    },
+    {
+      "epoch": 4.327693980160348,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.2951912623997824e-05,
+      "loss": 0.8349,
+      "step": 63695
+    },
+    {
+      "epoch": 4.32803370023101,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2949789373556192e-05,
+      "loss": 0.8584,
+      "step": 63700
+    },
+    {
+      "epoch": 4.328373420301672,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2947666123114556e-05,
+      "loss": 0.7862,
+      "step": 63705
+    },
+    {
+      "epoch": 4.328713140372333,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.2945542872672916e-05,
+      "loss": 0.8999,
+      "step": 63710
+    },
+    {
+      "epoch": 4.329052860442995,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2943419622231284e-05,
+      "loss": 0.8104,
+      "step": 63715
+    },
+    {
+      "epoch": 4.329392580513657,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2941296371789648e-05,
+      "loss": 0.9834,
+      "step": 63720
+    },
+    {
+      "epoch": 4.329732300584318,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.293917312134801e-05,
+      "loss": 0.9176,
+      "step": 63725
+    },
+    {
+      "epoch": 4.33007202065498,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2937049870906372e-05,
+      "loss": 0.8441,
+      "step": 63730
+    },
+    {
+      "epoch": 4.330411740725642,
+      "grad_norm": 8.125,
+      "learning_rate": 2.293492662046474e-05,
+      "loss": 0.8717,
+      "step": 63735
+    },
+    {
+      "epoch": 4.330751460796304,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.29328033700231e-05,
+      "loss": 0.8432,
+      "step": 63740
+    },
+    {
+      "epoch": 4.331091180866966,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.2930680119581464e-05,
+      "loss": 0.9117,
+      "step": 63745
+    },
+    {
+      "epoch": 4.331430900937628,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2928556869139832e-05,
+      "loss": 0.9032,
+      "step": 63750
+    },
+    {
+      "epoch": 4.331770621008289,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2926433618698193e-05,
+      "loss": 0.911,
+      "step": 63755
+    },
+    {
+      "epoch": 4.332110341078951,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.2924310368256557e-05,
+      "loss": 0.8703,
+      "step": 63760
+    },
+    {
+      "epoch": 4.332450061149613,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2922187117814924e-05,
+      "loss": 0.8332,
+      "step": 63765
+    },
+    {
+      "epoch": 4.332789781220274,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2920063867373285e-05,
+      "loss": 0.8458,
+      "step": 63770
+    },
+    {
+      "epoch": 4.333129501290936,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.291794061693165e-05,
+      "loss": 0.8273,
+      "step": 63775
+    },
+    {
+      "epoch": 4.333469221361598,
+      "grad_norm": 7.5,
+      "learning_rate": 2.2915817366490013e-05,
+      "loss": 0.818,
+      "step": 63780
+    },
+    {
+      "epoch": 4.33380894143226,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2913694116048377e-05,
+      "loss": 0.9022,
+      "step": 63785
+    },
+    {
+      "epoch": 4.334148661502922,
+      "grad_norm": 7.75,
+      "learning_rate": 2.291157086560674e-05,
+      "loss": 0.9456,
+      "step": 63790
+    },
+    {
+      "epoch": 4.334488381573584,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.2909447615165105e-05,
+      "loss": 0.7755,
+      "step": 63795
+    },
+    {
+      "epoch": 4.334828101644245,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2907324364723472e-05,
+      "loss": 0.8096,
+      "step": 63800
+    },
+    {
+      "epoch": 4.335167821714907,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2905201114281833e-05,
+      "loss": 0.9826,
+      "step": 63805
+    },
+    {
+      "epoch": 4.335507541785569,
+      "grad_norm": 7.5,
+      "learning_rate": 2.2903077863840197e-05,
+      "loss": 0.9044,
+      "step": 63810
+    },
+    {
+      "epoch": 4.33584726185623,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.290095461339856e-05,
+      "loss": 0.8804,
+      "step": 63815
+    },
+    {
+      "epoch": 4.336186981926892,
+      "grad_norm": 9.6875,
+      "learning_rate": 2.2898831362956925e-05,
+      "loss": 0.9729,
+      "step": 63820
+    },
+    {
+      "epoch": 4.336526701997554,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.289670811251529e-05,
+      "loss": 0.9848,
+      "step": 63825
+    },
+    {
+      "epoch": 4.336866422068216,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.2894584862073653e-05,
+      "loss": 0.8124,
+      "step": 63830
+    },
+    {
+      "epoch": 4.337206142138878,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2892461611632017e-05,
+      "loss": 0.8948,
+      "step": 63835
+    },
+    {
+      "epoch": 4.33754586220954,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.289033836119038e-05,
+      "loss": 0.8487,
+      "step": 63840
+    },
+    {
+      "epoch": 4.337885582280201,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.2888215110748745e-05,
+      "loss": 0.8964,
+      "step": 63845
+    },
+    {
+      "epoch": 4.338225302350863,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.288609186030711e-05,
+      "loss": 0.8085,
+      "step": 63850
+    },
+    {
+      "epoch": 4.338565022421525,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2883968609865473e-05,
+      "loss": 0.9449,
+      "step": 63855
+    },
+    {
+      "epoch": 4.338904742492186,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2881845359423837e-05,
+      "loss": 0.8657,
+      "step": 63860
+    },
+    {
+      "epoch": 4.339244462562848,
+      "grad_norm": 7.625,
+      "learning_rate": 2.28797221089822e-05,
+      "loss": 0.8525,
+      "step": 63865
+    },
+    {
+      "epoch": 4.33958418263351,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2877598858540565e-05,
+      "loss": 0.9076,
+      "step": 63870
+    },
+    {
+      "epoch": 4.339923902704172,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.287547560809893e-05,
+      "loss": 0.9039,
+      "step": 63875
+    },
+    {
+      "epoch": 4.340263622774834,
+      "grad_norm": 6.875,
+      "learning_rate": 2.287335235765729e-05,
+      "loss": 0.9816,
+      "step": 63880
+    },
+    {
+      "epoch": 4.340603342845496,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2871229107215657e-05,
+      "loss": 0.8706,
+      "step": 63885
+    },
+    {
+      "epoch": 4.340943062916157,
+      "grad_norm": 8.25,
+      "learning_rate": 2.286910585677402e-05,
+      "loss": 0.8954,
+      "step": 63890
+    },
+    {
+      "epoch": 4.341282782986819,
+      "grad_norm": 8.125,
+      "learning_rate": 2.286698260633238e-05,
+      "loss": 0.92,
+      "step": 63895
+    },
+    {
+      "epoch": 4.341622503057481,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.286485935589075e-05,
+      "loss": 0.8834,
+      "step": 63900
+    },
+    {
+      "epoch": 4.341962223128142,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.2862736105449113e-05,
+      "loss": 0.8423,
+      "step": 63905
+    },
+    {
+      "epoch": 4.342301943198804,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2860612855007473e-05,
+      "loss": 0.9384,
+      "step": 63910
+    },
+    {
+      "epoch": 4.3426416632694655,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2858489604565837e-05,
+      "loss": 0.9476,
+      "step": 63915
+    },
+    {
+      "epoch": 4.342981383340128,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2856366354124205e-05,
+      "loss": 0.9381,
+      "step": 63920
+    },
+    {
+      "epoch": 4.34332110341079,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.2854243103682565e-05,
+      "loss": 0.9463,
+      "step": 63925
+    },
+    {
+      "epoch": 4.343660823481451,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.285211985324093e-05,
+      "loss": 0.867,
+      "step": 63930
+    },
+    {
+      "epoch": 4.344000543552113,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2849996602799297e-05,
+      "loss": 1.043,
+      "step": 63935
+    },
+    {
+      "epoch": 4.344340263622775,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2847873352357657e-05,
+      "loss": 0.8877,
+      "step": 63940
+    },
+    {
+      "epoch": 4.344679983693436,
+      "grad_norm": 8.125,
+      "learning_rate": 2.284575010191602e-05,
+      "loss": 0.9062,
+      "step": 63945
+    },
+    {
+      "epoch": 4.345019703764098,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2843626851474385e-05,
+      "loss": 0.9607,
+      "step": 63950
+    },
+    {
+      "epoch": 4.34535942383476,
+      "grad_norm": 7.625,
+      "learning_rate": 2.284150360103275e-05,
+      "loss": 0.8609,
+      "step": 63955
+    },
+    {
+      "epoch": 4.345699143905422,
+      "grad_norm": 10.6875,
+      "learning_rate": 2.2839380350591113e-05,
+      "loss": 0.8631,
+      "step": 63960
+    },
+    {
+      "epoch": 4.346038863976084,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.2837257100149477e-05,
+      "loss": 0.9219,
+      "step": 63965
+    },
+    {
+      "epoch": 4.346378584046746,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.2835133849707845e-05,
+      "loss": 0.8292,
+      "step": 63970
+    },
+    {
+      "epoch": 4.346718304117407,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2833010599266205e-05,
+      "loss": 0.877,
+      "step": 63975
+    },
+    {
+      "epoch": 4.347058024188069,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.283088734882457e-05,
+      "loss": 0.9014,
+      "step": 63980
+    },
+    {
+      "epoch": 4.347397744258731,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2828764098382933e-05,
+      "loss": 0.913,
+      "step": 63985
+    },
+    {
+      "epoch": 4.347737464329392,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2826640847941297e-05,
+      "loss": 0.8424,
+      "step": 63990
+    },
+    {
+      "epoch": 4.348077184400054,
+      "grad_norm": 7.0,
+      "learning_rate": 2.282451759749966e-05,
+      "loss": 0.8915,
+      "step": 63995
+    },
+    {
+      "epoch": 4.348416904470716,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2822394347058025e-05,
+      "loss": 0.7645,
+      "step": 64000
+    },
+    {
+      "epoch": 4.348756624541378,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.282027109661639e-05,
+      "loss": 0.8715,
+      "step": 64005
+    },
+    {
+      "epoch": 4.34909634461204,
+      "grad_norm": 8.75,
+      "learning_rate": 2.2818147846174753e-05,
+      "loss": 0.9491,
+      "step": 64010
+    },
+    {
+      "epoch": 4.349436064682702,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2816024595733117e-05,
+      "loss": 0.8941,
+      "step": 64015
+    },
+    {
+      "epoch": 4.349775784753363,
+      "grad_norm": 8.0,
+      "learning_rate": 2.281390134529148e-05,
+      "loss": 0.7652,
+      "step": 64020
+    },
+    {
+      "epoch": 4.350115504824025,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2811778094849845e-05,
+      "loss": 0.849,
+      "step": 64025
+    },
+    {
+      "epoch": 4.350455224894687,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.280965484440821e-05,
+      "loss": 0.9756,
+      "step": 64030
+    },
+    {
+      "epoch": 4.350794944965348,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.2807531593966573e-05,
+      "loss": 0.8365,
+      "step": 64035
+    },
+    {
+      "epoch": 4.35113466503601,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2805408343524937e-05,
+      "loss": 0.8452,
+      "step": 64040
+    },
+    {
+      "epoch": 4.351474385106672,
+      "grad_norm": 8.875,
+      "learning_rate": 2.28032850930833e-05,
+      "loss": 0.9156,
+      "step": 64045
+    },
+    {
+      "epoch": 4.351814105177334,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.2801161842641662e-05,
+      "loss": 0.8417,
+      "step": 64050
+    },
+    {
+      "epoch": 4.352153825247996,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.279903859220003e-05,
+      "loss": 0.8571,
+      "step": 64055
+    },
+    {
+      "epoch": 4.352493545318658,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2796915341758393e-05,
+      "loss": 0.8342,
+      "step": 64060
+    },
+    {
+      "epoch": 4.352833265389319,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2794792091316754e-05,
+      "loss": 0.8764,
+      "step": 64065
+    },
+    {
+      "epoch": 4.353172985459981,
+      "grad_norm": 6.1875,
+      "learning_rate": 2.279266884087512e-05,
+      "loss": 0.8843,
+      "step": 64070
+    },
+    {
+      "epoch": 4.353512705530643,
+      "grad_norm": 7.0,
+      "learning_rate": 2.2790545590433485e-05,
+      "loss": 0.8809,
+      "step": 64075
+    },
+    {
+      "epoch": 4.353852425601304,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2788422339991846e-05,
+      "loss": 0.8554,
+      "step": 64080
+    },
+    {
+      "epoch": 4.354192145671966,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.278629908955021e-05,
+      "loss": 0.8319,
+      "step": 64085
+    },
+    {
+      "epoch": 4.354531865742628,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2784175839108577e-05,
+      "loss": 0.7701,
+      "step": 64090
+    },
+    {
+      "epoch": 4.35487158581329,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2782052588666938e-05,
+      "loss": 0.9429,
+      "step": 64095
+    },
+    {
+      "epoch": 4.355211305883952,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2779929338225302e-05,
+      "loss": 0.8169,
+      "step": 64100
+    },
+    {
+      "epoch": 4.355551025954614,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.277780608778367e-05,
+      "loss": 0.8719,
+      "step": 64105
+    },
+    {
+      "epoch": 4.355890746025275,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.277568283734203e-05,
+      "loss": 0.8583,
+      "step": 64110
+    },
+    {
+      "epoch": 4.356230466095937,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.2773559586900394e-05,
+      "loss": 0.7197,
+      "step": 64115
+    },
+    {
+      "epoch": 4.356570186166599,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2771436336458758e-05,
+      "loss": 0.9518,
+      "step": 64120
+    },
+    {
+      "epoch": 4.35690990623726,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2769313086017122e-05,
+      "loss": 0.886,
+      "step": 64125
+    },
+    {
+      "epoch": 4.357249626307922,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2767189835575486e-05,
+      "loss": 0.7982,
+      "step": 64130
+    },
+    {
+      "epoch": 4.357589346378584,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.276506658513385e-05,
+      "loss": 0.9301,
+      "step": 64135
+    },
+    {
+      "epoch": 4.357929066449246,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.2762943334692217e-05,
+      "loss": 0.9191,
+      "step": 64140
+    },
+    {
+      "epoch": 4.358268786519908,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2760820084250578e-05,
+      "loss": 0.8122,
+      "step": 64145
+    },
+    {
+      "epoch": 4.35860850659057,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2758696833808942e-05,
+      "loss": 0.814,
+      "step": 64150
+    },
+    {
+      "epoch": 4.358948226661231,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2756573583367306e-05,
+      "loss": 0.894,
+      "step": 64155
+    },
+    {
+      "epoch": 4.359287946731893,
+      "grad_norm": 7.625,
+      "learning_rate": 2.275445033292567e-05,
+      "loss": 0.9752,
+      "step": 64160
+    },
+    {
+      "epoch": 4.359627666802555,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2752327082484034e-05,
+      "loss": 0.8595,
+      "step": 64165
+    },
+    {
+      "epoch": 4.359967386873216,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2750203832042398e-05,
+      "loss": 0.7941,
+      "step": 64170
+    },
+    {
+      "epoch": 4.360307106943878,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.2748080581600762e-05,
+      "loss": 0.8535,
+      "step": 64175
+    },
+    {
+      "epoch": 4.36064682701454,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2745957331159126e-05,
+      "loss": 0.8738,
+      "step": 64180
+    },
+    {
+      "epoch": 4.360986547085202,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.274383408071749e-05,
+      "loss": 0.9666,
+      "step": 64185
+    },
+    {
+      "epoch": 4.361326267155864,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2741710830275854e-05,
+      "loss": 0.8209,
+      "step": 64190
+    },
+    {
+      "epoch": 4.361665987226525,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.2739587579834218e-05,
+      "loss": 0.9031,
+      "step": 64195
+    },
+    {
+      "epoch": 4.362005707297187,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.2737464329392582e-05,
+      "loss": 0.8268,
+      "step": 64200
+    },
+    {
+      "epoch": 4.362345427367849,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.2735341078950946e-05,
+      "loss": 0.8414,
+      "step": 64205
+    },
+    {
+      "epoch": 4.36268514743851,
+      "grad_norm": 7.125,
+      "learning_rate": 2.273321782850931e-05,
+      "loss": 0.8439,
+      "step": 64210
+    },
+    {
+      "epoch": 4.363024867509172,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2731094578067674e-05,
+      "loss": 0.897,
+      "step": 64215
+    },
+    {
+      "epoch": 4.363364587579834,
+      "grad_norm": 8.25,
+      "learning_rate": 2.2728971327626035e-05,
+      "loss": 0.9344,
+      "step": 64220
+    },
+    {
+      "epoch": 4.3637043076504956,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2726848077184402e-05,
+      "loss": 0.8286,
+      "step": 64225
+    },
+    {
+      "epoch": 4.364044027721158,
+      "grad_norm": 7.0,
+      "learning_rate": 2.2724724826742766e-05,
+      "loss": 0.9125,
+      "step": 64230
+    },
+    {
+      "epoch": 4.36438374779182,
+      "grad_norm": 8.625,
+      "learning_rate": 2.2722601576301127e-05,
+      "loss": 1.0522,
+      "step": 64235
+    },
+    {
+      "epoch": 4.364723467862481,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2720478325859494e-05,
+      "loss": 0.8838,
+      "step": 64240
+    },
+    {
+      "epoch": 4.365063187933143,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2718355075417858e-05,
+      "loss": 0.8476,
+      "step": 64245
+    },
+    {
+      "epoch": 4.365402908003805,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.271623182497622e-05,
+      "loss": 0.9043,
+      "step": 64250
+    },
+    {
+      "epoch": 4.365742628074466,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2714108574534583e-05,
+      "loss": 0.9001,
+      "step": 64255
+    },
+    {
+      "epoch": 4.366082348145128,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.271198532409295e-05,
+      "loss": 0.9221,
+      "step": 64260
+    },
+    {
+      "epoch": 4.36642206821579,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.270986207365131e-05,
+      "loss": 0.7874,
+      "step": 64265
+    },
+    {
+      "epoch": 4.366761788286452,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2707738823209675e-05,
+      "loss": 0.8906,
+      "step": 64270
+    },
+    {
+      "epoch": 4.367101508357114,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2705615572768042e-05,
+      "loss": 0.831,
+      "step": 64275
+    },
+    {
+      "epoch": 4.367441228427776,
+      "grad_norm": 6.875,
+      "learning_rate": 2.2703492322326403e-05,
+      "loss": 0.7666,
+      "step": 64280
+    },
+    {
+      "epoch": 4.367780948498437,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.2701369071884767e-05,
+      "loss": 0.8323,
+      "step": 64285
+    },
+    {
+      "epoch": 4.368120668569099,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2699245821443134e-05,
+      "loss": 0.8681,
+      "step": 64290
+    },
+    {
+      "epoch": 4.368460388639761,
+      "grad_norm": 8.875,
+      "learning_rate": 2.2697122571001495e-05,
+      "loss": 0.9495,
+      "step": 64295
+    },
+    {
+      "epoch": 4.368800108710422,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.269499932055986e-05,
+      "loss": 0.91,
+      "step": 64300
+    },
+    {
+      "epoch": 4.369139828781084,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.2692876070118223e-05,
+      "loss": 0.876,
+      "step": 64305
+    },
+    {
+      "epoch": 4.369479548851746,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.269075281967659e-05,
+      "loss": 0.869,
+      "step": 64310
+    },
+    {
+      "epoch": 4.369819268922408,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.268862956923495e-05,
+      "loss": 0.8596,
+      "step": 64315
+    },
+    {
+      "epoch": 4.37015898899307,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2686506318793315e-05,
+      "loss": 0.854,
+      "step": 64320
+    },
+    {
+      "epoch": 4.370498709063732,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2684383068351682e-05,
+      "loss": 0.7716,
+      "step": 64325
+    },
+    {
+      "epoch": 4.370838429134393,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.2682259817910043e-05,
+      "loss": 0.969,
+      "step": 64330
+    },
+    {
+      "epoch": 4.371178149205055,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.2680136567468407e-05,
+      "loss": 0.864,
+      "step": 64335
+    },
+    {
+      "epoch": 4.371517869275717,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.267801331702677e-05,
+      "loss": 0.8677,
+      "step": 64340
+    },
+    {
+      "epoch": 4.371857589346378,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2675890066585135e-05,
+      "loss": 0.9476,
+      "step": 64345
+    },
+    {
+      "epoch": 4.37219730941704,
+      "grad_norm": 7.375,
+      "learning_rate": 2.26737668161435e-05,
+      "loss": 0.8581,
+      "step": 64350
+    },
+    {
+      "epoch": 4.372537029487702,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.2671643565701863e-05,
+      "loss": 0.8922,
+      "step": 64355
+    },
+    {
+      "epoch": 4.372876749558364,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.2669520315260227e-05,
+      "loss": 0.8574,
+      "step": 64360
+    },
+    {
+      "epoch": 4.373216469629026,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.266739706481859e-05,
+      "loss": 0.9823,
+      "step": 64365
+    },
+    {
+      "epoch": 4.373556189699688,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2665273814376955e-05,
+      "loss": 0.8585,
+      "step": 64370
+    },
+    {
+      "epoch": 4.373895909770349,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.266315056393532e-05,
+      "loss": 0.8433,
+      "step": 64375
+    },
+    {
+      "epoch": 4.374235629841011,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.2661027313493683e-05,
+      "loss": 0.9576,
+      "step": 64380
+    },
+    {
+      "epoch": 4.374575349911673,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2658904063052047e-05,
+      "loss": 0.8264,
+      "step": 64385
+    },
+    {
+      "epoch": 4.374915069982334,
+      "grad_norm": 7.125,
+      "learning_rate": 2.265678081261041e-05,
+      "loss": 0.792,
+      "step": 64390
+    },
+    {
+      "epoch": 4.375254790052996,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.2654657562168775e-05,
+      "loss": 0.8753,
+      "step": 64395
+    },
+    {
+      "epoch": 4.375594510123658,
+      "grad_norm": 8.375,
+      "learning_rate": 2.265253431172714e-05,
+      "loss": 0.8073,
+      "step": 64400
+    },
+    {
+      "epoch": 4.37593423019432,
+      "grad_norm": 7.375,
+      "learning_rate": 2.26504110612855e-05,
+      "loss": 0.7955,
+      "step": 64405
+    },
+    {
+      "epoch": 4.376273950264982,
+      "grad_norm": 8.375,
+      "learning_rate": 2.2648287810843867e-05,
+      "loss": 0.8764,
+      "step": 64410
+    },
+    {
+      "epoch": 4.376613670335644,
+      "grad_norm": 7.25,
+      "learning_rate": 2.264616456040223e-05,
+      "loss": 0.78,
+      "step": 64415
+    },
+    {
+      "epoch": 4.376953390406305,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.264404130996059e-05,
+      "loss": 0.8685,
+      "step": 64420
+    },
+    {
+      "epoch": 4.377293110476967,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.264191805951896e-05,
+      "loss": 0.7593,
+      "step": 64425
+    },
+    {
+      "epoch": 4.377632830547629,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2639794809077323e-05,
+      "loss": 0.9833,
+      "step": 64430
+    },
+    {
+      "epoch": 4.37797255061829,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2637671558635683e-05,
+      "loss": 0.9012,
+      "step": 64435
+    },
+    {
+      "epoch": 4.378312270688952,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.2635548308194047e-05,
+      "loss": 0.8639,
+      "step": 64440
+    },
+    {
+      "epoch": 4.378651990759614,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2633425057752415e-05,
+      "loss": 0.888,
+      "step": 64445
+    },
+    {
+      "epoch": 4.378991710830276,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2631301807310775e-05,
+      "loss": 1.0289,
+      "step": 64450
+    },
+    {
+      "epoch": 4.379331430900938,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.262917855686914e-05,
+      "loss": 0.8844,
+      "step": 64455
+    },
+    {
+      "epoch": 4.3796711509716,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.2627055306427507e-05,
+      "loss": 0.9259,
+      "step": 64460
+    },
+    {
+      "epoch": 4.380010871042261,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2624932055985867e-05,
+      "loss": 0.8586,
+      "step": 64465
+    },
+    {
+      "epoch": 4.380350591112923,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.262280880554423e-05,
+      "loss": 0.8148,
+      "step": 64470
+    },
+    {
+      "epoch": 4.380690311183585,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2620685555102596e-05,
+      "loss": 0.8926,
+      "step": 64475
+    },
+    {
+      "epoch": 4.381030031254246,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2618562304660963e-05,
+      "loss": 0.9532,
+      "step": 64480
+    },
+    {
+      "epoch": 4.381369751324908,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.2616439054219324e-05,
+      "loss": 0.9676,
+      "step": 64485
+    },
+    {
+      "epoch": 4.38170947139557,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.2614315803777688e-05,
+      "loss": 0.8421,
+      "step": 64490
+    },
+    {
+      "epoch": 4.382049191466232,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.2612192553336055e-05,
+      "loss": 0.9271,
+      "step": 64495
+    },
+    {
+      "epoch": 4.382388911536894,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2610069302894416e-05,
+      "loss": 0.8369,
+      "step": 64500
+    },
+    {
+      "epoch": 4.382728631607556,
+      "grad_norm": 6.625,
+      "learning_rate": 2.260794605245278e-05,
+      "loss": 0.8046,
+      "step": 64505
+    },
+    {
+      "epoch": 4.383068351678217,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2605822802011144e-05,
+      "loss": 0.8861,
+      "step": 64510
+    },
+    {
+      "epoch": 4.383408071748879,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.2603699551569508e-05,
+      "loss": 0.8807,
+      "step": 64515
+    },
+    {
+      "epoch": 4.383747791819541,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.260157630112787e-05,
+      "loss": 0.8442,
+      "step": 64520
+    },
+    {
+      "epoch": 4.384087511890202,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.2599453050686236e-05,
+      "loss": 0.8799,
+      "step": 64525
+    },
+    {
+      "epoch": 4.384427231960864,
+      "grad_norm": 8.25,
+      "learning_rate": 2.25973298002446e-05,
+      "loss": 0.9535,
+      "step": 64530
+    },
+    {
+      "epoch": 4.384766952031526,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2595206549802964e-05,
+      "loss": 0.9705,
+      "step": 64535
+    },
+    {
+      "epoch": 4.385106672102188,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2593083299361328e-05,
+      "loss": 0.8558,
+      "step": 64540
+    },
+    {
+      "epoch": 4.38544639217285,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.259096004891969e-05,
+      "loss": 0.7884,
+      "step": 64545
+    },
+    {
+      "epoch": 4.385786112243512,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.2588836798478056e-05,
+      "loss": 0.8798,
+      "step": 64550
+    },
+    {
+      "epoch": 4.386125832314173,
+      "grad_norm": 8.375,
+      "learning_rate": 2.258671354803642e-05,
+      "loss": 0.825,
+      "step": 64555
+    },
+    {
+      "epoch": 4.386465552384835,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.2584590297594784e-05,
+      "loss": 0.937,
+      "step": 64560
+    },
+    {
+      "epoch": 4.386805272455497,
+      "grad_norm": 8.375,
+      "learning_rate": 2.2582467047153148e-05,
+      "loss": 0.9514,
+      "step": 64565
+    },
+    {
+      "epoch": 4.387144992526158,
+      "grad_norm": 8.125,
+      "learning_rate": 2.258034379671151e-05,
+      "loss": 0.8504,
+      "step": 64570
+    },
+    {
+      "epoch": 4.38748471259682,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2578220546269872e-05,
+      "loss": 0.8521,
+      "step": 64575
+    },
+    {
+      "epoch": 4.3878244326674825,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.257609729582824e-05,
+      "loss": 0.816,
+      "step": 64580
+    },
+    {
+      "epoch": 4.388164152738144,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.2573974045386604e-05,
+      "loss": 0.8797,
+      "step": 64585
+    },
+    {
+      "epoch": 4.388503872808806,
+      "grad_norm": 9.25,
+      "learning_rate": 2.2571850794944964e-05,
+      "loss": 0.8701,
+      "step": 64590
+    },
+    {
+      "epoch": 4.388843592879467,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.256972754450333e-05,
+      "loss": 0.9365,
+      "step": 64595
+    },
+    {
+      "epoch": 4.389183312950129,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2567604294061696e-05,
+      "loss": 0.8343,
+      "step": 64600
+    },
+    {
+      "epoch": 4.389523033020791,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2565481043620056e-05,
+      "loss": 0.9303,
+      "step": 64605
+    },
+    {
+      "epoch": 4.389862753091452,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.256335779317842e-05,
+      "loss": 0.8812,
+      "step": 64610
+    },
+    {
+      "epoch": 4.390202473162114,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2561234542736788e-05,
+      "loss": 0.9522,
+      "step": 64615
+    },
+    {
+      "epoch": 4.390542193232776,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2559111292295148e-05,
+      "loss": 0.9243,
+      "step": 64620
+    },
+    {
+      "epoch": 4.390881913303438,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2556988041853512e-05,
+      "loss": 0.803,
+      "step": 64625
+    },
+    {
+      "epoch": 4.3912216333741,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.255486479141188e-05,
+      "loss": 0.8269,
+      "step": 64630
+    },
+    {
+      "epoch": 4.391561353444762,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.255274154097024e-05,
+      "loss": 0.861,
+      "step": 64635
+    },
+    {
+      "epoch": 4.391901073515423,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.2550618290528604e-05,
+      "loss": 0.8463,
+      "step": 64640
+    },
+    {
+      "epoch": 4.392240793586085,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2548495040086968e-05,
+      "loss": 0.7766,
+      "step": 64645
+    },
+    {
+      "epoch": 4.392580513656747,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2546371789645336e-05,
+      "loss": 0.9372,
+      "step": 64650
+    },
+    {
+      "epoch": 4.392920233727408,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2544248539203696e-05,
+      "loss": 0.8369,
+      "step": 64655
+    },
+    {
+      "epoch": 4.39325995379807,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.254212528876206e-05,
+      "loss": 1.0006,
+      "step": 64660
+    },
+    {
+      "epoch": 4.393599673868732,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.2540002038320428e-05,
+      "loss": 0.8191,
+      "step": 64665
+    },
+    {
+      "epoch": 4.393939393939394,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.2537878787878788e-05,
+      "loss": 0.7809,
+      "step": 64670
+    },
+    {
+      "epoch": 4.394279114010056,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2535755537437152e-05,
+      "loss": 0.8471,
+      "step": 64675
+    },
+    {
+      "epoch": 4.394618834080718,
+      "grad_norm": 9.375,
+      "learning_rate": 2.253363228699552e-05,
+      "loss": 0.8411,
+      "step": 64680
+    },
+    {
+      "epoch": 4.394958554151379,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.253150903655388e-05,
+      "loss": 0.9574,
+      "step": 64685
+    },
+    {
+      "epoch": 4.395298274222041,
+      "grad_norm": 8.625,
+      "learning_rate": 2.2529385786112244e-05,
+      "loss": 0.795,
+      "step": 64690
+    },
+    {
+      "epoch": 4.395637994292703,
+      "grad_norm": 8.625,
+      "learning_rate": 2.2527262535670608e-05,
+      "loss": 0.9033,
+      "step": 64695
+    },
+    {
+      "epoch": 4.395977714363364,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2525139285228972e-05,
+      "loss": 0.8721,
+      "step": 64700
+    },
+    {
+      "epoch": 4.396317434434026,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.2523016034787336e-05,
+      "loss": 0.8675,
+      "step": 64705
+    },
+    {
+      "epoch": 4.396657154504688,
+      "grad_norm": 7.125,
+      "learning_rate": 2.25208927843457e-05,
+      "loss": 0.8386,
+      "step": 64710
+    },
+    {
+      "epoch": 4.39699687457535,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.2518769533904064e-05,
+      "loss": 0.8095,
+      "step": 64715
+    },
+    {
+      "epoch": 4.397336594646012,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2516646283462428e-05,
+      "loss": 0.8284,
+      "step": 64720
+    },
+    {
+      "epoch": 4.397676314716674,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.2514523033020792e-05,
+      "loss": 0.8408,
+      "step": 64725
+    },
+    {
+      "epoch": 4.398016034787335,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2512399782579156e-05,
+      "loss": 0.8009,
+      "step": 64730
+    },
+    {
+      "epoch": 4.398355754857997,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.251027653213752e-05,
+      "loss": 0.9083,
+      "step": 64735
+    },
+    {
+      "epoch": 4.398695474928659,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2508153281695884e-05,
+      "loss": 0.8056,
+      "step": 64740
+    },
+    {
+      "epoch": 4.39903519499932,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.2506030031254245e-05,
+      "loss": 0.8271,
+      "step": 64745
+    },
+    {
+      "epoch": 4.399374915069982,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.2503906780812612e-05,
+      "loss": 0.8182,
+      "step": 64750
+    },
+    {
+      "epoch": 4.399714635140644,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2501783530370976e-05,
+      "loss": 0.8595,
+      "step": 64755
+    },
+    {
+      "epoch": 4.400054355211306,
+      "grad_norm": 6.625,
+      "learning_rate": 2.2499660279929337e-05,
+      "loss": 0.8483,
+      "step": 64760
+    },
+    {
+      "epoch": 4.400394075281968,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2497537029487704e-05,
+      "loss": 0.8461,
+      "step": 64765
+    },
+    {
+      "epoch": 4.40073379535263,
+      "grad_norm": 8.25,
+      "learning_rate": 2.249541377904607e-05,
+      "loss": 0.77,
+      "step": 64770
+    },
+    {
+      "epoch": 4.401073515423291,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.249329052860443e-05,
+      "loss": 0.8262,
+      "step": 64775
+    },
+    {
+      "epoch": 4.401413235493953,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.2491167278162796e-05,
+      "loss": 0.826,
+      "step": 64780
+    },
+    {
+      "epoch": 4.401752955564615,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.248904402772116e-05,
+      "loss": 0.8345,
+      "step": 64785
+    },
+    {
+      "epoch": 4.402092675635276,
+      "grad_norm": 7.625,
+      "learning_rate": 2.248692077727952e-05,
+      "loss": 0.8694,
+      "step": 64790
+    },
+    {
+      "epoch": 4.402432395705938,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.2484797526837885e-05,
+      "loss": 0.8899,
+      "step": 64795
+    },
+    {
+      "epoch": 4.4027721157766,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.2482674276396252e-05,
+      "loss": 0.9093,
+      "step": 64800
+    },
+    {
+      "epoch": 4.403111835847262,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2480551025954613e-05,
+      "loss": 0.8828,
+      "step": 64805
+    },
+    {
+      "epoch": 4.403451555917924,
+      "grad_norm": 8.25,
+      "learning_rate": 2.2478427775512977e-05,
+      "loss": 0.7882,
+      "step": 64810
+    },
+    {
+      "epoch": 4.403791275988586,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.2476304525071344e-05,
+      "loss": 0.8937,
+      "step": 64815
+    },
+    {
+      "epoch": 4.404130996059247,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.247418127462971e-05,
+      "loss": 0.8638,
+      "step": 64820
+    },
+    {
+      "epoch": 4.404470716129909,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.247205802418807e-05,
+      "loss": 0.9343,
+      "step": 64825
+    },
+    {
+      "epoch": 4.404810436200571,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.2469934773746433e-05,
+      "loss": 0.9313,
+      "step": 64830
+    },
+    {
+      "epoch": 4.405150156271232,
+      "grad_norm": 8.375,
+      "learning_rate": 2.24678115233048e-05,
+      "loss": 0.8889,
+      "step": 64835
+    },
+    {
+      "epoch": 4.405489876341894,
+      "grad_norm": 7.625,
+      "learning_rate": 2.246568827286316e-05,
+      "loss": 0.8708,
+      "step": 64840
+    },
+    {
+      "epoch": 4.4058295964125564,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.2463565022421525e-05,
+      "loss": 0.8386,
+      "step": 64845
+    },
+    {
+      "epoch": 4.406169316483218,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2461441771979892e-05,
+      "loss": 0.813,
+      "step": 64850
+    },
+    {
+      "epoch": 4.40650903655388,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2459318521538253e-05,
+      "loss": 0.9125,
+      "step": 64855
+    },
+    {
+      "epoch": 4.406848756624542,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2457195271096617e-05,
+      "loss": 0.8311,
+      "step": 64860
+    },
+    {
+      "epoch": 4.407188476695203,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.245507202065498e-05,
+      "loss": 0.9243,
+      "step": 64865
+    },
+    {
+      "epoch": 4.407528196765865,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2452948770213345e-05,
+      "loss": 0.8501,
+      "step": 64870
+    },
+    {
+      "epoch": 4.407867916836526,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.245082551977171e-05,
+      "loss": 0.7873,
+      "step": 64875
+    },
+    {
+      "epoch": 4.408207636907188,
+      "grad_norm": 8.875,
+      "learning_rate": 2.2448702269330073e-05,
+      "loss": 0.8961,
+      "step": 64880
+    },
+    {
+      "epoch": 4.40854735697785,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.2446579018888437e-05,
+      "loss": 0.8217,
+      "step": 64885
+    },
+    {
+      "epoch": 4.408887077048512,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.24444557684468e-05,
+      "loss": 0.8422,
+      "step": 64890
+    },
+    {
+      "epoch": 4.409226797119174,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2442332518005165e-05,
+      "loss": 0.854,
+      "step": 64895
+    },
+    {
+      "epoch": 4.409566517189836,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.244020926756353e-05,
+      "loss": 0.9164,
+      "step": 64900
+    },
+    {
+      "epoch": 4.409906237260497,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2438086017121893e-05,
+      "loss": 0.8811,
+      "step": 64905
+    },
+    {
+      "epoch": 4.410245957331159,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2435962766680257e-05,
+      "loss": 0.9522,
+      "step": 64910
+    },
+    {
+      "epoch": 4.410585677401821,
+      "grad_norm": 7.75,
+      "learning_rate": 2.243383951623862e-05,
+      "loss": 0.8852,
+      "step": 64915
+    },
+    {
+      "epoch": 4.410925397472482,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.2431716265796985e-05,
+      "loss": 0.7422,
+      "step": 64920
+    },
+    {
+      "epoch": 4.411265117543144,
+      "grad_norm": 8.75,
+      "learning_rate": 2.242959301535535e-05,
+      "loss": 0.8153,
+      "step": 64925
+    },
+    {
+      "epoch": 4.411604837613806,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.242746976491371e-05,
+      "loss": 0.8238,
+      "step": 64930
+    },
+    {
+      "epoch": 4.411944557684468,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2425346514472077e-05,
+      "loss": 0.8306,
+      "step": 64935
+    },
+    {
+      "epoch": 4.41228427775513,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.242322326403044e-05,
+      "loss": 0.9403,
+      "step": 64940
+    },
+    {
+      "epoch": 4.412623997825792,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.2421100013588802e-05,
+      "loss": 0.872,
+      "step": 64945
+    },
+    {
+      "epoch": 4.412963717896453,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.241897676314717e-05,
+      "loss": 0.9173,
+      "step": 64950
+    },
+    {
+      "epoch": 4.413303437967115,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.2416853512705533e-05,
+      "loss": 0.8644,
+      "step": 64955
+    },
+    {
+      "epoch": 4.413643158037777,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.2414730262263894e-05,
+      "loss": 0.9021,
+      "step": 64960
+    },
+    {
+      "epoch": 4.413982878108438,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.2412607011822258e-05,
+      "loss": 0.803,
+      "step": 64965
+    },
+    {
+      "epoch": 4.4143225981791,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.2410483761380625e-05,
+      "loss": 0.8042,
+      "step": 64970
+    },
+    {
+      "epoch": 4.414662318249762,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.2408360510938986e-05,
+      "loss": 0.8965,
+      "step": 64975
+    },
+    {
+      "epoch": 4.415002038320424,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.240623726049735e-05,
+      "loss": 0.9206,
+      "step": 64980
+    },
+    {
+      "epoch": 4.415341758391086,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2404114010055717e-05,
+      "loss": 0.8512,
+      "step": 64985
+    },
+    {
+      "epoch": 4.415681478461748,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.240199075961408e-05,
+      "loss": 0.8888,
+      "step": 64990
+    },
+    {
+      "epoch": 4.416021198532409,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2399867509172442e-05,
+      "loss": 0.8518,
+      "step": 64995
+    },
+    {
+      "epoch": 4.416360918603071,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2397744258730806e-05,
+      "loss": 0.824,
+      "step": 65000
+    },
+    {
+      "epoch": 4.416700638673733,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2395621008289173e-05,
+      "loss": 0.7601,
+      "step": 65005
+    },
+    {
+      "epoch": 4.417040358744394,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2393497757847534e-05,
+      "loss": 0.8562,
+      "step": 65010
+    },
+    {
+      "epoch": 4.417380078815056,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.2391374507405898e-05,
+      "loss": 0.8643,
+      "step": 65015
+    },
+    {
+      "epoch": 4.417719798885718,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2389251256964265e-05,
+      "loss": 0.8229,
+      "step": 65020
+    },
+    {
+      "epoch": 4.41805951895638,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.2387128006522626e-05,
+      "loss": 0.8201,
+      "step": 65025
+    },
+    {
+      "epoch": 4.418399239027042,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.238500475608099e-05,
+      "loss": 0.8531,
+      "step": 65030
+    },
+    {
+      "epoch": 4.418738959097704,
+      "grad_norm": 8.375,
+      "learning_rate": 2.2382881505639354e-05,
+      "loss": 0.8336,
+      "step": 65035
+    },
+    {
+      "epoch": 4.419078679168365,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2380758255197718e-05,
+      "loss": 0.9146,
+      "step": 65040
+    },
+    {
+      "epoch": 4.419418399239027,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2378635004756082e-05,
+      "loss": 0.8006,
+      "step": 65045
+    },
+    {
+      "epoch": 4.419758119309689,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.2376511754314446e-05,
+      "loss": 0.9274,
+      "step": 65050
+    },
+    {
+      "epoch": 4.42009783938035,
+      "grad_norm": 7.5,
+      "learning_rate": 2.237438850387281e-05,
+      "loss": 0.8193,
+      "step": 65055
+    },
+    {
+      "epoch": 4.420437559451012,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2372265253431174e-05,
+      "loss": 0.8312,
+      "step": 65060
+    },
+    {
+      "epoch": 4.420777279521674,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.2370142002989538e-05,
+      "loss": 0.8487,
+      "step": 65065
+    },
+    {
+      "epoch": 4.421116999592336,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2368018752547902e-05,
+      "loss": 0.8645,
+      "step": 65070
+    },
+    {
+      "epoch": 4.421456719662998,
+      "grad_norm": 8.25,
+      "learning_rate": 2.2365895502106266e-05,
+      "loss": 0.9081,
+      "step": 65075
+    },
+    {
+      "epoch": 4.42179643973366,
+      "grad_norm": 7.75,
+      "learning_rate": 2.236377225166463e-05,
+      "loss": 0.8256,
+      "step": 65080
+    },
+    {
+      "epoch": 4.422136159804321,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2361649001222994e-05,
+      "loss": 0.8627,
+      "step": 65085
+    },
+    {
+      "epoch": 4.422475879874983,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2359525750781358e-05,
+      "loss": 0.8793,
+      "step": 65090
+    },
+    {
+      "epoch": 4.422815599945645,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2357402500339722e-05,
+      "loss": 0.9021,
+      "step": 65095
+    },
+    {
+      "epoch": 4.423155320016306,
+      "grad_norm": 8.75,
+      "learning_rate": 2.2355279249898082e-05,
+      "loss": 0.9326,
+      "step": 65100
+    },
+    {
+      "epoch": 4.423495040086968,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.235315599945645e-05,
+      "loss": 0.835,
+      "step": 65105
+    },
+    {
+      "epoch": 4.42383476015763,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2351032749014814e-05,
+      "loss": 0.9657,
+      "step": 65110
+    },
+    {
+      "epoch": 4.424174480228292,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2348909498573174e-05,
+      "loss": 0.8577,
+      "step": 65115
+    },
+    {
+      "epoch": 4.424514200298954,
+      "grad_norm": 8.375,
+      "learning_rate": 2.2346786248131542e-05,
+      "loss": 0.8862,
+      "step": 65120
+    },
+    {
+      "epoch": 4.424853920369616,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2344662997689906e-05,
+      "loss": 1.0027,
+      "step": 65125
+    },
+    {
+      "epoch": 4.425193640440277,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2342539747248266e-05,
+      "loss": 0.8474,
+      "step": 65130
+    },
+    {
+      "epoch": 4.425533360510939,
+      "grad_norm": 7.25,
+      "learning_rate": 2.234041649680663e-05,
+      "loss": 0.9118,
+      "step": 65135
+    },
+    {
+      "epoch": 4.425873080581601,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2338293246364998e-05,
+      "loss": 0.8151,
+      "step": 65140
+    },
+    {
+      "epoch": 4.426212800652262,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.233616999592336e-05,
+      "loss": 0.8951,
+      "step": 65145
+    },
+    {
+      "epoch": 4.426552520722924,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2334046745481722e-05,
+      "loss": 0.8178,
+      "step": 65150
+    },
+    {
+      "epoch": 4.4268922407935865,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.233192349504009e-05,
+      "loss": 0.8744,
+      "step": 65155
+    },
+    {
+      "epoch": 4.427231960864248,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2329800244598454e-05,
+      "loss": 0.9716,
+      "step": 65160
+    },
+    {
+      "epoch": 4.42757168093491,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2327676994156814e-05,
+      "loss": 0.9652,
+      "step": 65165
+    },
+    {
+      "epoch": 4.427911401005572,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.232555374371518e-05,
+      "loss": 0.9136,
+      "step": 65170
+    },
+    {
+      "epoch": 4.428251121076233,
+      "grad_norm": 6.625,
+      "learning_rate": 2.2323430493273546e-05,
+      "loss": 0.7931,
+      "step": 65175
+    },
+    {
+      "epoch": 4.428590841146895,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2321307242831906e-05,
+      "loss": 0.9604,
+      "step": 65180
+    },
+    {
+      "epoch": 4.428930561217557,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.231918399239027e-05,
+      "loss": 0.9395,
+      "step": 65185
+    },
+    {
+      "epoch": 4.429270281288218,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2317060741948638e-05,
+      "loss": 0.9427,
+      "step": 65190
+    },
+    {
+      "epoch": 4.42961000135888,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2314937491507e-05,
+      "loss": 0.9223,
+      "step": 65195
+    },
+    {
+      "epoch": 4.4299497214295425,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2312814241065363e-05,
+      "loss": 0.8885,
+      "step": 65200
+    },
+    {
+      "epoch": 4.430289441500204,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.231069099062373e-05,
+      "loss": 0.9079,
+      "step": 65205
+    },
+    {
+      "epoch": 4.430629161570866,
+      "grad_norm": 7.5,
+      "learning_rate": 2.230856774018209e-05,
+      "loss": 0.8469,
+      "step": 65210
+    },
+    {
+      "epoch": 4.430968881641528,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.2306444489740455e-05,
+      "loss": 0.7999,
+      "step": 65215
+    },
+    {
+      "epoch": 4.431308601712189,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.230432123929882e-05,
+      "loss": 0.9349,
+      "step": 65220
+    },
+    {
+      "epoch": 4.431648321782851,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.2302197988857183e-05,
+      "loss": 0.9741,
+      "step": 65225
+    },
+    {
+      "epoch": 4.431988041853513,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2300074738415547e-05,
+      "loss": 0.9049,
+      "step": 65230
+    },
+    {
+      "epoch": 4.432327761924174,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.229795148797391e-05,
+      "loss": 0.9102,
+      "step": 65235
+    },
+    {
+      "epoch": 4.432667481994836,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.2295828237532275e-05,
+      "loss": 0.8554,
+      "step": 65240
+    },
+    {
+      "epoch": 4.4330072020654985,
+      "grad_norm": 7.5,
+      "learning_rate": 2.229370498709064e-05,
+      "loss": 0.8309,
+      "step": 65245
+    },
+    {
+      "epoch": 4.43334692213616,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.2291581736649003e-05,
+      "loss": 0.9689,
+      "step": 65250
+    },
+    {
+      "epoch": 4.433686642206822,
+      "grad_norm": 8.875,
+      "learning_rate": 2.2289458486207367e-05,
+      "loss": 0.911,
+      "step": 65255
+    },
+    {
+      "epoch": 4.434026362277484,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.228733523576573e-05,
+      "loss": 0.8855,
+      "step": 65260
+    },
+    {
+      "epoch": 4.434366082348145,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2285211985324095e-05,
+      "loss": 0.8378,
+      "step": 65265
+    },
+    {
+      "epoch": 4.434705802418807,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2283088734882455e-05,
+      "loss": 0.8724,
+      "step": 65270
+    },
+    {
+      "epoch": 4.435045522489468,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2280965484440823e-05,
+      "loss": 0.8318,
+      "step": 65275
+    },
+    {
+      "epoch": 4.43538524256013,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2278842233999187e-05,
+      "loss": 0.8642,
+      "step": 65280
+    },
+    {
+      "epoch": 4.435724962630792,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2276718983557547e-05,
+      "loss": 0.9317,
+      "step": 65285
+    },
+    {
+      "epoch": 4.436064682701454,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.2274595733115915e-05,
+      "loss": 0.8239,
+      "step": 65290
+    },
+    {
+      "epoch": 4.436404402772116,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.227247248267428e-05,
+      "loss": 0.8394,
+      "step": 65295
+    },
+    {
+      "epoch": 4.436744122842778,
+      "grad_norm": 8.625,
+      "learning_rate": 2.227034923223264e-05,
+      "loss": 0.9622,
+      "step": 65300
+    },
+    {
+      "epoch": 4.437083842913439,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.2268225981791007e-05,
+      "loss": 0.8446,
+      "step": 65305
+    },
+    {
+      "epoch": 4.437423562984101,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.226610273134937e-05,
+      "loss": 0.9104,
+      "step": 65310
+    },
+    {
+      "epoch": 4.437763283054763,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.226397948090773e-05,
+      "loss": 0.7723,
+      "step": 65315
+    },
+    {
+      "epoch": 4.438103003125424,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.2261856230466095e-05,
+      "loss": 0.9294,
+      "step": 65320
+    },
+    {
+      "epoch": 4.438442723196086,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.2259732980024463e-05,
+      "loss": 0.887,
+      "step": 65325
+    },
+    {
+      "epoch": 4.438782443266748,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.2257609729582827e-05,
+      "loss": 0.8485,
+      "step": 65330
+    },
+    {
+      "epoch": 4.43912216333741,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.2255486479141187e-05,
+      "loss": 0.8981,
+      "step": 65335
+    },
+    {
+      "epoch": 4.439461883408072,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2253363228699555e-05,
+      "loss": 0.7678,
+      "step": 65340
+    },
+    {
+      "epoch": 4.439801603478734,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.225123997825792e-05,
+      "loss": 0.8246,
+      "step": 65345
+    },
+    {
+      "epoch": 4.440141323549395,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.224911672781628e-05,
+      "loss": 0.8908,
+      "step": 65350
+    },
+    {
+      "epoch": 4.440481043620057,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.2246993477374643e-05,
+      "loss": 0.8837,
+      "step": 65355
+    },
+    {
+      "epoch": 4.440820763690719,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.224487022693301e-05,
+      "loss": 0.8674,
+      "step": 65360
+    },
+    {
+      "epoch": 4.44116048376138,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.224274697649137e-05,
+      "loss": 0.8393,
+      "step": 65365
+    },
+    {
+      "epoch": 4.441500203832042,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2240623726049735e-05,
+      "loss": 0.8895,
+      "step": 65370
+    },
+    {
+      "epoch": 4.441839923902704,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.2238500475608103e-05,
+      "loss": 0.8522,
+      "step": 65375
+    },
+    {
+      "epoch": 4.442179643973366,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2236377225166463e-05,
+      "loss": 0.9169,
+      "step": 65380
+    },
+    {
+      "epoch": 4.442519364044028,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2234253974724827e-05,
+      "loss": 0.8258,
+      "step": 65385
+    },
+    {
+      "epoch": 4.44285908411469,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.223213072428319e-05,
+      "loss": 0.8832,
+      "step": 65390
+    },
+    {
+      "epoch": 4.443198804185351,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2230007473841555e-05,
+      "loss": 0.9084,
+      "step": 65395
+    },
+    {
+      "epoch": 4.443538524256013,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.222788422339992e-05,
+      "loss": 0.8696,
+      "step": 65400
+    },
+    {
+      "epoch": 4.443878244326675,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2225760972958283e-05,
+      "loss": 0.8116,
+      "step": 65405
+    },
+    {
+      "epoch": 4.444217964397336,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2223637722516647e-05,
+      "loss": 0.8187,
+      "step": 65410
+    },
+    {
+      "epoch": 4.444557684467998,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.222151447207501e-05,
+      "loss": 0.9058,
+      "step": 65415
+    },
+    {
+      "epoch": 4.4448974045386604,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2219391221633375e-05,
+      "loss": 0.9352,
+      "step": 65420
+    },
+    {
+      "epoch": 4.445237124609322,
+      "grad_norm": 7.5,
+      "learning_rate": 2.221726797119174e-05,
+      "loss": 0.9509,
+      "step": 65425
+    },
+    {
+      "epoch": 4.445576844679984,
+      "grad_norm": 7.5,
+      "learning_rate": 2.2215144720750103e-05,
+      "loss": 0.9253,
+      "step": 65430
+    },
+    {
+      "epoch": 4.445916564750646,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2213021470308467e-05,
+      "loss": 0.8803,
+      "step": 65435
+    },
+    {
+      "epoch": 4.446256284821307,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.221089821986683e-05,
+      "loss": 0.9216,
+      "step": 65440
+    },
+    {
+      "epoch": 4.446596004891969,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.2208774969425195e-05,
+      "loss": 0.8418,
+      "step": 65445
+    },
+    {
+      "epoch": 4.446935724962631,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.220665171898356e-05,
+      "loss": 0.9242,
+      "step": 65450
+    },
+    {
+      "epoch": 4.447275445033292,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.220452846854192e-05,
+      "loss": 0.9098,
+      "step": 65455
+    },
+    {
+      "epoch": 4.447615165103954,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2202405218100287e-05,
+      "loss": 0.9486,
+      "step": 65460
+    },
+    {
+      "epoch": 4.4479548851746165,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.220028196765865e-05,
+      "loss": 0.9217,
+      "step": 65465
+    },
+    {
+      "epoch": 4.448294605245278,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.2198158717217012e-05,
+      "loss": 0.8076,
+      "step": 65470
+    },
+    {
+      "epoch": 4.44863432531594,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.219603546677538e-05,
+      "loss": 0.845,
+      "step": 65475
+    },
+    {
+      "epoch": 4.448974045386602,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2193912216333743e-05,
+      "loss": 0.8794,
+      "step": 65480
+    },
+    {
+      "epoch": 4.449313765457263,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2191788965892104e-05,
+      "loss": 0.8287,
+      "step": 65485
+    },
+    {
+      "epoch": 4.449653485527925,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2189665715450468e-05,
+      "loss": 0.9456,
+      "step": 65490
+    },
+    {
+      "epoch": 4.449993205598587,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2187542465008835e-05,
+      "loss": 0.8955,
+      "step": 65495
+    },
+    {
+      "epoch": 4.450332925669248,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.21854192145672e-05,
+      "loss": 0.8273,
+      "step": 65500
+    },
+    {
+      "epoch": 4.45067264573991,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.218329596412556e-05,
+      "loss": 0.8365,
+      "step": 65505
+    },
+    {
+      "epoch": 4.4510123658105725,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2181172713683927e-05,
+      "loss": 0.8993,
+      "step": 65510
+    },
+    {
+      "epoch": 4.451352085881234,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.217904946324229e-05,
+      "loss": 0.9133,
+      "step": 65515
+    },
+    {
+      "epoch": 4.451691805951896,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2176926212800652e-05,
+      "loss": 0.8653,
+      "step": 65520
+    },
+    {
+      "epoch": 4.452031526022558,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.2174802962359016e-05,
+      "loss": 0.9168,
+      "step": 65525
+    },
+    {
+      "epoch": 4.452371246093219,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.2172679711917383e-05,
+      "loss": 0.824,
+      "step": 65530
+    },
+    {
+      "epoch": 4.452710966163881,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.2170556461475744e-05,
+      "loss": 0.7008,
+      "step": 65535
+    },
+    {
+      "epoch": 4.453050686234543,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.2168433211034108e-05,
+      "loss": 0.856,
+      "step": 65540
+    },
+    {
+      "epoch": 4.453390406305204,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.2166309960592475e-05,
+      "loss": 0.9237,
+      "step": 65545
+    },
+    {
+      "epoch": 4.453730126375866,
+      "grad_norm": 7.5,
+      "learning_rate": 2.2164186710150836e-05,
+      "loss": 0.8787,
+      "step": 65550
+    },
+    {
+      "epoch": 4.454069846446528,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.21620634597092e-05,
+      "loss": 0.9151,
+      "step": 65555
+    },
+    {
+      "epoch": 4.45440956651719,
+      "grad_norm": 7.875,
+      "learning_rate": 2.2159940209267564e-05,
+      "loss": 0.8619,
+      "step": 65560
+    },
+    {
+      "epoch": 4.454749286587852,
+      "grad_norm": 8.25,
+      "learning_rate": 2.2157816958825928e-05,
+      "loss": 0.8215,
+      "step": 65565
+    },
+    {
+      "epoch": 4.455089006658513,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2155693708384292e-05,
+      "loss": 0.9017,
+      "step": 65570
+    },
+    {
+      "epoch": 4.455428726729175,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2153570457942656e-05,
+      "loss": 0.9007,
+      "step": 65575
+    },
+    {
+      "epoch": 4.455768446799837,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.215144720750102e-05,
+      "loss": 0.866,
+      "step": 65580
+    },
+    {
+      "epoch": 4.456108166870498,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.2149323957059384e-05,
+      "loss": 0.7797,
+      "step": 65585
+    },
+    {
+      "epoch": 4.45644788694116,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.2147200706617748e-05,
+      "loss": 0.751,
+      "step": 65590
+    },
+    {
+      "epoch": 4.456787607011822,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2145077456176112e-05,
+      "loss": 0.7541,
+      "step": 65595
+    },
+    {
+      "epoch": 4.457127327082484,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.2142954205734476e-05,
+      "loss": 0.8303,
+      "step": 65600
+    },
+    {
+      "epoch": 4.457467047153146,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.214083095529284e-05,
+      "loss": 0.8952,
+      "step": 65605
+    },
+    {
+      "epoch": 4.457806767223808,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.2138707704851204e-05,
+      "loss": 0.9342,
+      "step": 65610
+    },
+    {
+      "epoch": 4.458146487294469,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2136584454409568e-05,
+      "loss": 0.8879,
+      "step": 65615
+    },
+    {
+      "epoch": 4.458486207365131,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2134461203967932e-05,
+      "loss": 0.8865,
+      "step": 65620
+    },
+    {
+      "epoch": 4.458825927435793,
+      "grad_norm": 9.5625,
+      "learning_rate": 2.2132337953526293e-05,
+      "loss": 0.8351,
+      "step": 65625
+    },
+    {
+      "epoch": 4.459165647506454,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.213021470308466e-05,
+      "loss": 0.8246,
+      "step": 65630
+    },
+    {
+      "epoch": 4.459505367577116,
+      "grad_norm": 9.125,
+      "learning_rate": 2.2128091452643024e-05,
+      "loss": 0.8875,
+      "step": 65635
+    },
+    {
+      "epoch": 4.459845087647778,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2125968202201385e-05,
+      "loss": 0.8603,
+      "step": 65640
+    },
+    {
+      "epoch": 4.46018480771844,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2123844951759752e-05,
+      "loss": 0.7763,
+      "step": 65645
+    },
+    {
+      "epoch": 4.460524527789102,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2121721701318116e-05,
+      "loss": 0.8228,
+      "step": 65650
+    },
+    {
+      "epoch": 4.460864247859764,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.2119598450876477e-05,
+      "loss": 0.8884,
+      "step": 65655
+    },
+    {
+      "epoch": 4.461203967930425,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.211747520043484e-05,
+      "loss": 0.8152,
+      "step": 65660
+    },
+    {
+      "epoch": 4.461543688001087,
+      "grad_norm": 8.875,
+      "learning_rate": 2.2115351949993208e-05,
+      "loss": 0.905,
+      "step": 65665
+    },
+    {
+      "epoch": 4.461883408071749,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.2113228699551572e-05,
+      "loss": 0.8401,
+      "step": 65670
+    },
+    {
+      "epoch": 4.46222312814241,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.2111105449109933e-05,
+      "loss": 0.9509,
+      "step": 65675
+    },
+    {
+      "epoch": 4.462562848213072,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.21089821986683e-05,
+      "loss": 0.9026,
+      "step": 65680
+    },
+    {
+      "epoch": 4.462902568283734,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2106858948226664e-05,
+      "loss": 0.8261,
+      "step": 65685
+    },
+    {
+      "epoch": 4.463242288354396,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.2104735697785025e-05,
+      "loss": 0.9324,
+      "step": 65690
+    },
+    {
+      "epoch": 4.463582008425058,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2102612447343392e-05,
+      "loss": 0.8905,
+      "step": 65695
+    },
+    {
+      "epoch": 4.46392172849572,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.2100489196901756e-05,
+      "loss": 0.8277,
+      "step": 65700
+    },
+    {
+      "epoch": 4.464261448566381,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2098365946460117e-05,
+      "loss": 0.9041,
+      "step": 65705
+    },
+    {
+      "epoch": 4.464601168637043,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.209624269601848e-05,
+      "loss": 0.9106,
+      "step": 65710
+    },
+    {
+      "epoch": 4.464940888707705,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.2094119445576848e-05,
+      "loss": 0.8461,
+      "step": 65715
+    },
+    {
+      "epoch": 4.465280608778366,
+      "grad_norm": 7.875,
+      "learning_rate": 2.209199619513521e-05,
+      "loss": 0.82,
+      "step": 65720
+    },
+    {
+      "epoch": 4.465620328849028,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2089872944693573e-05,
+      "loss": 0.8805,
+      "step": 65725
+    },
+    {
+      "epoch": 4.4659600489196905,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.208774969425194e-05,
+      "loss": 0.8696,
+      "step": 65730
+    },
+    {
+      "epoch": 4.466299768990352,
+      "grad_norm": 8.125,
+      "learning_rate": 2.20856264438103e-05,
+      "loss": 0.8483,
+      "step": 65735
+    },
+    {
+      "epoch": 4.466639489061014,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.2083503193368665e-05,
+      "loss": 0.7796,
+      "step": 65740
+    },
+    {
+      "epoch": 4.466979209131676,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.208137994292703e-05,
+      "loss": 0.8793,
+      "step": 65745
+    },
+    {
+      "epoch": 4.467318929202337,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2079256692485393e-05,
+      "loss": 0.7761,
+      "step": 65750
+    },
+    {
+      "epoch": 4.467658649272999,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.2077133442043757e-05,
+      "loss": 0.8747,
+      "step": 65755
+    },
+    {
+      "epoch": 4.467998369343661,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.207501019160212e-05,
+      "loss": 0.8172,
+      "step": 65760
+    },
+    {
+      "epoch": 4.468338089414322,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2072886941160485e-05,
+      "loss": 0.8814,
+      "step": 65765
+    },
+    {
+      "epoch": 4.468677809484984,
+      "grad_norm": 7.5,
+      "learning_rate": 2.207076369071885e-05,
+      "loss": 0.8931,
+      "step": 65770
+    },
+    {
+      "epoch": 4.4690175295556465,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2068640440277213e-05,
+      "loss": 0.855,
+      "step": 65775
+    },
+    {
+      "epoch": 4.469357249626308,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.2066517189835577e-05,
+      "loss": 0.821,
+      "step": 65780
+    },
+    {
+      "epoch": 4.46969696969697,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.206439393939394e-05,
+      "loss": 0.8675,
+      "step": 65785
+    },
+    {
+      "epoch": 4.470036689767632,
+      "grad_norm": 9.625,
+      "learning_rate": 2.2062270688952305e-05,
+      "loss": 0.9227,
+      "step": 65790
+    },
+    {
+      "epoch": 4.470376409838293,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.206014743851067e-05,
+      "loss": 0.8944,
+      "step": 65795
+    },
+    {
+      "epoch": 4.470716129908955,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.2058024188069033e-05,
+      "loss": 0.8992,
+      "step": 65800
+    },
+    {
+      "epoch": 4.471055849979617,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.2055900937627397e-05,
+      "loss": 0.8472,
+      "step": 65805
+    },
+    {
+      "epoch": 4.471395570050278,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2053777687185757e-05,
+      "loss": 0.8222,
+      "step": 65810
+    },
+    {
+      "epoch": 4.47173529012094,
+      "grad_norm": 6.09375,
+      "learning_rate": 2.2051654436744125e-05,
+      "loss": 0.8256,
+      "step": 65815
+    },
+    {
+      "epoch": 4.4720750101916025,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.204953118630249e-05,
+      "loss": 0.9594,
+      "step": 65820
+    },
+    {
+      "epoch": 4.472414730262264,
+      "grad_norm": 8.625,
+      "learning_rate": 2.204740793586085e-05,
+      "loss": 0.8486,
+      "step": 65825
+    },
+    {
+      "epoch": 4.472754450332926,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.2045284685419217e-05,
+      "loss": 0.8347,
+      "step": 65830
+    },
+    {
+      "epoch": 4.473094170403588,
+      "grad_norm": 8.125,
+      "learning_rate": 2.204316143497758e-05,
+      "loss": 0.7759,
+      "step": 65835
+    },
+    {
+      "epoch": 4.473433890474249,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2041038184535945e-05,
+      "loss": 0.8034,
+      "step": 65840
+    },
+    {
+      "epoch": 4.473773610544911,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.2038914934094305e-05,
+      "loss": 0.8976,
+      "step": 65845
+    },
+    {
+      "epoch": 4.474113330615573,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2036791683652673e-05,
+      "loss": 0.8202,
+      "step": 65850
+    },
+    {
+      "epoch": 4.474453050686234,
+      "grad_norm": 9.125,
+      "learning_rate": 2.2034668433211037e-05,
+      "loss": 0.8741,
+      "step": 65855
+    },
+    {
+      "epoch": 4.474792770756896,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.2032545182769397e-05,
+      "loss": 0.7987,
+      "step": 65860
+    },
+    {
+      "epoch": 4.4751324908275585,
+      "grad_norm": 8.75,
+      "learning_rate": 2.2030421932327765e-05,
+      "loss": 0.9534,
+      "step": 65865
+    },
+    {
+      "epoch": 4.47547221089822,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.202829868188613e-05,
+      "loss": 0.8873,
+      "step": 65870
+    },
+    {
+      "epoch": 4.475811930968882,
+      "grad_norm": 7.875,
+      "learning_rate": 2.202617543144449e-05,
+      "loss": 0.9147,
+      "step": 65875
+    },
+    {
+      "epoch": 4.476151651039544,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2024052181002853e-05,
+      "loss": 0.8169,
+      "step": 65880
+    },
+    {
+      "epoch": 4.476491371110205,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.202192893056122e-05,
+      "loss": 0.8115,
+      "step": 65885
+    },
+    {
+      "epoch": 4.476831091180867,
+      "grad_norm": 8.0,
+      "learning_rate": 2.201980568011958e-05,
+      "loss": 0.8725,
+      "step": 65890
+    },
+    {
+      "epoch": 4.477170811251529,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2017682429677945e-05,
+      "loss": 0.8759,
+      "step": 65895
+    },
+    {
+      "epoch": 4.47751053132219,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2015559179236313e-05,
+      "loss": 0.7796,
+      "step": 65900
+    },
+    {
+      "epoch": 4.477850251392852,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2013435928794674e-05,
+      "loss": 0.8188,
+      "step": 65905
+    },
+    {
+      "epoch": 4.4781899714635145,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2011312678353038e-05,
+      "loss": 0.8968,
+      "step": 65910
+    },
+    {
+      "epoch": 4.478529691534176,
+      "grad_norm": 7.5,
+      "learning_rate": 2.20091894279114e-05,
+      "loss": 0.8199,
+      "step": 65915
+    },
+    {
+      "epoch": 4.478869411604838,
+      "grad_norm": 7.125,
+      "learning_rate": 2.2007066177469766e-05,
+      "loss": 0.9126,
+      "step": 65920
+    },
+    {
+      "epoch": 4.4792091316755,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.200494292702813e-05,
+      "loss": 0.9122,
+      "step": 65925
+    },
+    {
+      "epoch": 4.479548851746161,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.2002819676586494e-05,
+      "loss": 0.8569,
+      "step": 65930
+    },
+    {
+      "epoch": 4.479888571816823,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.2000696426144858e-05,
+      "loss": 0.8556,
+      "step": 65935
+    },
+    {
+      "epoch": 4.480228291887485,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.199857317570322e-05,
+      "loss": 0.9254,
+      "step": 65940
+    },
+    {
+      "epoch": 4.480568011958146,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1996449925261586e-05,
+      "loss": 0.8525,
+      "step": 65945
+    },
+    {
+      "epoch": 4.480907732028808,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.199432667481995e-05,
+      "loss": 0.8186,
+      "step": 65950
+    },
+    {
+      "epoch": 4.4812474520994705,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.1992203424378314e-05,
+      "loss": 0.902,
+      "step": 65955
+    },
+    {
+      "epoch": 4.481587172170132,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1990080173936678e-05,
+      "loss": 0.888,
+      "step": 65960
+    },
+    {
+      "epoch": 4.481926892240794,
+      "grad_norm": 7.875,
+      "learning_rate": 2.198795692349504e-05,
+      "loss": 0.9042,
+      "step": 65965
+    },
+    {
+      "epoch": 4.482266612311455,
+      "grad_norm": 7.75,
+      "learning_rate": 2.1985833673053406e-05,
+      "loss": 0.8493,
+      "step": 65970
+    },
+    {
+      "epoch": 4.482606332382117,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.198371042261177e-05,
+      "loss": 0.8333,
+      "step": 65975
+    },
+    {
+      "epoch": 4.482946052452779,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.198158717217013e-05,
+      "loss": 0.8587,
+      "step": 65980
+    },
+    {
+      "epoch": 4.48328577252344,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.1979463921728498e-05,
+      "loss": 0.9144,
+      "step": 65985
+    },
+    {
+      "epoch": 4.483625492594102,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.197734067128686e-05,
+      "loss": 0.904,
+      "step": 65990
+    },
+    {
+      "epoch": 4.483965212664764,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1975217420845222e-05,
+      "loss": 0.8476,
+      "step": 65995
+    },
+    {
+      "epoch": 4.484304932735426,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.197309417040359e-05,
+      "loss": 0.7731,
+      "step": 66000
+    },
+    {
+      "epoch": 4.484644652806088,
+      "grad_norm": 6.875,
+      "learning_rate": 2.1970970919961954e-05,
+      "loss": 0.8516,
+      "step": 66005
+    },
+    {
+      "epoch": 4.48498437287675,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1968847669520318e-05,
+      "loss": 0.7975,
+      "step": 66010
+    },
+    {
+      "epoch": 4.485324092947411,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1966724419078678e-05,
+      "loss": 0.9566,
+      "step": 66015
+    },
+    {
+      "epoch": 4.485663813018073,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1964601168637046e-05,
+      "loss": 0.8767,
+      "step": 66020
+    },
+    {
+      "epoch": 4.486003533088735,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.196247791819541e-05,
+      "loss": 0.8129,
+      "step": 66025
+    },
+    {
+      "epoch": 4.486343253159396,
+      "grad_norm": 7.875,
+      "learning_rate": 2.196035466775377e-05,
+      "loss": 0.8355,
+      "step": 66030
+    },
+    {
+      "epoch": 4.486682973230058,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.1958231417312138e-05,
+      "loss": 0.8879,
+      "step": 66035
+    },
+    {
+      "epoch": 4.4870226933007205,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.19561081668705e-05,
+      "loss": 0.8976,
+      "step": 66040
+    },
+    {
+      "epoch": 4.487362413371382,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1953984916428862e-05,
+      "loss": 0.9125,
+      "step": 66045
+    },
+    {
+      "epoch": 4.487702133442044,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1951861665987226e-05,
+      "loss": 0.882,
+      "step": 66050
+    },
+    {
+      "epoch": 4.488041853512706,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1949738415545594e-05,
+      "loss": 0.8584,
+      "step": 66055
+    },
+    {
+      "epoch": 4.488381573583367,
+      "grad_norm": 6.875,
+      "learning_rate": 2.1947615165103954e-05,
+      "loss": 0.8199,
+      "step": 66060
+    },
+    {
+      "epoch": 4.488721293654029,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1945491914662318e-05,
+      "loss": 0.8315,
+      "step": 66065
+    },
+    {
+      "epoch": 4.489061013724691,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.1943368664220686e-05,
+      "loss": 0.9147,
+      "step": 66070
+    },
+    {
+      "epoch": 4.489400733795352,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1941245413779046e-05,
+      "loss": 0.8927,
+      "step": 66075
+    },
+    {
+      "epoch": 4.489740453866014,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.193912216333741e-05,
+      "loss": 0.8261,
+      "step": 66080
+    },
+    {
+      "epoch": 4.4900801739366765,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1936998912895774e-05,
+      "loss": 0.8248,
+      "step": 66085
+    },
+    {
+      "epoch": 4.490419894007338,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1934875662454138e-05,
+      "loss": 0.9851,
+      "step": 66090
+    },
+    {
+      "epoch": 4.490759614078,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1932752412012502e-05,
+      "loss": 0.7925,
+      "step": 66095
+    },
+    {
+      "epoch": 4.491099334148662,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.1930629161570866e-05,
+      "loss": 0.8438,
+      "step": 66100
+    },
+    {
+      "epoch": 4.491439054219323,
+      "grad_norm": 8.625,
+      "learning_rate": 2.192850591112923e-05,
+      "loss": 0.9036,
+      "step": 66105
+    },
+    {
+      "epoch": 4.491778774289985,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1926382660687594e-05,
+      "loss": 0.7529,
+      "step": 66110
+    },
+    {
+      "epoch": 4.492118494360647,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1924259410245958e-05,
+      "loss": 0.9384,
+      "step": 66115
+    },
+    {
+      "epoch": 4.492458214431308,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.1922136159804322e-05,
+      "loss": 0.8783,
+      "step": 66120
+    },
+    {
+      "epoch": 4.49279793450197,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1920012909362686e-05,
+      "loss": 0.8339,
+      "step": 66125
+    },
+    {
+      "epoch": 4.4931376545726325,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.191788965892105e-05,
+      "loss": 0.8154,
+      "step": 66130
+    },
+    {
+      "epoch": 4.493477374643294,
+      "grad_norm": 9.25,
+      "learning_rate": 2.1915766408479414e-05,
+      "loss": 0.7742,
+      "step": 66135
+    },
+    {
+      "epoch": 4.493817094713956,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.1913643158037778e-05,
+      "loss": 0.908,
+      "step": 66140
+    },
+    {
+      "epoch": 4.494156814784618,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.1911519907596142e-05,
+      "loss": 0.903,
+      "step": 66145
+    },
+    {
+      "epoch": 4.494496534855279,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1909396657154503e-05,
+      "loss": 0.8814,
+      "step": 66150
+    },
+    {
+      "epoch": 4.494836254925941,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.190727340671287e-05,
+      "loss": 0.9511,
+      "step": 66155
+    },
+    {
+      "epoch": 4.495175974996603,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.1905150156271234e-05,
+      "loss": 0.8368,
+      "step": 66160
+    },
+    {
+      "epoch": 4.495515695067264,
+      "grad_norm": 7.25,
+      "learning_rate": 2.1903026905829595e-05,
+      "loss": 0.7998,
+      "step": 66165
+    },
+    {
+      "epoch": 4.495855415137926,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.1900903655387962e-05,
+      "loss": 0.902,
+      "step": 66170
+    },
+    {
+      "epoch": 4.4961951352085885,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1898780404946326e-05,
+      "loss": 0.938,
+      "step": 66175
+    },
+    {
+      "epoch": 4.49653485527925,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.189665715450469e-05,
+      "loss": 0.8659,
+      "step": 66180
+    },
+    {
+      "epoch": 4.496874575349912,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.189453390406305e-05,
+      "loss": 0.8119,
+      "step": 66185
+    },
+    {
+      "epoch": 4.497214295420574,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.189241065362142e-05,
+      "loss": 0.9104,
+      "step": 66190
+    },
+    {
+      "epoch": 4.497554015491235,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1890287403179782e-05,
+      "loss": 0.8236,
+      "step": 66195
+    },
+    {
+      "epoch": 4.497893735561897,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.1888164152738143e-05,
+      "loss": 0.9309,
+      "step": 66200
+    },
+    {
+      "epoch": 4.498233455632559,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.188604090229651e-05,
+      "loss": 0.8234,
+      "step": 66205
+    },
+    {
+      "epoch": 4.49857317570322,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.1883917651854874e-05,
+      "loss": 0.902,
+      "step": 66210
+    },
+    {
+      "epoch": 4.498912895773882,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.1881794401413235e-05,
+      "loss": 0.9589,
+      "step": 66215
+    },
+    {
+      "epoch": 4.4992526158445445,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1879671150971602e-05,
+      "loss": 1.0103,
+      "step": 66220
+    },
+    {
+      "epoch": 4.499592335915206,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.1877547900529966e-05,
+      "loss": 0.8621,
+      "step": 66225
+    },
+    {
+      "epoch": 4.499932055985868,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1875424650088327e-05,
+      "loss": 0.8022,
+      "step": 66230
+    },
+    {
+      "epoch": 4.500271776056529,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.187330139964669e-05,
+      "loss": 0.9026,
+      "step": 66235
+    },
+    {
+      "epoch": 4.500611496127191,
+      "grad_norm": 7.5,
+      "learning_rate": 2.187117814920506e-05,
+      "loss": 0.8199,
+      "step": 66240
+    },
+    {
+      "epoch": 4.500951216197853,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.186905489876342e-05,
+      "loss": 0.7704,
+      "step": 66245
+    },
+    {
+      "epoch": 4.501290936268514,
+      "grad_norm": 6.875,
+      "learning_rate": 2.1866931648321783e-05,
+      "loss": 1.1415,
+      "step": 66250
+    },
+    {
+      "epoch": 4.501630656339176,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.186480839788015e-05,
+      "loss": 0.843,
+      "step": 66255
+    },
+    {
+      "epoch": 4.501970376409838,
+      "grad_norm": 8.625,
+      "learning_rate": 2.186268514743851e-05,
+      "loss": 0.8833,
+      "step": 66260
+    },
+    {
+      "epoch": 4.5023100964805,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1860561896996875e-05,
+      "loss": 0.8425,
+      "step": 66265
+    },
+    {
+      "epoch": 4.502649816551162,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.185843864655524e-05,
+      "loss": 0.9397,
+      "step": 66270
+    },
+    {
+      "epoch": 4.502989536621824,
+      "grad_norm": 9.0,
+      "learning_rate": 2.1856315396113603e-05,
+      "loss": 0.8667,
+      "step": 66275
+    },
+    {
+      "epoch": 4.503329256692485,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1854192145671967e-05,
+      "loss": 0.9485,
+      "step": 66280
+    },
+    {
+      "epoch": 4.503668976763147,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.185206889523033e-05,
+      "loss": 0.872,
+      "step": 66285
+    },
+    {
+      "epoch": 4.504008696833809,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1849945644788695e-05,
+      "loss": 0.9424,
+      "step": 66290
+    },
+    {
+      "epoch": 4.50434841690447,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.184782239434706e-05,
+      "loss": 0.878,
+      "step": 66295
+    },
+    {
+      "epoch": 4.504688136975132,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1845699143905423e-05,
+      "loss": 0.8635,
+      "step": 66300
+    },
+    {
+      "epoch": 4.5050278570457944,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.1843575893463787e-05,
+      "loss": 0.9557,
+      "step": 66305
+    },
+    {
+      "epoch": 4.505367577116456,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.184145264302215e-05,
+      "loss": 0.832,
+      "step": 66310
+    },
+    {
+      "epoch": 4.505707297187118,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.1839329392580515e-05,
+      "loss": 0.8247,
+      "step": 66315
+    },
+    {
+      "epoch": 4.50604701725778,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.183720614213888e-05,
+      "loss": 0.7786,
+      "step": 66320
+    },
+    {
+      "epoch": 4.506386737328441,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1835082891697243e-05,
+      "loss": 1.0091,
+      "step": 66325
+    },
+    {
+      "epoch": 4.506726457399103,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1832959641255607e-05,
+      "loss": 0.7914,
+      "step": 66330
+    },
+    {
+      "epoch": 4.507066177469765,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.1830836390813968e-05,
+      "loss": 0.9033,
+      "step": 66335
+    },
+    {
+      "epoch": 4.507405897540426,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1828713140372335e-05,
+      "loss": 0.7573,
+      "step": 66340
+    },
+    {
+      "epoch": 4.507745617611088,
+      "grad_norm": 8.125,
+      "learning_rate": 2.18265898899307e-05,
+      "loss": 0.8771,
+      "step": 66345
+    },
+    {
+      "epoch": 4.5080853376817505,
+      "grad_norm": 7.875,
+      "learning_rate": 2.182446663948906e-05,
+      "loss": 0.8695,
+      "step": 66350
+    },
+    {
+      "epoch": 4.508425057752412,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1822343389047427e-05,
+      "loss": 0.8413,
+      "step": 66355
+    },
+    {
+      "epoch": 4.508764777823074,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.182022013860579e-05,
+      "loss": 0.9854,
+      "step": 66360
+    },
+    {
+      "epoch": 4.509104497893736,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1818096888164155e-05,
+      "loss": 0.8913,
+      "step": 66365
+    },
+    {
+      "epoch": 4.509444217964397,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1815973637722516e-05,
+      "loss": 0.7795,
+      "step": 66370
+    },
+    {
+      "epoch": 4.509783938035059,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1813850387280883e-05,
+      "loss": 0.7847,
+      "step": 66375
+    },
+    {
+      "epoch": 4.510123658105721,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1811727136839247e-05,
+      "loss": 0.8657,
+      "step": 66380
+    },
+    {
+      "epoch": 4.510463378176382,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.1809603886397608e-05,
+      "loss": 0.8474,
+      "step": 66385
+    },
+    {
+      "epoch": 4.510803098247044,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.1807480635955975e-05,
+      "loss": 0.8722,
+      "step": 66390
+    },
+    {
+      "epoch": 4.5111428183177065,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.180535738551434e-05,
+      "loss": 0.8725,
+      "step": 66395
+    },
+    {
+      "epoch": 4.511482538388368,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.18032341350727e-05,
+      "loss": 0.8263,
+      "step": 66400
+    },
+    {
+      "epoch": 4.51182225845903,
+      "grad_norm": 7.875,
+      "learning_rate": 2.1801110884631064e-05,
+      "loss": 0.9181,
+      "step": 66405
+    },
+    {
+      "epoch": 4.512161978529692,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.179898763418943e-05,
+      "loss": 0.8172,
+      "step": 66410
+    },
+    {
+      "epoch": 4.512501698600353,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.1796864383747792e-05,
+      "loss": 0.901,
+      "step": 66415
+    },
+    {
+      "epoch": 4.512841418671015,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1794741133306156e-05,
+      "loss": 1.151,
+      "step": 66420
+    },
+    {
+      "epoch": 4.513181138741677,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.1792617882864523e-05,
+      "loss": 0.8259,
+      "step": 66425
+    },
+    {
+      "epoch": 4.513520858812338,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1790494632422884e-05,
+      "loss": 0.8776,
+      "step": 66430
+    },
+    {
+      "epoch": 4.513860578883,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1788371381981248e-05,
+      "loss": 0.873,
+      "step": 66435
+    },
+    {
+      "epoch": 4.5142002989536625,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1786248131539612e-05,
+      "loss": 0.9402,
+      "step": 66440
+    },
+    {
+      "epoch": 4.514540019024324,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.1784124881097976e-05,
+      "loss": 0.8283,
+      "step": 66445
+    },
+    {
+      "epoch": 4.514879739094986,
+      "grad_norm": 7.5,
+      "learning_rate": 2.178200163065634e-05,
+      "loss": 0.7982,
+      "step": 66450
+    },
+    {
+      "epoch": 4.515219459165648,
+      "grad_norm": 7.125,
+      "learning_rate": 2.1779878380214704e-05,
+      "loss": 0.8834,
+      "step": 66455
+    },
+    {
+      "epoch": 4.515559179236309,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1777755129773068e-05,
+      "loss": 0.9623,
+      "step": 66460
+    },
+    {
+      "epoch": 4.515898899306971,
+      "grad_norm": 7.25,
+      "learning_rate": 2.1775631879331432e-05,
+      "loss": 0.8872,
+      "step": 66465
+    },
+    {
+      "epoch": 4.516238619377633,
+      "grad_norm": 7.0,
+      "learning_rate": 2.1773508628889796e-05,
+      "loss": 0.9069,
+      "step": 66470
+    },
+    {
+      "epoch": 4.516578339448294,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.177138537844816e-05,
+      "loss": 0.7873,
+      "step": 66475
+    },
+    {
+      "epoch": 4.516918059518956,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1769262128006524e-05,
+      "loss": 0.7957,
+      "step": 66480
+    },
+    {
+      "epoch": 4.5172577795896185,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.1767138877564888e-05,
+      "loss": 0.7873,
+      "step": 66485
+    },
+    {
+      "epoch": 4.51759749966028,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.1765015627123252e-05,
+      "loss": 0.8547,
+      "step": 66490
+    },
+    {
+      "epoch": 4.517937219730942,
+      "grad_norm": 7.875,
+      "learning_rate": 2.1762892376681616e-05,
+      "loss": 0.88,
+      "step": 66495
+    },
+    {
+      "epoch": 4.518276939801604,
+      "grad_norm": 8.875,
+      "learning_rate": 2.176076912623998e-05,
+      "loss": 0.8657,
+      "step": 66500
+    },
+    {
+      "epoch": 4.518616659872265,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.175864587579834e-05,
+      "loss": 0.9388,
+      "step": 66505
+    },
+    {
+      "epoch": 4.518956379942927,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.1756522625356708e-05,
+      "loss": 0.8474,
+      "step": 66510
+    },
+    {
+      "epoch": 4.519296100013589,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.1754399374915072e-05,
+      "loss": 0.8637,
+      "step": 66515
+    },
+    {
+      "epoch": 4.51963582008425,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1752276124473432e-05,
+      "loss": 0.8519,
+      "step": 66520
+    },
+    {
+      "epoch": 4.519975540154912,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.17501528740318e-05,
+      "loss": 0.8678,
+      "step": 66525
+    },
+    {
+      "epoch": 4.5203152602255745,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1748029623590164e-05,
+      "loss": 0.9236,
+      "step": 66530
+    },
+    {
+      "epoch": 4.520654980296236,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.1745906373148528e-05,
+      "loss": 0.8276,
+      "step": 66535
+    },
+    {
+      "epoch": 4.520994700366898,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.174378312270689e-05,
+      "loss": 0.8005,
+      "step": 66540
+    },
+    {
+      "epoch": 4.52133442043756,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1741659872265256e-05,
+      "loss": 0.8097,
+      "step": 66545
+    },
+    {
+      "epoch": 4.521674140508221,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.173953662182362e-05,
+      "loss": 0.8746,
+      "step": 66550
+    },
+    {
+      "epoch": 4.522013860578883,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.173741337138198e-05,
+      "loss": 0.8377,
+      "step": 66555
+    },
+    {
+      "epoch": 4.522353580649545,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.1735290120940348e-05,
+      "loss": 0.797,
+      "step": 66560
+    },
+    {
+      "epoch": 4.522693300720206,
+      "grad_norm": 9.0,
+      "learning_rate": 2.1733166870498712e-05,
+      "loss": 0.9252,
+      "step": 66565
+    },
+    {
+      "epoch": 4.523033020790868,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1731043620057072e-05,
+      "loss": 1.0279,
+      "step": 66570
+    },
+    {
+      "epoch": 4.5233727408615305,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.1728920369615436e-05,
+      "loss": 0.7468,
+      "step": 66575
+    },
+    {
+      "epoch": 4.523712460932192,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1726797119173804e-05,
+      "loss": 0.9392,
+      "step": 66580
+    },
+    {
+      "epoch": 4.524052181002854,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.1724673868732164e-05,
+      "loss": 0.9107,
+      "step": 66585
+    },
+    {
+      "epoch": 4.524391901073516,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.172255061829053e-05,
+      "loss": 0.8898,
+      "step": 66590
+    },
+    {
+      "epoch": 4.524731621144177,
+      "grad_norm": 7.0,
+      "learning_rate": 2.1720427367848896e-05,
+      "loss": 0.8988,
+      "step": 66595
+    },
+    {
+      "epoch": 4.525071341214839,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1718304117407256e-05,
+      "loss": 0.8882,
+      "step": 66600
+    },
+    {
+      "epoch": 4.525411061285501,
+      "grad_norm": 7.375,
+      "learning_rate": 2.171618086696562e-05,
+      "loss": 0.9365,
+      "step": 66605
+    },
+    {
+      "epoch": 4.525750781356162,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1714057616523984e-05,
+      "loss": 0.8499,
+      "step": 66610
+    },
+    {
+      "epoch": 4.5260905014268245,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.171193436608235e-05,
+      "loss": 0.9034,
+      "step": 66615
+    },
+    {
+      "epoch": 4.5264302214974865,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1709811115640713e-05,
+      "loss": 0.8638,
+      "step": 66620
+    },
+    {
+      "epoch": 4.526769941568148,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.1707687865199077e-05,
+      "loss": 0.8983,
+      "step": 66625
+    },
+    {
+      "epoch": 4.52710966163881,
+      "grad_norm": 7.5,
+      "learning_rate": 2.170556461475744e-05,
+      "loss": 0.8697,
+      "step": 66630
+    },
+    {
+      "epoch": 4.527449381709472,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.1703441364315805e-05,
+      "loss": 0.8578,
+      "step": 66635
+    },
+    {
+      "epoch": 4.527789101780133,
+      "grad_norm": 8.5,
+      "learning_rate": 2.170131811387417e-05,
+      "loss": 0.8754,
+      "step": 66640
+    },
+    {
+      "epoch": 4.528128821850795,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.1699194863432533e-05,
+      "loss": 0.8374,
+      "step": 66645
+    },
+    {
+      "epoch": 4.528468541921457,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.1697071612990897e-05,
+      "loss": 0.9088,
+      "step": 66650
+    },
+    {
+      "epoch": 4.528808261992118,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.169494836254926e-05,
+      "loss": 0.8914,
+      "step": 66655
+    },
+    {
+      "epoch": 4.5291479820627805,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1692825112107625e-05,
+      "loss": 0.8838,
+      "step": 66660
+    },
+    {
+      "epoch": 4.5294877021334425,
+      "grad_norm": 8.375,
+      "learning_rate": 2.169070186166599e-05,
+      "loss": 0.7539,
+      "step": 66665
+    },
+    {
+      "epoch": 4.529827422204104,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.1688578611224353e-05,
+      "loss": 0.7998,
+      "step": 66670
+    },
+    {
+      "epoch": 4.530167142274766,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.1686455360782713e-05,
+      "loss": 0.8918,
+      "step": 66675
+    },
+    {
+      "epoch": 4.530506862345427,
+      "grad_norm": 6.09375,
+      "learning_rate": 2.168433211034108e-05,
+      "loss": 0.8966,
+      "step": 66680
+    },
+    {
+      "epoch": 4.530846582416089,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.1682208859899445e-05,
+      "loss": 0.9016,
+      "step": 66685
+    },
+    {
+      "epoch": 4.531186302486751,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1680085609457805e-05,
+      "loss": 0.7693,
+      "step": 66690
+    },
+    {
+      "epoch": 4.531526022557412,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1677962359016173e-05,
+      "loss": 0.9152,
+      "step": 66695
+    },
+    {
+      "epoch": 4.531865742628074,
+      "grad_norm": 7.25,
+      "learning_rate": 2.1675839108574537e-05,
+      "loss": 0.8796,
+      "step": 66700
+    },
+    {
+      "epoch": 4.5322054626987365,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.16737158581329e-05,
+      "loss": 0.8809,
+      "step": 66705
+    },
+    {
+      "epoch": 4.532545182769398,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.167159260769126e-05,
+      "loss": 0.8751,
+      "step": 66710
+    },
+    {
+      "epoch": 4.53288490284006,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.166946935724963e-05,
+      "loss": 0.8732,
+      "step": 66715
+    },
+    {
+      "epoch": 4.533224622910722,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.1667346106807993e-05,
+      "loss": 0.8851,
+      "step": 66720
+    },
+    {
+      "epoch": 4.533564342981383,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1665222856366353e-05,
+      "loss": 0.869,
+      "step": 66725
+    },
+    {
+      "epoch": 4.533904063052045,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.166309960592472e-05,
+      "loss": 0.8492,
+      "step": 66730
+    },
+    {
+      "epoch": 4.534243783122707,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1660976355483085e-05,
+      "loss": 0.8766,
+      "step": 66735
+    },
+    {
+      "epoch": 4.534583503193368,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.1658853105041445e-05,
+      "loss": 0.8674,
+      "step": 66740
+    },
+    {
+      "epoch": 4.53492322326403,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1656729854599813e-05,
+      "loss": 0.8836,
+      "step": 66745
+    },
+    {
+      "epoch": 4.5352629433346925,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1654606604158177e-05,
+      "loss": 0.8889,
+      "step": 66750
+    },
+    {
+      "epoch": 4.535602663405354,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1652483353716537e-05,
+      "loss": 0.8848,
+      "step": 66755
+    },
+    {
+      "epoch": 4.535942383476016,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.16503601032749e-05,
+      "loss": 0.7954,
+      "step": 66760
+    },
+    {
+      "epoch": 4.536282103546678,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.164823685283327e-05,
+      "loss": 0.7909,
+      "step": 66765
+    },
+    {
+      "epoch": 4.536621823617339,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.164611360239163e-05,
+      "loss": 0.8676,
+      "step": 66770
+    },
+    {
+      "epoch": 4.536961543688001,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1643990351949993e-05,
+      "loss": 0.8978,
+      "step": 66775
+    },
+    {
+      "epoch": 4.537301263758663,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.164186710150836e-05,
+      "loss": 0.7993,
+      "step": 66780
+    },
+    {
+      "epoch": 4.537640983829324,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.163974385106672e-05,
+      "loss": 0.8216,
+      "step": 66785
+    },
+    {
+      "epoch": 4.537980703899986,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1637620600625085e-05,
+      "loss": 0.8052,
+      "step": 66790
+    },
+    {
+      "epoch": 4.5383204239706485,
+      "grad_norm": 8.25,
+      "learning_rate": 2.163549735018345e-05,
+      "loss": 0.9023,
+      "step": 66795
+    },
+    {
+      "epoch": 4.53866014404131,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1633374099741813e-05,
+      "loss": 0.8342,
+      "step": 66800
+    },
+    {
+      "epoch": 4.538999864111972,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1631250849300177e-05,
+      "loss": 0.9238,
+      "step": 66805
+    },
+    {
+      "epoch": 4.539339584182634,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.162912759885854e-05,
+      "loss": 0.9146,
+      "step": 66810
+    },
+    {
+      "epoch": 4.539679304253295,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1627004348416905e-05,
+      "loss": 0.9394,
+      "step": 66815
+    },
+    {
+      "epoch": 4.540019024323957,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.162488109797527e-05,
+      "loss": 0.8625,
+      "step": 66820
+    },
+    {
+      "epoch": 4.540358744394619,
+      "grad_norm": 8.75,
+      "learning_rate": 2.1622757847533633e-05,
+      "loss": 0.9437,
+      "step": 66825
+    },
+    {
+      "epoch": 4.54069846446528,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1620634597091997e-05,
+      "loss": 0.8636,
+      "step": 66830
+    },
+    {
+      "epoch": 4.541038184535942,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.161851134665036e-05,
+      "loss": 0.929,
+      "step": 66835
+    },
+    {
+      "epoch": 4.5413779046066045,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1616388096208725e-05,
+      "loss": 0.9429,
+      "step": 66840
+    },
+    {
+      "epoch": 4.541717624677266,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.161426484576709e-05,
+      "loss": 0.835,
+      "step": 66845
+    },
+    {
+      "epoch": 4.542057344747928,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.1612141595325453e-05,
+      "loss": 0.8922,
+      "step": 66850
+    },
+    {
+      "epoch": 4.54239706481859,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.1610018344883817e-05,
+      "loss": 0.798,
+      "step": 66855
+    },
+    {
+      "epoch": 4.542736784889251,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1607895094442178e-05,
+      "loss": 0.8684,
+      "step": 66860
+    },
+    {
+      "epoch": 4.543076504959913,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.1605771844000545e-05,
+      "loss": 0.8789,
+      "step": 66865
+    },
+    {
+      "epoch": 4.543416225030575,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.160364859355891e-05,
+      "loss": 0.9584,
+      "step": 66870
+    },
+    {
+      "epoch": 4.543755945101236,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.1601525343117273e-05,
+      "loss": 0.8306,
+      "step": 66875
+    },
+    {
+      "epoch": 4.5440956651718984,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.1599402092675637e-05,
+      "loss": 0.892,
+      "step": 66880
+    },
+    {
+      "epoch": 4.54443538524256,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1597278842234e-05,
+      "loss": 0.8848,
+      "step": 66885
+    },
+    {
+      "epoch": 4.544775105313222,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.1595155591792365e-05,
+      "loss": 0.8403,
+      "step": 66890
+    },
+    {
+      "epoch": 4.545114825383884,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1593032341350726e-05,
+      "loss": 0.9355,
+      "step": 66895
+    },
+    {
+      "epoch": 4.545454545454545,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.1590909090909093e-05,
+      "loss": 0.8271,
+      "step": 66900
+    },
+    {
+      "epoch": 4.545794265525207,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1588785840467457e-05,
+      "loss": 0.8735,
+      "step": 66905
+    },
+    {
+      "epoch": 4.546133985595869,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1586662590025818e-05,
+      "loss": 0.8105,
+      "step": 66910
+    },
+    {
+      "epoch": 4.54647370566653,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.1584539339584185e-05,
+      "loss": 0.9104,
+      "step": 66915
+    },
+    {
+      "epoch": 4.546813425737192,
+      "grad_norm": 8.25,
+      "learning_rate": 2.158241608914255e-05,
+      "loss": 0.9523,
+      "step": 66920
+    },
+    {
+      "epoch": 4.5471531458078545,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.158029283870091e-05,
+      "loss": 1.0186,
+      "step": 66925
+    },
+    {
+      "epoch": 4.547492865878516,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1578169588259274e-05,
+      "loss": 0.9402,
+      "step": 66930
+    },
+    {
+      "epoch": 4.547832585949178,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.157604633781764e-05,
+      "loss": 0.8935,
+      "step": 66935
+    },
+    {
+      "epoch": 4.54817230601984,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.1573923087376002e-05,
+      "loss": 0.8656,
+      "step": 66940
+    },
+    {
+      "epoch": 4.548512026090501,
+      "grad_norm": 7.125,
+      "learning_rate": 2.1571799836934366e-05,
+      "loss": 0.8348,
+      "step": 66945
+    },
+    {
+      "epoch": 4.548851746161163,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.1569676586492733e-05,
+      "loss": 0.8079,
+      "step": 66950
+    },
+    {
+      "epoch": 4.549191466231825,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1567553336051094e-05,
+      "loss": 0.9017,
+      "step": 66955
+    },
+    {
+      "epoch": 4.549531186302486,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.1565430085609458e-05,
+      "loss": 0.9894,
+      "step": 66960
+    },
+    {
+      "epoch": 4.549870906373148,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1563306835167822e-05,
+      "loss": 0.8884,
+      "step": 66965
+    },
+    {
+      "epoch": 4.5502106264438105,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1561183584726186e-05,
+      "loss": 0.8596,
+      "step": 66970
+    },
+    {
+      "epoch": 4.550550346514472,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.155906033428455e-05,
+      "loss": 0.8614,
+      "step": 66975
+    },
+    {
+      "epoch": 4.550890066585134,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.1556937083842914e-05,
+      "loss": 0.8334,
+      "step": 66980
+    },
+    {
+      "epoch": 4.551229786655796,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.1554813833401278e-05,
+      "loss": 0.8949,
+      "step": 66985
+    },
+    {
+      "epoch": 4.551569506726457,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1552690582959642e-05,
+      "loss": 0.8484,
+      "step": 66990
+    },
+    {
+      "epoch": 4.551909226797119,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.1550567332518006e-05,
+      "loss": 0.8398,
+      "step": 66995
+    },
+    {
+      "epoch": 4.552248946867781,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.154844408207637e-05,
+      "loss": 0.8763,
+      "step": 67000
+    },
+    {
+      "epoch": 4.552588666938442,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.1546320831634734e-05,
+      "loss": 0.7938,
+      "step": 67005
+    },
+    {
+      "epoch": 4.552928387009104,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1544197581193098e-05,
+      "loss": 0.8297,
+      "step": 67010
+    },
+    {
+      "epoch": 4.5532681070797665,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1542074330751462e-05,
+      "loss": 0.7892,
+      "step": 67015
+    },
+    {
+      "epoch": 4.553607827150428,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1539951080309826e-05,
+      "loss": 0.8466,
+      "step": 67020
+    },
+    {
+      "epoch": 4.55394754722109,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.153782782986819e-05,
+      "loss": 0.9001,
+      "step": 67025
+    },
+    {
+      "epoch": 4.554287267291752,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.153570457942655e-05,
+      "loss": 0.8427,
+      "step": 67030
+    },
+    {
+      "epoch": 4.554626987362413,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1533581328984918e-05,
+      "loss": 0.7461,
+      "step": 67035
+    },
+    {
+      "epoch": 4.554966707433075,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1531458078543282e-05,
+      "loss": 0.8771,
+      "step": 67040
+    },
+    {
+      "epoch": 4.555306427503737,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1529334828101646e-05,
+      "loss": 0.8805,
+      "step": 67045
+    },
+    {
+      "epoch": 4.555646147574398,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.152721157766001e-05,
+      "loss": 0.8565,
+      "step": 67050
+    },
+    {
+      "epoch": 4.55598586764506,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1525088327218374e-05,
+      "loss": 0.9254,
+      "step": 67055
+    },
+    {
+      "epoch": 4.5563255877157225,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.1522965076776738e-05,
+      "loss": 0.837,
+      "step": 67060
+    },
+    {
+      "epoch": 4.556665307786384,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.15208418263351e-05,
+      "loss": 0.8651,
+      "step": 67065
+    },
+    {
+      "epoch": 4.557005027857046,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1518718575893466e-05,
+      "loss": 0.8603,
+      "step": 67070
+    },
+    {
+      "epoch": 4.557344747927708,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.151659532545183e-05,
+      "loss": 0.9219,
+      "step": 67075
+    },
+    {
+      "epoch": 4.557684467998369,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.151447207501019e-05,
+      "loss": 0.8082,
+      "step": 67080
+    },
+    {
+      "epoch": 4.558024188069031,
+      "grad_norm": 7.125,
+      "learning_rate": 2.1512348824568558e-05,
+      "loss": 0.9023,
+      "step": 67085
+    },
+    {
+      "epoch": 4.558363908139693,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1510225574126922e-05,
+      "loss": 0.9444,
+      "step": 67090
+    },
+    {
+      "epoch": 4.558703628210354,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.1508102323685283e-05,
+      "loss": 0.8189,
+      "step": 67095
+    },
+    {
+      "epoch": 4.559043348281016,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.1505979073243647e-05,
+      "loss": 0.7873,
+      "step": 67100
+    },
+    {
+      "epoch": 4.5593830683516785,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.1503855822802014e-05,
+      "loss": 0.8058,
+      "step": 67105
+    },
+    {
+      "epoch": 4.55972278842234,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1501732572360375e-05,
+      "loss": 0.8704,
+      "step": 67110
+    },
+    {
+      "epoch": 4.560062508493002,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.149960932191874e-05,
+      "loss": 0.9273,
+      "step": 67115
+    },
+    {
+      "epoch": 4.560402228563664,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.1497486071477106e-05,
+      "loss": 0.9373,
+      "step": 67120
+    },
+    {
+      "epoch": 4.560741948634325,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.1495362821035467e-05,
+      "loss": 0.8757,
+      "step": 67125
+    },
+    {
+      "epoch": 4.561081668704987,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.149323957059383e-05,
+      "loss": 0.8737,
+      "step": 67130
+    },
+    {
+      "epoch": 4.561421388775649,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1491116320152198e-05,
+      "loss": 0.9341,
+      "step": 67135
+    },
+    {
+      "epoch": 4.56176110884631,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.148899306971056e-05,
+      "loss": 0.8634,
+      "step": 67140
+    },
+    {
+      "epoch": 4.562100828916972,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.1486869819268923e-05,
+      "loss": 0.8057,
+      "step": 67145
+    },
+    {
+      "epoch": 4.5624405489876345,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1484746568827287e-05,
+      "loss": 0.8537,
+      "step": 67150
+    },
+    {
+      "epoch": 4.562780269058296,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.148262331838565e-05,
+      "loss": 0.9149,
+      "step": 67155
+    },
+    {
+      "epoch": 4.563119989128958,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1480500067944015e-05,
+      "loss": 0.8825,
+      "step": 67160
+    },
+    {
+      "epoch": 4.56345970919962,
+      "grad_norm": 6.875,
+      "learning_rate": 2.147837681750238e-05,
+      "loss": 0.8541,
+      "step": 67165
+    },
+    {
+      "epoch": 4.563799429270281,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1476253567060743e-05,
+      "loss": 0.8977,
+      "step": 67170
+    },
+    {
+      "epoch": 4.564139149340943,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1474130316619107e-05,
+      "loss": 0.885,
+      "step": 67175
+    },
+    {
+      "epoch": 4.564478869411605,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.147200706617747e-05,
+      "loss": 0.8454,
+      "step": 67180
+    },
+    {
+      "epoch": 4.564818589482266,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.1469883815735835e-05,
+      "loss": 0.8849,
+      "step": 67185
+    },
+    {
+      "epoch": 4.5651583095529285,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.14677605652942e-05,
+      "loss": 0.8481,
+      "step": 67190
+    },
+    {
+      "epoch": 4.5654980296235905,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1465637314852563e-05,
+      "loss": 0.8796,
+      "step": 67195
+    },
+    {
+      "epoch": 4.565837749694252,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.1463514064410923e-05,
+      "loss": 0.783,
+      "step": 67200
+    },
+    {
+      "epoch": 4.566177469764914,
+      "grad_norm": 7.625,
+      "learning_rate": 2.146139081396929e-05,
+      "loss": 0.8661,
+      "step": 67205
+    },
+    {
+      "epoch": 4.566517189835576,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1459267563527655e-05,
+      "loss": 0.8552,
+      "step": 67210
+    },
+    {
+      "epoch": 4.566856909906237,
+      "grad_norm": 7.75,
+      "learning_rate": 2.145714431308602e-05,
+      "loss": 0.8503,
+      "step": 67215
+    },
+    {
+      "epoch": 4.567196629976899,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1455021062644383e-05,
+      "loss": 0.9178,
+      "step": 67220
+    },
+    {
+      "epoch": 4.567536350047561,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1452897812202747e-05,
+      "loss": 0.8865,
+      "step": 67225
+    },
+    {
+      "epoch": 4.567876070118222,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.145077456176111e-05,
+      "loss": 0.8158,
+      "step": 67230
+    },
+    {
+      "epoch": 4.5682157901888845,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.1448651311319475e-05,
+      "loss": 0.8174,
+      "step": 67235
+    },
+    {
+      "epoch": 4.5685555102595465,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.144652806087784e-05,
+      "loss": 0.886,
+      "step": 67240
+    },
+    {
+      "epoch": 4.568895230330208,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1444404810436203e-05,
+      "loss": 0.9309,
+      "step": 67245
+    },
+    {
+      "epoch": 4.56923495040087,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.1442281559994563e-05,
+      "loss": 0.8408,
+      "step": 67250
+    },
+    {
+      "epoch": 4.569574670471532,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.144015830955293e-05,
+      "loss": 0.7962,
+      "step": 67255
+    },
+    {
+      "epoch": 4.569914390542193,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.1438035059111295e-05,
+      "loss": 0.8045,
+      "step": 67260
+    },
+    {
+      "epoch": 4.570254110612855,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.1435911808669655e-05,
+      "loss": 0.856,
+      "step": 67265
+    },
+    {
+      "epoch": 4.570593830683517,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1433788558228023e-05,
+      "loss": 0.9025,
+      "step": 67270
+    },
+    {
+      "epoch": 4.570933550754178,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.1431665307786387e-05,
+      "loss": 0.9273,
+      "step": 67275
+    },
+    {
+      "epoch": 4.5712732708248405,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1429542057344747e-05,
+      "loss": 0.7922,
+      "step": 67280
+    },
+    {
+      "epoch": 4.5716129908955025,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.142741880690311e-05,
+      "loss": 0.8296,
+      "step": 67285
+    },
+    {
+      "epoch": 4.571952710966164,
+      "grad_norm": 7.5,
+      "learning_rate": 2.142529555646148e-05,
+      "loss": 0.8362,
+      "step": 67290
+    },
+    {
+      "epoch": 4.572292431036826,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.142317230601984e-05,
+      "loss": 0.8086,
+      "step": 67295
+    },
+    {
+      "epoch": 4.572632151107488,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1421049055578203e-05,
+      "loss": 0.847,
+      "step": 67300
+    },
+    {
+      "epoch": 4.572971871178149,
+      "grad_norm": 6.5,
+      "learning_rate": 2.141892580513657e-05,
+      "loss": 0.7844,
+      "step": 67305
+    },
+    {
+      "epoch": 4.573311591248811,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.141680255469493e-05,
+      "loss": 0.9138,
+      "step": 67310
+    },
+    {
+      "epoch": 4.573651311319473,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.1414679304253295e-05,
+      "loss": 0.9318,
+      "step": 67315
+    },
+    {
+      "epoch": 4.573991031390134,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.141255605381166e-05,
+      "loss": 0.9116,
+      "step": 67320
+    },
+    {
+      "epoch": 4.5743307514607965,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.1410432803370023e-05,
+      "loss": 0.8465,
+      "step": 67325
+    },
+    {
+      "epoch": 4.574670471531459,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1408309552928387e-05,
+      "loss": 0.8485,
+      "step": 67330
+    },
+    {
+      "epoch": 4.57501019160212,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.140618630248675e-05,
+      "loss": 0.8121,
+      "step": 67335
+    },
+    {
+      "epoch": 4.575349911672782,
+      "grad_norm": 7.0,
+      "learning_rate": 2.1404063052045116e-05,
+      "loss": 0.8143,
+      "step": 67340
+    },
+    {
+      "epoch": 4.575689631743444,
+      "grad_norm": 7.375,
+      "learning_rate": 2.140193980160348e-05,
+      "loss": 0.8606,
+      "step": 67345
+    },
+    {
+      "epoch": 4.576029351814105,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1399816551161844e-05,
+      "loss": 0.9102,
+      "step": 67350
+    },
+    {
+      "epoch": 4.576369071884767,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.1397693300720208e-05,
+      "loss": 0.8264,
+      "step": 67355
+    },
+    {
+      "epoch": 4.576708791955428,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.139557005027857e-05,
+      "loss": 0.8521,
+      "step": 67360
+    },
+    {
+      "epoch": 4.57704851202609,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1393446799836936e-05,
+      "loss": 0.8721,
+      "step": 67365
+    },
+    {
+      "epoch": 4.5773882320967525,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.13913235493953e-05,
+      "loss": 0.8927,
+      "step": 67370
+    },
+    {
+      "epoch": 4.577727952167414,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1389200298953664e-05,
+      "loss": 0.8405,
+      "step": 67375
+    },
+    {
+      "epoch": 4.578067672238076,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1387077048512028e-05,
+      "loss": 0.8972,
+      "step": 67380
+    },
+    {
+      "epoch": 4.578407392308738,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.138495379807039e-05,
+      "loss": 0.9056,
+      "step": 67385
+    },
+    {
+      "epoch": 4.578747112379399,
+      "grad_norm": 7.75,
+      "learning_rate": 2.1382830547628756e-05,
+      "loss": 0.8796,
+      "step": 67390
+    },
+    {
+      "epoch": 4.579086832450061,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.138070729718712e-05,
+      "loss": 0.8542,
+      "step": 67395
+    },
+    {
+      "epoch": 4.579426552520723,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.1378584046745484e-05,
+      "loss": 0.9463,
+      "step": 67400
+    },
+    {
+      "epoch": 4.579766272591384,
+      "grad_norm": 7.375,
+      "learning_rate": 2.1376460796303848e-05,
+      "loss": 0.8713,
+      "step": 67405
+    },
+    {
+      "epoch": 4.580105992662046,
+      "grad_norm": 8.375,
+      "learning_rate": 2.137433754586221e-05,
+      "loss": 0.9197,
+      "step": 67410
+    },
+    {
+      "epoch": 4.5804457127327085,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.1372214295420576e-05,
+      "loss": 0.8564,
+      "step": 67415
+    },
+    {
+      "epoch": 4.58078543280337,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.1370091044978936e-05,
+      "loss": 0.8101,
+      "step": 67420
+    },
+    {
+      "epoch": 4.581125152874032,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1367967794537304e-05,
+      "loss": 1.0198,
+      "step": 67425
+    },
+    {
+      "epoch": 4.581464872944694,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1365844544095668e-05,
+      "loss": 0.7931,
+      "step": 67430
+    },
+    {
+      "epoch": 4.581804593015355,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1363721293654028e-05,
+      "loss": 0.85,
+      "step": 67435
+    },
+    {
+      "epoch": 4.582144313086017,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1361598043212396e-05,
+      "loss": 0.7191,
+      "step": 67440
+    },
+    {
+      "epoch": 4.582484033156679,
+      "grad_norm": 7.625,
+      "learning_rate": 2.135947479277076e-05,
+      "loss": 0.8586,
+      "step": 67445
+    },
+    {
+      "epoch": 4.58282375322734,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.135735154232912e-05,
+      "loss": 0.8597,
+      "step": 67450
+    },
+    {
+      "epoch": 4.583163473298002,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.1355228291887484e-05,
+      "loss": 0.7416,
+      "step": 67455
+    },
+    {
+      "epoch": 4.5835031933686645,
+      "grad_norm": 8.125,
+      "learning_rate": 2.135310504144585e-05,
+      "loss": 0.9006,
+      "step": 67460
+    },
+    {
+      "epoch": 4.583842913439326,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.1350981791004212e-05,
+      "loss": 0.8248,
+      "step": 67465
+    },
+    {
+      "epoch": 4.584182633509988,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1348858540562576e-05,
+      "loss": 0.8941,
+      "step": 67470
+    },
+    {
+      "epoch": 4.58452235358065,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1346735290120944e-05,
+      "loss": 0.7922,
+      "step": 67475
+    },
+    {
+      "epoch": 4.584862073651311,
+      "grad_norm": 7.125,
+      "learning_rate": 2.1344612039679304e-05,
+      "loss": 0.7971,
+      "step": 67480
+    },
+    {
+      "epoch": 4.585201793721973,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.1342488789237668e-05,
+      "loss": 0.8577,
+      "step": 67485
+    },
+    {
+      "epoch": 4.585541513792635,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1340365538796032e-05,
+      "loss": 0.838,
+      "step": 67490
+    },
+    {
+      "epoch": 4.585881233863296,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.1338242288354396e-05,
+      "loss": 0.8505,
+      "step": 67495
+    },
+    {
+      "epoch": 4.5862209539339585,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.133611903791276e-05,
+      "loss": 0.8243,
+      "step": 67500
+    },
+    {
+      "epoch": 4.5865606740046205,
+      "grad_norm": 7.25,
+      "learning_rate": 2.1333995787471124e-05,
+      "loss": 0.8327,
+      "step": 67505
+    },
+    {
+      "epoch": 4.586900394075282,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1331872537029488e-05,
+      "loss": 0.8948,
+      "step": 67510
+    },
+    {
+      "epoch": 4.587240114145944,
+      "grad_norm": 8.875,
+      "learning_rate": 2.1329749286587852e-05,
+      "loss": 0.8675,
+      "step": 67515
+    },
+    {
+      "epoch": 4.587579834216606,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1327626036146216e-05,
+      "loss": 0.8764,
+      "step": 67520
+    },
+    {
+      "epoch": 4.587919554287267,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.132550278570458e-05,
+      "loss": 0.8381,
+      "step": 67525
+    },
+    {
+      "epoch": 4.588259274357929,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1323379535262944e-05,
+      "loss": 0.8431,
+      "step": 67530
+    },
+    {
+      "epoch": 4.588598994428591,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1321256284821308e-05,
+      "loss": 0.8449,
+      "step": 67535
+    },
+    {
+      "epoch": 4.588938714499252,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1319133034379672e-05,
+      "loss": 0.7771,
+      "step": 67540
+    },
+    {
+      "epoch": 4.5892784345699145,
+      "grad_norm": 8.875,
+      "learning_rate": 2.1317009783938036e-05,
+      "loss": 0.8936,
+      "step": 67545
+    },
+    {
+      "epoch": 4.5896181546405765,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.13148865334964e-05,
+      "loss": 0.8228,
+      "step": 67550
+    },
+    {
+      "epoch": 4.589957874711238,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1312763283054764e-05,
+      "loss": 0.7851,
+      "step": 67555
+    },
+    {
+      "epoch": 4.5902975947819,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1310640032613128e-05,
+      "loss": 0.937,
+      "step": 67560
+    },
+    {
+      "epoch": 4.590637314852561,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.1308516782171492e-05,
+      "loss": 0.8556,
+      "step": 67565
+    },
+    {
+      "epoch": 4.590977034923223,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.1306393531729856e-05,
+      "loss": 0.8066,
+      "step": 67570
+    },
+    {
+      "epoch": 4.591316754993885,
+      "grad_norm": 8.0,
+      "learning_rate": 2.130427028128822e-05,
+      "loss": 0.9525,
+      "step": 67575
+    },
+    {
+      "epoch": 4.591656475064546,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1302147030846584e-05,
+      "loss": 0.766,
+      "step": 67580
+    },
+    {
+      "epoch": 4.591996195135208,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.130002378040495e-05,
+      "loss": 0.8506,
+      "step": 67585
+    },
+    {
+      "epoch": 4.5923359152058705,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.129790052996331e-05,
+      "loss": 0.8496,
+      "step": 67590
+    },
+    {
+      "epoch": 4.592675635276532,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1295777279521676e-05,
+      "loss": 0.8355,
+      "step": 67595
+    },
+    {
+      "epoch": 4.593015355347194,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.129365402908004e-05,
+      "loss": 0.808,
+      "step": 67600
+    },
+    {
+      "epoch": 4.593355075417856,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.12915307786384e-05,
+      "loss": 0.9428,
+      "step": 67605
+    },
+    {
+      "epoch": 4.593694795488517,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.128940752819677e-05,
+      "loss": 0.9437,
+      "step": 67610
+    },
+    {
+      "epoch": 4.594034515559179,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1287284277755132e-05,
+      "loss": 0.8811,
+      "step": 67615
+    },
+    {
+      "epoch": 4.594374235629841,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1285161027313493e-05,
+      "loss": 0.8321,
+      "step": 67620
+    },
+    {
+      "epoch": 4.594713955700502,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.1283037776871857e-05,
+      "loss": 0.8682,
+      "step": 67625
+    },
+    {
+      "epoch": 4.595053675771164,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.1280914526430224e-05,
+      "loss": 0.8354,
+      "step": 67630
+    },
+    {
+      "epoch": 4.5953933958418265,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1278791275988585e-05,
+      "loss": 0.9416,
+      "step": 67635
+    },
+    {
+      "epoch": 4.595733115912488,
+      "grad_norm": 8.0,
+      "learning_rate": 2.127666802554695e-05,
+      "loss": 0.8282,
+      "step": 67640
+    },
+    {
+      "epoch": 4.59607283598315,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.1274544775105316e-05,
+      "loss": 0.9471,
+      "step": 67645
+    },
+    {
+      "epoch": 4.596412556053812,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.1272421524663677e-05,
+      "loss": 0.8329,
+      "step": 67650
+    },
+    {
+      "epoch": 4.596752276124473,
+      "grad_norm": 7.875,
+      "learning_rate": 2.127029827422204e-05,
+      "loss": 0.8565,
+      "step": 67655
+    },
+    {
+      "epoch": 4.597091996195135,
+      "grad_norm": 8.125,
+      "learning_rate": 2.126817502378041e-05,
+      "loss": 0.8233,
+      "step": 67660
+    },
+    {
+      "epoch": 4.597431716265797,
+      "grad_norm": 7.25,
+      "learning_rate": 2.126605177333877e-05,
+      "loss": 0.8423,
+      "step": 67665
+    },
+    {
+      "epoch": 4.597771436336458,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1263928522897133e-05,
+      "loss": 0.9099,
+      "step": 67670
+    },
+    {
+      "epoch": 4.59811115640712,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1261805272455497e-05,
+      "loss": 0.8727,
+      "step": 67675
+    },
+    {
+      "epoch": 4.5984508764777825,
+      "grad_norm": 7.625,
+      "learning_rate": 2.125968202201386e-05,
+      "loss": 0.8131,
+      "step": 67680
+    },
+    {
+      "epoch": 4.598790596548444,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1257558771572225e-05,
+      "loss": 0.7821,
+      "step": 67685
+    },
+    {
+      "epoch": 4.599130316619106,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.125543552113059e-05,
+      "loss": 0.9725,
+      "step": 67690
+    },
+    {
+      "epoch": 4.599470036689768,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.1253312270688953e-05,
+      "loss": 0.7792,
+      "step": 67695
+    },
+    {
+      "epoch": 4.599809756760429,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1251189020247317e-05,
+      "loss": 0.8603,
+      "step": 67700
+    },
+    {
+      "epoch": 4.600149476831091,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.124906576980568e-05,
+      "loss": 0.8986,
+      "step": 67705
+    },
+    {
+      "epoch": 4.600489196901753,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.1246942519364045e-05,
+      "loss": 0.8399,
+      "step": 67710
+    },
+    {
+      "epoch": 4.600828916972414,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.124481926892241e-05,
+      "loss": 0.8341,
+      "step": 67715
+    },
+    {
+      "epoch": 4.601168637043076,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1242696018480773e-05,
+      "loss": 0.8627,
+      "step": 67720
+    },
+    {
+      "epoch": 4.6015083571137385,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1240572768039137e-05,
+      "loss": 0.7818,
+      "step": 67725
+    },
+    {
+      "epoch": 4.6018480771844,
+      "grad_norm": 7.125,
+      "learning_rate": 2.12384495175975e-05,
+      "loss": 0.8353,
+      "step": 67730
+    },
+    {
+      "epoch": 4.602187797255062,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.1236326267155865e-05,
+      "loss": 0.9067,
+      "step": 67735
+    },
+    {
+      "epoch": 4.602527517325724,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.123420301671423e-05,
+      "loss": 1.0085,
+      "step": 67740
+    },
+    {
+      "epoch": 4.602867237396385,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1232079766272593e-05,
+      "loss": 0.9378,
+      "step": 67745
+    },
+    {
+      "epoch": 4.603206957467047,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1229956515830957e-05,
+      "loss": 0.8688,
+      "step": 67750
+    },
+    {
+      "epoch": 4.603546677537709,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.122783326538932e-05,
+      "loss": 0.799,
+      "step": 67755
+    },
+    {
+      "epoch": 4.60388639760837,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1225710014947685e-05,
+      "loss": 0.8968,
+      "step": 67760
+    },
+    {
+      "epoch": 4.6042261176790324,
+      "grad_norm": 7.5,
+      "learning_rate": 2.122358676450605e-05,
+      "loss": 0.8587,
+      "step": 67765
+    },
+    {
+      "epoch": 4.6045658377496945,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.1221463514064413e-05,
+      "loss": 0.875,
+      "step": 67770
+    },
+    {
+      "epoch": 4.604905557820356,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1219340263622774e-05,
+      "loss": 0.8631,
+      "step": 67775
+    },
+    {
+      "epoch": 4.605245277891018,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.121721701318114e-05,
+      "loss": 0.9166,
+      "step": 67780
+    },
+    {
+      "epoch": 4.60558499796168,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1215093762739505e-05,
+      "loss": 0.9195,
+      "step": 67785
+    },
+    {
+      "epoch": 4.605924718032341,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.1212970512297866e-05,
+      "loss": 0.8552,
+      "step": 67790
+    },
+    {
+      "epoch": 4.606264438103003,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.1210847261856233e-05,
+      "loss": 0.9127,
+      "step": 67795
+    },
+    {
+      "epoch": 4.606604158173665,
+      "grad_norm": 9.875,
+      "learning_rate": 2.1208724011414597e-05,
+      "loss": 0.8672,
+      "step": 67800
+    },
+    {
+      "epoch": 4.606943878244326,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.1206600760972958e-05,
+      "loss": 0.9392,
+      "step": 67805
+    },
+    {
+      "epoch": 4.6072835983149885,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1204477510531322e-05,
+      "loss": 0.9048,
+      "step": 67810
+    },
+    {
+      "epoch": 4.6076233183856505,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.120235426008969e-05,
+      "loss": 0.8259,
+      "step": 67815
+    },
+    {
+      "epoch": 4.607963038456312,
+      "grad_norm": 7.75,
+      "learning_rate": 2.120023100964805e-05,
+      "loss": 0.7826,
+      "step": 67820
+    },
+    {
+      "epoch": 4.608302758526974,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.1198107759206414e-05,
+      "loss": 0.9102,
+      "step": 67825
+    },
+    {
+      "epoch": 4.608642478597636,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.119598450876478e-05,
+      "loss": 0.9588,
+      "step": 67830
+    },
+    {
+      "epoch": 4.608982198668297,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1193861258323142e-05,
+      "loss": 0.871,
+      "step": 67835
+    },
+    {
+      "epoch": 4.609321918738959,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1191738007881506e-05,
+      "loss": 0.8235,
+      "step": 67840
+    },
+    {
+      "epoch": 4.609661638809621,
+      "grad_norm": 8.375,
+      "learning_rate": 2.118961475743987e-05,
+      "loss": 0.9523,
+      "step": 67845
+    },
+    {
+      "epoch": 4.610001358880282,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.1187491506998234e-05,
+      "loss": 0.9145,
+      "step": 67850
+    },
+    {
+      "epoch": 4.6103410789509445,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1185368256556598e-05,
+      "loss": 0.9061,
+      "step": 67855
+    },
+    {
+      "epoch": 4.6106807990216065,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.1183245006114962e-05,
+      "loss": 0.8619,
+      "step": 67860
+    },
+    {
+      "epoch": 4.611020519092268,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.1181121755673326e-05,
+      "loss": 0.8111,
+      "step": 67865
+    },
+    {
+      "epoch": 4.61136023916293,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.117899850523169e-05,
+      "loss": 0.7752,
+      "step": 67870
+    },
+    {
+      "epoch": 4.611699959233592,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1176875254790054e-05,
+      "loss": 0.975,
+      "step": 67875
+    },
+    {
+      "epoch": 4.612039679304253,
+      "grad_norm": 7.5,
+      "learning_rate": 2.1174752004348418e-05,
+      "loss": 0.8753,
+      "step": 67880
+    },
+    {
+      "epoch": 4.612379399374915,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1172628753906782e-05,
+      "loss": 0.8915,
+      "step": 67885
+    },
+    {
+      "epoch": 4.612719119445577,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1170505503465146e-05,
+      "loss": 0.8942,
+      "step": 67890
+    },
+    {
+      "epoch": 4.613058839516238,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.116838225302351e-05,
+      "loss": 0.8551,
+      "step": 67895
+    },
+    {
+      "epoch": 4.6133985595869005,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1166259002581874e-05,
+      "loss": 0.9381,
+      "step": 67900
+    },
+    {
+      "epoch": 4.613738279657563,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1164135752140238e-05,
+      "loss": 0.9675,
+      "step": 67905
+    },
+    {
+      "epoch": 4.614077999728224,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.1162012501698602e-05,
+      "loss": 0.8108,
+      "step": 67910
+    },
+    {
+      "epoch": 4.614417719798886,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.1159889251256966e-05,
+      "loss": 0.9092,
+      "step": 67915
+    },
+    {
+      "epoch": 4.614757439869548,
+      "grad_norm": 8.125,
+      "learning_rate": 2.115776600081533e-05,
+      "loss": 0.8549,
+      "step": 67920
+    },
+    {
+      "epoch": 4.615097159940209,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1155642750373694e-05,
+      "loss": 0.9398,
+      "step": 67925
+    },
+    {
+      "epoch": 4.615436880010871,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.1153519499932058e-05,
+      "loss": 0.9529,
+      "step": 67930
+    },
+    {
+      "epoch": 4.615776600081533,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1151396249490422e-05,
+      "loss": 0.9684,
+      "step": 67935
+    },
+    {
+      "epoch": 4.616116320152194,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1149272999048786e-05,
+      "loss": 0.8025,
+      "step": 67940
+    },
+    {
+      "epoch": 4.6164560402228565,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.1147149748607146e-05,
+      "loss": 0.9525,
+      "step": 67945
+    },
+    {
+      "epoch": 4.616795760293519,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1145026498165514e-05,
+      "loss": 0.8911,
+      "step": 67950
+    },
+    {
+      "epoch": 4.61713548036418,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.1142903247723878e-05,
+      "loss": 0.8677,
+      "step": 67955
+    },
+    {
+      "epoch": 4.617475200434842,
+      "grad_norm": 7.5,
+      "learning_rate": 2.114077999728224e-05,
+      "loss": 0.8708,
+      "step": 67960
+    },
+    {
+      "epoch": 4.617814920505504,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.1138656746840606e-05,
+      "loss": 0.9385,
+      "step": 67965
+    },
+    {
+      "epoch": 4.618154640576165,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.113653349639897e-05,
+      "loss": 0.8115,
+      "step": 67970
+    },
+    {
+      "epoch": 4.618494360646827,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.113441024595733e-05,
+      "loss": 0.8653,
+      "step": 67975
+    },
+    {
+      "epoch": 4.618834080717489,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1132286995515694e-05,
+      "loss": 0.927,
+      "step": 67980
+    },
+    {
+      "epoch": 4.61917380078815,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.1130163745074062e-05,
+      "loss": 0.8937,
+      "step": 67985
+    },
+    {
+      "epoch": 4.6195135208588125,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1128040494632422e-05,
+      "loss": 0.8292,
+      "step": 67990
+    },
+    {
+      "epoch": 4.619853240929475,
+      "grad_norm": 9.875,
+      "learning_rate": 2.1125917244190786e-05,
+      "loss": 0.8882,
+      "step": 67995
+    },
+    {
+      "epoch": 4.620192961000136,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1123793993749154e-05,
+      "loss": 0.8506,
+      "step": 68000
+    },
+    {
+      "epoch": 4.620532681070798,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.1121670743307514e-05,
+      "loss": 0.9437,
+      "step": 68005
+    },
+    {
+      "epoch": 4.62087240114146,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.111954749286588e-05,
+      "loss": 0.8621,
+      "step": 68010
+    },
+    {
+      "epoch": 4.621212121212121,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1117424242424242e-05,
+      "loss": 0.9113,
+      "step": 68015
+    },
+    {
+      "epoch": 4.621551841282783,
+      "grad_norm": 7.0,
+      "learning_rate": 2.1115300991982606e-05,
+      "loss": 0.7786,
+      "step": 68020
+    },
+    {
+      "epoch": 4.621891561353445,
+      "grad_norm": 7.75,
+      "learning_rate": 2.111317774154097e-05,
+      "loss": 0.8519,
+      "step": 68025
+    },
+    {
+      "epoch": 4.622231281424106,
+      "grad_norm": 7.5,
+      "learning_rate": 2.1111054491099334e-05,
+      "loss": 0.8287,
+      "step": 68030
+    },
+    {
+      "epoch": 4.6225710014947685,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.11089312406577e-05,
+      "loss": 0.8197,
+      "step": 68035
+    },
+    {
+      "epoch": 4.62291072156543,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1106807990216062e-05,
+      "loss": 0.8811,
+      "step": 68040
+    },
+    {
+      "epoch": 4.623250441636092,
+      "grad_norm": 8.75,
+      "learning_rate": 2.1104684739774426e-05,
+      "loss": 0.8959,
+      "step": 68045
+    },
+    {
+      "epoch": 4.623590161706754,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.110256148933279e-05,
+      "loss": 0.9619,
+      "step": 68050
+    },
+    {
+      "epoch": 4.623929881777415,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.1100438238891155e-05,
+      "loss": 0.9026,
+      "step": 68055
+    },
+    {
+      "epoch": 4.624269601848077,
+      "grad_norm": 8.5,
+      "learning_rate": 2.109831498844952e-05,
+      "loss": 0.8085,
+      "step": 68060
+    },
+    {
+      "epoch": 4.624609321918739,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1096191738007883e-05,
+      "loss": 0.8749,
+      "step": 68065
+    },
+    {
+      "epoch": 4.6249490419894,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1094068487566247e-05,
+      "loss": 0.946,
+      "step": 68070
+    },
+    {
+      "epoch": 4.6252887620600625,
+      "grad_norm": 6.5,
+      "learning_rate": 2.109194523712461e-05,
+      "loss": 0.9664,
+      "step": 68075
+    },
+    {
+      "epoch": 4.6256284821307245,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.1089821986682975e-05,
+      "loss": 0.8742,
+      "step": 68080
+    },
+    {
+      "epoch": 4.625968202201386,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.108769873624134e-05,
+      "loss": 0.8686,
+      "step": 68085
+    },
+    {
+      "epoch": 4.626307922272048,
+      "grad_norm": 7.625,
+      "learning_rate": 2.1085575485799703e-05,
+      "loss": 0.8534,
+      "step": 68090
+    },
+    {
+      "epoch": 4.62664764234271,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.1083452235358067e-05,
+      "loss": 0.9261,
+      "step": 68095
+    },
+    {
+      "epoch": 4.626987362413371,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.108132898491643e-05,
+      "loss": 0.9432,
+      "step": 68100
+    },
+    {
+      "epoch": 4.627327082484033,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1079205734474795e-05,
+      "loss": 0.9154,
+      "step": 68105
+    },
+    {
+      "epoch": 4.627666802554695,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.107708248403316e-05,
+      "loss": 0.8322,
+      "step": 68110
+    },
+    {
+      "epoch": 4.628006522625356,
+      "grad_norm": 8.25,
+      "learning_rate": 2.107495923359152e-05,
+      "loss": 0.9519,
+      "step": 68115
+    },
+    {
+      "epoch": 4.6283462426960185,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1072835983149887e-05,
+      "loss": 0.895,
+      "step": 68120
+    },
+    {
+      "epoch": 4.6286859627666805,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.107071273270825e-05,
+      "loss": 0.9112,
+      "step": 68125
+    },
+    {
+      "epoch": 4.629025682837342,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.106858948226661e-05,
+      "loss": 0.8804,
+      "step": 68130
+    },
+    {
+      "epoch": 4.629365402908004,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.106646623182498e-05,
+      "loss": 0.9721,
+      "step": 68135
+    },
+    {
+      "epoch": 4.629705122978666,
+      "grad_norm": 7.125,
+      "learning_rate": 2.1064342981383343e-05,
+      "loss": 0.8422,
+      "step": 68140
+    },
+    {
+      "epoch": 4.630044843049327,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.1062219730941703e-05,
+      "loss": 0.8798,
+      "step": 68145
+    },
+    {
+      "epoch": 4.630384563119989,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.106009648050007e-05,
+      "loss": 0.8039,
+      "step": 68150
+    },
+    {
+      "epoch": 4.630724283190651,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.1057973230058435e-05,
+      "loss": 0.7736,
+      "step": 68155
+    },
+    {
+      "epoch": 4.631064003261312,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.1055849979616795e-05,
+      "loss": 0.9463,
+      "step": 68160
+    },
+    {
+      "epoch": 4.6314037233319745,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.105372672917516e-05,
+      "loss": 0.8319,
+      "step": 68165
+    },
+    {
+      "epoch": 4.6317434434026366,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.1051603478733527e-05,
+      "loss": 0.8408,
+      "step": 68170
+    },
+    {
+      "epoch": 4.632083163473298,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.1049480228291887e-05,
+      "loss": 0.7954,
+      "step": 68175
+    },
+    {
+      "epoch": 4.63242288354396,
+      "grad_norm": 8.25,
+      "learning_rate": 2.104735697785025e-05,
+      "loss": 0.9064,
+      "step": 68180
+    },
+    {
+      "epoch": 4.632762603614622,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.104523372740862e-05,
+      "loss": 0.8515,
+      "step": 68185
+    },
+    {
+      "epoch": 4.633102323685283,
+      "grad_norm": 8.5,
+      "learning_rate": 2.104311047696698e-05,
+      "loss": 0.8885,
+      "step": 68190
+    },
+    {
+      "epoch": 4.633442043755945,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.1040987226525343e-05,
+      "loss": 0.942,
+      "step": 68195
+    },
+    {
+      "epoch": 4.633781763826607,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.1038863976083707e-05,
+      "loss": 0.8252,
+      "step": 68200
+    },
+    {
+      "epoch": 4.634121483897268,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.103674072564207e-05,
+      "loss": 0.8629,
+      "step": 68205
+    },
+    {
+      "epoch": 4.6344612039679305,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.1034617475200435e-05,
+      "loss": 0.8331,
+      "step": 68210
+    },
+    {
+      "epoch": 4.634800924038593,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.10324942247588e-05,
+      "loss": 0.9752,
+      "step": 68215
+    },
+    {
+      "epoch": 4.635140644109254,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1030370974317163e-05,
+      "loss": 0.8422,
+      "step": 68220
+    },
+    {
+      "epoch": 4.635480364179916,
+      "grad_norm": 7.875,
+      "learning_rate": 2.1028247723875527e-05,
+      "loss": 0.8404,
+      "step": 68225
+    },
+    {
+      "epoch": 4.635820084250578,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.102612447343389e-05,
+      "loss": 0.9,
+      "step": 68230
+    },
+    {
+      "epoch": 4.636159804321239,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.1024001222992255e-05,
+      "loss": 0.7597,
+      "step": 68235
+    },
+    {
+      "epoch": 4.636499524391901,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.102187797255062e-05,
+      "loss": 0.9042,
+      "step": 68240
+    },
+    {
+      "epoch": 4.636839244462563,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1019754722108983e-05,
+      "loss": 0.8782,
+      "step": 68245
+    },
+    {
+      "epoch": 4.637178964533224,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1017631471667347e-05,
+      "loss": 0.8122,
+      "step": 68250
+    },
+    {
+      "epoch": 4.6375186846038865,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.101550822122571e-05,
+      "loss": 0.8462,
+      "step": 68255
+    },
+    {
+      "epoch": 4.637858404674548,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.1013384970784075e-05,
+      "loss": 0.8511,
+      "step": 68260
+    },
+    {
+      "epoch": 4.63819812474521,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.101126172034244e-05,
+      "loss": 0.8489,
+      "step": 68265
+    },
+    {
+      "epoch": 4.638537844815872,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.1009138469900803e-05,
+      "loss": 0.8393,
+      "step": 68270
+    },
+    {
+      "epoch": 4.638877564886533,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1007015219459167e-05,
+      "loss": 0.845,
+      "step": 68275
+    },
+    {
+      "epoch": 4.639217284957195,
+      "grad_norm": 7.875,
+      "learning_rate": 2.100489196901753e-05,
+      "loss": 0.9218,
+      "step": 68280
+    },
+    {
+      "epoch": 4.639557005027857,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.1002768718575895e-05,
+      "loss": 0.769,
+      "step": 68285
+    },
+    {
+      "epoch": 4.639896725098518,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.100064546813426e-05,
+      "loss": 0.9969,
+      "step": 68290
+    },
+    {
+      "epoch": 4.64023644516918,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.0998522217692623e-05,
+      "loss": 0.8832,
+      "step": 68295
+    },
+    {
+      "epoch": 4.6405761652398425,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0996398967250984e-05,
+      "loss": 0.817,
+      "step": 68300
+    },
+    {
+      "epoch": 4.640915885310504,
+      "grad_norm": 5.5,
+      "learning_rate": 2.099427571680935e-05,
+      "loss": 0.8013,
+      "step": 68305
+    },
+    {
+      "epoch": 4.641255605381166,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0992152466367715e-05,
+      "loss": 0.8818,
+      "step": 68310
+    },
+    {
+      "epoch": 4.641595325451828,
+      "grad_norm": 10.0625,
+      "learning_rate": 2.0990029215926076e-05,
+      "loss": 0.7844,
+      "step": 68315
+    },
+    {
+      "epoch": 4.641935045522489,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.0987905965484443e-05,
+      "loss": 0.9581,
+      "step": 68320
+    },
+    {
+      "epoch": 4.642274765593151,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0985782715042807e-05,
+      "loss": 0.8634,
+      "step": 68325
+    },
+    {
+      "epoch": 4.642614485663813,
+      "grad_norm": 7.125,
+      "learning_rate": 2.0983659464601168e-05,
+      "loss": 0.7901,
+      "step": 68330
+    },
+    {
+      "epoch": 4.642954205734474,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0981536214159532e-05,
+      "loss": 0.8744,
+      "step": 68335
+    },
+    {
+      "epoch": 4.6432939258051364,
+      "grad_norm": 7.0,
+      "learning_rate": 2.09794129637179e-05,
+      "loss": 0.8128,
+      "step": 68340
+    },
+    {
+      "epoch": 4.6436336458757985,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.097728971327626e-05,
+      "loss": 0.7542,
+      "step": 68345
+    },
+    {
+      "epoch": 4.64397336594646,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0975166462834624e-05,
+      "loss": 0.7688,
+      "step": 68350
+    },
+    {
+      "epoch": 4.644313086017122,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.097304321239299e-05,
+      "loss": 0.8639,
+      "step": 68355
+    },
+    {
+      "epoch": 4.644652806087784,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0970919961951352e-05,
+      "loss": 0.8295,
+      "step": 68360
+    },
+    {
+      "epoch": 4.644992526158445,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.0968796711509716e-05,
+      "loss": 0.853,
+      "step": 68365
+    },
+    {
+      "epoch": 4.645332246229107,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.096667346106808e-05,
+      "loss": 0.9479,
+      "step": 68370
+    },
+    {
+      "epoch": 4.645671966299769,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.0964550210626444e-05,
+      "loss": 0.8761,
+      "step": 68375
+    },
+    {
+      "epoch": 4.64601168637043,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0962426960184808e-05,
+      "loss": 0.8904,
+      "step": 68380
+    },
+    {
+      "epoch": 4.6463514064410925,
+      "grad_norm": 7.0,
+      "learning_rate": 2.0960303709743172e-05,
+      "loss": 0.8886,
+      "step": 68385
+    },
+    {
+      "epoch": 4.6466911265117545,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0958180459301536e-05,
+      "loss": 0.9097,
+      "step": 68390
+    },
+    {
+      "epoch": 4.647030846582416,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.09560572088599e-05,
+      "loss": 0.9237,
+      "step": 68395
+    },
+    {
+      "epoch": 4.647370566653078,
+      "grad_norm": 8.75,
+      "learning_rate": 2.0953933958418264e-05,
+      "loss": 0.8942,
+      "step": 68400
+    },
+    {
+      "epoch": 4.64771028672374,
+      "grad_norm": 6.25,
+      "learning_rate": 2.0951810707976628e-05,
+      "loss": 0.8857,
+      "step": 68405
+    },
+    {
+      "epoch": 4.648050006794401,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0949687457534992e-05,
+      "loss": 0.8787,
+      "step": 68410
+    },
+    {
+      "epoch": 4.648389726865063,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0947564207093356e-05,
+      "loss": 0.9779,
+      "step": 68415
+    },
+    {
+      "epoch": 4.648729446935725,
+      "grad_norm": 7.875,
+      "learning_rate": 2.094544095665172e-05,
+      "loss": 0.8695,
+      "step": 68420
+    },
+    {
+      "epoch": 4.649069167006386,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0943317706210084e-05,
+      "loss": 0.9572,
+      "step": 68425
+    },
+    {
+      "epoch": 4.6494088870770485,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0941194455768448e-05,
+      "loss": 0.8874,
+      "step": 68430
+    },
+    {
+      "epoch": 4.6497486071477105,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0939071205326812e-05,
+      "loss": 0.8676,
+      "step": 68435
+    },
+    {
+      "epoch": 4.650088327218372,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0936947954885176e-05,
+      "loss": 0.9206,
+      "step": 68440
+    },
+    {
+      "epoch": 4.650428047289034,
+      "grad_norm": 7.125,
+      "learning_rate": 2.093482470444354e-05,
+      "loss": 0.9401,
+      "step": 68445
+    },
+    {
+      "epoch": 4.650767767359696,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0932701454001904e-05,
+      "loss": 0.8558,
+      "step": 68450
+    },
+    {
+      "epoch": 4.651107487430357,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0930578203560268e-05,
+      "loss": 0.8519,
+      "step": 68455
+    },
+    {
+      "epoch": 4.651447207501019,
+      "grad_norm": 6.875,
+      "learning_rate": 2.0928454953118632e-05,
+      "loss": 0.9171,
+      "step": 68460
+    },
+    {
+      "epoch": 4.651786927571681,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.0926331702676996e-05,
+      "loss": 0.7842,
+      "step": 68465
+    },
+    {
+      "epoch": 4.652126647642342,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.0924208452235357e-05,
+      "loss": 0.8524,
+      "step": 68470
+    },
+    {
+      "epoch": 4.6524663677130045,
+      "grad_norm": 7.5,
+      "learning_rate": 2.0922085201793724e-05,
+      "loss": 0.8882,
+      "step": 68475
+    },
+    {
+      "epoch": 4.6528060877836666,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0919961951352088e-05,
+      "loss": 0.846,
+      "step": 68480
+    },
+    {
+      "epoch": 4.653145807854328,
+      "grad_norm": 8.0,
+      "learning_rate": 2.091783870091045e-05,
+      "loss": 0.9283,
+      "step": 68485
+    },
+    {
+      "epoch": 4.65348552792499,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0915715450468816e-05,
+      "loss": 0.8986,
+      "step": 68490
+    },
+    {
+      "epoch": 4.653825247995652,
+      "grad_norm": 7.75,
+      "learning_rate": 2.091359220002718e-05,
+      "loss": 0.8577,
+      "step": 68495
+    },
+    {
+      "epoch": 4.654164968066313,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.091146894958554e-05,
+      "loss": 0.9143,
+      "step": 68500
+    },
+    {
+      "epoch": 4.654504688136975,
+      "grad_norm": 7.125,
+      "learning_rate": 2.0909345699143905e-05,
+      "loss": 0.856,
+      "step": 68505
+    },
+    {
+      "epoch": 4.654844408207637,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0907222448702272e-05,
+      "loss": 0.8603,
+      "step": 68510
+    },
+    {
+      "epoch": 4.655184128278298,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0905099198260633e-05,
+      "loss": 0.922,
+      "step": 68515
+    },
+    {
+      "epoch": 4.6555238483489605,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0902975947818997e-05,
+      "loss": 0.8444,
+      "step": 68520
+    },
+    {
+      "epoch": 4.655863568419623,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0900852697377364e-05,
+      "loss": 0.8962,
+      "step": 68525
+    },
+    {
+      "epoch": 4.656203288490284,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0898729446935725e-05,
+      "loss": 0.7998,
+      "step": 68530
+    },
+    {
+      "epoch": 4.656543008560946,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.089660619649409e-05,
+      "loss": 0.8452,
+      "step": 68535
+    },
+    {
+      "epoch": 4.656882728631608,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0894482946052453e-05,
+      "loss": 0.8413,
+      "step": 68540
+    },
+    {
+      "epoch": 4.657222448702269,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0892359695610817e-05,
+      "loss": 0.809,
+      "step": 68545
+    },
+    {
+      "epoch": 4.657562168772931,
+      "grad_norm": 8.75,
+      "learning_rate": 2.089023644516918e-05,
+      "loss": 0.8636,
+      "step": 68550
+    },
+    {
+      "epoch": 4.657901888843593,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0888113194727545e-05,
+      "loss": 0.8952,
+      "step": 68555
+    },
+    {
+      "epoch": 4.658241608914254,
+      "grad_norm": 6.75,
+      "learning_rate": 2.088598994428591e-05,
+      "loss": 0.8875,
+      "step": 68560
+    },
+    {
+      "epoch": 4.6585813289849165,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0883866693844273e-05,
+      "loss": 0.8443,
+      "step": 68565
+    },
+    {
+      "epoch": 4.658921049055579,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0881743443402637e-05,
+      "loss": 0.8285,
+      "step": 68570
+    },
+    {
+      "epoch": 4.65926076912624,
+      "grad_norm": 8.875,
+      "learning_rate": 2.0879620192961004e-05,
+      "loss": 0.8462,
+      "step": 68575
+    },
+    {
+      "epoch": 4.659600489196902,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0877496942519365e-05,
+      "loss": 0.9016,
+      "step": 68580
+    },
+    {
+      "epoch": 4.659940209267564,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.087537369207773e-05,
+      "loss": 0.8847,
+      "step": 68585
+    },
+    {
+      "epoch": 4.660279929338225,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0873250441636093e-05,
+      "loss": 0.9031,
+      "step": 68590
+    },
+    {
+      "epoch": 4.660619649408887,
+      "grad_norm": 7.125,
+      "learning_rate": 2.0871127191194457e-05,
+      "loss": 0.8179,
+      "step": 68595
+    },
+    {
+      "epoch": 4.660959369479549,
+      "grad_norm": 7.75,
+      "learning_rate": 2.086900394075282e-05,
+      "loss": 0.701,
+      "step": 68600
+    },
+    {
+      "epoch": 4.66129908955021,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0866880690311185e-05,
+      "loss": 0.8521,
+      "step": 68605
+    },
+    {
+      "epoch": 4.6616388096208725,
+      "grad_norm": 7.125,
+      "learning_rate": 2.086475743986955e-05,
+      "loss": 1.1135,
+      "step": 68610
+    },
+    {
+      "epoch": 4.661978529691535,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0862634189427913e-05,
+      "loss": 0.9077,
+      "step": 68615
+    },
+    {
+      "epoch": 4.662318249762196,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.0860510938986277e-05,
+      "loss": 0.8474,
+      "step": 68620
+    },
+    {
+      "epoch": 4.662657969832858,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.085838768854464e-05,
+      "loss": 0.8314,
+      "step": 68625
+    },
+    {
+      "epoch": 4.66299768990352,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0856264438103005e-05,
+      "loss": 0.8734,
+      "step": 68630
+    },
+    {
+      "epoch": 4.663337409974181,
+      "grad_norm": 7.625,
+      "learning_rate": 2.085414118766137e-05,
+      "loss": 0.8101,
+      "step": 68635
+    },
+    {
+      "epoch": 4.663677130044843,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.085201793721973e-05,
+      "loss": 0.7836,
+      "step": 68640
+    },
+    {
+      "epoch": 4.664016850115505,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0849894686778097e-05,
+      "loss": 0.8992,
+      "step": 68645
+    },
+    {
+      "epoch": 4.6643565701861665,
+      "grad_norm": 8.0,
+      "learning_rate": 2.084777143633646e-05,
+      "loss": 0.8705,
+      "step": 68650
+    },
+    {
+      "epoch": 4.6646962902568285,
+      "grad_norm": 8.0,
+      "learning_rate": 2.084564818589482e-05,
+      "loss": 0.8519,
+      "step": 68655
+    },
+    {
+      "epoch": 4.665036010327491,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.084352493545319e-05,
+      "loss": 0.862,
+      "step": 68660
+    },
+    {
+      "epoch": 4.665375730398152,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0841401685011553e-05,
+      "loss": 0.902,
+      "step": 68665
+    },
+    {
+      "epoch": 4.665715450468814,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0839278434569913e-05,
+      "loss": 0.7339,
+      "step": 68670
+    },
+    {
+      "epoch": 4.666055170539476,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.083715518412828e-05,
+      "loss": 0.8916,
+      "step": 68675
+    },
+    {
+      "epoch": 4.666394890610137,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.0835031933686645e-05,
+      "loss": 0.856,
+      "step": 68680
+    },
+    {
+      "epoch": 4.666734610680799,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0832908683245005e-05,
+      "loss": 0.8429,
+      "step": 68685
+    },
+    {
+      "epoch": 4.667074330751461,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.083078543280337e-05,
+      "loss": 0.8061,
+      "step": 68690
+    },
+    {
+      "epoch": 4.6674140508221225,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0828662182361737e-05,
+      "loss": 0.8191,
+      "step": 68695
+    },
+    {
+      "epoch": 4.6677537708927845,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0826538931920097e-05,
+      "loss": 0.9112,
+      "step": 68700
+    },
+    {
+      "epoch": 4.668093490963447,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.082441568147846e-05,
+      "loss": 0.8652,
+      "step": 68705
+    },
+    {
+      "epoch": 4.668433211034108,
+      "grad_norm": 7.625,
+      "learning_rate": 2.082229243103683e-05,
+      "loss": 0.8783,
+      "step": 68710
+    },
+    {
+      "epoch": 4.66877293110477,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.082016918059519e-05,
+      "loss": 0.9237,
+      "step": 68715
+    },
+    {
+      "epoch": 4.669112651175431,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.0818045930153553e-05,
+      "loss": 0.8084,
+      "step": 68720
+    },
+    {
+      "epoch": 4.669452371246093,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0815922679711917e-05,
+      "loss": 0.8393,
+      "step": 68725
+    },
+    {
+      "epoch": 4.669792091316755,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.081379942927028e-05,
+      "loss": 0.842,
+      "step": 68730
+    },
+    {
+      "epoch": 4.670131811387416,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.0811676178828645e-05,
+      "loss": 0.7819,
+      "step": 68735
+    },
+    {
+      "epoch": 4.6704715314580785,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.080955292838701e-05,
+      "loss": 0.9426,
+      "step": 68740
+    },
+    {
+      "epoch": 4.6708112515287405,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0807429677945377e-05,
+      "loss": 0.9133,
+      "step": 68745
+    },
+    {
+      "epoch": 4.671150971599402,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.0805306427503737e-05,
+      "loss": 0.8851,
+      "step": 68750
+    },
+    {
+      "epoch": 4.671490691670064,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.08031831770621e-05,
+      "loss": 0.9715,
+      "step": 68755
+    },
+    {
+      "epoch": 4.671830411740726,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0801059926620465e-05,
+      "loss": 0.8575,
+      "step": 68760
+    },
+    {
+      "epoch": 4.672170131811387,
+      "grad_norm": 8.25,
+      "learning_rate": 2.079893667617883e-05,
+      "loss": 0.8632,
+      "step": 68765
+    },
+    {
+      "epoch": 4.672509851882049,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0796813425737194e-05,
+      "loss": 0.7597,
+      "step": 68770
+    },
+    {
+      "epoch": 4.672849571952711,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.0794690175295558e-05,
+      "loss": 0.9323,
+      "step": 68775
+    },
+    {
+      "epoch": 4.673189292023372,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.079256692485392e-05,
+      "loss": 0.9533,
+      "step": 68780
+    },
+    {
+      "epoch": 4.6735290120940345,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0790443674412286e-05,
+      "loss": 0.857,
+      "step": 68785
+    },
+    {
+      "epoch": 4.673868732164697,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.078832042397065e-05,
+      "loss": 0.9361,
+      "step": 68790
+    },
+    {
+      "epoch": 4.674208452235358,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0786197173529014e-05,
+      "loss": 0.8533,
+      "step": 68795
+    },
+    {
+      "epoch": 4.67454817230602,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0784073923087378e-05,
+      "loss": 0.9012,
+      "step": 68800
+    },
+    {
+      "epoch": 4.674887892376682,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.078195067264574e-05,
+      "loss": 0.9495,
+      "step": 68805
+    },
+    {
+      "epoch": 4.675227612447343,
+      "grad_norm": 7.5,
+      "learning_rate": 2.0779827422204106e-05,
+      "loss": 0.8854,
+      "step": 68810
+    },
+    {
+      "epoch": 4.675567332518005,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.077770417176247e-05,
+      "loss": 0.8423,
+      "step": 68815
+    },
+    {
+      "epoch": 4.675907052588667,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.0775580921320834e-05,
+      "loss": 0.884,
+      "step": 68820
+    },
+    {
+      "epoch": 4.676246772659328,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0773457670879194e-05,
+      "loss": 0.7794,
+      "step": 68825
+    },
+    {
+      "epoch": 4.6765864927299905,
+      "grad_norm": 8.375,
+      "learning_rate": 2.077133442043756e-05,
+      "loss": 0.8828,
+      "step": 68830
+    },
+    {
+      "epoch": 4.676926212800653,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0769211169995926e-05,
+      "loss": 0.9627,
+      "step": 68835
+    },
+    {
+      "epoch": 4.677265932871314,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0767087919554286e-05,
+      "loss": 0.8174,
+      "step": 68840
+    },
+    {
+      "epoch": 4.677605652941976,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0764964669112654e-05,
+      "loss": 0.806,
+      "step": 68845
+    },
+    {
+      "epoch": 4.677945373012638,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0762841418671018e-05,
+      "loss": 0.8278,
+      "step": 68850
+    },
+    {
+      "epoch": 4.678285093083299,
+      "grad_norm": 8.75,
+      "learning_rate": 2.0760718168229378e-05,
+      "loss": 0.9275,
+      "step": 68855
+    },
+    {
+      "epoch": 4.678624813153961,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0758594917787742e-05,
+      "loss": 0.8654,
+      "step": 68860
+    },
+    {
+      "epoch": 4.678964533224623,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.075647166734611e-05,
+      "loss": 0.7694,
+      "step": 68865
+    },
+    {
+      "epoch": 4.679304253295284,
+      "grad_norm": 7.75,
+      "learning_rate": 2.075434841690447e-05,
+      "loss": 0.8952,
+      "step": 68870
+    },
+    {
+      "epoch": 4.6796439733659465,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0752225166462834e-05,
+      "loss": 0.9501,
+      "step": 68875
+    },
+    {
+      "epoch": 4.679983693436609,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.07501019160212e-05,
+      "loss": 0.8896,
+      "step": 68880
+    },
+    {
+      "epoch": 4.68032341350727,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0747978665579562e-05,
+      "loss": 0.8411,
+      "step": 68885
+    },
+    {
+      "epoch": 4.680663133577932,
+      "grad_norm": 6.75,
+      "learning_rate": 2.0745855415137926e-05,
+      "loss": 0.898,
+      "step": 68890
+    },
+    {
+      "epoch": 4.681002853648594,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.074373216469629e-05,
+      "loss": 0.9251,
+      "step": 68895
+    },
+    {
+      "epoch": 4.681342573719255,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0741608914254654e-05,
+      "loss": 0.9558,
+      "step": 68900
+    },
+    {
+      "epoch": 4.681682293789917,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0739485663813018e-05,
+      "loss": 0.9695,
+      "step": 68905
+    },
+    {
+      "epoch": 4.682022013860579,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0737362413371382e-05,
+      "loss": 0.9255,
+      "step": 68910
+    },
+    {
+      "epoch": 4.68236173393124,
+      "grad_norm": 8.5,
+      "learning_rate": 2.073523916292975e-05,
+      "loss": 0.9395,
+      "step": 68915
+    },
+    {
+      "epoch": 4.6827014540019025,
+      "grad_norm": 8.75,
+      "learning_rate": 2.073311591248811e-05,
+      "loss": 0.9403,
+      "step": 68920
+    },
+    {
+      "epoch": 4.683041174072565,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0730992662046474e-05,
+      "loss": 0.9343,
+      "step": 68925
+    },
+    {
+      "epoch": 4.683380894143226,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.0728869411604838e-05,
+      "loss": 0.822,
+      "step": 68930
+    },
+    {
+      "epoch": 4.683720614213888,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.0726746161163202e-05,
+      "loss": 0.92,
+      "step": 68935
+    },
+    {
+      "epoch": 4.684060334284549,
+      "grad_norm": 7.5,
+      "learning_rate": 2.0724622910721566e-05,
+      "loss": 0.973,
+      "step": 68940
+    },
+    {
+      "epoch": 4.684400054355211,
+      "grad_norm": 7.625,
+      "learning_rate": 2.072249966027993e-05,
+      "loss": 0.8562,
+      "step": 68945
+    },
+    {
+      "epoch": 4.684739774425873,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0720376409838294e-05,
+      "loss": 0.8465,
+      "step": 68950
+    },
+    {
+      "epoch": 4.685079494496534,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.0718253159396658e-05,
+      "loss": 0.9824,
+      "step": 68955
+    },
+    {
+      "epoch": 4.6854192145671965,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0716129908955022e-05,
+      "loss": 0.8078,
+      "step": 68960
+    },
+    {
+      "epoch": 4.6857589346378585,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.0714006658513386e-05,
+      "loss": 0.8778,
+      "step": 68965
+    },
+    {
+      "epoch": 4.68609865470852,
+      "grad_norm": 8.75,
+      "learning_rate": 2.071188340807175e-05,
+      "loss": 0.7753,
+      "step": 68970
+    },
+    {
+      "epoch": 4.686438374779182,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0709760157630114e-05,
+      "loss": 0.8177,
+      "step": 68975
+    },
+    {
+      "epoch": 4.686778094849844,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0707636907188478e-05,
+      "loss": 0.9184,
+      "step": 68980
+    },
+    {
+      "epoch": 4.687117814920505,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0705513656746842e-05,
+      "loss": 0.8347,
+      "step": 68985
+    },
+    {
+      "epoch": 4.687457534991167,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0703390406305206e-05,
+      "loss": 0.8835,
+      "step": 68990
+    },
+    {
+      "epoch": 4.687797255061829,
+      "grad_norm": 5.96875,
+      "learning_rate": 2.0701267155863567e-05,
+      "loss": 0.9809,
+      "step": 68995
+    },
+    {
+      "epoch": 4.68813697513249,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.0699143905421934e-05,
+      "loss": 0.897,
+      "step": 69000
+    },
+    {
+      "epoch": 4.6884766952031525,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0697020654980298e-05,
+      "loss": 0.8784,
+      "step": 69005
+    },
+    {
+      "epoch": 4.6888164152738145,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.069489740453866e-05,
+      "loss": 0.8494,
+      "step": 69010
+    },
+    {
+      "epoch": 4.689156135344476,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0692774154097026e-05,
+      "loss": 0.899,
+      "step": 69015
+    },
+    {
+      "epoch": 4.689495855415138,
+      "grad_norm": 6.46875,
+      "learning_rate": 2.069065090365539e-05,
+      "loss": 0.8725,
+      "step": 69020
+    },
+    {
+      "epoch": 4.6898355754858,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.068852765321375e-05,
+      "loss": 0.7822,
+      "step": 69025
+    },
+    {
+      "epoch": 4.690175295556461,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0686404402772115e-05,
+      "loss": 0.9054,
+      "step": 69030
+    },
+    {
+      "epoch": 4.690515015627123,
+      "grad_norm": 7.25,
+      "learning_rate": 2.0684281152330482e-05,
+      "loss": 0.9168,
+      "step": 69035
+    },
+    {
+      "epoch": 4.690854735697785,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0682157901888843e-05,
+      "loss": 0.8794,
+      "step": 69040
+    },
+    {
+      "epoch": 4.691194455768446,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0680034651447207e-05,
+      "loss": 0.7735,
+      "step": 69045
+    },
+    {
+      "epoch": 4.6915341758391085,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0677911401005574e-05,
+      "loss": 0.8417,
+      "step": 69050
+    },
+    {
+      "epoch": 4.6918738959097706,
+      "grad_norm": 7.375,
+      "learning_rate": 2.0675788150563935e-05,
+      "loss": 0.8151,
+      "step": 69055
+    },
+    {
+      "epoch": 4.692213615980432,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.06736649001223e-05,
+      "loss": 0.876,
+      "step": 69060
+    },
+    {
+      "epoch": 4.692553336051094,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0671541649680663e-05,
+      "loss": 0.8842,
+      "step": 69065
+    },
+    {
+      "epoch": 4.692893056121756,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0669418399239027e-05,
+      "loss": 0.9684,
+      "step": 69070
+    },
+    {
+      "epoch": 4.693232776192417,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.066729514879739e-05,
+      "loss": 0.9392,
+      "step": 69075
+    },
+    {
+      "epoch": 4.693572496263079,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0665171898355755e-05,
+      "loss": 1.055,
+      "step": 69080
+    },
+    {
+      "epoch": 4.693912216333741,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0663048647914122e-05,
+      "loss": 0.9013,
+      "step": 69085
+    },
+    {
+      "epoch": 4.694251936404402,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0660925397472483e-05,
+      "loss": 0.8634,
+      "step": 69090
+    },
+    {
+      "epoch": 4.6945916564750645,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0658802147030847e-05,
+      "loss": 0.8943,
+      "step": 69095
+    },
+    {
+      "epoch": 4.694931376545727,
+      "grad_norm": 5.71875,
+      "learning_rate": 2.0656678896589214e-05,
+      "loss": 0.9118,
+      "step": 69100
+    },
+    {
+      "epoch": 4.695271096616388,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0654555646147575e-05,
+      "loss": 0.951,
+      "step": 69105
+    },
+    {
+      "epoch": 4.69561081668705,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.065243239570594e-05,
+      "loss": 0.9061,
+      "step": 69110
+    },
+    {
+      "epoch": 4.695950536757712,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0650309145264303e-05,
+      "loss": 0.9204,
+      "step": 69115
+    },
+    {
+      "epoch": 4.696290256828373,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0648185894822667e-05,
+      "loss": 0.8697,
+      "step": 69120
+    },
+    {
+      "epoch": 4.696629976899035,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.064606264438103e-05,
+      "loss": 0.757,
+      "step": 69125
+    },
+    {
+      "epoch": 4.696969696969697,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0643939393939395e-05,
+      "loss": 0.8827,
+      "step": 69130
+    },
+    {
+      "epoch": 4.697309417040358,
+      "grad_norm": 8.25,
+      "learning_rate": 2.064181614349776e-05,
+      "loss": 0.9237,
+      "step": 69135
+    },
+    {
+      "epoch": 4.6976491371110205,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.0639692893056123e-05,
+      "loss": 0.8258,
+      "step": 69140
+    },
+    {
+      "epoch": 4.697988857181683,
+      "grad_norm": 7.5,
+      "learning_rate": 2.0637569642614487e-05,
+      "loss": 0.9316,
+      "step": 69145
+    },
+    {
+      "epoch": 4.698328577252344,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.063544639217285e-05,
+      "loss": 0.9259,
+      "step": 69150
+    },
+    {
+      "epoch": 4.698668297323006,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0633323141731215e-05,
+      "loss": 0.9278,
+      "step": 69155
+    },
+    {
+      "epoch": 4.699008017393668,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.063119989128958e-05,
+      "loss": 0.9077,
+      "step": 69160
+    },
+    {
+      "epoch": 4.699347737464329,
+      "grad_norm": 7.625,
+      "learning_rate": 2.062907664084794e-05,
+      "loss": 0.8128,
+      "step": 69165
+    },
+    {
+      "epoch": 4.699687457534991,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0626953390406307e-05,
+      "loss": 0.9503,
+      "step": 69170
+    },
+    {
+      "epoch": 4.700027177605653,
+      "grad_norm": 8.0,
+      "learning_rate": 2.062483013996467e-05,
+      "loss": 0.8981,
+      "step": 69175
+    },
+    {
+      "epoch": 4.700366897676314,
+      "grad_norm": 8.375,
+      "learning_rate": 2.062270688952303e-05,
+      "loss": 0.8595,
+      "step": 69180
+    },
+    {
+      "epoch": 4.7007066177469765,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.06205836390814e-05,
+      "loss": 0.8926,
+      "step": 69185
+    },
+    {
+      "epoch": 4.701046337817639,
+      "grad_norm": 8.625,
+      "learning_rate": 2.0618460388639763e-05,
+      "loss": 0.7738,
+      "step": 69190
+    },
+    {
+      "epoch": 4.7013860578883,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0616337138198124e-05,
+      "loss": 0.9289,
+      "step": 69195
+    },
+    {
+      "epoch": 4.701725777958962,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.061421388775649e-05,
+      "loss": 0.8589,
+      "step": 69200
+    },
+    {
+      "epoch": 4.702065498029624,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0612090637314855e-05,
+      "loss": 0.9604,
+      "step": 69205
+    },
+    {
+      "epoch": 4.702405218100285,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0609967386873216e-05,
+      "loss": 0.8597,
+      "step": 69210
+    },
+    {
+      "epoch": 4.702744938170947,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.060784413643158e-05,
+      "loss": 0.7996,
+      "step": 69215
+    },
+    {
+      "epoch": 4.703084658241609,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.0605720885989947e-05,
+      "loss": 0.853,
+      "step": 69220
+    },
+    {
+      "epoch": 4.7034243783122704,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0603597635548308e-05,
+      "loss": 0.874,
+      "step": 69225
+    },
+    {
+      "epoch": 4.7037640983829325,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.0601474385106672e-05,
+      "loss": 0.9112,
+      "step": 69230
+    },
+    {
+      "epoch": 4.704103818453595,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.059935113466504e-05,
+      "loss": 0.9117,
+      "step": 69235
+    },
+    {
+      "epoch": 4.704443538524256,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.05972278842234e-05,
+      "loss": 0.8761,
+      "step": 69240
+    },
+    {
+      "epoch": 4.704783258594918,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0595104633781764e-05,
+      "loss": 0.8487,
+      "step": 69245
+    },
+    {
+      "epoch": 4.70512297866558,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0592981383340128e-05,
+      "loss": 0.8959,
+      "step": 69250
+    },
+    {
+      "epoch": 4.705462698736241,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0590858132898495e-05,
+      "loss": 0.835,
+      "step": 69255
+    },
+    {
+      "epoch": 4.705802418806903,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0588734882456856e-05,
+      "loss": 1.0017,
+      "step": 69260
+    },
+    {
+      "epoch": 4.706142138877565,
+      "grad_norm": 9.25,
+      "learning_rate": 2.058661163201522e-05,
+      "loss": 0.8653,
+      "step": 69265
+    },
+    {
+      "epoch": 4.7064818589482265,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0584488381573587e-05,
+      "loss": 0.8242,
+      "step": 69270
+    },
+    {
+      "epoch": 4.7068215790188885,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.0582365131131948e-05,
+      "loss": 0.8846,
+      "step": 69275
+    },
+    {
+      "epoch": 4.707161299089551,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0580241880690312e-05,
+      "loss": 0.8268,
+      "step": 69280
+    },
+    {
+      "epoch": 4.707501019160212,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0578118630248676e-05,
+      "loss": 0.7468,
+      "step": 69285
+    },
+    {
+      "epoch": 4.707840739230874,
+      "grad_norm": 7.375,
+      "learning_rate": 2.057599537980704e-05,
+      "loss": 0.8849,
+      "step": 69290
+    },
+    {
+      "epoch": 4.708180459301536,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0573872129365404e-05,
+      "loss": 0.9009,
+      "step": 69295
+    },
+    {
+      "epoch": 4.708520179372197,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0571748878923768e-05,
+      "loss": 0.975,
+      "step": 69300
+    },
+    {
+      "epoch": 4.708859899442859,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0569625628482132e-05,
+      "loss": 0.9109,
+      "step": 69305
+    },
+    {
+      "epoch": 4.709199619513521,
+      "grad_norm": 7.0,
+      "learning_rate": 2.0567502378040496e-05,
+      "loss": 0.7974,
+      "step": 69310
+    },
+    {
+      "epoch": 4.7095393395841825,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.056537912759886e-05,
+      "loss": 0.9064,
+      "step": 69315
+    },
+    {
+      "epoch": 4.7098790596548445,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0563255877157224e-05,
+      "loss": 0.8723,
+      "step": 69320
+    },
+    {
+      "epoch": 4.710218779725507,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0561132626715588e-05,
+      "loss": 0.9116,
+      "step": 69325
+    },
+    {
+      "epoch": 4.710558499796168,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0559009376273952e-05,
+      "loss": 0.8546,
+      "step": 69330
+    },
+    {
+      "epoch": 4.71089821986683,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0556886125832316e-05,
+      "loss": 0.8499,
+      "step": 69335
+    },
+    {
+      "epoch": 4.711237939937492,
+      "grad_norm": 8.0,
+      "learning_rate": 2.055476287539068e-05,
+      "loss": 0.9684,
+      "step": 69340
+    },
+    {
+      "epoch": 4.711577660008153,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0552639624949044e-05,
+      "loss": 0.8701,
+      "step": 69345
+    },
+    {
+      "epoch": 4.711917380078815,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.0550516374507404e-05,
+      "loss": 0.8189,
+      "step": 69350
+    },
+    {
+      "epoch": 4.712257100149477,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.0548393124065772e-05,
+      "loss": 0.76,
+      "step": 69355
+    },
+    {
+      "epoch": 4.7125968202201385,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0546269873624136e-05,
+      "loss": 0.813,
+      "step": 69360
+    },
+    {
+      "epoch": 4.712936540290801,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.0544146623182496e-05,
+      "loss": 0.8803,
+      "step": 69365
+    },
+    {
+      "epoch": 4.713276260361463,
+      "grad_norm": 8.625,
+      "learning_rate": 2.0542023372740864e-05,
+      "loss": 0.9133,
+      "step": 69370
+    },
+    {
+      "epoch": 4.713615980432124,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.0539900122299228e-05,
+      "loss": 0.8378,
+      "step": 69375
+    },
+    {
+      "epoch": 4.713955700502786,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.053777687185759e-05,
+      "loss": 0.9101,
+      "step": 69380
+    },
+    {
+      "epoch": 4.714295420573448,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0535653621415952e-05,
+      "loss": 0.8063,
+      "step": 69385
+    },
+    {
+      "epoch": 4.714635140644109,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.053353037097432e-05,
+      "loss": 0.9007,
+      "step": 69390
+    },
+    {
+      "epoch": 4.714974860714771,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.053140712053268e-05,
+      "loss": 0.9082,
+      "step": 69395
+    },
+    {
+      "epoch": 4.715314580785432,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0529283870091044e-05,
+      "loss": 0.9224,
+      "step": 69400
+    },
+    {
+      "epoch": 4.7156543008560945,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.0527160619649412e-05,
+      "loss": 0.8593,
+      "step": 69405
+    },
+    {
+      "epoch": 4.715994020926757,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.0525037369207772e-05,
+      "loss": 0.857,
+      "step": 69410
+    },
+    {
+      "epoch": 4.716333740997418,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0522914118766136e-05,
+      "loss": 0.8387,
+      "step": 69415
+    },
+    {
+      "epoch": 4.71667346106808,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.05207908683245e-05,
+      "loss": 0.8361,
+      "step": 69420
+    },
+    {
+      "epoch": 4.717013181138742,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0518667617882868e-05,
+      "loss": 0.8715,
+      "step": 69425
+    },
+    {
+      "epoch": 4.717352901209403,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.051654436744123e-05,
+      "loss": 0.8696,
+      "step": 69430
+    },
+    {
+      "epoch": 4.717692621280065,
+      "grad_norm": 8.75,
+      "learning_rate": 2.0514421116999592e-05,
+      "loss": 0.8945,
+      "step": 69435
+    },
+    {
+      "epoch": 4.718032341350727,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.051229786655796e-05,
+      "loss": 0.8372,
+      "step": 69440
+    },
+    {
+      "epoch": 4.718372061421388,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.051017461611632e-05,
+      "loss": 0.8731,
+      "step": 69445
+    },
+    {
+      "epoch": 4.7187117814920505,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0508051365674684e-05,
+      "loss": 0.9479,
+      "step": 69450
+    },
+    {
+      "epoch": 4.719051501562713,
+      "grad_norm": 9.5,
+      "learning_rate": 2.050592811523305e-05,
+      "loss": 0.8062,
+      "step": 69455
+    },
+    {
+      "epoch": 4.719391221633374,
+      "grad_norm": 5.875,
+      "learning_rate": 2.0503804864791412e-05,
+      "loss": 0.7833,
+      "step": 69460
+    },
+    {
+      "epoch": 4.719730941704036,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.0501681614349776e-05,
+      "loss": 0.8985,
+      "step": 69465
+    },
+    {
+      "epoch": 4.720070661774698,
+      "grad_norm": 7.625,
+      "learning_rate": 2.049955836390814e-05,
+      "loss": 0.8849,
+      "step": 69470
+    },
+    {
+      "epoch": 4.720410381845359,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.0497435113466504e-05,
+      "loss": 0.9238,
+      "step": 69475
+    },
+    {
+      "epoch": 4.720750101916021,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.049531186302487e-05,
+      "loss": 0.8209,
+      "step": 69480
+    },
+    {
+      "epoch": 4.721089821986683,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0493188612583233e-05,
+      "loss": 0.9407,
+      "step": 69485
+    },
+    {
+      "epoch": 4.721429542057344,
+      "grad_norm": 7.375,
+      "learning_rate": 2.0491065362141597e-05,
+      "loss": 0.8163,
+      "step": 69490
+    },
+    {
+      "epoch": 4.7217692621280065,
+      "grad_norm": 6.71875,
+      "learning_rate": 2.048894211169996e-05,
+      "loss": 0.8789,
+      "step": 69495
+    },
+    {
+      "epoch": 4.722108982198669,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0486818861258325e-05,
+      "loss": 0.9025,
+      "step": 69500
+    },
+    {
+      "epoch": 4.72244870226933,
+      "grad_norm": 8.25,
+      "learning_rate": 2.048469561081669e-05,
+      "loss": 0.9701,
+      "step": 69505
+    },
+    {
+      "epoch": 4.722788422339992,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.0482572360375053e-05,
+      "loss": 0.7899,
+      "step": 69510
+    },
+    {
+      "epoch": 4.723128142410654,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0480449109933417e-05,
+      "loss": 0.8203,
+      "step": 69515
+    },
+    {
+      "epoch": 4.723467862481315,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0478325859491777e-05,
+      "loss": 0.8452,
+      "step": 69520
+    },
+    {
+      "epoch": 4.723807582551977,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0476202609050145e-05,
+      "loss": 0.9415,
+      "step": 69525
+    },
+    {
+      "epoch": 4.724147302622639,
+      "grad_norm": 8.25,
+      "learning_rate": 2.047407935860851e-05,
+      "loss": 0.8932,
+      "step": 69530
+    },
+    {
+      "epoch": 4.7244870226933005,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.047195610816687e-05,
+      "loss": 0.7979,
+      "step": 69535
+    },
+    {
+      "epoch": 4.7248267427639625,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0469832857725237e-05,
+      "loss": 0.89,
+      "step": 69540
+    },
+    {
+      "epoch": 4.725166462834625,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.04677096072836e-05,
+      "loss": 0.9363,
+      "step": 69545
+    },
+    {
+      "epoch": 4.725506182905286,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.046558635684196e-05,
+      "loss": 0.7965,
+      "step": 69550
+    },
+    {
+      "epoch": 4.725845902975948,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.0463463106400325e-05,
+      "loss": 0.8619,
+      "step": 69555
+    },
+    {
+      "epoch": 4.72618562304661,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.0461339855958693e-05,
+      "loss": 0.7971,
+      "step": 69560
+    },
+    {
+      "epoch": 4.726525343117271,
+      "grad_norm": 7.5,
+      "learning_rate": 2.0459216605517053e-05,
+      "loss": 0.8794,
+      "step": 69565
+    },
+    {
+      "epoch": 4.726865063187933,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.0457093355075417e-05,
+      "loss": 0.8729,
+      "step": 69570
+    },
+    {
+      "epoch": 4.727204783258595,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.0454970104633785e-05,
+      "loss": 0.8593,
+      "step": 69575
+    },
+    {
+      "epoch": 4.7275445033292565,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0452846854192145e-05,
+      "loss": 0.7994,
+      "step": 69580
+    },
+    {
+      "epoch": 4.7278842233999185,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.045072360375051e-05,
+      "loss": 0.8183,
+      "step": 69585
+    },
+    {
+      "epoch": 4.728223943470581,
+      "grad_norm": 6.75,
+      "learning_rate": 2.0448600353308877e-05,
+      "loss": 0.9074,
+      "step": 69590
+    },
+    {
+      "epoch": 4.728563663541242,
+      "grad_norm": 8.625,
+      "learning_rate": 2.044647710286724e-05,
+      "loss": 0.863,
+      "step": 69595
+    },
+    {
+      "epoch": 4.728903383611904,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.04443538524256e-05,
+      "loss": 0.8701,
+      "step": 69600
+    },
+    {
+      "epoch": 4.729243103682566,
+      "grad_norm": 8.625,
+      "learning_rate": 2.0442230601983965e-05,
+      "loss": 0.9099,
+      "step": 69605
+    },
+    {
+      "epoch": 4.729582823753227,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.0440107351542333e-05,
+      "loss": 0.9523,
+      "step": 69610
+    },
+    {
+      "epoch": 4.729922543823889,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.0437984101100693e-05,
+      "loss": 0.8912,
+      "step": 69615
+    },
+    {
+      "epoch": 4.73026226389455,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0435860850659057e-05,
+      "loss": 0.9446,
+      "step": 69620
+    },
+    {
+      "epoch": 4.7306019839652125,
+      "grad_norm": 6.15625,
+      "learning_rate": 2.0433737600217425e-05,
+      "loss": 0.8533,
+      "step": 69625
+    },
+    {
+      "epoch": 4.7309417040358746,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0431614349775785e-05,
+      "loss": 0.7989,
+      "step": 69630
+    },
+    {
+      "epoch": 4.731281424106536,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.042949109933415e-05,
+      "loss": 0.8226,
+      "step": 69635
+    },
+    {
+      "epoch": 4.731621144177198,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0427367848892513e-05,
+      "loss": 0.9506,
+      "step": 69640
+    },
+    {
+      "epoch": 4.73196086424786,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0425244598450877e-05,
+      "loss": 0.8662,
+      "step": 69645
+    },
+    {
+      "epoch": 4.732300584318521,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.042312134800924e-05,
+      "loss": 0.902,
+      "step": 69650
+    },
+    {
+      "epoch": 4.732640304389183,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0420998097567605e-05,
+      "loss": 0.8268,
+      "step": 69655
+    },
+    {
+      "epoch": 4.732980024459845,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.041887484712597e-05,
+      "loss": 0.8151,
+      "step": 69660
+    },
+    {
+      "epoch": 4.733319744530506,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0416751596684333e-05,
+      "loss": 0.7573,
+      "step": 69665
+    },
+    {
+      "epoch": 4.7336594646011685,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0414628346242697e-05,
+      "loss": 0.8442,
+      "step": 69670
+    },
+    {
+      "epoch": 4.733999184671831,
+      "grad_norm": 8.625,
+      "learning_rate": 2.041250509580106e-05,
+      "loss": 0.9163,
+      "step": 69675
+    },
+    {
+      "epoch": 4.734338904742492,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0410381845359425e-05,
+      "loss": 0.8389,
+      "step": 69680
+    },
+    {
+      "epoch": 4.734678624813154,
+      "grad_norm": 8.25,
+      "learning_rate": 2.040825859491779e-05,
+      "loss": 0.8763,
+      "step": 69685
+    },
+    {
+      "epoch": 4.735018344883816,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0406135344476153e-05,
+      "loss": 0.8823,
+      "step": 69690
+    },
+    {
+      "epoch": 4.735358064954477,
+      "grad_norm": 8.75,
+      "learning_rate": 2.0404012094034517e-05,
+      "loss": 0.8343,
+      "step": 69695
+    },
+    {
+      "epoch": 4.735697785025139,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.040188884359288e-05,
+      "loss": 0.7834,
+      "step": 69700
+    },
+    {
+      "epoch": 4.736037505095801,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.0399765593151242e-05,
+      "loss": 0.8737,
+      "step": 69705
+    },
+    {
+      "epoch": 4.736377225166462,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.039764234270961e-05,
+      "loss": 0.9499,
+      "step": 69710
+    },
+    {
+      "epoch": 4.7367169452371245,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.0395519092267973e-05,
+      "loss": 0.9143,
+      "step": 69715
+    },
+    {
+      "epoch": 4.737056665307787,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0393395841826334e-05,
+      "loss": 0.7703,
+      "step": 69720
+    },
+    {
+      "epoch": 4.737396385378448,
+      "grad_norm": 8.0,
+      "learning_rate": 2.03912725913847e-05,
+      "loss": 0.8709,
+      "step": 69725
+    },
+    {
+      "epoch": 4.73773610544911,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.0389149340943065e-05,
+      "loss": 0.8497,
+      "step": 69730
+    },
+    {
+      "epoch": 4.738075825519772,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0387026090501426e-05,
+      "loss": 0.8682,
+      "step": 69735
+    },
+    {
+      "epoch": 4.738415545590433,
+      "grad_norm": 8.5,
+      "learning_rate": 2.038490284005979e-05,
+      "loss": 0.8212,
+      "step": 69740
+    },
+    {
+      "epoch": 4.738755265661095,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0382779589618157e-05,
+      "loss": 0.8149,
+      "step": 69745
+    },
+    {
+      "epoch": 4.739094985731757,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.0380656339176518e-05,
+      "loss": 0.8496,
+      "step": 69750
+    },
+    {
+      "epoch": 4.739434705802418,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0378533088734882e-05,
+      "loss": 0.8897,
+      "step": 69755
+    },
+    {
+      "epoch": 4.7397744258730805,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.037640983829325e-05,
+      "loss": 0.8034,
+      "step": 69760
+    },
+    {
+      "epoch": 4.740114145943743,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0374286587851613e-05,
+      "loss": 0.8206,
+      "step": 69765
+    },
+    {
+      "epoch": 4.740453866014404,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0372163337409974e-05,
+      "loss": 0.9109,
+      "step": 69770
+    },
+    {
+      "epoch": 4.740793586085066,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.0370040086968338e-05,
+      "loss": 0.8293,
+      "step": 69775
+    },
+    {
+      "epoch": 4.741133306155728,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0367916836526705e-05,
+      "loss": 0.7615,
+      "step": 69780
+    },
+    {
+      "epoch": 4.741473026226389,
+      "grad_norm": 7.875,
+      "learning_rate": 2.0365793586085066e-05,
+      "loss": 0.8639,
+      "step": 69785
+    },
+    {
+      "epoch": 4.741812746297051,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.036367033564343e-05,
+      "loss": 0.8999,
+      "step": 69790
+    },
+    {
+      "epoch": 4.742152466367713,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.0361547085201797e-05,
+      "loss": 0.9193,
+      "step": 69795
+    },
+    {
+      "epoch": 4.7424921864383744,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0359423834760158e-05,
+      "loss": 0.8716,
+      "step": 69800
+    },
+    {
+      "epoch": 4.7428319065090365,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.0357300584318522e-05,
+      "loss": 0.8444,
+      "step": 69805
+    },
+    {
+      "epoch": 4.743171626579699,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0355177333876886e-05,
+      "loss": 0.9478,
+      "step": 69810
+    },
+    {
+      "epoch": 4.74351134665036,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.035305408343525e-05,
+      "loss": 0.771,
+      "step": 69815
+    },
+    {
+      "epoch": 4.743851066721022,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0350930832993614e-05,
+      "loss": 0.8872,
+      "step": 69820
+    },
+    {
+      "epoch": 4.744190786791684,
+      "grad_norm": 9.125,
+      "learning_rate": 2.0348807582551978e-05,
+      "loss": 0.9545,
+      "step": 69825
+    },
+    {
+      "epoch": 4.744530506862345,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0346684332110342e-05,
+      "loss": 0.9143,
+      "step": 69830
+    },
+    {
+      "epoch": 4.744870226933007,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0344561081668706e-05,
+      "loss": 0.8566,
+      "step": 69835
+    },
+    {
+      "epoch": 4.745209947003669,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.034243783122707e-05,
+      "loss": 0.7886,
+      "step": 69840
+    },
+    {
+      "epoch": 4.7455496670743305,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.0340314580785434e-05,
+      "loss": 0.9258,
+      "step": 69845
+    },
+    {
+      "epoch": 4.7458893871449925,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0338191330343798e-05,
+      "loss": 0.768,
+      "step": 69850
+    },
+    {
+      "epoch": 4.746229107215655,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0336068079902162e-05,
+      "loss": 0.8547,
+      "step": 69855
+    },
+    {
+      "epoch": 4.746568827286316,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.0333944829460526e-05,
+      "loss": 0.8342,
+      "step": 69860
+    },
+    {
+      "epoch": 4.746908547356978,
+      "grad_norm": 8.125,
+      "learning_rate": 2.033182157901889e-05,
+      "loss": 0.7953,
+      "step": 69865
+    },
+    {
+      "epoch": 4.74724826742764,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0329698328577254e-05,
+      "loss": 0.8175,
+      "step": 69870
+    },
+    {
+      "epoch": 4.747587987498301,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.0327575078135615e-05,
+      "loss": 0.9078,
+      "step": 69875
+    },
+    {
+      "epoch": 4.747927707568963,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.0325451827693982e-05,
+      "loss": 0.8829,
+      "step": 69880
+    },
+    {
+      "epoch": 4.748267427639625,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0323328577252346e-05,
+      "loss": 0.9188,
+      "step": 69885
+    },
+    {
+      "epoch": 4.7486071477102865,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0321205326810707e-05,
+      "loss": 0.9506,
+      "step": 69890
+    },
+    {
+      "epoch": 4.7489468677809485,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0319082076369074e-05,
+      "loss": 0.7503,
+      "step": 69895
+    },
+    {
+      "epoch": 4.749286587851611,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0316958825927438e-05,
+      "loss": 0.9322,
+      "step": 69900
+    },
+    {
+      "epoch": 4.749626307922272,
+      "grad_norm": 6.5625,
+      "learning_rate": 2.03148355754858e-05,
+      "loss": 0.862,
+      "step": 69905
+    },
+    {
+      "epoch": 4.749966027992934,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.0312712325044163e-05,
+      "loss": 0.9701,
+      "step": 69910
+    },
+    {
+      "epoch": 4.750305748063596,
+      "grad_norm": 8.5,
+      "learning_rate": 2.031058907460253e-05,
+      "loss": 0.9955,
+      "step": 69915
+    },
+    {
+      "epoch": 4.750645468134257,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.030846582416089e-05,
+      "loss": 0.9678,
+      "step": 69920
+    },
+    {
+      "epoch": 4.750985188204919,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0306342573719255e-05,
+      "loss": 0.8758,
+      "step": 69925
+    },
+    {
+      "epoch": 4.751324908275581,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0304219323277622e-05,
+      "loss": 0.8482,
+      "step": 69930
+    },
+    {
+      "epoch": 4.7516646283462425,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0302096072835986e-05,
+      "loss": 0.8516,
+      "step": 69935
+    },
+    {
+      "epoch": 4.7520043484169046,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0299972822394347e-05,
+      "loss": 0.8766,
+      "step": 69940
+    },
+    {
+      "epoch": 4.752344068487567,
+      "grad_norm": 7.625,
+      "learning_rate": 2.029784957195271e-05,
+      "loss": 1.0018,
+      "step": 69945
+    },
+    {
+      "epoch": 4.752683788558228,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0295726321511078e-05,
+      "loss": 0.8193,
+      "step": 69950
+    },
+    {
+      "epoch": 4.75302350862889,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.029360307106944e-05,
+      "loss": 0.7991,
+      "step": 69955
+    },
+    {
+      "epoch": 4.753363228699552,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0291479820627803e-05,
+      "loss": 0.8606,
+      "step": 69960
+    },
+    {
+      "epoch": 4.753702948770213,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.028935657018617e-05,
+      "loss": 0.8893,
+      "step": 69965
+    },
+    {
+      "epoch": 4.754042668840875,
+      "grad_norm": 8.0,
+      "learning_rate": 2.028723331974453e-05,
+      "loss": 0.8969,
+      "step": 69970
+    },
+    {
+      "epoch": 4.754382388911537,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0285110069302895e-05,
+      "loss": 0.844,
+      "step": 69975
+    },
+    {
+      "epoch": 4.7547221089821985,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.028298681886126e-05,
+      "loss": 0.8792,
+      "step": 69980
+    },
+    {
+      "epoch": 4.755061829052861,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0280863568419623e-05,
+      "loss": 0.8392,
+      "step": 69985
+    },
+    {
+      "epoch": 4.755401549123523,
+      "grad_norm": 7.5,
+      "learning_rate": 2.0278740317977987e-05,
+      "loss": 0.9101,
+      "step": 69990
+    },
+    {
+      "epoch": 4.755741269194184,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.027661706753635e-05,
+      "loss": 0.8259,
+      "step": 69995
+    },
+    {
+      "epoch": 4.756080989264846,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0274493817094715e-05,
+      "loss": 0.9363,
+      "step": 70000
+    },
+    {
+      "epoch": 4.756420709335508,
+      "grad_norm": 7.25,
+      "learning_rate": 2.027237056665308e-05,
+      "loss": 0.8451,
+      "step": 70005
+    },
+    {
+      "epoch": 4.756760429406169,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.0270247316211443e-05,
+      "loss": 0.8828,
+      "step": 70010
+    },
+    {
+      "epoch": 4.757100149476831,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.0268124065769807e-05,
+      "loss": 0.8089,
+      "step": 70015
+    },
+    {
+      "epoch": 4.757439869547493,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.026600081532817e-05,
+      "loss": 0.8906,
+      "step": 70020
+    },
+    {
+      "epoch": 4.7577795896181545,
+      "grad_norm": 6.75,
+      "learning_rate": 2.0263877564886535e-05,
+      "loss": 0.7877,
+      "step": 70025
+    },
+    {
+      "epoch": 4.758119309688817,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.02617543144449e-05,
+      "loss": 0.9253,
+      "step": 70030
+    },
+    {
+      "epoch": 4.758459029759479,
+      "grad_norm": 5.875,
+      "learning_rate": 2.0259631064003263e-05,
+      "loss": 0.8565,
+      "step": 70035
+    },
+    {
+      "epoch": 4.75879874983014,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0257507813561627e-05,
+      "loss": 0.8297,
+      "step": 70040
+    },
+    {
+      "epoch": 4.759138469900802,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0255384563119987e-05,
+      "loss": 0.9166,
+      "step": 70045
+    },
+    {
+      "epoch": 4.759478189971464,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.0253261312678355e-05,
+      "loss": 0.7962,
+      "step": 70050
+    },
+    {
+      "epoch": 4.759817910042125,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.025113806223672e-05,
+      "loss": 0.8089,
+      "step": 70055
+    },
+    {
+      "epoch": 4.760157630112787,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.024901481179508e-05,
+      "loss": 0.8217,
+      "step": 70060
+    },
+    {
+      "epoch": 4.760497350183449,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0246891561353447e-05,
+      "loss": 0.9273,
+      "step": 70065
+    },
+    {
+      "epoch": 4.7608370702541105,
+      "grad_norm": 7.75,
+      "learning_rate": 2.024476831091181e-05,
+      "loss": 0.9028,
+      "step": 70070
+    },
+    {
+      "epoch": 4.761176790324773,
+      "grad_norm": 7.875,
+      "learning_rate": 2.024264506047017e-05,
+      "loss": 0.9091,
+      "step": 70075
+    },
+    {
+      "epoch": 4.761516510395434,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.0240521810028535e-05,
+      "loss": 0.8409,
+      "step": 70080
+    },
+    {
+      "epoch": 4.761856230466096,
+      "grad_norm": 7.0,
+      "learning_rate": 2.0238398559586903e-05,
+      "loss": 0.8095,
+      "step": 70085
+    },
+    {
+      "epoch": 4.762195950536758,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0236275309145263e-05,
+      "loss": 0.7903,
+      "step": 70090
+    },
+    {
+      "epoch": 4.762535670607419,
+      "grad_norm": 7.25,
+      "learning_rate": 2.0234152058703627e-05,
+      "loss": 0.7699,
+      "step": 70095
+    },
+    {
+      "epoch": 4.762875390678081,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.0232028808261995e-05,
+      "loss": 0.8802,
+      "step": 70100
+    },
+    {
+      "epoch": 4.763215110748743,
+      "grad_norm": 11.9375,
+      "learning_rate": 2.022990555782036e-05,
+      "loss": 0.8987,
+      "step": 70105
+    },
+    {
+      "epoch": 4.7635548308194045,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.022778230737872e-05,
+      "loss": 0.8411,
+      "step": 70110
+    },
+    {
+      "epoch": 4.7638945508900665,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0225659056937087e-05,
+      "loss": 0.8599,
+      "step": 70115
+    },
+    {
+      "epoch": 4.764234270960729,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.022353580649545e-05,
+      "loss": 0.9214,
+      "step": 70120
+    },
+    {
+      "epoch": 4.76457399103139,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.022141255605381e-05,
+      "loss": 0.9365,
+      "step": 70125
+    },
+    {
+      "epoch": 4.764913711102052,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0219289305612175e-05,
+      "loss": 0.8387,
+      "step": 70130
+    },
+    {
+      "epoch": 4.765253431172714,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0217166055170543e-05,
+      "loss": 0.8887,
+      "step": 70135
+    },
+    {
+      "epoch": 4.765593151243375,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0215042804728903e-05,
+      "loss": 0.8602,
+      "step": 70140
+    },
+    {
+      "epoch": 4.765932871314037,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0212919554287267e-05,
+      "loss": 0.9229,
+      "step": 70145
+    },
+    {
+      "epoch": 4.766272591384699,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0210796303845635e-05,
+      "loss": 0.8492,
+      "step": 70150
+    },
+    {
+      "epoch": 4.7666123114553605,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0208673053403995e-05,
+      "loss": 0.8991,
+      "step": 70155
+    },
+    {
+      "epoch": 4.7669520315260225,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.020654980296236e-05,
+      "loss": 0.9372,
+      "step": 70160
+    },
+    {
+      "epoch": 4.767291751596685,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0204426552520723e-05,
+      "loss": 0.9261,
+      "step": 70165
+    },
+    {
+      "epoch": 4.767631471667346,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.0202303302079087e-05,
+      "loss": 0.8456,
+      "step": 70170
+    },
+    {
+      "epoch": 4.767971191738008,
+      "grad_norm": 7.25,
+      "learning_rate": 2.020018005163745e-05,
+      "loss": 0.9089,
+      "step": 70175
+    },
+    {
+      "epoch": 4.76831091180867,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.0198056801195815e-05,
+      "loss": 0.8306,
+      "step": 70180
+    },
+    {
+      "epoch": 4.768650631879331,
+      "grad_norm": 7.875,
+      "learning_rate": 2.019593355075418e-05,
+      "loss": 0.9411,
+      "step": 70185
+    },
+    {
+      "epoch": 4.768990351949993,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0193810300312543e-05,
+      "loss": 0.8767,
+      "step": 70190
+    },
+    {
+      "epoch": 4.769330072020655,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0191687049870907e-05,
+      "loss": 0.8757,
+      "step": 70195
+    },
+    {
+      "epoch": 4.7696697920913165,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.018956379942927e-05,
+      "loss": 0.8347,
+      "step": 70200
+    },
+    {
+      "epoch": 4.7700095121619785,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0187440548987636e-05,
+      "loss": 0.9138,
+      "step": 70205
+    },
+    {
+      "epoch": 4.770349232232641,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.0185317298546e-05,
+      "loss": 0.9356,
+      "step": 70210
+    },
+    {
+      "epoch": 4.770688952303302,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0183194048104364e-05,
+      "loss": 0.9,
+      "step": 70215
+    },
+    {
+      "epoch": 4.771028672373964,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0181070797662728e-05,
+      "loss": 0.8618,
+      "step": 70220
+    },
+    {
+      "epoch": 4.771368392444626,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.017894754722109e-05,
+      "loss": 0.798,
+      "step": 70225
+    },
+    {
+      "epoch": 4.771708112515287,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0176824296779452e-05,
+      "loss": 0.9176,
+      "step": 70230
+    },
+    {
+      "epoch": 4.772047832585949,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.017470104633782e-05,
+      "loss": 0.9005,
+      "step": 70235
+    },
+    {
+      "epoch": 4.772387552656611,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0172577795896184e-05,
+      "loss": 0.8479,
+      "step": 70240
+    },
+    {
+      "epoch": 4.7727272727272725,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0170454545454544e-05,
+      "loss": 0.8847,
+      "step": 70245
+    },
+    {
+      "epoch": 4.773066992797935,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.016833129501291e-05,
+      "loss": 0.9059,
+      "step": 70250
+    },
+    {
+      "epoch": 4.773406712868597,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.0166208044571276e-05,
+      "loss": 0.8872,
+      "step": 70255
+    },
+    {
+      "epoch": 4.773746432939258,
+      "grad_norm": 7.375,
+      "learning_rate": 2.0164084794129636e-05,
+      "loss": 0.9356,
+      "step": 70260
+    },
+    {
+      "epoch": 4.77408615300992,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0161961543688e-05,
+      "loss": 0.874,
+      "step": 70265
+    },
+    {
+      "epoch": 4.774425873080582,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0159838293246368e-05,
+      "loss": 0.841,
+      "step": 70270
+    },
+    {
+      "epoch": 4.774765593151243,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.015771504280473e-05,
+      "loss": 0.8897,
+      "step": 70275
+    },
+    {
+      "epoch": 4.775105313221905,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0155591792363092e-05,
+      "loss": 0.8689,
+      "step": 70280
+    },
+    {
+      "epoch": 4.775445033292567,
+      "grad_norm": 7.25,
+      "learning_rate": 2.015346854192146e-05,
+      "loss": 0.8871,
+      "step": 70285
+    },
+    {
+      "epoch": 4.7757847533632285,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.0151345291479824e-05,
+      "loss": 0.9253,
+      "step": 70290
+    },
+    {
+      "epoch": 4.776124473433891,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0149222041038184e-05,
+      "loss": 0.815,
+      "step": 70295
+    },
+    {
+      "epoch": 4.776464193504552,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.0147098790596548e-05,
+      "loss": 0.8625,
+      "step": 70300
+    },
+    {
+      "epoch": 4.776803913575214,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.0144975540154916e-05,
+      "loss": 0.9523,
+      "step": 70305
+    },
+    {
+      "epoch": 4.777143633645876,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0142852289713276e-05,
+      "loss": 0.9095,
+      "step": 70310
+    },
+    {
+      "epoch": 4.777483353716537,
+      "grad_norm": 7.75,
+      "learning_rate": 2.014072903927164e-05,
+      "loss": 0.8434,
+      "step": 70315
+    },
+    {
+      "epoch": 4.777823073787199,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0138605788830008e-05,
+      "loss": 0.9171,
+      "step": 70320
+    },
+    {
+      "epoch": 4.778162793857861,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0136482538388368e-05,
+      "loss": 0.8227,
+      "step": 70325
+    },
+    {
+      "epoch": 4.778502513928522,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0134359287946732e-05,
+      "loss": 0.906,
+      "step": 70330
+    },
+    {
+      "epoch": 4.7788422339991845,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.0132236037505096e-05,
+      "loss": 0.7988,
+      "step": 70335
+    },
+    {
+      "epoch": 4.779181954069847,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.013011278706346e-05,
+      "loss": 0.7973,
+      "step": 70340
+    },
+    {
+      "epoch": 4.779521674140508,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0127989536621824e-05,
+      "loss": 0.8849,
+      "step": 70345
+    },
+    {
+      "epoch": 4.77986139421117,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.0125866286180188e-05,
+      "loss": 0.8955,
+      "step": 70350
+    },
+    {
+      "epoch": 4.780201114281832,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0123743035738552e-05,
+      "loss": 0.8874,
+      "step": 70355
+    },
+    {
+      "epoch": 4.780540834352493,
+      "grad_norm": 7.375,
+      "learning_rate": 2.0121619785296916e-05,
+      "loss": 0.8874,
+      "step": 70360
+    },
+    {
+      "epoch": 4.780880554423155,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.011949653485528e-05,
+      "loss": 0.9275,
+      "step": 70365
+    },
+    {
+      "epoch": 4.781220274493817,
+      "grad_norm": 7.0,
+      "learning_rate": 2.0117373284413644e-05,
+      "loss": 0.9067,
+      "step": 70370
+    },
+    {
+      "epoch": 4.7815599945644784,
+      "grad_norm": 8.625,
+      "learning_rate": 2.0115250033972008e-05,
+      "loss": 0.9355,
+      "step": 70375
+    },
+    {
+      "epoch": 4.7818997146351405,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0113126783530372e-05,
+      "loss": 0.9413,
+      "step": 70380
+    },
+    {
+      "epoch": 4.782239434705803,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0111003533088736e-05,
+      "loss": 0.8111,
+      "step": 70385
+    },
+    {
+      "epoch": 4.782579154776464,
+      "grad_norm": 7.875,
+      "learning_rate": 2.01088802826471e-05,
+      "loss": 0.8618,
+      "step": 70390
+    },
+    {
+      "epoch": 4.782918874847126,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.0106757032205464e-05,
+      "loss": 0.9096,
+      "step": 70395
+    },
+    {
+      "epoch": 4.783258594917788,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0104633781763825e-05,
+      "loss": 0.8093,
+      "step": 70400
+    },
+    {
+      "epoch": 4.783598314988449,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.0102510531322192e-05,
+      "loss": 0.8253,
+      "step": 70405
+    },
+    {
+      "epoch": 4.783938035059111,
+      "grad_norm": 7.75,
+      "learning_rate": 2.0100387280880556e-05,
+      "loss": 0.8473,
+      "step": 70410
+    },
+    {
+      "epoch": 4.784277755129773,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0098264030438917e-05,
+      "loss": 0.9055,
+      "step": 70415
+    },
+    {
+      "epoch": 4.7846174752004345,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0096140779997284e-05,
+      "loss": 0.9034,
+      "step": 70420
+    },
+    {
+      "epoch": 4.7849571952710965,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0094017529555648e-05,
+      "loss": 0.8161,
+      "step": 70425
+    },
+    {
+      "epoch": 4.785296915341759,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.009189427911401e-05,
+      "loss": 0.8093,
+      "step": 70430
+    },
+    {
+      "epoch": 4.78563663541242,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0089771028672373e-05,
+      "loss": 0.997,
+      "step": 70435
+    },
+    {
+      "epoch": 4.785976355483082,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.008764777823074e-05,
+      "loss": 0.7842,
+      "step": 70440
+    },
+    {
+      "epoch": 4.786316075553744,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0085524527789104e-05,
+      "loss": 0.8894,
+      "step": 70445
+    },
+    {
+      "epoch": 4.786655795624405,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.0083401277347465e-05,
+      "loss": 0.8418,
+      "step": 70450
+    },
+    {
+      "epoch": 4.786995515695067,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.0081278026905832e-05,
+      "loss": 0.8023,
+      "step": 70455
+    },
+    {
+      "epoch": 4.787335235765729,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.0079154776464196e-05,
+      "loss": 0.9433,
+      "step": 70460
+    },
+    {
+      "epoch": 4.7876749558363905,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.0077031526022557e-05,
+      "loss": 0.8138,
+      "step": 70465
+    },
+    {
+      "epoch": 4.7880146759070525,
+      "grad_norm": 8.0,
+      "learning_rate": 2.007490827558092e-05,
+      "loss": 0.8465,
+      "step": 70470
+    },
+    {
+      "epoch": 4.788354395977715,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.007278502513929e-05,
+      "loss": 0.9239,
+      "step": 70475
+    },
+    {
+      "epoch": 4.788694116048376,
+      "grad_norm": 7.125,
+      "learning_rate": 2.007066177469765e-05,
+      "loss": 0.8249,
+      "step": 70480
+    },
+    {
+      "epoch": 4.789033836119038,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0068538524256013e-05,
+      "loss": 0.8177,
+      "step": 70485
+    },
+    {
+      "epoch": 4.7893735561897,
+      "grad_norm": 8.875,
+      "learning_rate": 2.006641527381438e-05,
+      "loss": 0.9126,
+      "step": 70490
+    },
+    {
+      "epoch": 4.789713276260361,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.006429202337274e-05,
+      "loss": 0.8345,
+      "step": 70495
+    },
+    {
+      "epoch": 4.790052996331023,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0062168772931105e-05,
+      "loss": 1.0193,
+      "step": 70500
+    },
+    {
+      "epoch": 4.790392716401685,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0060045522489472e-05,
+      "loss": 0.8489,
+      "step": 70505
+    },
+    {
+      "epoch": 4.7907324364723465,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0057922272047833e-05,
+      "loss": 0.8484,
+      "step": 70510
+    },
+    {
+      "epoch": 4.7910721565430086,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0055799021606197e-05,
+      "loss": 0.8678,
+      "step": 70515
+    },
+    {
+      "epoch": 4.791411876613671,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.005367577116456e-05,
+      "loss": 0.933,
+      "step": 70520
+    },
+    {
+      "epoch": 4.791751596684332,
+      "grad_norm": 8.75,
+      "learning_rate": 2.0051552520722925e-05,
+      "loss": 0.9096,
+      "step": 70525
+    },
+    {
+      "epoch": 4.792091316754994,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.004942927028129e-05,
+      "loss": 0.7823,
+      "step": 70530
+    },
+    {
+      "epoch": 4.792431036825656,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.0047306019839653e-05,
+      "loss": 0.8235,
+      "step": 70535
+    },
+    {
+      "epoch": 4.792770756896317,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.0045182769398017e-05,
+      "loss": 0.869,
+      "step": 70540
+    },
+    {
+      "epoch": 4.793110476966979,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.004305951895638e-05,
+      "loss": 0.9071,
+      "step": 70545
+    },
+    {
+      "epoch": 4.793450197037641,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0040936268514745e-05,
+      "loss": 0.8912,
+      "step": 70550
+    },
+    {
+      "epoch": 4.7937899171083025,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.003881301807311e-05,
+      "loss": 0.939,
+      "step": 70555
+    },
+    {
+      "epoch": 4.794129637178965,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.0036689767631473e-05,
+      "loss": 0.8405,
+      "step": 70560
+    },
+    {
+      "epoch": 4.794469357249627,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0034566517189837e-05,
+      "loss": 0.8026,
+      "step": 70565
+    },
+    {
+      "epoch": 4.794809077320288,
+      "grad_norm": 6.84375,
+      "learning_rate": 2.0032443266748198e-05,
+      "loss": 0.9772,
+      "step": 70570
+    },
+    {
+      "epoch": 4.79514879739095,
+      "grad_norm": 8.875,
+      "learning_rate": 2.0030320016306565e-05,
+      "loss": 0.8498,
+      "step": 70575
+    },
+    {
+      "epoch": 4.795488517461612,
+      "grad_norm": 8.375,
+      "learning_rate": 2.002819676586493e-05,
+      "loss": 0.8607,
+      "step": 70580
+    },
+    {
+      "epoch": 4.795828237532273,
+      "grad_norm": 7.375,
+      "learning_rate": 2.002607351542329e-05,
+      "loss": 0.8945,
+      "step": 70585
+    },
+    {
+      "epoch": 4.796167957602935,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0023950264981657e-05,
+      "loss": 0.894,
+      "step": 70590
+    },
+    {
+      "epoch": 4.796507677673597,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.002182701454002e-05,
+      "loss": 0.9555,
+      "step": 70595
+    },
+    {
+      "epoch": 4.7968473977442585,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.001970376409838e-05,
+      "loss": 0.9065,
+      "step": 70600
+    },
+    {
+      "epoch": 4.797187117814921,
+      "grad_norm": 8.875,
+      "learning_rate": 2.001758051365675e-05,
+      "loss": 0.8157,
+      "step": 70605
+    },
+    {
+      "epoch": 4.797526837885583,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0015457263215113e-05,
+      "loss": 0.8514,
+      "step": 70610
+    },
+    {
+      "epoch": 4.797866557956244,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.0013334012773477e-05,
+      "loss": 0.9322,
+      "step": 70615
+    },
+    {
+      "epoch": 4.798206278026906,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0011210762331838e-05,
+      "loss": 0.8857,
+      "step": 70620
+    },
+    {
+      "epoch": 4.798545998097568,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.0009087511890205e-05,
+      "loss": 0.918,
+      "step": 70625
+    },
+    {
+      "epoch": 4.798885718168229,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.000696426144857e-05,
+      "loss": 0.901,
+      "step": 70630
+    },
+    {
+      "epoch": 4.799225438238891,
+      "grad_norm": 8.25,
+      "learning_rate": 2.000484101100693e-05,
+      "loss": 0.9159,
+      "step": 70635
+    },
+    {
+      "epoch": 4.799565158309553,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0002717760565297e-05,
+      "loss": 0.9169,
+      "step": 70640
+    },
+    {
+      "epoch": 4.7999048783802145,
+      "grad_norm": 6.40625,
+      "learning_rate": 2.000059451012366e-05,
+      "loss": 0.8056,
+      "step": 70645
+    },
+    {
+      "epoch": 4.800244598450877,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.999847125968202e-05,
+      "loss": 0.8765,
+      "step": 70650
+    },
+    {
+      "epoch": 4.800584318521539,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9996348009240386e-05,
+      "loss": 0.8429,
+      "step": 70655
+    },
+    {
+      "epoch": 4.8009240385922,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9994224758798753e-05,
+      "loss": 0.9151,
+      "step": 70660
+    },
+    {
+      "epoch": 4.801263758662862,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9992101508357114e-05,
+      "loss": 0.8902,
+      "step": 70665
+    },
+    {
+      "epoch": 4.801603478733524,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9989978257915478e-05,
+      "loss": 0.8179,
+      "step": 70670
+    },
+    {
+      "epoch": 4.801943198804185,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9987855007473845e-05,
+      "loss": 1.0818,
+      "step": 70675
+    },
+    {
+      "epoch": 4.802282918874847,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9985731757032206e-05,
+      "loss": 0.8859,
+      "step": 70680
+    },
+    {
+      "epoch": 4.802622638945509,
+      "grad_norm": 8.375,
+      "learning_rate": 1.998360850659057e-05,
+      "loss": 0.9039,
+      "step": 70685
+    },
+    {
+      "epoch": 4.8029623590161705,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9981485256148934e-05,
+      "loss": 0.827,
+      "step": 70690
+    },
+    {
+      "epoch": 4.803302079086833,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9979362005707298e-05,
+      "loss": 0.8528,
+      "step": 70695
+    },
+    {
+      "epoch": 4.803641799157495,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9977238755265662e-05,
+      "loss": 0.8492,
+      "step": 70700
+    },
+    {
+      "epoch": 4.803981519228156,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.9975115504824026e-05,
+      "loss": 0.8442,
+      "step": 70705
+    },
+    {
+      "epoch": 4.804321239298818,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.997299225438239e-05,
+      "loss": 0.8895,
+      "step": 70710
+    },
+    {
+      "epoch": 4.80466095936948,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9970869003940754e-05,
+      "loss": 0.8222,
+      "step": 70715
+    },
+    {
+      "epoch": 4.805000679440141,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9968745753499118e-05,
+      "loss": 0.8584,
+      "step": 70720
+    },
+    {
+      "epoch": 4.805340399510803,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9966622503057482e-05,
+      "loss": 0.879,
+      "step": 70725
+    },
+    {
+      "epoch": 4.805680119581465,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9964499252615846e-05,
+      "loss": 0.8281,
+      "step": 70730
+    },
+    {
+      "epoch": 4.8060198396521265,
+      "grad_norm": 8.0,
+      "learning_rate": 1.996237600217421e-05,
+      "loss": 1.0023,
+      "step": 70735
+    },
+    {
+      "epoch": 4.806359559722789,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9960252751732574e-05,
+      "loss": 0.8444,
+      "step": 70740
+    },
+    {
+      "epoch": 4.806699279793451,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9958129501290938e-05,
+      "loss": 0.7762,
+      "step": 70745
+    },
+    {
+      "epoch": 4.807038999864112,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9956006250849302e-05,
+      "loss": 0.8613,
+      "step": 70750
+    },
+    {
+      "epoch": 4.807378719934774,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.9953883000407662e-05,
+      "loss": 0.8761,
+      "step": 70755
+    },
+    {
+      "epoch": 4.807718440005435,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.995175974996603e-05,
+      "loss": 0.8507,
+      "step": 70760
+    },
+    {
+      "epoch": 4.808058160076097,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9949636499524394e-05,
+      "loss": 0.9982,
+      "step": 70765
+    },
+    {
+      "epoch": 4.808397880146759,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9947513249082754e-05,
+      "loss": 0.8542,
+      "step": 70770
+    },
+    {
+      "epoch": 4.8087376002174205,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9945389998641122e-05,
+      "loss": 0.9194,
+      "step": 70775
+    },
+    {
+      "epoch": 4.8090773202880825,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9943266748199486e-05,
+      "loss": 0.7863,
+      "step": 70780
+    },
+    {
+      "epoch": 4.809417040358745,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.994114349775785e-05,
+      "loss": 0.8773,
+      "step": 70785
+    },
+    {
+      "epoch": 4.809756760429406,
+      "grad_norm": 7.375,
+      "learning_rate": 1.993902024731621e-05,
+      "loss": 0.8486,
+      "step": 70790
+    },
+    {
+      "epoch": 4.810096480500068,
+      "grad_norm": 7.125,
+      "learning_rate": 1.9936896996874578e-05,
+      "loss": 0.8972,
+      "step": 70795
+    },
+    {
+      "epoch": 4.81043620057073,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9934773746432942e-05,
+      "loss": 0.9294,
+      "step": 70800
+    },
+    {
+      "epoch": 4.810775920641391,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9932650495991302e-05,
+      "loss": 0.8522,
+      "step": 70805
+    },
+    {
+      "epoch": 4.811115640712053,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.993052724554967e-05,
+      "loss": 0.8684,
+      "step": 70810
+    },
+    {
+      "epoch": 4.811455360782715,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.9928403995108034e-05,
+      "loss": 0.7717,
+      "step": 70815
+    },
+    {
+      "epoch": 4.8117950808533765,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9926280744666394e-05,
+      "loss": 0.8327,
+      "step": 70820
+    },
+    {
+      "epoch": 4.812134800924039,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.992415749422476e-05,
+      "loss": 0.8644,
+      "step": 70825
+    },
+    {
+      "epoch": 4.812474520994701,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9922034243783126e-05,
+      "loss": 0.8929,
+      "step": 70830
+    },
+    {
+      "epoch": 4.812814241065362,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9919910993341486e-05,
+      "loss": 0.9761,
+      "step": 70835
+    },
+    {
+      "epoch": 4.813153961136024,
+      "grad_norm": 7.625,
+      "learning_rate": 1.991778774289985e-05,
+      "loss": 0.92,
+      "step": 70840
+    },
+    {
+      "epoch": 4.813493681206686,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9915664492458218e-05,
+      "loss": 0.9223,
+      "step": 70845
+    },
+    {
+      "epoch": 4.813833401277347,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.991354124201658e-05,
+      "loss": 0.8725,
+      "step": 70850
+    },
+    {
+      "epoch": 4.814173121348009,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9911417991574942e-05,
+      "loss": 0.8304,
+      "step": 70855
+    },
+    {
+      "epoch": 4.814512841418671,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9909294741133306e-05,
+      "loss": 0.907,
+      "step": 70860
+    },
+    {
+      "epoch": 4.8148525614893325,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.990717149069167e-05,
+      "loss": 0.8077,
+      "step": 70865
+    },
+    {
+      "epoch": 4.815192281559995,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9905048240250034e-05,
+      "loss": 0.8492,
+      "step": 70870
+    },
+    {
+      "epoch": 4.815532001630657,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.99029249898084e-05,
+      "loss": 0.8673,
+      "step": 70875
+    },
+    {
+      "epoch": 4.815871721701318,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9900801739366762e-05,
+      "loss": 0.8602,
+      "step": 70880
+    },
+    {
+      "epoch": 4.81621144177198,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.9898678488925126e-05,
+      "loss": 0.8202,
+      "step": 70885
+    },
+    {
+      "epoch": 4.816551161842642,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.989655523848349e-05,
+      "loss": 0.9693,
+      "step": 70890
+    },
+    {
+      "epoch": 4.816890881913303,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9894431988041854e-05,
+      "loss": 0.8988,
+      "step": 70895
+    },
+    {
+      "epoch": 4.817230601983965,
+      "grad_norm": 8.5,
+      "learning_rate": 1.989230873760022e-05,
+      "loss": 0.8492,
+      "step": 70900
+    },
+    {
+      "epoch": 4.817570322054627,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9890185487158582e-05,
+      "loss": 0.8649,
+      "step": 70905
+    },
+    {
+      "epoch": 4.8179100421252885,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9888062236716946e-05,
+      "loss": 0.8636,
+      "step": 70910
+    },
+    {
+      "epoch": 4.818249762195951,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.988593898627531e-05,
+      "loss": 0.811,
+      "step": 70915
+    },
+    {
+      "epoch": 4.818589482266613,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.9883815735833675e-05,
+      "loss": 0.8605,
+      "step": 70920
+    },
+    {
+      "epoch": 4.818929202337274,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9881692485392035e-05,
+      "loss": 0.8863,
+      "step": 70925
+    },
+    {
+      "epoch": 4.819268922407936,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9879569234950403e-05,
+      "loss": 0.8398,
+      "step": 70930
+    },
+    {
+      "epoch": 4.819608642478598,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.9877445984508767e-05,
+      "loss": 0.8632,
+      "step": 70935
+    },
+    {
+      "epoch": 4.819948362549259,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9875322734067127e-05,
+      "loss": 0.8779,
+      "step": 70940
+    },
+    {
+      "epoch": 4.820288082619921,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9873199483625495e-05,
+      "loss": 0.7876,
+      "step": 70945
+    },
+    {
+      "epoch": 4.820627802690583,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.987107623318386e-05,
+      "loss": 0.875,
+      "step": 70950
+    },
+    {
+      "epoch": 4.8209675227612445,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9868952982742223e-05,
+      "loss": 0.918,
+      "step": 70955
+    },
+    {
+      "epoch": 4.821307242831907,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9866829732300583e-05,
+      "loss": 0.9031,
+      "step": 70960
+    },
+    {
+      "epoch": 4.821646962902569,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.986470648185895e-05,
+      "loss": 0.8483,
+      "step": 70965
+    },
+    {
+      "epoch": 4.82198668297323,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.9862583231417315e-05,
+      "loss": 0.988,
+      "step": 70970
+    },
+    {
+      "epoch": 4.822326403043892,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.9860459980975675e-05,
+      "loss": 0.8795,
+      "step": 70975
+    },
+    {
+      "epoch": 4.822666123114553,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.9858336730534043e-05,
+      "loss": 0.8759,
+      "step": 70980
+    },
+    {
+      "epoch": 4.823005843185215,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.9856213480092407e-05,
+      "loss": 0.9077,
+      "step": 70985
+    },
+    {
+      "epoch": 4.823345563255877,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9854090229650767e-05,
+      "loss": 0.864,
+      "step": 70990
+    },
+    {
+      "epoch": 4.8236852833265385,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.985196697920913e-05,
+      "loss": 0.8078,
+      "step": 70995
+    },
+    {
+      "epoch": 4.8240250033972005,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.98498437287675e-05,
+      "loss": 0.9254,
+      "step": 71000
+    },
+    {
+      "epoch": 4.824364723467863,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.984772047832586e-05,
+      "loss": 0.9716,
+      "step": 71005
+    },
+    {
+      "epoch": 4.824704443538524,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.9845597227884223e-05,
+      "loss": 0.8546,
+      "step": 71010
+    },
+    {
+      "epoch": 4.825044163609186,
+      "grad_norm": 8.25,
+      "learning_rate": 1.984347397744259e-05,
+      "loss": 0.8103,
+      "step": 71015
+    },
+    {
+      "epoch": 4.825383883679848,
+      "grad_norm": 8.0,
+      "learning_rate": 1.984135072700095e-05,
+      "loss": 0.8493,
+      "step": 71020
+    },
+    {
+      "epoch": 4.825723603750509,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9839227476559315e-05,
+      "loss": 0.9107,
+      "step": 71025
+    },
+    {
+      "epoch": 4.826063323821171,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9837104226117683e-05,
+      "loss": 0.852,
+      "step": 71030
+    },
+    {
+      "epoch": 4.826403043891833,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9834980975676043e-05,
+      "loss": 0.8297,
+      "step": 71035
+    },
+    {
+      "epoch": 4.8267427639624945,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.9832857725234407e-05,
+      "loss": 0.8978,
+      "step": 71040
+    },
+    {
+      "epoch": 4.8270824840331565,
+      "grad_norm": 8.125,
+      "learning_rate": 1.983073447479277e-05,
+      "loss": 0.8681,
+      "step": 71045
+    },
+    {
+      "epoch": 4.827422204103819,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9828611224351135e-05,
+      "loss": 0.846,
+      "step": 71050
+    },
+    {
+      "epoch": 4.82776192417448,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.98264879739095e-05,
+      "loss": 0.8445,
+      "step": 71055
+    },
+    {
+      "epoch": 4.828101644245142,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9824364723467863e-05,
+      "loss": 0.8856,
+      "step": 71060
+    },
+    {
+      "epoch": 4.828441364315804,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9822241473026227e-05,
+      "loss": 0.9096,
+      "step": 71065
+    },
+    {
+      "epoch": 4.828781084386465,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.982011822258459e-05,
+      "loss": 0.8162,
+      "step": 71070
+    },
+    {
+      "epoch": 4.829120804457127,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.9817994972142955e-05,
+      "loss": 0.7896,
+      "step": 71075
+    },
+    {
+      "epoch": 4.829460524527789,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.981587172170132e-05,
+      "loss": 0.9253,
+      "step": 71080
+    },
+    {
+      "epoch": 4.8298002445984505,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.9813748471259683e-05,
+      "loss": 0.8371,
+      "step": 71085
+    },
+    {
+      "epoch": 4.8301399646691126,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9811625220818047e-05,
+      "loss": 0.8542,
+      "step": 71090
+    },
+    {
+      "epoch": 4.830479684739775,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.9809501970376408e-05,
+      "loss": 0.8745,
+      "step": 71095
+    },
+    {
+      "epoch": 4.830819404810436,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9807378719934775e-05,
+      "loss": 0.8654,
+      "step": 71100
+    },
+    {
+      "epoch": 4.831159124881098,
+      "grad_norm": 8.125,
+      "learning_rate": 1.980525546949314e-05,
+      "loss": 0.8667,
+      "step": 71105
+    },
+    {
+      "epoch": 4.83149884495176,
+      "grad_norm": 8.375,
+      "learning_rate": 1.98031322190515e-05,
+      "loss": 0.9316,
+      "step": 71110
+    },
+    {
+      "epoch": 4.831838565022421,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9801008968609867e-05,
+      "loss": 0.9069,
+      "step": 71115
+    },
+    {
+      "epoch": 4.832178285093083,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.979888571816823e-05,
+      "loss": 0.9261,
+      "step": 71120
+    },
+    {
+      "epoch": 4.832518005163745,
+      "grad_norm": 7.125,
+      "learning_rate": 1.9796762467726595e-05,
+      "loss": 0.936,
+      "step": 71125
+    },
+    {
+      "epoch": 4.8328577252344065,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.979463921728496e-05,
+      "loss": 0.844,
+      "step": 71130
+    },
+    {
+      "epoch": 4.833197445305069,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9792515966843323e-05,
+      "loss": 0.9606,
+      "step": 71135
+    },
+    {
+      "epoch": 4.833537165375731,
+      "grad_norm": 8.875,
+      "learning_rate": 1.9790392716401687e-05,
+      "loss": 0.9324,
+      "step": 71140
+    },
+    {
+      "epoch": 4.833876885446392,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9788269465960048e-05,
+      "loss": 0.852,
+      "step": 71145
+    },
+    {
+      "epoch": 4.834216605517054,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9786146215518415e-05,
+      "loss": 0.8858,
+      "step": 71150
+    },
+    {
+      "epoch": 4.834556325587716,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.978402296507678e-05,
+      "loss": 0.9427,
+      "step": 71155
+    },
+    {
+      "epoch": 4.834896045658377,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.978189971463514e-05,
+      "loss": 0.9219,
+      "step": 71160
+    },
+    {
+      "epoch": 4.835235765729039,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9779776464193507e-05,
+      "loss": 0.8709,
+      "step": 71165
+    },
+    {
+      "epoch": 4.835575485799701,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.977765321375187e-05,
+      "loss": 0.9064,
+      "step": 71170
+    },
+    {
+      "epoch": 4.8359152058703625,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.9775529963310232e-05,
+      "loss": 0.9276,
+      "step": 71175
+    },
+    {
+      "epoch": 4.836254925941025,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9773406712868596e-05,
+      "loss": 0.8448,
+      "step": 71180
+    },
+    {
+      "epoch": 4.836594646011687,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.9771283462426963e-05,
+      "loss": 0.8267,
+      "step": 71185
+    },
+    {
+      "epoch": 4.836934366082348,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.9769160211985324e-05,
+      "loss": 0.8049,
+      "step": 71190
+    },
+    {
+      "epoch": 4.83727408615301,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9767036961543688e-05,
+      "loss": 0.8127,
+      "step": 71195
+    },
+    {
+      "epoch": 4.837613806223672,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9764913711102055e-05,
+      "loss": 0.8823,
+      "step": 71200
+    },
+    {
+      "epoch": 4.837953526294333,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9762790460660416e-05,
+      "loss": 0.8369,
+      "step": 71205
+    },
+    {
+      "epoch": 4.838293246364995,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.976066721021878e-05,
+      "loss": 0.8805,
+      "step": 71210
+    },
+    {
+      "epoch": 4.838632966435657,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9758543959777144e-05,
+      "loss": 0.8888,
+      "step": 71215
+    },
+    {
+      "epoch": 4.8389726865063185,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9756420709335508e-05,
+      "loss": 0.8504,
+      "step": 71220
+    },
+    {
+      "epoch": 4.839312406576981,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9754297458893872e-05,
+      "loss": 0.9004,
+      "step": 71225
+    },
+    {
+      "epoch": 4.839652126647643,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.9752174208452236e-05,
+      "loss": 0.8454,
+      "step": 71230
+    },
+    {
+      "epoch": 4.839991846718304,
+      "grad_norm": 8.375,
+      "learning_rate": 1.97500509580106e-05,
+      "loss": 0.8368,
+      "step": 71235
+    },
+    {
+      "epoch": 4.840331566788966,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9747927707568964e-05,
+      "loss": 0.9004,
+      "step": 71240
+    },
+    {
+      "epoch": 4.840671286859628,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9745804457127328e-05,
+      "loss": 0.8923,
+      "step": 71245
+    },
+    {
+      "epoch": 4.841011006930289,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9743681206685692e-05,
+      "loss": 0.8551,
+      "step": 71250
+    },
+    {
+      "epoch": 4.841350727000951,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9741557956244056e-05,
+      "loss": 0.9157,
+      "step": 71255
+    },
+    {
+      "epoch": 4.841690447071613,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.973943470580242e-05,
+      "loss": 0.8954,
+      "step": 71260
+    },
+    {
+      "epoch": 4.8420301671422745,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9737311455360784e-05,
+      "loss": 0.884,
+      "step": 71265
+    },
+    {
+      "epoch": 4.842369887212937,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.9735188204919148e-05,
+      "loss": 0.8246,
+      "step": 71270
+    },
+    {
+      "epoch": 4.842709607283599,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9733064954477512e-05,
+      "loss": 0.8136,
+      "step": 71275
+    },
+    {
+      "epoch": 4.84304932735426,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9730941704035873e-05,
+      "loss": 0.9692,
+      "step": 71280
+    },
+    {
+      "epoch": 4.843389047424922,
+      "grad_norm": 7.25,
+      "learning_rate": 1.972881845359424e-05,
+      "loss": 0.8544,
+      "step": 71285
+    },
+    {
+      "epoch": 4.843728767495584,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9726695203152604e-05,
+      "loss": 0.8251,
+      "step": 71290
+    },
+    {
+      "epoch": 4.844068487566245,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9724571952710968e-05,
+      "loss": 0.8503,
+      "step": 71295
+    },
+    {
+      "epoch": 4.844408207636907,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9722448702269332e-05,
+      "loss": 0.8222,
+      "step": 71300
+    },
+    {
+      "epoch": 4.844747927707569,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9720325451827696e-05,
+      "loss": 0.9783,
+      "step": 71305
+    },
+    {
+      "epoch": 4.8450876477782305,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.971820220138606e-05,
+      "loss": 0.8567,
+      "step": 71310
+    },
+    {
+      "epoch": 4.845427367848893,
+      "grad_norm": 8.125,
+      "learning_rate": 1.971607895094442e-05,
+      "loss": 0.9363,
+      "step": 71315
+    },
+    {
+      "epoch": 4.845767087919555,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9713955700502788e-05,
+      "loss": 0.9331,
+      "step": 71320
+    },
+    {
+      "epoch": 4.846106807990216,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9711832450061152e-05,
+      "loss": 0.9208,
+      "step": 71325
+    },
+    {
+      "epoch": 4.846446528060878,
+      "grad_norm": 8.75,
+      "learning_rate": 1.9709709199619513e-05,
+      "loss": 0.8564,
+      "step": 71330
+    },
+    {
+      "epoch": 4.84678624813154,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.970758594917788e-05,
+      "loss": 0.8692,
+      "step": 71335
+    },
+    {
+      "epoch": 4.847125968202201,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9705462698736244e-05,
+      "loss": 0.8481,
+      "step": 71340
+    },
+    {
+      "epoch": 4.847465688272863,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9703339448294605e-05,
+      "loss": 0.8733,
+      "step": 71345
+    },
+    {
+      "epoch": 4.847805408343525,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.970121619785297e-05,
+      "loss": 0.8738,
+      "step": 71350
+    },
+    {
+      "epoch": 4.8481451284141865,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.9699092947411336e-05,
+      "loss": 0.9011,
+      "step": 71355
+    },
+    {
+      "epoch": 4.848484848484849,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9696969696969697e-05,
+      "loss": 0.8449,
+      "step": 71360
+    },
+    {
+      "epoch": 4.848824568555511,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.969484644652806e-05,
+      "loss": 0.872,
+      "step": 71365
+    },
+    {
+      "epoch": 4.849164288626172,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.9692723196086428e-05,
+      "loss": 0.8216,
+      "step": 71370
+    },
+    {
+      "epoch": 4.849504008696834,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.969059994564479e-05,
+      "loss": 0.9061,
+      "step": 71375
+    },
+    {
+      "epoch": 4.849843728767496,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9688476695203153e-05,
+      "loss": 0.8524,
+      "step": 71380
+    },
+    {
+      "epoch": 4.850183448838157,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9686353444761517e-05,
+      "loss": 0.8903,
+      "step": 71385
+    },
+    {
+      "epoch": 4.850523168908819,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.968423019431988e-05,
+      "loss": 0.8376,
+      "step": 71390
+    },
+    {
+      "epoch": 4.850862888979481,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9682106943878245e-05,
+      "loss": 0.8715,
+      "step": 71395
+    },
+    {
+      "epoch": 4.8512026090501426,
+      "grad_norm": 7.25,
+      "learning_rate": 1.967998369343661e-05,
+      "loss": 0.9183,
+      "step": 71400
+    },
+    {
+      "epoch": 4.851542329120805,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.9677860442994973e-05,
+      "loss": 0.8181,
+      "step": 71405
+    },
+    {
+      "epoch": 4.851882049191467,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.9675737192553337e-05,
+      "loss": 0.8388,
+      "step": 71410
+    },
+    {
+      "epoch": 4.852221769262128,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.96736139421117e-05,
+      "loss": 0.8892,
+      "step": 71415
+    },
+    {
+      "epoch": 4.85256148933279,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9671490691670065e-05,
+      "loss": 0.8847,
+      "step": 71420
+    },
+    {
+      "epoch": 4.852901209403452,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.966936744122843e-05,
+      "loss": 0.859,
+      "step": 71425
+    },
+    {
+      "epoch": 4.853240929474113,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9667244190786793e-05,
+      "loss": 1.0077,
+      "step": 71430
+    },
+    {
+      "epoch": 4.853580649544775,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9665120940345157e-05,
+      "loss": 0.8526,
+      "step": 71435
+    },
+    {
+      "epoch": 4.8539203696154365,
+      "grad_norm": 6.875,
+      "learning_rate": 1.966299768990352e-05,
+      "loss": 0.8607,
+      "step": 71440
+    },
+    {
+      "epoch": 4.854260089686099,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.9660874439461885e-05,
+      "loss": 0.8604,
+      "step": 71445
+    },
+    {
+      "epoch": 4.854599809756761,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.9658751189020245e-05,
+      "loss": 0.8734,
+      "step": 71450
+    },
+    {
+      "epoch": 4.854939529827422,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9656627938578613e-05,
+      "loss": 0.9097,
+      "step": 71455
+    },
+    {
+      "epoch": 4.855279249898084,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9654504688136977e-05,
+      "loss": 0.867,
+      "step": 71460
+    },
+    {
+      "epoch": 4.855618969968746,
+      "grad_norm": 7.5,
+      "learning_rate": 1.965238143769534e-05,
+      "loss": 0.8207,
+      "step": 71465
+    },
+    {
+      "epoch": 4.855958690039407,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9650258187253705e-05,
+      "loss": 0.8889,
+      "step": 71470
+    },
+    {
+      "epoch": 4.856298410110069,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.964813493681207e-05,
+      "loss": 0.8561,
+      "step": 71475
+    },
+    {
+      "epoch": 4.856638130180731,
+      "grad_norm": 7.75,
+      "learning_rate": 1.9646011686370433e-05,
+      "loss": 0.8929,
+      "step": 71480
+    },
+    {
+      "epoch": 4.8569778502513925,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9643888435928793e-05,
+      "loss": 0.8502,
+      "step": 71485
+    },
+    {
+      "epoch": 4.857317570322055,
+      "grad_norm": 7.875,
+      "learning_rate": 1.964176518548716e-05,
+      "loss": 0.9034,
+      "step": 71490
+    },
+    {
+      "epoch": 4.857657290392717,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9639641935045525e-05,
+      "loss": 0.84,
+      "step": 71495
+    },
+    {
+      "epoch": 4.857997010463378,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.9637518684603885e-05,
+      "loss": 0.9605,
+      "step": 71500
+    },
+    {
+      "epoch": 4.85833673053404,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9635395434162253e-05,
+      "loss": 0.8023,
+      "step": 71505
+    },
+    {
+      "epoch": 4.858676450604702,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9633272183720617e-05,
+      "loss": 0.7972,
+      "step": 71510
+    },
+    {
+      "epoch": 4.859016170675363,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.9631148933278977e-05,
+      "loss": 0.8533,
+      "step": 71515
+    },
+    {
+      "epoch": 4.859355890746025,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.9629025682837345e-05,
+      "loss": 0.8637,
+      "step": 71520
+    },
+    {
+      "epoch": 4.859695610816687,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.962690243239571e-05,
+      "loss": 0.8932,
+      "step": 71525
+    },
+    {
+      "epoch": 4.8600353308873485,
+      "grad_norm": 8.5,
+      "learning_rate": 1.962477918195407e-05,
+      "loss": 1.0626,
+      "step": 71530
+    },
+    {
+      "epoch": 4.860375050958011,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9622655931512433e-05,
+      "loss": 0.9188,
+      "step": 71535
+    },
+    {
+      "epoch": 4.860714771028673,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.96205326810708e-05,
+      "loss": 0.8413,
+      "step": 71540
+    },
+    {
+      "epoch": 4.861054491099334,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.961840943062916e-05,
+      "loss": 1.022,
+      "step": 71545
+    },
+    {
+      "epoch": 4.861394211169996,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.9616286180187525e-05,
+      "loss": 0.908,
+      "step": 71550
+    },
+    {
+      "epoch": 4.861733931240658,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9614162929745893e-05,
+      "loss": 0.8903,
+      "step": 71555
+    },
+    {
+      "epoch": 4.862073651311319,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9612039679304253e-05,
+      "loss": 0.896,
+      "step": 71560
+    },
+    {
+      "epoch": 4.862413371381981,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9609916428862617e-05,
+      "loss": 0.7811,
+      "step": 71565
+    },
+    {
+      "epoch": 4.862753091452643,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.960779317842098e-05,
+      "loss": 0.8293,
+      "step": 71570
+    },
+    {
+      "epoch": 4.8630928115233045,
+      "grad_norm": 7.75,
+      "learning_rate": 1.9605669927979345e-05,
+      "loss": 0.8626,
+      "step": 71575
+    },
+    {
+      "epoch": 4.863432531593967,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.960354667753771e-05,
+      "loss": 0.905,
+      "step": 71580
+    },
+    {
+      "epoch": 4.863772251664629,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9601423427096073e-05,
+      "loss": 1.0289,
+      "step": 71585
+    },
+    {
+      "epoch": 4.86411197173529,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.9599300176654437e-05,
+      "loss": 0.7939,
+      "step": 71590
+    },
+    {
+      "epoch": 4.864451691805952,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.95971769262128e-05,
+      "loss": 0.9442,
+      "step": 71595
+    },
+    {
+      "epoch": 4.864791411876614,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9595053675771165e-05,
+      "loss": 0.9435,
+      "step": 71600
+    },
+    {
+      "epoch": 4.865131131947275,
+      "grad_norm": 8.375,
+      "learning_rate": 1.959293042532953e-05,
+      "loss": 0.8422,
+      "step": 71605
+    },
+    {
+      "epoch": 4.865470852017937,
+      "grad_norm": 6.875,
+      "learning_rate": 1.9590807174887893e-05,
+      "loss": 0.8371,
+      "step": 71610
+    },
+    {
+      "epoch": 4.865810572088599,
+      "grad_norm": 6.375,
+      "learning_rate": 1.9588683924446257e-05,
+      "loss": 0.8075,
+      "step": 71615
+    },
+    {
+      "epoch": 4.8661502921592605,
+      "grad_norm": 7.0,
+      "learning_rate": 1.958656067400462e-05,
+      "loss": 0.8485,
+      "step": 71620
+    },
+    {
+      "epoch": 4.866490012229923,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.9584437423562985e-05,
+      "loss": 0.892,
+      "step": 71625
+    },
+    {
+      "epoch": 4.866829732300585,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.958231417312135e-05,
+      "loss": 0.8501,
+      "step": 71630
+    },
+    {
+      "epoch": 4.867169452371246,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.9580190922679714e-05,
+      "loss": 0.8756,
+      "step": 71635
+    },
+    {
+      "epoch": 4.867509172441908,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9578067672238078e-05,
+      "loss": 0.877,
+      "step": 71640
+    },
+    {
+      "epoch": 4.86784889251257,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.957594442179644e-05,
+      "loss": 0.9984,
+      "step": 71645
+    },
+    {
+      "epoch": 4.868188612583231,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9573821171354806e-05,
+      "loss": 0.919,
+      "step": 71650
+    },
+    {
+      "epoch": 4.868528332653893,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.957169792091317e-05,
+      "loss": 0.857,
+      "step": 71655
+    },
+    {
+      "epoch": 4.8688680527245545,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9569574670471534e-05,
+      "loss": 0.8524,
+      "step": 71660
+    },
+    {
+      "epoch": 4.8692077727952165,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9567451420029898e-05,
+      "loss": 0.8179,
+      "step": 71665
+    },
+    {
+      "epoch": 4.869547492865879,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9565328169588258e-05,
+      "loss": 1.0393,
+      "step": 71670
+    },
+    {
+      "epoch": 4.86988721293654,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9563204919146626e-05,
+      "loss": 0.9278,
+      "step": 71675
+    },
+    {
+      "epoch": 4.870226933007202,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.956108166870499e-05,
+      "loss": 0.898,
+      "step": 71680
+    },
+    {
+      "epoch": 4.870566653077864,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.955895841826335e-05,
+      "loss": 0.7974,
+      "step": 71685
+    },
+    {
+      "epoch": 4.870906373148525,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.9556835167821718e-05,
+      "loss": 0.7994,
+      "step": 71690
+    },
+    {
+      "epoch": 4.871246093219187,
+      "grad_norm": 8.5,
+      "learning_rate": 1.955471191738008e-05,
+      "loss": 0.8548,
+      "step": 71695
+    },
+    {
+      "epoch": 4.871585813289849,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9552588666938442e-05,
+      "loss": 0.8146,
+      "step": 71700
+    },
+    {
+      "epoch": 4.8719255333605105,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9550465416496806e-05,
+      "loss": 0.8613,
+      "step": 71705
+    },
+    {
+      "epoch": 4.872265253431173,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9548342166055174e-05,
+      "loss": 0.8064,
+      "step": 71710
+    },
+    {
+      "epoch": 4.872604973501835,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9546218915613534e-05,
+      "loss": 0.8236,
+      "step": 71715
+    },
+    {
+      "epoch": 4.872944693572496,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9544095665171898e-05,
+      "loss": 0.9172,
+      "step": 71720
+    },
+    {
+      "epoch": 4.873284413643158,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9541972414730266e-05,
+      "loss": 0.8663,
+      "step": 71725
+    },
+    {
+      "epoch": 4.87362413371382,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9539849164288626e-05,
+      "loss": 0.9464,
+      "step": 71730
+    },
+    {
+      "epoch": 4.873963853784481,
+      "grad_norm": 7.875,
+      "learning_rate": 1.953772591384699e-05,
+      "loss": 0.8881,
+      "step": 71735
+    },
+    {
+      "epoch": 4.874303573855143,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.9535602663405354e-05,
+      "loss": 0.903,
+      "step": 71740
+    },
+    {
+      "epoch": 4.874643293925805,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9533479412963718e-05,
+      "loss": 0.9346,
+      "step": 71745
+    },
+    {
+      "epoch": 4.8749830139964665,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9531356162522082e-05,
+      "loss": 0.8841,
+      "step": 71750
+    },
+    {
+      "epoch": 4.875322734067129,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9529232912080446e-05,
+      "loss": 0.9033,
+      "step": 71755
+    },
+    {
+      "epoch": 4.875662454137791,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.952710966163881e-05,
+      "loss": 0.9961,
+      "step": 71760
+    },
+    {
+      "epoch": 4.876002174208452,
+      "grad_norm": 8.875,
+      "learning_rate": 1.9524986411197174e-05,
+      "loss": 0.9358,
+      "step": 71765
+    },
+    {
+      "epoch": 4.876341894279114,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.9522863160755538e-05,
+      "loss": 0.8035,
+      "step": 71770
+    },
+    {
+      "epoch": 4.876681614349776,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9520739910313902e-05,
+      "loss": 0.8426,
+      "step": 71775
+    },
+    {
+      "epoch": 4.877021334420437,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9518616659872266e-05,
+      "loss": 0.8345,
+      "step": 71780
+    },
+    {
+      "epoch": 4.877361054491099,
+      "grad_norm": 8.875,
+      "learning_rate": 1.951649340943063e-05,
+      "loss": 0.907,
+      "step": 71785
+    },
+    {
+      "epoch": 4.877700774561761,
+      "grad_norm": 8.75,
+      "learning_rate": 1.9514370158988994e-05,
+      "loss": 0.8171,
+      "step": 71790
+    },
+    {
+      "epoch": 4.8780404946324225,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9512246908547358e-05,
+      "loss": 0.8651,
+      "step": 71795
+    },
+    {
+      "epoch": 4.878380214703085,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.9510123658105722e-05,
+      "loss": 0.9037,
+      "step": 71800
+    },
+    {
+      "epoch": 4.878719934773747,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9508000407664086e-05,
+      "loss": 0.8223,
+      "step": 71805
+    },
+    {
+      "epoch": 4.879059654844408,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.950587715722245e-05,
+      "loss": 0.8217,
+      "step": 71810
+    },
+    {
+      "epoch": 4.87939937491507,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9503753906780814e-05,
+      "loss": 0.9357,
+      "step": 71815
+    },
+    {
+      "epoch": 4.879739094985732,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9501630656339178e-05,
+      "loss": 0.8344,
+      "step": 71820
+    },
+    {
+      "epoch": 4.880078815056393,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9499507405897542e-05,
+      "loss": 0.885,
+      "step": 71825
+    },
+    {
+      "epoch": 4.880418535127055,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9497384155455906e-05,
+      "loss": 0.7952,
+      "step": 71830
+    },
+    {
+      "epoch": 4.880758255197717,
+      "grad_norm": 8.125,
+      "learning_rate": 1.949526090501427e-05,
+      "loss": 0.9072,
+      "step": 71835
+    },
+    {
+      "epoch": 4.8810979752683785,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.949313765457263e-05,
+      "loss": 0.8749,
+      "step": 71840
+    },
+    {
+      "epoch": 4.881437695339041,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9491014404130998e-05,
+      "loss": 0.8848,
+      "step": 71845
+    },
+    {
+      "epoch": 4.881777415409703,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.9488891153689362e-05,
+      "loss": 0.8361,
+      "step": 71850
+    },
+    {
+      "epoch": 4.882117135480364,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9486767903247723e-05,
+      "loss": 0.8926,
+      "step": 71855
+    },
+    {
+      "epoch": 4.882456855551026,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.948464465280609e-05,
+      "loss": 0.9364,
+      "step": 71860
+    },
+    {
+      "epoch": 4.882796575621688,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.9482521402364454e-05,
+      "loss": 0.9085,
+      "step": 71865
+    },
+    {
+      "epoch": 4.883136295692349,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9480398151922815e-05,
+      "loss": 0.9661,
+      "step": 71870
+    },
+    {
+      "epoch": 4.883476015763011,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.947827490148118e-05,
+      "loss": 0.8326,
+      "step": 71875
+    },
+    {
+      "epoch": 4.883815735833673,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9476151651039546e-05,
+      "loss": 0.8598,
+      "step": 71880
+    },
+    {
+      "epoch": 4.8841554559043345,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9474028400597907e-05,
+      "loss": 0.8639,
+      "step": 71885
+    },
+    {
+      "epoch": 4.884495175974997,
+      "grad_norm": 7.5,
+      "learning_rate": 1.947190515015627e-05,
+      "loss": 0.8393,
+      "step": 71890
+    },
+    {
+      "epoch": 4.884834896045659,
+      "grad_norm": 8.375,
+      "learning_rate": 1.946978189971464e-05,
+      "loss": 0.903,
+      "step": 71895
+    },
+    {
+      "epoch": 4.88517461611632,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.9467658649273e-05,
+      "loss": 0.8161,
+      "step": 71900
+    },
+    {
+      "epoch": 4.885514336186982,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9465535398831363e-05,
+      "loss": 0.9318,
+      "step": 71905
+    },
+    {
+      "epoch": 4.885854056257644,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9463412148389727e-05,
+      "loss": 0.8219,
+      "step": 71910
+    },
+    {
+      "epoch": 4.886193776328305,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.946128889794809e-05,
+      "loss": 0.7853,
+      "step": 71915
+    },
+    {
+      "epoch": 4.886533496398967,
+      "grad_norm": 7.5,
+      "learning_rate": 1.9459165647506455e-05,
+      "loss": 0.8339,
+      "step": 71920
+    },
+    {
+      "epoch": 4.886873216469629,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.945704239706482e-05,
+      "loss": 0.879,
+      "step": 71925
+    },
+    {
+      "epoch": 4.8872129365402905,
+      "grad_norm": 12.875,
+      "learning_rate": 1.9454919146623183e-05,
+      "loss": 0.7966,
+      "step": 71930
+    },
+    {
+      "epoch": 4.887552656610953,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9452795896181547e-05,
+      "loss": 0.9104,
+      "step": 71935
+    },
+    {
+      "epoch": 4.887892376681615,
+      "grad_norm": 7.625,
+      "learning_rate": 1.945067264573991e-05,
+      "loss": 0.9144,
+      "step": 71940
+    },
+    {
+      "epoch": 4.888232096752276,
+      "grad_norm": 6.75,
+      "learning_rate": 1.9448549395298275e-05,
+      "loss": 0.878,
+      "step": 71945
+    },
+    {
+      "epoch": 4.888571816822938,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.944642614485664e-05,
+      "loss": 0.8255,
+      "step": 71950
+    },
+    {
+      "epoch": 4.8889115368936,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9444302894415003e-05,
+      "loss": 0.8557,
+      "step": 71955
+    },
+    {
+      "epoch": 4.889251256964261,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9442179643973367e-05,
+      "loss": 0.8525,
+      "step": 71960
+    },
+    {
+      "epoch": 4.889590977034923,
+      "grad_norm": 8.125,
+      "learning_rate": 1.944005639353173e-05,
+      "loss": 0.7952,
+      "step": 71965
+    },
+    {
+      "epoch": 4.889930697105585,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9437933143090095e-05,
+      "loss": 0.8826,
+      "step": 71970
+    },
+    {
+      "epoch": 4.8902704171762466,
+      "grad_norm": 7.375,
+      "learning_rate": 1.943580989264846e-05,
+      "loss": 0.8644,
+      "step": 71975
+    },
+    {
+      "epoch": 4.890610137246909,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9433686642206823e-05,
+      "loss": 0.8175,
+      "step": 71980
+    },
+    {
+      "epoch": 4.890949857317571,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9431563391765187e-05,
+      "loss": 0.8802,
+      "step": 71985
+    },
+    {
+      "epoch": 4.891289577388232,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.942944014132355e-05,
+      "loss": 0.8776,
+      "step": 71990
+    },
+    {
+      "epoch": 4.891629297458894,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.9427316890881915e-05,
+      "loss": 0.928,
+      "step": 71995
+    },
+    {
+      "epoch": 4.891969017529556,
+      "grad_norm": 8.0,
+      "learning_rate": 1.942519364044028e-05,
+      "loss": 0.8633,
+      "step": 72000
+    },
+    {
+      "epoch": 4.892308737600217,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9423070389998643e-05,
+      "loss": 0.7798,
+      "step": 72005
+    },
+    {
+      "epoch": 4.892648457670879,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.9420947139557004e-05,
+      "loss": 0.7778,
+      "step": 72010
+    },
+    {
+      "epoch": 4.892988177741541,
+      "grad_norm": 8.0,
+      "learning_rate": 1.941882388911537e-05,
+      "loss": 0.956,
+      "step": 72015
+    },
+    {
+      "epoch": 4.893327897812203,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9416700638673735e-05,
+      "loss": 0.9167,
+      "step": 72020
+    },
+    {
+      "epoch": 4.893667617882865,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9414577388232096e-05,
+      "loss": 0.8818,
+      "step": 72025
+    },
+    {
+      "epoch": 4.894007337953527,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9412454137790463e-05,
+      "loss": 0.939,
+      "step": 72030
+    },
+    {
+      "epoch": 4.894347058024188,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.9410330887348827e-05,
+      "loss": 0.8497,
+      "step": 72035
+    },
+    {
+      "epoch": 4.89468677809485,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9408207636907188e-05,
+      "loss": 0.9525,
+      "step": 72040
+    },
+    {
+      "epoch": 4.895026498165512,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9406084386465555e-05,
+      "loss": 0.9235,
+      "step": 72045
+    },
+    {
+      "epoch": 4.895366218236173,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.940396113602392e-05,
+      "loss": 0.8904,
+      "step": 72050
+    },
+    {
+      "epoch": 4.895705938306835,
+      "grad_norm": 8.375,
+      "learning_rate": 1.940183788558228e-05,
+      "loss": 0.8466,
+      "step": 72055
+    },
+    {
+      "epoch": 4.896045658377497,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9399714635140644e-05,
+      "loss": 0.817,
+      "step": 72060
+    },
+    {
+      "epoch": 4.896385378448159,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.939759138469901e-05,
+      "loss": 0.9703,
+      "step": 72065
+    },
+    {
+      "epoch": 4.896725098518821,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.939546813425737e-05,
+      "loss": 0.9042,
+      "step": 72070
+    },
+    {
+      "epoch": 4.897064818589483,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9393344883815736e-05,
+      "loss": 0.9,
+      "step": 72075
+    },
+    {
+      "epoch": 4.897404538660144,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.9391221633374103e-05,
+      "loss": 0.9151,
+      "step": 72080
+    },
+    {
+      "epoch": 4.897744258730806,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9389098382932464e-05,
+      "loss": 0.9147,
+      "step": 72085
+    },
+    {
+      "epoch": 4.898083978801468,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.9386975132490828e-05,
+      "loss": 0.8314,
+      "step": 72090
+    },
+    {
+      "epoch": 4.898423698872129,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9384851882049192e-05,
+      "loss": 0.8755,
+      "step": 72095
+    },
+    {
+      "epoch": 4.898763418942791,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9382728631607556e-05,
+      "loss": 0.89,
+      "step": 72100
+    },
+    {
+      "epoch": 4.899103139013453,
+      "grad_norm": 8.375,
+      "learning_rate": 1.938060538116592e-05,
+      "loss": 0.8945,
+      "step": 72105
+    },
+    {
+      "epoch": 4.899442859084115,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9378482130724284e-05,
+      "loss": 0.8398,
+      "step": 72110
+    },
+    {
+      "epoch": 4.899782579154777,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9376358880282648e-05,
+      "loss": 0.8061,
+      "step": 72115
+    },
+    {
+      "epoch": 4.900122299225439,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9374235629841012e-05,
+      "loss": 0.9034,
+      "step": 72120
+    },
+    {
+      "epoch": 4.9004620192961,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9372112379399376e-05,
+      "loss": 0.7961,
+      "step": 72125
+    },
+    {
+      "epoch": 4.900801739366762,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.936998912895774e-05,
+      "loss": 0.8381,
+      "step": 72130
+    },
+    {
+      "epoch": 4.901141459437423,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9367865878516104e-05,
+      "loss": 0.9069,
+      "step": 72135
+    },
+    {
+      "epoch": 4.901481179508085,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.9365742628074468e-05,
+      "loss": 0.8591,
+      "step": 72140
+    },
+    {
+      "epoch": 4.901820899578747,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9363619377632832e-05,
+      "loss": 0.8813,
+      "step": 72145
+    },
+    {
+      "epoch": 4.9021606196494085,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9361496127191196e-05,
+      "loss": 0.8998,
+      "step": 72150
+    },
+    {
+      "epoch": 4.902500339720071,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.935937287674956e-05,
+      "loss": 0.8662,
+      "step": 72155
+    },
+    {
+      "epoch": 4.902840059790733,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.9357249626307924e-05,
+      "loss": 0.8434,
+      "step": 72160
+    },
+    {
+      "epoch": 4.903179779861394,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9355126375866288e-05,
+      "loss": 0.9697,
+      "step": 72165
+    },
+    {
+      "epoch": 4.903519499932056,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9353003125424652e-05,
+      "loss": 0.8296,
+      "step": 72170
+    },
+    {
+      "epoch": 4.903859220002718,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9350879874983016e-05,
+      "loss": 0.9017,
+      "step": 72175
+    },
+    {
+      "epoch": 4.904198940073379,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.934875662454138e-05,
+      "loss": 0.9861,
+      "step": 72180
+    },
+    {
+      "epoch": 4.904538660144041,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.9346633374099744e-05,
+      "loss": 0.8755,
+      "step": 72185
+    },
+    {
+      "epoch": 4.904878380214703,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9344510123658108e-05,
+      "loss": 0.9701,
+      "step": 72190
+    },
+    {
+      "epoch": 4.9052181002853645,
+      "grad_norm": 5.71875,
+      "learning_rate": 1.934238687321647e-05,
+      "loss": 0.887,
+      "step": 72195
+    },
+    {
+      "epoch": 4.905557820356027,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9340263622774836e-05,
+      "loss": 0.8569,
+      "step": 72200
+    },
+    {
+      "epoch": 4.905897540426689,
+      "grad_norm": 8.625,
+      "learning_rate": 1.93381403723332e-05,
+      "loss": 0.9029,
+      "step": 72205
+    },
+    {
+      "epoch": 4.90623726049735,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.933601712189156e-05,
+      "loss": 0.9141,
+      "step": 72210
+    },
+    {
+      "epoch": 4.906576980568012,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.9333893871449928e-05,
+      "loss": 0.8781,
+      "step": 72215
+    },
+    {
+      "epoch": 4.906916700638674,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9331770621008292e-05,
+      "loss": 0.9035,
+      "step": 72220
+    },
+    {
+      "epoch": 4.907256420709335,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9329647370566652e-05,
+      "loss": 0.9001,
+      "step": 72225
+    },
+    {
+      "epoch": 4.907596140779997,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9327524120125016e-05,
+      "loss": 0.7843,
+      "step": 72230
+    },
+    {
+      "epoch": 4.907935860850659,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9325400869683384e-05,
+      "loss": 0.8962,
+      "step": 72235
+    },
+    {
+      "epoch": 4.9082755809213205,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9323277619241744e-05,
+      "loss": 0.8603,
+      "step": 72240
+    },
+    {
+      "epoch": 4.908615300991983,
+      "grad_norm": 8.25,
+      "learning_rate": 1.932115436880011e-05,
+      "loss": 0.885,
+      "step": 72245
+    },
+    {
+      "epoch": 4.908955021062645,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9319031118358476e-05,
+      "loss": 0.9095,
+      "step": 72250
+    },
+    {
+      "epoch": 4.909294741133306,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9316907867916836e-05,
+      "loss": 0.8605,
+      "step": 72255
+    },
+    {
+      "epoch": 4.909634461203968,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.93147846174752e-05,
+      "loss": 0.9431,
+      "step": 72260
+    },
+    {
+      "epoch": 4.90997418127463,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9312661367033564e-05,
+      "loss": 0.8653,
+      "step": 72265
+    },
+    {
+      "epoch": 4.910313901345291,
+      "grad_norm": 7.375,
+      "learning_rate": 1.931053811659193e-05,
+      "loss": 0.8488,
+      "step": 72270
+    },
+    {
+      "epoch": 4.910653621415953,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9308414866150292e-05,
+      "loss": 0.9284,
+      "step": 72275
+    },
+    {
+      "epoch": 4.910993341486615,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9306291615708656e-05,
+      "loss": 0.882,
+      "step": 72280
+    },
+    {
+      "epoch": 4.911333061557277,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.930416836526702e-05,
+      "loss": 0.9323,
+      "step": 72285
+    },
+    {
+      "epoch": 4.911672781627939,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9302045114825384e-05,
+      "loss": 0.8485,
+      "step": 72290
+    },
+    {
+      "epoch": 4.912012501698601,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.929992186438375e-05,
+      "loss": 0.8335,
+      "step": 72295
+    },
+    {
+      "epoch": 4.912352221769262,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9297798613942112e-05,
+      "loss": 0.8357,
+      "step": 72300
+    },
+    {
+      "epoch": 4.912691941839924,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9295675363500476e-05,
+      "loss": 0.8585,
+      "step": 72305
+    },
+    {
+      "epoch": 4.913031661910586,
+      "grad_norm": 8.125,
+      "learning_rate": 1.929355211305884e-05,
+      "loss": 0.8693,
+      "step": 72310
+    },
+    {
+      "epoch": 4.913371381981247,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.9291428862617204e-05,
+      "loss": 0.867,
+      "step": 72315
+    },
+    {
+      "epoch": 4.913711102051909,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.928930561217557e-05,
+      "loss": 0.7905,
+      "step": 72320
+    },
+    {
+      "epoch": 4.914050822122571,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.9287182361733932e-05,
+      "loss": 0.8726,
+      "step": 72325
+    },
+    {
+      "epoch": 4.914390542193233,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9285059111292296e-05,
+      "loss": 0.8869,
+      "step": 72330
+    },
+    {
+      "epoch": 4.914730262263895,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.928293586085066e-05,
+      "loss": 0.8326,
+      "step": 72335
+    },
+    {
+      "epoch": 4.915069982334556,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.9280812610409024e-05,
+      "loss": 0.8156,
+      "step": 72340
+    },
+    {
+      "epoch": 4.915409702405218,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.927868935996739e-05,
+      "loss": 0.8542,
+      "step": 72345
+    },
+    {
+      "epoch": 4.91574942247588,
+      "grad_norm": 7.0,
+      "learning_rate": 1.9276566109525753e-05,
+      "loss": 0.8932,
+      "step": 72350
+    },
+    {
+      "epoch": 4.916089142546541,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9274442859084117e-05,
+      "loss": 0.8711,
+      "step": 72355
+    },
+    {
+      "epoch": 4.916428862617203,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.927231960864248e-05,
+      "loss": 0.8631,
+      "step": 72360
+    },
+    {
+      "epoch": 4.916768582687865,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.927019635820084e-05,
+      "loss": 0.7763,
+      "step": 72365
+    },
+    {
+      "epoch": 4.9171083027585265,
+      "grad_norm": 7.375,
+      "learning_rate": 1.926807310775921e-05,
+      "loss": 0.8549,
+      "step": 72370
+    },
+    {
+      "epoch": 4.917448022829189,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9265949857317573e-05,
+      "loss": 0.8606,
+      "step": 72375
+    },
+    {
+      "epoch": 4.917787742899851,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9263826606875933e-05,
+      "loss": 0.8773,
+      "step": 72380
+    },
+    {
+      "epoch": 4.918127462970512,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.92617033564343e-05,
+      "loss": 0.7762,
+      "step": 72385
+    },
+    {
+      "epoch": 4.918467183041174,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9259580105992665e-05,
+      "loss": 0.8924,
+      "step": 72390
+    },
+    {
+      "epoch": 4.918806903111836,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9257456855551025e-05,
+      "loss": 0.9115,
+      "step": 72395
+    },
+    {
+      "epoch": 4.919146623182497,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.925533360510939e-05,
+      "loss": 0.8937,
+      "step": 72400
+    },
+    {
+      "epoch": 4.919486343253159,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.9253210354667757e-05,
+      "loss": 0.7857,
+      "step": 72405
+    },
+    {
+      "epoch": 4.919826063323821,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9251087104226117e-05,
+      "loss": 0.8767,
+      "step": 72410
+    },
+    {
+      "epoch": 4.9201657833944825,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.924896385378448e-05,
+      "loss": 0.8346,
+      "step": 72415
+    },
+    {
+      "epoch": 4.920505503465145,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.924684060334285e-05,
+      "loss": 0.8898,
+      "step": 72420
+    },
+    {
+      "epoch": 4.920845223535807,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.924471735290121e-05,
+      "loss": 0.878,
+      "step": 72425
+    },
+    {
+      "epoch": 4.921184943606468,
+      "grad_norm": 8.875,
+      "learning_rate": 1.9242594102459573e-05,
+      "loss": 0.8974,
+      "step": 72430
+    },
+    {
+      "epoch": 4.92152466367713,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.9240470852017937e-05,
+      "loss": 0.8325,
+      "step": 72435
+    },
+    {
+      "epoch": 4.921864383747792,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.92383476015763e-05,
+      "loss": 0.8529,
+      "step": 72440
+    },
+    {
+      "epoch": 4.922204103818453,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.9236224351134665e-05,
+      "loss": 0.887,
+      "step": 72445
+    },
+    {
+      "epoch": 4.922543823889115,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.923410110069303e-05,
+      "loss": 0.8374,
+      "step": 72450
+    },
+    {
+      "epoch": 4.922883543959777,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.9231977850251393e-05,
+      "loss": 0.7476,
+      "step": 72455
+    },
+    {
+      "epoch": 4.9232232640304385,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9229854599809757e-05,
+      "loss": 0.8527,
+      "step": 72460
+    },
+    {
+      "epoch": 4.923562984101101,
+      "grad_norm": 8.375,
+      "learning_rate": 1.922773134936812e-05,
+      "loss": 0.9203,
+      "step": 72465
+    },
+    {
+      "epoch": 4.923902704171763,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9225608098926485e-05,
+      "loss": 0.9322,
+      "step": 72470
+    },
+    {
+      "epoch": 4.924242424242424,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.922348484848485e-05,
+      "loss": 0.8221,
+      "step": 72475
+    },
+    {
+      "epoch": 4.924582144313086,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.9221361598043213e-05,
+      "loss": 0.8768,
+      "step": 72480
+    },
+    {
+      "epoch": 4.924921864383748,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9219238347601577e-05,
+      "loss": 0.8786,
+      "step": 72485
+    },
+    {
+      "epoch": 4.925261584454409,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.921711509715994e-05,
+      "loss": 0.8238,
+      "step": 72490
+    },
+    {
+      "epoch": 4.925601304525071,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.9214991846718305e-05,
+      "loss": 0.8412,
+      "step": 72495
+    },
+    {
+      "epoch": 4.925941024595733,
+      "grad_norm": 8.25,
+      "learning_rate": 1.921286859627667e-05,
+      "loss": 0.7662,
+      "step": 72500
+    },
+    {
+      "epoch": 4.9262807446663945,
+      "grad_norm": 10.0,
+      "learning_rate": 1.9210745345835033e-05,
+      "loss": 0.8219,
+      "step": 72505
+    },
+    {
+      "epoch": 4.926620464737057,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9208622095393397e-05,
+      "loss": 0.9178,
+      "step": 72510
+    },
+    {
+      "epoch": 4.926960184807719,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.920649884495176e-05,
+      "loss": 0.8617,
+      "step": 72515
+    },
+    {
+      "epoch": 4.92729990487838,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9204375594510125e-05,
+      "loss": 0.925,
+      "step": 72520
+    },
+    {
+      "epoch": 4.927639624949042,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.920225234406849e-05,
+      "loss": 0.8972,
+      "step": 72525
+    },
+    {
+      "epoch": 4.927979345019704,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9200129093626853e-05,
+      "loss": 0.8163,
+      "step": 72530
+    },
+    {
+      "epoch": 4.928319065090365,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9198005843185214e-05,
+      "loss": 0.9115,
+      "step": 72535
+    },
+    {
+      "epoch": 4.928658785161027,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.919588259274358e-05,
+      "loss": 0.8872,
+      "step": 72540
+    },
+    {
+      "epoch": 4.928998505231689,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.9193759342301945e-05,
+      "loss": 0.9266,
+      "step": 72545
+    },
+    {
+      "epoch": 4.9293382253023506,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9191636091860306e-05,
+      "loss": 0.8221,
+      "step": 72550
+    },
+    {
+      "epoch": 4.929677945373013,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9189512841418673e-05,
+      "loss": 0.866,
+      "step": 72555
+    },
+    {
+      "epoch": 4.930017665443675,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9187389590977037e-05,
+      "loss": 0.9062,
+      "step": 72560
+    },
+    {
+      "epoch": 4.930357385514336,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9185266340535398e-05,
+      "loss": 0.8373,
+      "step": 72565
+    },
+    {
+      "epoch": 4.930697105584998,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9183143090093765e-05,
+      "loss": 0.8929,
+      "step": 72570
+    },
+    {
+      "epoch": 4.93103682565566,
+      "grad_norm": 8.25,
+      "learning_rate": 1.918101983965213e-05,
+      "loss": 0.8351,
+      "step": 72575
+    },
+    {
+      "epoch": 4.931376545726321,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.917889658921049e-05,
+      "loss": 0.9202,
+      "step": 72580
+    },
+    {
+      "epoch": 4.931716265796983,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.9176773338768854e-05,
+      "loss": 0.9361,
+      "step": 72585
+    },
+    {
+      "epoch": 4.932055985867645,
+      "grad_norm": 8.25,
+      "learning_rate": 1.917465008832722e-05,
+      "loss": 0.8877,
+      "step": 72590
+    },
+    {
+      "epoch": 4.932395705938307,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.9172526837885582e-05,
+      "loss": 0.9005,
+      "step": 72595
+    },
+    {
+      "epoch": 4.932735426008969,
+      "grad_norm": 7.5,
+      "learning_rate": 1.9170403587443946e-05,
+      "loss": 0.8326,
+      "step": 72600
+    },
+    {
+      "epoch": 4.933075146079631,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9168280337002313e-05,
+      "loss": 0.9144,
+      "step": 72605
+    },
+    {
+      "epoch": 4.933414866150292,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9166157086560674e-05,
+      "loss": 0.9308,
+      "step": 72610
+    },
+    {
+      "epoch": 4.933754586220954,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9164033836119038e-05,
+      "loss": 0.8954,
+      "step": 72615
+    },
+    {
+      "epoch": 4.934094306291616,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9161910585677402e-05,
+      "loss": 0.9494,
+      "step": 72620
+    },
+    {
+      "epoch": 4.934434026362277,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9159787335235766e-05,
+      "loss": 0.8497,
+      "step": 72625
+    },
+    {
+      "epoch": 4.934773746432939,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.915766408479413e-05,
+      "loss": 0.957,
+      "step": 72630
+    },
+    {
+      "epoch": 4.935113466503601,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.9155540834352494e-05,
+      "loss": 0.7729,
+      "step": 72635
+    },
+    {
+      "epoch": 4.935453186574263,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9153417583910858e-05,
+      "loss": 0.9002,
+      "step": 72640
+    },
+    {
+      "epoch": 4.935792906644925,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9151294333469222e-05,
+      "loss": 0.8354,
+      "step": 72645
+    },
+    {
+      "epoch": 4.936132626715587,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9149171083027586e-05,
+      "loss": 0.7971,
+      "step": 72650
+    },
+    {
+      "epoch": 4.936472346786248,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.914704783258595e-05,
+      "loss": 0.8594,
+      "step": 72655
+    },
+    {
+      "epoch": 4.93681206685691,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9144924582144314e-05,
+      "loss": 0.9027,
+      "step": 72660
+    },
+    {
+      "epoch": 4.937151786927572,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9142801331702678e-05,
+      "loss": 0.8045,
+      "step": 72665
+    },
+    {
+      "epoch": 4.937491506998233,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9140678081261042e-05,
+      "loss": 0.8986,
+      "step": 72670
+    },
+    {
+      "epoch": 4.937831227068895,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.9138554830819406e-05,
+      "loss": 0.8477,
+      "step": 72675
+    },
+    {
+      "epoch": 4.938170947139557,
+      "grad_norm": 6.75,
+      "learning_rate": 1.913643158037777e-05,
+      "loss": 0.7933,
+      "step": 72680
+    },
+    {
+      "epoch": 4.938510667210219,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9134308329936134e-05,
+      "loss": 0.7966,
+      "step": 72685
+    },
+    {
+      "epoch": 4.938850387280881,
+      "grad_norm": 7.5,
+      "learning_rate": 1.9132185079494498e-05,
+      "loss": 0.8814,
+      "step": 72690
+    },
+    {
+      "epoch": 4.939190107351543,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.9130061829052862e-05,
+      "loss": 0.8745,
+      "step": 72695
+    },
+    {
+      "epoch": 4.939529827422204,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.9127938578611226e-05,
+      "loss": 0.8265,
+      "step": 72700
+    },
+    {
+      "epoch": 4.939869547492866,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.912581532816959e-05,
+      "loss": 0.8229,
+      "step": 72705
+    },
+    {
+      "epoch": 4.940209267563528,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9123692077727954e-05,
+      "loss": 0.8518,
+      "step": 72710
+    },
+    {
+      "epoch": 4.940548987634189,
+      "grad_norm": 8.625,
+      "learning_rate": 1.9121568827286318e-05,
+      "loss": 0.9101,
+      "step": 72715
+    },
+    {
+      "epoch": 4.940888707704851,
+      "grad_norm": 7.875,
+      "learning_rate": 1.911944557684468e-05,
+      "loss": 0.9029,
+      "step": 72720
+    },
+    {
+      "epoch": 4.941228427775513,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9117322326403046e-05,
+      "loss": 0.8101,
+      "step": 72725
+    },
+    {
+      "epoch": 4.941568147846175,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.911519907596141e-05,
+      "loss": 0.9032,
+      "step": 72730
+    },
+    {
+      "epoch": 4.941907867916837,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.911307582551977e-05,
+      "loss": 0.8545,
+      "step": 72735
+    },
+    {
+      "epoch": 4.942247587987499,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.9110952575078138e-05,
+      "loss": 0.8411,
+      "step": 72740
+    },
+    {
+      "epoch": 4.94258730805816,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9108829324636502e-05,
+      "loss": 0.9492,
+      "step": 72745
+    },
+    {
+      "epoch": 4.942927028128822,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9106706074194863e-05,
+      "loss": 0.9255,
+      "step": 72750
+    },
+    {
+      "epoch": 4.943266748199484,
+      "grad_norm": 7.625,
+      "learning_rate": 1.9104582823753227e-05,
+      "loss": 0.8965,
+      "step": 72755
+    },
+    {
+      "epoch": 4.943606468270145,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9102459573311594e-05,
+      "loss": 0.8015,
+      "step": 72760
+    },
+    {
+      "epoch": 4.943946188340807,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9100336322869955e-05,
+      "loss": 0.9438,
+      "step": 72765
+    },
+    {
+      "epoch": 4.944285908411469,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.909821307242832e-05,
+      "loss": 0.8595,
+      "step": 72770
+    },
+    {
+      "epoch": 4.944625628482131,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.9096089821986686e-05,
+      "loss": 0.9111,
+      "step": 72775
+    },
+    {
+      "epoch": 4.944965348552793,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.9093966571545047e-05,
+      "loss": 0.8852,
+      "step": 72780
+    },
+    {
+      "epoch": 4.945305068623455,
+      "grad_norm": 7.5,
+      "learning_rate": 1.909184332110341e-05,
+      "loss": 0.9149,
+      "step": 72785
+    },
+    {
+      "epoch": 4.945644788694116,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9089720070661775e-05,
+      "loss": 0.9577,
+      "step": 72790
+    },
+    {
+      "epoch": 4.945984508764778,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.908759682022014e-05,
+      "loss": 0.8789,
+      "step": 72795
+    },
+    {
+      "epoch": 4.94632422883544,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9085473569778503e-05,
+      "loss": 0.852,
+      "step": 72800
+    },
+    {
+      "epoch": 4.946663948906101,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9083350319336867e-05,
+      "loss": 0.8088,
+      "step": 72805
+    },
+    {
+      "epoch": 4.947003668976763,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.908122706889523e-05,
+      "loss": 0.8317,
+      "step": 72810
+    },
+    {
+      "epoch": 4.9473433890474245,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9079103818453595e-05,
+      "loss": 0.9102,
+      "step": 72815
+    },
+    {
+      "epoch": 4.947683109118087,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.907698056801196e-05,
+      "loss": 0.8693,
+      "step": 72820
+    },
+    {
+      "epoch": 4.948022829188749,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9074857317570323e-05,
+      "loss": 0.9353,
+      "step": 72825
+    },
+    {
+      "epoch": 4.94836254925941,
+      "grad_norm": 7.25,
+      "learning_rate": 1.9072734067128687e-05,
+      "loss": 0.8612,
+      "step": 72830
+    },
+    {
+      "epoch": 4.948702269330072,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.907061081668705e-05,
+      "loss": 0.9205,
+      "step": 72835
+    },
+    {
+      "epoch": 4.949041989400734,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9068487566245415e-05,
+      "loss": 0.8439,
+      "step": 72840
+    },
+    {
+      "epoch": 4.949381709471395,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.906636431580378e-05,
+      "loss": 0.873,
+      "step": 72845
+    },
+    {
+      "epoch": 4.949721429542057,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9064241065362143e-05,
+      "loss": 0.9029,
+      "step": 72850
+    },
+    {
+      "epoch": 4.950061149612719,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9062117814920507e-05,
+      "loss": 0.7907,
+      "step": 72855
+    },
+    {
+      "epoch": 4.9504008696833806,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.905999456447887e-05,
+      "loss": 0.8596,
+      "step": 72860
+    },
+    {
+      "epoch": 4.950740589754043,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.9057871314037235e-05,
+      "loss": 0.9198,
+      "step": 72865
+    },
+    {
+      "epoch": 4.951080309824705,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.90557480635956e-05,
+      "loss": 0.8778,
+      "step": 72870
+    },
+    {
+      "epoch": 4.951420029895366,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9053624813153963e-05,
+      "loss": 0.8681,
+      "step": 72875
+    },
+    {
+      "epoch": 4.951759749966028,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9051501562712327e-05,
+      "loss": 0.9163,
+      "step": 72880
+    },
+    {
+      "epoch": 4.95209947003669,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.904937831227069e-05,
+      "loss": 0.87,
+      "step": 72885
+    },
+    {
+      "epoch": 4.952439190107351,
+      "grad_norm": 7.75,
+      "learning_rate": 1.904725506182905e-05,
+      "loss": 0.8831,
+      "step": 72890
+    },
+    {
+      "epoch": 4.952778910178013,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.904513181138742e-05,
+      "loss": 0.8837,
+      "step": 72895
+    },
+    {
+      "epoch": 4.953118630248675,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.9043008560945783e-05,
+      "loss": 0.7952,
+      "step": 72900
+    },
+    {
+      "epoch": 4.953458350319337,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9040885310504143e-05,
+      "loss": 0.856,
+      "step": 72905
+    },
+    {
+      "epoch": 4.953798070389999,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.903876206006251e-05,
+      "loss": 0.8986,
+      "step": 72910
+    },
+    {
+      "epoch": 4.954137790460661,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9036638809620875e-05,
+      "loss": 0.7939,
+      "step": 72915
+    },
+    {
+      "epoch": 4.954477510531322,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.9034515559179235e-05,
+      "loss": 0.9352,
+      "step": 72920
+    },
+    {
+      "epoch": 4.954817230601984,
+      "grad_norm": 6.875,
+      "learning_rate": 1.90323923087376e-05,
+      "loss": 0.9243,
+      "step": 72925
+    },
+    {
+      "epoch": 4.955156950672646,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.9030269058295967e-05,
+      "loss": 0.9322,
+      "step": 72930
+    },
+    {
+      "epoch": 4.955496670743307,
+      "grad_norm": 6.03125,
+      "learning_rate": 1.9028145807854327e-05,
+      "loss": 0.75,
+      "step": 72935
+    },
+    {
+      "epoch": 4.955836390813969,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.902602255741269e-05,
+      "loss": 0.9351,
+      "step": 72940
+    },
+    {
+      "epoch": 4.956176110884631,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.902389930697106e-05,
+      "loss": 0.8586,
+      "step": 72945
+    },
+    {
+      "epoch": 4.956515830955293,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.902177605652942e-05,
+      "loss": 0.8428,
+      "step": 72950
+    },
+    {
+      "epoch": 4.956855551025955,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9019652806087783e-05,
+      "loss": 0.8437,
+      "step": 72955
+    },
+    {
+      "epoch": 4.957195271096617,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.901752955564615e-05,
+      "loss": 0.9683,
+      "step": 72960
+    },
+    {
+      "epoch": 4.957534991167278,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.901540630520451e-05,
+      "loss": 0.9027,
+      "step": 72965
+    },
+    {
+      "epoch": 4.95787471123794,
+      "grad_norm": 7.75,
+      "learning_rate": 1.9013283054762875e-05,
+      "loss": 0.8941,
+      "step": 72970
+    },
+    {
+      "epoch": 4.958214431308602,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.901115980432124e-05,
+      "loss": 0.9594,
+      "step": 72975
+    },
+    {
+      "epoch": 4.958554151379263,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9009036553879603e-05,
+      "loss": 0.8165,
+      "step": 72980
+    },
+    {
+      "epoch": 4.958893871449925,
+      "grad_norm": 6.875,
+      "learning_rate": 1.9006913303437967e-05,
+      "loss": 0.8794,
+      "step": 72985
+    },
+    {
+      "epoch": 4.959233591520587,
+      "grad_norm": 6.75,
+      "learning_rate": 1.900479005299633e-05,
+      "loss": 0.879,
+      "step": 72990
+    },
+    {
+      "epoch": 4.959573311591249,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.90026668025547e-05,
+      "loss": 0.9861,
+      "step": 72995
+    },
+    {
+      "epoch": 4.959913031661911,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.900054355211306e-05,
+      "loss": 0.8214,
+      "step": 73000
+    },
+    {
+      "epoch": 4.960252751732573,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8998420301671423e-05,
+      "loss": 0.9393,
+      "step": 73005
+    },
+    {
+      "epoch": 4.960592471803234,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8996297051229787e-05,
+      "loss": 0.9133,
+      "step": 73010
+    },
+    {
+      "epoch": 4.960932191873896,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.899417380078815e-05,
+      "loss": 0.8152,
+      "step": 73015
+    },
+    {
+      "epoch": 4.961271911944557,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8992050550346515e-05,
+      "loss": 0.8774,
+      "step": 73020
+    },
+    {
+      "epoch": 4.961611632015219,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.898992729990488e-05,
+      "loss": 0.7842,
+      "step": 73025
+    },
+    {
+      "epoch": 4.961951352085881,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8987804049463243e-05,
+      "loss": 0.9155,
+      "step": 73030
+    },
+    {
+      "epoch": 4.9622910721565425,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8985680799021607e-05,
+      "loss": 0.8152,
+      "step": 73035
+    },
+    {
+      "epoch": 4.962630792227205,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.898355754857997e-05,
+      "loss": 0.8523,
+      "step": 73040
+    },
+    {
+      "epoch": 4.962970512297867,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8981434298138335e-05,
+      "loss": 0.8516,
+      "step": 73045
+    },
+    {
+      "epoch": 4.963310232368528,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.89793110476967e-05,
+      "loss": 0.8797,
+      "step": 73050
+    },
+    {
+      "epoch": 4.96364995243919,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8977187797255063e-05,
+      "loss": 0.9619,
+      "step": 73055
+    },
+    {
+      "epoch": 4.963989672509852,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8975064546813428e-05,
+      "loss": 0.8995,
+      "step": 73060
+    },
+    {
+      "epoch": 4.964329392580513,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.897294129637179e-05,
+      "loss": 0.8445,
+      "step": 73065
+    },
+    {
+      "epoch": 4.964669112651175,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8970818045930156e-05,
+      "loss": 0.9233,
+      "step": 73070
+    },
+    {
+      "epoch": 4.965008832721837,
+      "grad_norm": 7.125,
+      "learning_rate": 1.8968694795488516e-05,
+      "loss": 0.8563,
+      "step": 73075
+    },
+    {
+      "epoch": 4.9653485527924985,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8966571545046884e-05,
+      "loss": 0.8543,
+      "step": 73080
+    },
+    {
+      "epoch": 4.965688272863161,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.8964448294605248e-05,
+      "loss": 0.8557,
+      "step": 73085
+    },
+    {
+      "epoch": 4.966027992933823,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8962325044163608e-05,
+      "loss": 0.8973,
+      "step": 73090
+    },
+    {
+      "epoch": 4.966367713004484,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8960201793721976e-05,
+      "loss": 0.8975,
+      "step": 73095
+    },
+    {
+      "epoch": 4.966707433075146,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.895807854328034e-05,
+      "loss": 0.7614,
+      "step": 73100
+    },
+    {
+      "epoch": 4.967047153145808,
+      "grad_norm": 7.75,
+      "learning_rate": 1.89559552928387e-05,
+      "loss": 0.834,
+      "step": 73105
+    },
+    {
+      "epoch": 4.967386873216469,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8953832042397064e-05,
+      "loss": 0.8908,
+      "step": 73110
+    },
+    {
+      "epoch": 4.967726593287131,
+      "grad_norm": 7.25,
+      "learning_rate": 1.895170879195543e-05,
+      "loss": 0.8266,
+      "step": 73115
+    },
+    {
+      "epoch": 4.968066313357793,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.8949585541513792e-05,
+      "loss": 0.9125,
+      "step": 73120
+    },
+    {
+      "epoch": 4.9684060334284545,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8947462291072156e-05,
+      "loss": 0.9778,
+      "step": 73125
+    },
+    {
+      "epoch": 4.968745753499117,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.8945339040630524e-05,
+      "loss": 0.829,
+      "step": 73130
+    },
+    {
+      "epoch": 4.969085473569779,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8943215790188884e-05,
+      "loss": 0.7354,
+      "step": 73135
+    },
+    {
+      "epoch": 4.96942519364044,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8941092539747248e-05,
+      "loss": 0.9306,
+      "step": 73140
+    },
+    {
+      "epoch": 4.969764913711102,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8938969289305612e-05,
+      "loss": 0.8076,
+      "step": 73145
+    },
+    {
+      "epoch": 4.970104633781764,
+      "grad_norm": 6.25,
+      "learning_rate": 1.8936846038863976e-05,
+      "loss": 0.8785,
+      "step": 73150
+    },
+    {
+      "epoch": 4.970444353852425,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.893472278842234e-05,
+      "loss": 0.8639,
+      "step": 73155
+    },
+    {
+      "epoch": 4.970784073923087,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8932599537980704e-05,
+      "loss": 0.9634,
+      "step": 73160
+    },
+    {
+      "epoch": 4.971123793993749,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.893047628753907e-05,
+      "loss": 0.9158,
+      "step": 73165
+    },
+    {
+      "epoch": 4.971463514064411,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8928353037097432e-05,
+      "loss": 0.9323,
+      "step": 73170
+    },
+    {
+      "epoch": 4.971803234135073,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8926229786655796e-05,
+      "loss": 0.9228,
+      "step": 73175
+    },
+    {
+      "epoch": 4.972142954205735,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.892410653621416e-05,
+      "loss": 0.8965,
+      "step": 73180
+    },
+    {
+      "epoch": 4.972482674276396,
+      "grad_norm": 7.125,
+      "learning_rate": 1.8921983285772524e-05,
+      "loss": 0.8166,
+      "step": 73185
+    },
+    {
+      "epoch": 4.972822394347058,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8919860035330888e-05,
+      "loss": 0.8871,
+      "step": 73190
+    },
+    {
+      "epoch": 4.97316211441772,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8917736784889252e-05,
+      "loss": 0.878,
+      "step": 73195
+    },
+    {
+      "epoch": 4.973501834488381,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.8915613534447616e-05,
+      "loss": 0.813,
+      "step": 73200
+    },
+    {
+      "epoch": 4.973841554559043,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.891349028400598e-05,
+      "loss": 0.8627,
+      "step": 73205
+    },
+    {
+      "epoch": 4.974181274629705,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.8911367033564344e-05,
+      "loss": 0.9246,
+      "step": 73210
+    },
+    {
+      "epoch": 4.974520994700367,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8909243783122708e-05,
+      "loss": 0.8322,
+      "step": 73215
+    },
+    {
+      "epoch": 4.974860714771029,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8907120532681072e-05,
+      "loss": 0.8455,
+      "step": 73220
+    },
+    {
+      "epoch": 4.975200434841691,
+      "grad_norm": 7.25,
+      "learning_rate": 1.8904997282239436e-05,
+      "loss": 0.8583,
+      "step": 73225
+    },
+    {
+      "epoch": 4.975540154912352,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.89028740317978e-05,
+      "loss": 0.8162,
+      "step": 73230
+    },
+    {
+      "epoch": 4.975879874983014,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8900750781356164e-05,
+      "loss": 0.9338,
+      "step": 73235
+    },
+    {
+      "epoch": 4.976219595053676,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8898627530914528e-05,
+      "loss": 0.8499,
+      "step": 73240
+    },
+    {
+      "epoch": 4.976559315124337,
+      "grad_norm": 8.0,
+      "learning_rate": 1.889650428047289e-05,
+      "loss": 0.8477,
+      "step": 73245
+    },
+    {
+      "epoch": 4.976899035194999,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.8894381030031256e-05,
+      "loss": 0.9407,
+      "step": 73250
+    },
+    {
+      "epoch": 4.977238755265661,
+      "grad_norm": 7.5,
+      "learning_rate": 1.889225777958962e-05,
+      "loss": 0.92,
+      "step": 73255
+    },
+    {
+      "epoch": 4.977578475336323,
+      "grad_norm": 7.0,
+      "learning_rate": 1.889013452914798e-05,
+      "loss": 0.8369,
+      "step": 73260
+    },
+    {
+      "epoch": 4.977918195406985,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8888011278706348e-05,
+      "loss": 0.9504,
+      "step": 73265
+    },
+    {
+      "epoch": 4.978257915477647,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8885888028264712e-05,
+      "loss": 0.8414,
+      "step": 73270
+    },
+    {
+      "epoch": 4.978597635548308,
+      "grad_norm": 7.25,
+      "learning_rate": 1.8883764777823073e-05,
+      "loss": 0.8529,
+      "step": 73275
+    },
+    {
+      "epoch": 4.97893735561897,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8881641527381437e-05,
+      "loss": 0.9237,
+      "step": 73280
+    },
+    {
+      "epoch": 4.979277075689632,
+      "grad_norm": 9.125,
+      "learning_rate": 1.8879518276939804e-05,
+      "loss": 0.8892,
+      "step": 73285
+    },
+    {
+      "epoch": 4.979616795760293,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8877395026498165e-05,
+      "loss": 0.8124,
+      "step": 73290
+    },
+    {
+      "epoch": 4.979956515830955,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.887527177605653e-05,
+      "loss": 0.9135,
+      "step": 73295
+    },
+    {
+      "epoch": 4.980296235901617,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8873148525614896e-05,
+      "loss": 0.8137,
+      "step": 73300
+    },
+    {
+      "epoch": 4.980635955972279,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8871025275173257e-05,
+      "loss": 0.8967,
+      "step": 73305
+    },
+    {
+      "epoch": 4.980975676042941,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.886890202473162e-05,
+      "loss": 0.738,
+      "step": 73310
+    },
+    {
+      "epoch": 4.981315396113603,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8866778774289985e-05,
+      "loss": 0.8685,
+      "step": 73315
+    },
+    {
+      "epoch": 4.981655116184264,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.886465552384835e-05,
+      "loss": 0.8555,
+      "step": 73320
+    },
+    {
+      "epoch": 4.981994836254926,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8862532273406713e-05,
+      "loss": 0.8665,
+      "step": 73325
+    },
+    {
+      "epoch": 4.982334556325588,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8860409022965077e-05,
+      "loss": 0.803,
+      "step": 73330
+    },
+    {
+      "epoch": 4.982674276396249,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8858285772523444e-05,
+      "loss": 0.95,
+      "step": 73335
+    },
+    {
+      "epoch": 4.983013996466911,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8856162522081805e-05,
+      "loss": 0.8874,
+      "step": 73340
+    },
+    {
+      "epoch": 4.983353716537573,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.885403927164017e-05,
+      "loss": 0.8632,
+      "step": 73345
+    },
+    {
+      "epoch": 4.983693436608235,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8851916021198533e-05,
+      "loss": 0.7725,
+      "step": 73350
+    },
+    {
+      "epoch": 4.984033156678897,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.8849792770756897e-05,
+      "loss": 0.7749,
+      "step": 73355
+    },
+    {
+      "epoch": 4.984372876749559,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.884766952031526e-05,
+      "loss": 0.7941,
+      "step": 73360
+    },
+    {
+      "epoch": 4.98471259682022,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8845546269873625e-05,
+      "loss": 0.9145,
+      "step": 73365
+    },
+    {
+      "epoch": 4.985052316890882,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.884342301943199e-05,
+      "loss": 0.9501,
+      "step": 73370
+    },
+    {
+      "epoch": 4.985392036961544,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8841299768990353e-05,
+      "loss": 0.818,
+      "step": 73375
+    },
+    {
+      "epoch": 4.985731757032205,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.8839176518548717e-05,
+      "loss": 0.8826,
+      "step": 73380
+    },
+    {
+      "epoch": 4.986071477102867,
+      "grad_norm": 8.5,
+      "learning_rate": 1.883705326810708e-05,
+      "loss": 0.9178,
+      "step": 73385
+    },
+    {
+      "epoch": 4.986411197173529,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8834930017665445e-05,
+      "loss": 0.89,
+      "step": 73390
+    },
+    {
+      "epoch": 4.986750917244191,
+      "grad_norm": 7.125,
+      "learning_rate": 1.883280676722381e-05,
+      "loss": 0.859,
+      "step": 73395
+    },
+    {
+      "epoch": 4.987090637314853,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8830683516782173e-05,
+      "loss": 0.8543,
+      "step": 73400
+    },
+    {
+      "epoch": 4.987430357385515,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8828560266340537e-05,
+      "loss": 0.954,
+      "step": 73405
+    },
+    {
+      "epoch": 4.987770077456176,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.88264370158989e-05,
+      "loss": 0.815,
+      "step": 73410
+    },
+    {
+      "epoch": 4.988109797526838,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.882431376545726e-05,
+      "loss": 0.9063,
+      "step": 73415
+    },
+    {
+      "epoch": 4.9884495175975,
+      "grad_norm": 8.25,
+      "learning_rate": 1.882219051501563e-05,
+      "loss": 0.8776,
+      "step": 73420
+    },
+    {
+      "epoch": 4.988789237668161,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8820067264573993e-05,
+      "loss": 0.8648,
+      "step": 73425
+    },
+    {
+      "epoch": 4.989128957738823,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8817944014132354e-05,
+      "loss": 0.8616,
+      "step": 73430
+    },
+    {
+      "epoch": 4.989468677809485,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.881582076369072e-05,
+      "loss": 0.9649,
+      "step": 73435
+    },
+    {
+      "epoch": 4.989808397880147,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8813697513249085e-05,
+      "loss": 0.8887,
+      "step": 73440
+    },
+    {
+      "epoch": 4.990148117950809,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.8811574262807446e-05,
+      "loss": 0.8561,
+      "step": 73445
+    },
+    {
+      "epoch": 4.990487838021471,
+      "grad_norm": 8.0,
+      "learning_rate": 1.880945101236581e-05,
+      "loss": 0.8419,
+      "step": 73450
+    },
+    {
+      "epoch": 4.990827558092132,
+      "grad_norm": 7.25,
+      "learning_rate": 1.8807327761924177e-05,
+      "loss": 0.9657,
+      "step": 73455
+    },
+    {
+      "epoch": 4.991167278162794,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8805204511482538e-05,
+      "loss": 0.8675,
+      "step": 73460
+    },
+    {
+      "epoch": 4.991506998233456,
+      "grad_norm": 6.625,
+      "learning_rate": 1.88030812610409e-05,
+      "loss": 0.8346,
+      "step": 73465
+    },
+    {
+      "epoch": 4.991846718304117,
+      "grad_norm": 7.75,
+      "learning_rate": 1.880095801059927e-05,
+      "loss": 0.9071,
+      "step": 73470
+    },
+    {
+      "epoch": 4.992186438374779,
+      "grad_norm": 7.625,
+      "learning_rate": 1.879883476015763e-05,
+      "loss": 0.8916,
+      "step": 73475
+    },
+    {
+      "epoch": 4.9925261584454415,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8796711509715994e-05,
+      "loss": 0.9483,
+      "step": 73480
+    },
+    {
+      "epoch": 4.992865878516103,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.879458825927436e-05,
+      "loss": 0.8967,
+      "step": 73485
+    },
+    {
+      "epoch": 4.993205598586765,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.879246500883272e-05,
+      "loss": 0.9044,
+      "step": 73490
+    },
+    {
+      "epoch": 4.993545318657426,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8790341758391086e-05,
+      "loss": 0.9256,
+      "step": 73495
+    },
+    {
+      "epoch": 4.993885038728088,
+      "grad_norm": 6.375,
+      "learning_rate": 1.878821850794945e-05,
+      "loss": 0.9051,
+      "step": 73500
+    },
+    {
+      "epoch": 4.99422475879875,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8786095257507817e-05,
+      "loss": 0.8237,
+      "step": 73505
+    },
+    {
+      "epoch": 4.994564478869411,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8783972007066178e-05,
+      "loss": 0.8581,
+      "step": 73510
+    },
+    {
+      "epoch": 4.994904198940073,
+      "grad_norm": 8.25,
+      "learning_rate": 1.878184875662454e-05,
+      "loss": 0.9558,
+      "step": 73515
+    },
+    {
+      "epoch": 4.995243919010735,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.877972550618291e-05,
+      "loss": 0.8895,
+      "step": 73520
+    },
+    {
+      "epoch": 4.995583639081397,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.877760225574127e-05,
+      "loss": 0.8682,
+      "step": 73525
+    },
+    {
+      "epoch": 4.995923359152059,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8775479005299634e-05,
+      "loss": 0.8855,
+      "step": 73530
+    },
+    {
+      "epoch": 4.996263079222721,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8773355754857998e-05,
+      "loss": 0.8859,
+      "step": 73535
+    },
+    {
+      "epoch": 4.996602799293382,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8771232504416362e-05,
+      "loss": 0.8833,
+      "step": 73540
+    },
+    {
+      "epoch": 4.996942519364044,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8769109253974726e-05,
+      "loss": 0.9004,
+      "step": 73545
+    },
+    {
+      "epoch": 4.997282239434706,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.876698600353309e-05,
+      "loss": 0.8644,
+      "step": 73550
+    },
+    {
+      "epoch": 4.997621959505367,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8764862753091454e-05,
+      "loss": 0.8394,
+      "step": 73555
+    },
+    {
+      "epoch": 4.997961679576029,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.8762739502649818e-05,
+      "loss": 0.8657,
+      "step": 73560
+    },
+    {
+      "epoch": 4.998301399646691,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8760616252208182e-05,
+      "loss": 0.8426,
+      "step": 73565
+    },
+    {
+      "epoch": 4.998641119717353,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8758493001766546e-05,
+      "loss": 0.8793,
+      "step": 73570
+    },
+    {
+      "epoch": 4.998980839788015,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.875636975132491e-05,
+      "loss": 0.8473,
+      "step": 73575
+    },
+    {
+      "epoch": 4.999320559858677,
+      "grad_norm": 8.875,
+      "learning_rate": 1.8754246500883274e-05,
+      "loss": 0.8933,
+      "step": 73580
+    },
+    {
+      "epoch": 4.999660279929338,
+      "grad_norm": 5.875,
+      "learning_rate": 1.8752123250441638e-05,
+      "loss": 0.8617,
+      "step": 73585
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 12.875,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.8173,
+      "step": 73590
+    },
+    {
+      "epoch": 5.0,
+      "eval_loss": 1.7338234186172485,
+      "eval_runtime": 146.9108,
+      "eval_samples_per_second": 70.24,
+      "eval_steps_per_second": 8.781,
+      "step": 73590
+    },
+    {
+      "epoch": 5.000339720070662,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8747876749558366e-05,
+      "loss": 0.8442,
+      "step": 73595
+    },
+    {
+      "epoch": 5.000679440141323,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8745753499116726e-05,
+      "loss": 0.8109,
+      "step": 73600
+    },
+    {
+      "epoch": 5.001019160211985,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8743630248675094e-05,
+      "loss": 0.8117,
+      "step": 73605
+    },
+    {
+      "epoch": 5.001358880282647,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8741506998233458e-05,
+      "loss": 0.8933,
+      "step": 73610
+    },
+    {
+      "epoch": 5.001698600353309,
+      "grad_norm": 7.75,
+      "learning_rate": 1.873938374779182e-05,
+      "loss": 0.7942,
+      "step": 73615
+    },
+    {
+      "epoch": 5.002038320423971,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8737260497350186e-05,
+      "loss": 0.8363,
+      "step": 73620
+    },
+    {
+      "epoch": 5.002378040494633,
+      "grad_norm": 8.0,
+      "learning_rate": 1.873513724690855e-05,
+      "loss": 0.8563,
+      "step": 73625
+    },
+    {
+      "epoch": 5.002717760565294,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.873301399646691e-05,
+      "loss": 0.8466,
+      "step": 73630
+    },
+    {
+      "epoch": 5.003057480635956,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8730890746025274e-05,
+      "loss": 0.8692,
+      "step": 73635
+    },
+    {
+      "epoch": 5.003397200706618,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8728767495583642e-05,
+      "loss": 0.7699,
+      "step": 73640
+    },
+    {
+      "epoch": 5.003736920777279,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.8726644245142002e-05,
+      "loss": 0.804,
+      "step": 73645
+    },
+    {
+      "epoch": 5.004076640847941,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8724520994700366e-05,
+      "loss": 0.7984,
+      "step": 73650
+    },
+    {
+      "epoch": 5.004416360918603,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.8722397744258734e-05,
+      "loss": 0.9153,
+      "step": 73655
+    },
+    {
+      "epoch": 5.004756080989265,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8720274493817094e-05,
+      "loss": 0.8829,
+      "step": 73660
+    },
+    {
+      "epoch": 5.005095801059927,
+      "grad_norm": 9.25,
+      "learning_rate": 1.871815124337546e-05,
+      "loss": 0.8777,
+      "step": 73665
+    },
+    {
+      "epoch": 5.005435521130589,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.8716027992933822e-05,
+      "loss": 0.9122,
+      "step": 73670
+    },
+    {
+      "epoch": 5.00577524120125,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8713904742492186e-05,
+      "loss": 0.8341,
+      "step": 73675
+    },
+    {
+      "epoch": 5.006114961271912,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.871178149205055e-05,
+      "loss": 0.8365,
+      "step": 73680
+    },
+    {
+      "epoch": 5.006454681342574,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8709658241608914e-05,
+      "loss": 0.8902,
+      "step": 73685
+    },
+    {
+      "epoch": 5.006794401413235,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8707534991167282e-05,
+      "loss": 0.926,
+      "step": 73690
+    },
+    {
+      "epoch": 5.007134121483897,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.8705411740725642e-05,
+      "loss": 0.8166,
+      "step": 73695
+    },
+    {
+      "epoch": 5.007473841554559,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8703288490284006e-05,
+      "loss": 0.9193,
+      "step": 73700
+    },
+    {
+      "epoch": 5.007813561625221,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.870116523984237e-05,
+      "loss": 0.8651,
+      "step": 73705
+    },
+    {
+      "epoch": 5.008153281695883,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.8699041989400734e-05,
+      "loss": 0.8084,
+      "step": 73710
+    },
+    {
+      "epoch": 5.008493001766545,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.86969187389591e-05,
+      "loss": 0.8753,
+      "step": 73715
+    },
+    {
+      "epoch": 5.008832721837206,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8694795488517462e-05,
+      "loss": 0.8568,
+      "step": 73720
+    },
+    {
+      "epoch": 5.009172441907868,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8692672238075826e-05,
+      "loss": 0.8406,
+      "step": 73725
+    },
+    {
+      "epoch": 5.00951216197853,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.869054898763419e-05,
+      "loss": 0.8008,
+      "step": 73730
+    },
+    {
+      "epoch": 5.009851882049191,
+      "grad_norm": 6.875,
+      "learning_rate": 1.8688425737192554e-05,
+      "loss": 0.8058,
+      "step": 73735
+    },
+    {
+      "epoch": 5.010191602119853,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.868630248675092e-05,
+      "loss": 0.81,
+      "step": 73740
+    },
+    {
+      "epoch": 5.0105313221905154,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8684179236309282e-05,
+      "loss": 0.8168,
+      "step": 73745
+    },
+    {
+      "epoch": 5.010871042261177,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8682055985867646e-05,
+      "loss": 0.7932,
+      "step": 73750
+    },
+    {
+      "epoch": 5.011210762331839,
+      "grad_norm": 7.75,
+      "learning_rate": 1.867993273542601e-05,
+      "loss": 0.8777,
+      "step": 73755
+    },
+    {
+      "epoch": 5.0115504824025,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8677809484984374e-05,
+      "loss": 0.8253,
+      "step": 73760
+    },
+    {
+      "epoch": 5.011890202473162,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.867568623454274e-05,
+      "loss": 0.8351,
+      "step": 73765
+    },
+    {
+      "epoch": 5.012229922543824,
+      "grad_norm": 6.625,
+      "learning_rate": 1.86735629841011e-05,
+      "loss": 0.7779,
+      "step": 73770
+    },
+    {
+      "epoch": 5.012569642614485,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8671439733659467e-05,
+      "loss": 0.9497,
+      "step": 73775
+    },
+    {
+      "epoch": 5.012909362685147,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.866931648321783e-05,
+      "loss": 0.863,
+      "step": 73780
+    },
+    {
+      "epoch": 5.013249082755809,
+      "grad_norm": 8.125,
+      "learning_rate": 1.866719323277619e-05,
+      "loss": 0.9002,
+      "step": 73785
+    },
+    {
+      "epoch": 5.013588802826471,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.866506998233456e-05,
+      "loss": 0.7665,
+      "step": 73790
+    },
+    {
+      "epoch": 5.013928522897133,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8662946731892923e-05,
+      "loss": 0.8557,
+      "step": 73795
+    },
+    {
+      "epoch": 5.014268242967795,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8660823481451283e-05,
+      "loss": 0.8201,
+      "step": 73800
+    },
+    {
+      "epoch": 5.014607963038456,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8658700231009647e-05,
+      "loss": 0.8392,
+      "step": 73805
+    },
+    {
+      "epoch": 5.014947683109118,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8656576980568015e-05,
+      "loss": 0.8613,
+      "step": 73810
+    },
+    {
+      "epoch": 5.01528740317978,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8654453730126375e-05,
+      "loss": 0.8563,
+      "step": 73815
+    },
+    {
+      "epoch": 5.015627123250441,
+      "grad_norm": 8.75,
+      "learning_rate": 1.865233047968474e-05,
+      "loss": 0.9272,
+      "step": 73820
+    },
+    {
+      "epoch": 5.015966843321103,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8650207229243107e-05,
+      "loss": 0.8707,
+      "step": 73825
+    },
+    {
+      "epoch": 5.016306563391765,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.8648083978801467e-05,
+      "loss": 0.9041,
+      "step": 73830
+    },
+    {
+      "epoch": 5.016646283462427,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.864596072835983e-05,
+      "loss": 0.7661,
+      "step": 73835
+    },
+    {
+      "epoch": 5.016986003533089,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.8643837477918195e-05,
+      "loss": 0.8666,
+      "step": 73840
+    },
+    {
+      "epoch": 5.017325723603751,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.864171422747656e-05,
+      "loss": 0.7756,
+      "step": 73845
+    },
+    {
+      "epoch": 5.017665443674412,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8639590977034923e-05,
+      "loss": 0.8615,
+      "step": 73850
+    },
+    {
+      "epoch": 5.018005163745074,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8637467726593287e-05,
+      "loss": 0.8276,
+      "step": 73855
+    },
+    {
+      "epoch": 5.018344883815736,
+      "grad_norm": 7.25,
+      "learning_rate": 1.8635344476151655e-05,
+      "loss": 0.8135,
+      "step": 73860
+    },
+    {
+      "epoch": 5.018684603886397,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8633221225710015e-05,
+      "loss": 0.8708,
+      "step": 73865
+    },
+    {
+      "epoch": 5.019024323957059,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.863109797526838e-05,
+      "loss": 0.9076,
+      "step": 73870
+    },
+    {
+      "epoch": 5.019364044027721,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8628974724826747e-05,
+      "loss": 0.9046,
+      "step": 73875
+    },
+    {
+      "epoch": 5.019703764098383,
+      "grad_norm": 7.125,
+      "learning_rate": 1.8626851474385107e-05,
+      "loss": 0.7579,
+      "step": 73880
+    },
+    {
+      "epoch": 5.020043484169045,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.862472822394347e-05,
+      "loss": 0.9182,
+      "step": 73885
+    },
+    {
+      "epoch": 5.020383204239707,
+      "grad_norm": 7.125,
+      "learning_rate": 1.8622604973501835e-05,
+      "loss": 0.8718,
+      "step": 73890
+    },
+    {
+      "epoch": 5.020722924310368,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.86204817230602e-05,
+      "loss": 0.8165,
+      "step": 73895
+    },
+    {
+      "epoch": 5.02106264438103,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8618358472618563e-05,
+      "loss": 0.8957,
+      "step": 73900
+    },
+    {
+      "epoch": 5.021402364451692,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8616235222176927e-05,
+      "loss": 0.8936,
+      "step": 73905
+    },
+    {
+      "epoch": 5.021742084522353,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.861411197173529e-05,
+      "loss": 1.1092,
+      "step": 73910
+    },
+    {
+      "epoch": 5.022081804593015,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.8611988721293655e-05,
+      "loss": 0.8601,
+      "step": 73915
+    },
+    {
+      "epoch": 5.022421524663677,
+      "grad_norm": 8.25,
+      "learning_rate": 1.860986547085202e-05,
+      "loss": 0.8868,
+      "step": 73920
+    },
+    {
+      "epoch": 5.022761244734339,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8607742220410383e-05,
+      "loss": 0.8715,
+      "step": 73925
+    },
+    {
+      "epoch": 5.023100964805001,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8605618969968747e-05,
+      "loss": 0.8477,
+      "step": 73930
+    },
+    {
+      "epoch": 5.023440684875663,
+      "grad_norm": 7.5,
+      "learning_rate": 1.860349571952711e-05,
+      "loss": 0.9142,
+      "step": 73935
+    },
+    {
+      "epoch": 5.023780404946324,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8601372469085472e-05,
+      "loss": 0.8786,
+      "step": 73940
+    },
+    {
+      "epoch": 5.024120125016986,
+      "grad_norm": 7.375,
+      "learning_rate": 1.859924921864384e-05,
+      "loss": 0.8496,
+      "step": 73945
+    },
+    {
+      "epoch": 5.024459845087648,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8597125968202203e-05,
+      "loss": 0.7522,
+      "step": 73950
+    },
+    {
+      "epoch": 5.024799565158309,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8595002717760564e-05,
+      "loss": 0.7558,
+      "step": 73955
+    },
+    {
+      "epoch": 5.025139285228971,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.859287946731893e-05,
+      "loss": 0.9318,
+      "step": 73960
+    },
+    {
+      "epoch": 5.025479005299633,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8590756216877295e-05,
+      "loss": 0.8323,
+      "step": 73965
+    },
+    {
+      "epoch": 5.025818725370295,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8588632966435656e-05,
+      "loss": 0.9891,
+      "step": 73970
+    },
+    {
+      "epoch": 5.026158445440957,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8586509715994023e-05,
+      "loss": 0.8815,
+      "step": 73975
+    },
+    {
+      "epoch": 5.026498165511619,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8584386465552387e-05,
+      "loss": 0.8325,
+      "step": 73980
+    },
+    {
+      "epoch": 5.02683788558228,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8582263215110748e-05,
+      "loss": 0.8577,
+      "step": 73985
+    },
+    {
+      "epoch": 5.027177605652942,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8580139964669112e-05,
+      "loss": 0.8477,
+      "step": 73990
+    },
+    {
+      "epoch": 5.027517325723604,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.857801671422748e-05,
+      "loss": 0.8157,
+      "step": 73995
+    },
+    {
+      "epoch": 5.027857045794265,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.857589346378584e-05,
+      "loss": 0.8853,
+      "step": 74000
+    },
+    {
+      "epoch": 5.028196765864927,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.8573770213344204e-05,
+      "loss": 0.8724,
+      "step": 74005
+    },
+    {
+      "epoch": 5.028536485935589,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.857164696290257e-05,
+      "loss": 0.8355,
+      "step": 74010
+    },
+    {
+      "epoch": 5.028876206006251,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8569523712460932e-05,
+      "loss": 0.838,
+      "step": 74015
+    },
+    {
+      "epoch": 5.029215926076913,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8567400462019296e-05,
+      "loss": 0.8897,
+      "step": 74020
+    },
+    {
+      "epoch": 5.029555646147575,
+      "grad_norm": 8.625,
+      "learning_rate": 1.856527721157766e-05,
+      "loss": 0.9173,
+      "step": 74025
+    },
+    {
+      "epoch": 5.029895366218236,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8563153961136027e-05,
+      "loss": 0.9786,
+      "step": 74030
+    },
+    {
+      "epoch": 5.030235086288898,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8561030710694388e-05,
+      "loss": 0.8567,
+      "step": 74035
+    },
+    {
+      "epoch": 5.03057480635956,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8558907460252752e-05,
+      "loss": 0.8278,
+      "step": 74040
+    },
+    {
+      "epoch": 5.030914526430221,
+      "grad_norm": 6.875,
+      "learning_rate": 1.855678420981112e-05,
+      "loss": 0.8413,
+      "step": 74045
+    },
+    {
+      "epoch": 5.031254246500883,
+      "grad_norm": 6.75,
+      "learning_rate": 1.855466095936948e-05,
+      "loss": 0.9496,
+      "step": 74050
+    },
+    {
+      "epoch": 5.0315939665715455,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.8552537708927844e-05,
+      "loss": 0.7788,
+      "step": 74055
+    },
+    {
+      "epoch": 5.031933686642207,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8550414458486208e-05,
+      "loss": 0.859,
+      "step": 74060
+    },
+    {
+      "epoch": 5.032273406712869,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.8548291208044572e-05,
+      "loss": 0.8194,
+      "step": 74065
+    },
+    {
+      "epoch": 5.032613126783531,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.8546167957602936e-05,
+      "loss": 0.8542,
+      "step": 74070
+    },
+    {
+      "epoch": 5.032952846854192,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.85440447071613e-05,
+      "loss": 0.8946,
+      "step": 74075
+    },
+    {
+      "epoch": 5.033292566924854,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8541921456719664e-05,
+      "loss": 0.807,
+      "step": 74080
+    },
+    {
+      "epoch": 5.033632286995516,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8539798206278028e-05,
+      "loss": 0.7731,
+      "step": 74085
+    },
+    {
+      "epoch": 5.033972007066177,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8537674955836392e-05,
+      "loss": 0.8781,
+      "step": 74090
+    },
+    {
+      "epoch": 5.034311727136839,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8535551705394756e-05,
+      "loss": 0.8919,
+      "step": 74095
+    },
+    {
+      "epoch": 5.0346514472075015,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.853342845495312e-05,
+      "loss": 0.8802,
+      "step": 74100
+    },
+    {
+      "epoch": 5.034991167278163,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8531305204511484e-05,
+      "loss": 0.7944,
+      "step": 74105
+    },
+    {
+      "epoch": 5.035330887348825,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.8529181954069848e-05,
+      "loss": 0.8489,
+      "step": 74110
+    },
+    {
+      "epoch": 5.035670607419486,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8527058703628212e-05,
+      "loss": 0.8851,
+      "step": 74115
+    },
+    {
+      "epoch": 5.036010327490148,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8524935453186576e-05,
+      "loss": 0.7806,
+      "step": 74120
+    },
+    {
+      "epoch": 5.03635004756081,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8522812202744937e-05,
+      "loss": 0.9621,
+      "step": 74125
+    },
+    {
+      "epoch": 5.036689767631471,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.8520688952303304e-05,
+      "loss": 0.9422,
+      "step": 74130
+    },
+    {
+      "epoch": 5.037029487702133,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8518565701861668e-05,
+      "loss": 0.8672,
+      "step": 74135
+    },
+    {
+      "epoch": 5.037369207772795,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.851644245142003e-05,
+      "loss": 0.7413,
+      "step": 74140
+    },
+    {
+      "epoch": 5.037708927843457,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8514319200978396e-05,
+      "loss": 0.8167,
+      "step": 74145
+    },
+    {
+      "epoch": 5.038048647914119,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.851219595053676e-05,
+      "loss": 0.8615,
+      "step": 74150
+    },
+    {
+      "epoch": 5.038388367984781,
+      "grad_norm": 8.5,
+      "learning_rate": 1.851007270009512e-05,
+      "loss": 0.8661,
+      "step": 74155
+    },
+    {
+      "epoch": 5.038728088055442,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8507949449653485e-05,
+      "loss": 0.855,
+      "step": 74160
+    },
+    {
+      "epoch": 5.039067808126104,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.8505826199211852e-05,
+      "loss": 0.821,
+      "step": 74165
+    },
+    {
+      "epoch": 5.039407528196766,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.8503702948770213e-05,
+      "loss": 0.8492,
+      "step": 74170
+    },
+    {
+      "epoch": 5.039747248267427,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8501579698328577e-05,
+      "loss": 0.8088,
+      "step": 74175
+    },
+    {
+      "epoch": 5.040086968338089,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8499456447886944e-05,
+      "loss": 0.8632,
+      "step": 74180
+    },
+    {
+      "epoch": 5.040426688408751,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8497333197445305e-05,
+      "loss": 0.8461,
+      "step": 74185
+    },
+    {
+      "epoch": 5.040766408479413,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.849520994700367e-05,
+      "loss": 0.7855,
+      "step": 74190
+    },
+    {
+      "epoch": 5.041106128550075,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8493086696562033e-05,
+      "loss": 0.8022,
+      "step": 74195
+    },
+    {
+      "epoch": 5.041445848620737,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.84909634461204e-05,
+      "loss": 0.8198,
+      "step": 74200
+    },
+    {
+      "epoch": 5.041785568691398,
+      "grad_norm": 7.625,
+      "learning_rate": 1.848884019567876e-05,
+      "loss": 0.8297,
+      "step": 74205
+    },
+    {
+      "epoch": 5.04212528876206,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8486716945237125e-05,
+      "loss": 1.0143,
+      "step": 74210
+    },
+    {
+      "epoch": 5.042465008832722,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.8484593694795492e-05,
+      "loss": 0.8067,
+      "step": 74215
+    },
+    {
+      "epoch": 5.042804728903383,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8482470444353853e-05,
+      "loss": 0.8126,
+      "step": 74220
+    },
+    {
+      "epoch": 5.043144448974045,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.8480347193912217e-05,
+      "loss": 0.8058,
+      "step": 74225
+    },
+    {
+      "epoch": 5.043484169044707,
+      "grad_norm": 8.75,
+      "learning_rate": 1.847822394347058e-05,
+      "loss": 0.8883,
+      "step": 74230
+    },
+    {
+      "epoch": 5.043823889115369,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8476100693028945e-05,
+      "loss": 0.9085,
+      "step": 74235
+    },
+    {
+      "epoch": 5.044163609186031,
+      "grad_norm": 8.625,
+      "learning_rate": 1.847397744258731e-05,
+      "loss": 0.7877,
+      "step": 74240
+    },
+    {
+      "epoch": 5.044503329256693,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.8471854192145673e-05,
+      "loss": 0.8405,
+      "step": 74245
+    },
+    {
+      "epoch": 5.044843049327354,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8469730941704037e-05,
+      "loss": 0.8363,
+      "step": 74250
+    },
+    {
+      "epoch": 5.045182769398016,
+      "grad_norm": 7.875,
+      "learning_rate": 1.84676076912624e-05,
+      "loss": 0.8843,
+      "step": 74255
+    },
+    {
+      "epoch": 5.045522489468678,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8465484440820765e-05,
+      "loss": 0.8013,
+      "step": 74260
+    },
+    {
+      "epoch": 5.045862209539339,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.846336119037913e-05,
+      "loss": 0.8433,
+      "step": 74265
+    },
+    {
+      "epoch": 5.046201929610001,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8461237939937493e-05,
+      "loss": 0.9232,
+      "step": 74270
+    },
+    {
+      "epoch": 5.046541649680663,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8459114689495857e-05,
+      "loss": 0.8776,
+      "step": 74275
+    },
+    {
+      "epoch": 5.046881369751325,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.845699143905422e-05,
+      "loss": 0.8269,
+      "step": 74280
+    },
+    {
+      "epoch": 5.047221089821987,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.8454868188612585e-05,
+      "loss": 0.7985,
+      "step": 74285
+    },
+    {
+      "epoch": 5.047560809892649,
+      "grad_norm": 7.75,
+      "learning_rate": 1.845274493817095e-05,
+      "loss": 0.8648,
+      "step": 74290
+    },
+    {
+      "epoch": 5.04790052996331,
+      "grad_norm": 8.25,
+      "learning_rate": 1.845062168772931e-05,
+      "loss": 0.8651,
+      "step": 74295
+    },
+    {
+      "epoch": 5.048240250033972,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8448498437287677e-05,
+      "loss": 0.8499,
+      "step": 74300
+    },
+    {
+      "epoch": 5.048579970104634,
+      "grad_norm": 8.625,
+      "learning_rate": 1.844637518684604e-05,
+      "loss": 0.8244,
+      "step": 74305
+    },
+    {
+      "epoch": 5.048919690175295,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.84442519364044e-05,
+      "loss": 0.8618,
+      "step": 74310
+    },
+    {
+      "epoch": 5.049259410245957,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.844212868596277e-05,
+      "loss": 0.8187,
+      "step": 74315
+    },
+    {
+      "epoch": 5.0495991303166194,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8440005435521133e-05,
+      "loss": 0.8404,
+      "step": 74320
+    },
+    {
+      "epoch": 5.049938850387281,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.8437882185079493e-05,
+      "loss": 0.9171,
+      "step": 74325
+    },
+    {
+      "epoch": 5.050278570457943,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8435758934637857e-05,
+      "loss": 0.8332,
+      "step": 74330
+    },
+    {
+      "epoch": 5.050618290528605,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8433635684196225e-05,
+      "loss": 0.6901,
+      "step": 74335
+    },
+    {
+      "epoch": 5.050958010599266,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.8431512433754585e-05,
+      "loss": 0.7372,
+      "step": 74340
+    },
+    {
+      "epoch": 5.051297730669928,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.842938918331295e-05,
+      "loss": 0.8457,
+      "step": 74345
+    },
+    {
+      "epoch": 5.05163745074059,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8427265932871317e-05,
+      "loss": 0.821,
+      "step": 74350
+    },
+    {
+      "epoch": 5.051977170811251,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.8425142682429677e-05,
+      "loss": 0.8988,
+      "step": 74355
+    },
+    {
+      "epoch": 5.052316890881913,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.842301943198804e-05,
+      "loss": 0.9043,
+      "step": 74360
+    },
+    {
+      "epoch": 5.0526566109525755,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8420896181546405e-05,
+      "loss": 0.8322,
+      "step": 74365
+    },
+    {
+      "epoch": 5.052996331023237,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8418772931104773e-05,
+      "loss": 0.7852,
+      "step": 74370
+    },
+    {
+      "epoch": 5.053336051093899,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.8416649680663133e-05,
+      "loss": 0.8396,
+      "step": 74375
+    },
+    {
+      "epoch": 5.053675771164561,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8414526430221497e-05,
+      "loss": 0.7433,
+      "step": 74380
+    },
+    {
+      "epoch": 5.054015491235222,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8412403179779865e-05,
+      "loss": 0.9011,
+      "step": 74385
+    },
+    {
+      "epoch": 5.054355211305884,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.8410279929338225e-05,
+      "loss": 0.8296,
+      "step": 74390
+    },
+    {
+      "epoch": 5.054694931376546,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.840815667889659e-05,
+      "loss": 0.9018,
+      "step": 74395
+    },
+    {
+      "epoch": 5.055034651447207,
+      "grad_norm": 8.5,
+      "learning_rate": 1.8406033428454957e-05,
+      "loss": 0.7039,
+      "step": 74400
+    },
+    {
+      "epoch": 5.055374371517869,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8403910178013317e-05,
+      "loss": 0.8211,
+      "step": 74405
+    },
+    {
+      "epoch": 5.0557140915885315,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.840178692757168e-05,
+      "loss": 0.9245,
+      "step": 74410
+    },
+    {
+      "epoch": 5.056053811659193,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8399663677130045e-05,
+      "loss": 0.8076,
+      "step": 74415
+    },
+    {
+      "epoch": 5.056393531729855,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.839754042668841e-05,
+      "loss": 0.8598,
+      "step": 74420
+    },
+    {
+      "epoch": 5.056733251800517,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8395417176246773e-05,
+      "loss": 0.867,
+      "step": 74425
+    },
+    {
+      "epoch": 5.057072971871178,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8393293925805137e-05,
+      "loss": 0.7775,
+      "step": 74430
+    },
+    {
+      "epoch": 5.05741269194184,
+      "grad_norm": 7.875,
+      "learning_rate": 1.83911706753635e-05,
+      "loss": 0.8976,
+      "step": 74435
+    },
+    {
+      "epoch": 5.057752412012501,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8389047424921865e-05,
+      "loss": 0.8054,
+      "step": 74440
+    },
+    {
+      "epoch": 5.058092132083163,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.838692417448023e-05,
+      "loss": 0.7689,
+      "step": 74445
+    },
+    {
+      "epoch": 5.058431852153825,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.8384800924038593e-05,
+      "loss": 0.7072,
+      "step": 74450
+    },
+    {
+      "epoch": 5.058771572224487,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8382677673596957e-05,
+      "loss": 0.766,
+      "step": 74455
+    },
+    {
+      "epoch": 5.059111292295149,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.838055442315532e-05,
+      "loss": 0.7986,
+      "step": 74460
+    },
+    {
+      "epoch": 5.059451012365811,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8378431172713682e-05,
+      "loss": 0.9331,
+      "step": 74465
+    },
+    {
+      "epoch": 5.059790732436472,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.837630792227205e-05,
+      "loss": 0.8537,
+      "step": 74470
+    },
+    {
+      "epoch": 5.060130452507134,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8374184671830413e-05,
+      "loss": 0.8575,
+      "step": 74475
+    },
+    {
+      "epoch": 5.060470172577796,
+      "grad_norm": 7.0,
+      "learning_rate": 1.8372061421388774e-05,
+      "loss": 0.7969,
+      "step": 74480
+    },
+    {
+      "epoch": 5.060809892648457,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.836993817094714e-05,
+      "loss": 0.8145,
+      "step": 74485
+    },
+    {
+      "epoch": 5.061149612719119,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.8367814920505505e-05,
+      "loss": 0.7466,
+      "step": 74490
+    },
+    {
+      "epoch": 5.061489332789781,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8365691670063866e-05,
+      "loss": 0.7504,
+      "step": 74495
+    },
+    {
+      "epoch": 5.061829052860443,
+      "grad_norm": 6.5,
+      "learning_rate": 1.8363568419622234e-05,
+      "loss": 0.8107,
+      "step": 74500
+    },
+    {
+      "epoch": 5.062168772931105,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8361445169180598e-05,
+      "loss": 0.773,
+      "step": 74505
+    },
+    {
+      "epoch": 5.062508493001767,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8359321918738958e-05,
+      "loss": 0.8854,
+      "step": 74510
+    },
+    {
+      "epoch": 5.062848213072428,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.8357198668297322e-05,
+      "loss": 0.7808,
+      "step": 74515
+    },
+    {
+      "epoch": 5.06318793314309,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.835507541785569e-05,
+      "loss": 0.9044,
+      "step": 74520
+    },
+    {
+      "epoch": 5.063527653213752,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.835295216741405e-05,
+      "loss": 0.7815,
+      "step": 74525
+    },
+    {
+      "epoch": 5.063867373284413,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.8350828916972414e-05,
+      "loss": 0.8126,
+      "step": 74530
+    },
+    {
+      "epoch": 5.064207093355075,
+      "grad_norm": 6.125,
+      "learning_rate": 1.834870566653078e-05,
+      "loss": 0.809,
+      "step": 74535
+    },
+    {
+      "epoch": 5.064546813425737,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.8346582416089146e-05,
+      "loss": 0.7829,
+      "step": 74540
+    },
+    {
+      "epoch": 5.064886533496399,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8344459165647506e-05,
+      "loss": 0.8649,
+      "step": 74545
+    },
+    {
+      "epoch": 5.065226253567061,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.834233591520587e-05,
+      "loss": 0.9172,
+      "step": 74550
+    },
+    {
+      "epoch": 5.065565973637723,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8340212664764238e-05,
+      "loss": 0.8537,
+      "step": 74555
+    },
+    {
+      "epoch": 5.065905693708384,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8338089414322598e-05,
+      "loss": 0.8964,
+      "step": 74560
+    },
+    {
+      "epoch": 5.066245413779046,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8335966163880962e-05,
+      "loss": 0.8989,
+      "step": 74565
+    },
+    {
+      "epoch": 5.066585133849708,
+      "grad_norm": 7.25,
+      "learning_rate": 1.833384291343933e-05,
+      "loss": 0.8579,
+      "step": 74570
+    },
+    {
+      "epoch": 5.066924853920369,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.833171966299769e-05,
+      "loss": 0.8052,
+      "step": 74575
+    },
+    {
+      "epoch": 5.067264573991031,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8329596412556054e-05,
+      "loss": 0.773,
+      "step": 74580
+    },
+    {
+      "epoch": 5.067604294061693,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8327473162114418e-05,
+      "loss": 0.8274,
+      "step": 74585
+    },
+    {
+      "epoch": 5.067944014132355,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8325349911672782e-05,
+      "loss": 0.7694,
+      "step": 74590
+    },
+    {
+      "epoch": 5.068283734203017,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.8323226661231146e-05,
+      "loss": 0.8742,
+      "step": 74595
+    },
+    {
+      "epoch": 5.068623454273679,
+      "grad_norm": 8.0,
+      "learning_rate": 1.832110341078951e-05,
+      "loss": 0.8195,
+      "step": 74600
+    },
+    {
+      "epoch": 5.06896317434434,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8318980160347874e-05,
+      "loss": 0.9256,
+      "step": 74605
+    },
+    {
+      "epoch": 5.069302894415002,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8316856909906238e-05,
+      "loss": 0.8657,
+      "step": 74610
+    },
+    {
+      "epoch": 5.069642614485664,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8314733659464602e-05,
+      "loss": 0.8173,
+      "step": 74615
+    },
+    {
+      "epoch": 5.069982334556325,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8312610409022966e-05,
+      "loss": 0.8584,
+      "step": 74620
+    },
+    {
+      "epoch": 5.070322054626987,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.831048715858133e-05,
+      "loss": 0.7794,
+      "step": 74625
+    },
+    {
+      "epoch": 5.0706617746976494,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8308363908139694e-05,
+      "loss": 0.8749,
+      "step": 74630
+    },
+    {
+      "epoch": 5.071001494768311,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8306240657698058e-05,
+      "loss": 0.747,
+      "step": 74635
+    },
+    {
+      "epoch": 5.071341214838973,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.8304117407256422e-05,
+      "loss": 0.9157,
+      "step": 74640
+    },
+    {
+      "epoch": 5.071680934909635,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.8301994156814786e-05,
+      "loss": 0.8345,
+      "step": 74645
+    },
+    {
+      "epoch": 5.072020654980296,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.8299870906373147e-05,
+      "loss": 0.8463,
+      "step": 74650
+    },
+    {
+      "epoch": 5.072360375050958,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8297747655931514e-05,
+      "loss": 0.805,
+      "step": 74655
+    },
+    {
+      "epoch": 5.07270009512162,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.8295624405489878e-05,
+      "loss": 0.8333,
+      "step": 74660
+    },
+    {
+      "epoch": 5.073039815192281,
+      "grad_norm": 8.125,
+      "learning_rate": 1.829350115504824e-05,
+      "loss": 0.8678,
+      "step": 74665
+    },
+    {
+      "epoch": 5.073379535262943,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.8291377904606606e-05,
+      "loss": 0.843,
+      "step": 74670
+    },
+    {
+      "epoch": 5.0737192553336055,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.828925465416497e-05,
+      "loss": 0.8276,
+      "step": 74675
+    },
+    {
+      "epoch": 5.074058975404267,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.828713140372333e-05,
+      "loss": 0.8699,
+      "step": 74680
+    },
+    {
+      "epoch": 5.074398695474929,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.8285008153281695e-05,
+      "loss": 0.9141,
+      "step": 74685
+    },
+    {
+      "epoch": 5.074738415545591,
+      "grad_norm": 8.875,
+      "learning_rate": 1.8282884902840062e-05,
+      "loss": 0.8287,
+      "step": 74690
+    },
+    {
+      "epoch": 5.075078135616252,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.8280761652398423e-05,
+      "loss": 0.7652,
+      "step": 74695
+    },
+    {
+      "epoch": 5.075417855686914,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8278638401956787e-05,
+      "loss": 0.9213,
+      "step": 74700
+    },
+    {
+      "epoch": 5.075757575757576,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8276515151515154e-05,
+      "loss": 0.7944,
+      "step": 74705
+    },
+    {
+      "epoch": 5.076097295828237,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8274391901073518e-05,
+      "loss": 0.8641,
+      "step": 74710
+    },
+    {
+      "epoch": 5.076437015898899,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.827226865063188e-05,
+      "loss": 0.8084,
+      "step": 74715
+    },
+    {
+      "epoch": 5.0767767359695615,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8270145400190243e-05,
+      "loss": 0.8871,
+      "step": 74720
+    },
+    {
+      "epoch": 5.077116456040223,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.826802214974861e-05,
+      "loss": 0.8034,
+      "step": 74725
+    },
+    {
+      "epoch": 5.077456176110885,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.826589889930697e-05,
+      "loss": 0.892,
+      "step": 74730
+    },
+    {
+      "epoch": 5.077795896181547,
+      "grad_norm": 7.25,
+      "learning_rate": 1.8263775648865335e-05,
+      "loss": 0.8017,
+      "step": 74735
+    },
+    {
+      "epoch": 5.078135616252208,
+      "grad_norm": 7.125,
+      "learning_rate": 1.8261652398423702e-05,
+      "loss": 0.8928,
+      "step": 74740
+    },
+    {
+      "epoch": 5.07847533632287,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.8259529147982063e-05,
+      "loss": 0.8087,
+      "step": 74745
+    },
+    {
+      "epoch": 5.078815056393532,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8257405897540427e-05,
+      "loss": 0.8155,
+      "step": 74750
+    },
+    {
+      "epoch": 5.079154776464193,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.825528264709879e-05,
+      "loss": 0.9182,
+      "step": 74755
+    },
+    {
+      "epoch": 5.079494496534855,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8253159396657155e-05,
+      "loss": 0.8653,
+      "step": 74760
+    },
+    {
+      "epoch": 5.0798342166055175,
+      "grad_norm": 6.75,
+      "learning_rate": 1.825103614621552e-05,
+      "loss": 0.9107,
+      "step": 74765
+    },
+    {
+      "epoch": 5.080173936676179,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.8248912895773883e-05,
+      "loss": 0.8383,
+      "step": 74770
+    },
+    {
+      "epoch": 5.080513656746841,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8246789645332247e-05,
+      "loss": 0.9421,
+      "step": 74775
+    },
+    {
+      "epoch": 5.080853376817503,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.824466639489061e-05,
+      "loss": 0.8075,
+      "step": 74780
+    },
+    {
+      "epoch": 5.081193096888164,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.8242543144448975e-05,
+      "loss": 0.8543,
+      "step": 74785
+    },
+    {
+      "epoch": 5.081532816958826,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.824041989400734e-05,
+      "loss": 0.863,
+      "step": 74790
+    },
+    {
+      "epoch": 5.081872537029487,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8238296643565703e-05,
+      "loss": 0.8199,
+      "step": 74795
+    },
+    {
+      "epoch": 5.082212257100149,
+      "grad_norm": 8.5,
+      "learning_rate": 1.8236173393124067e-05,
+      "loss": 0.8655,
+      "step": 74800
+    },
+    {
+      "epoch": 5.082551977170811,
+      "grad_norm": 7.125,
+      "learning_rate": 1.823405014268243e-05,
+      "loss": 0.7868,
+      "step": 74805
+    },
+    {
+      "epoch": 5.082891697241473,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8231926892240795e-05,
+      "loss": 0.8136,
+      "step": 74810
+    },
+    {
+      "epoch": 5.083231417312135,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.822980364179916e-05,
+      "loss": 0.8563,
+      "step": 74815
+    },
+    {
+      "epoch": 5.083571137382797,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.822768039135752e-05,
+      "loss": 0.8153,
+      "step": 74820
+    },
+    {
+      "epoch": 5.083910857453458,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.8225557140915887e-05,
+      "loss": 0.8834,
+      "step": 74825
+    },
+    {
+      "epoch": 5.08425057752412,
+      "grad_norm": 8.0,
+      "learning_rate": 1.822343389047425e-05,
+      "loss": 0.8757,
+      "step": 74830
+    },
+    {
+      "epoch": 5.084590297594782,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.822131064003261e-05,
+      "loss": 0.8535,
+      "step": 74835
+    },
+    {
+      "epoch": 5.084930017665443,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.821918738959098e-05,
+      "loss": 0.8151,
+      "step": 74840
+    },
+    {
+      "epoch": 5.085269737736105,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8217064139149343e-05,
+      "loss": 0.9318,
+      "step": 74845
+    },
+    {
+      "epoch": 5.085609457806767,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8214940888707704e-05,
+      "loss": 0.8382,
+      "step": 74850
+    },
+    {
+      "epoch": 5.085949177877429,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8212817638266068e-05,
+      "loss": 0.8425,
+      "step": 74855
+    },
+    {
+      "epoch": 5.086288897948091,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.8210694387824435e-05,
+      "loss": 0.8176,
+      "step": 74860
+    },
+    {
+      "epoch": 5.086628618018753,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8208571137382796e-05,
+      "loss": 0.8065,
+      "step": 74865
+    },
+    {
+      "epoch": 5.086968338089414,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.820644788694116e-05,
+      "loss": 0.8155,
+      "step": 74870
+    },
+    {
+      "epoch": 5.087308058160076,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.8204324636499527e-05,
+      "loss": 0.8119,
+      "step": 74875
+    },
+    {
+      "epoch": 5.087647778230738,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.820220138605789e-05,
+      "loss": 0.8747,
+      "step": 74880
+    },
+    {
+      "epoch": 5.087987498301399,
+      "grad_norm": 7.75,
+      "learning_rate": 1.820007813561625e-05,
+      "loss": 0.8776,
+      "step": 74885
+    },
+    {
+      "epoch": 5.088327218372061,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8197954885174616e-05,
+      "loss": 0.8395,
+      "step": 74890
+    },
+    {
+      "epoch": 5.088666938442723,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8195831634732983e-05,
+      "loss": 0.8549,
+      "step": 74895
+    },
+    {
+      "epoch": 5.089006658513385,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8193708384291344e-05,
+      "loss": 0.8348,
+      "step": 74900
+    },
+    {
+      "epoch": 5.089346378584047,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.8191585133849708e-05,
+      "loss": 0.834,
+      "step": 74905
+    },
+    {
+      "epoch": 5.089686098654709,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8189461883408075e-05,
+      "loss": 0.7889,
+      "step": 74910
+    },
+    {
+      "epoch": 5.09002581872537,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8187338632966436e-05,
+      "loss": 0.8514,
+      "step": 74915
+    },
+    {
+      "epoch": 5.090365538796032,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.81852153825248e-05,
+      "loss": 0.8901,
+      "step": 74920
+    },
+    {
+      "epoch": 5.090705258866694,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8183092132083167e-05,
+      "loss": 0.905,
+      "step": 74925
+    },
+    {
+      "epoch": 5.091044978937355,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8180968881641528e-05,
+      "loss": 0.886,
+      "step": 74930
+    },
+    {
+      "epoch": 5.091384699008017,
+      "grad_norm": 8.125,
+      "learning_rate": 1.817884563119989e-05,
+      "loss": 0.847,
+      "step": 74935
+    },
+    {
+      "epoch": 5.0917244190786795,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8176722380758256e-05,
+      "loss": 0.9346,
+      "step": 74940
+    },
+    {
+      "epoch": 5.092064139149341,
+      "grad_norm": 7.75,
+      "learning_rate": 1.817459913031662e-05,
+      "loss": 0.8574,
+      "step": 74945
+    },
+    {
+      "epoch": 5.092403859220003,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8172475879874984e-05,
+      "loss": 0.8631,
+      "step": 74950
+    },
+    {
+      "epoch": 5.092743579290665,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8170352629433348e-05,
+      "loss": 0.8573,
+      "step": 74955
+    },
+    {
+      "epoch": 5.093083299361326,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.8168229378991712e-05,
+      "loss": 0.8114,
+      "step": 74960
+    },
+    {
+      "epoch": 5.093423019431988,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8166106128550076e-05,
+      "loss": 0.8818,
+      "step": 74965
+    },
+    {
+      "epoch": 5.09376273950265,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.816398287810844e-05,
+      "loss": 0.8686,
+      "step": 74970
+    },
+    {
+      "epoch": 5.094102459573311,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8161859627666804e-05,
+      "loss": 0.908,
+      "step": 74975
+    },
+    {
+      "epoch": 5.094442179643973,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8159736377225168e-05,
+      "loss": 0.8901,
+      "step": 74980
+    },
+    {
+      "epoch": 5.0947818997146355,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8157613126783532e-05,
+      "loss": 0.909,
+      "step": 74985
+    },
+    {
+      "epoch": 5.095121619785297,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.8155489876341892e-05,
+      "loss": 0.8054,
+      "step": 74990
+    },
+    {
+      "epoch": 5.095461339855959,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.815336662590026e-05,
+      "loss": 0.8976,
+      "step": 74995
+    },
+    {
+      "epoch": 5.095801059926621,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8151243375458624e-05,
+      "loss": 0.9343,
+      "step": 75000
+    },
+    {
+      "epoch": 5.096140779997282,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8149120125016984e-05,
+      "loss": 0.844,
+      "step": 75005
+    },
+    {
+      "epoch": 5.096480500067944,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8146996874575352e-05,
+      "loss": 0.9793,
+      "step": 75010
+    },
+    {
+      "epoch": 5.096820220138606,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8144873624133716e-05,
+      "loss": 0.7744,
+      "step": 75015
+    },
+    {
+      "epoch": 5.097159940209267,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8142750373692076e-05,
+      "loss": 0.8684,
+      "step": 75020
+    },
+    {
+      "epoch": 5.097499660279929,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8140627123250444e-05,
+      "loss": 0.904,
+      "step": 75025
+    },
+    {
+      "epoch": 5.0978393803505915,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8138503872808808e-05,
+      "loss": 0.8246,
+      "step": 75030
+    },
+    {
+      "epoch": 5.098179100421253,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.813638062236717e-05,
+      "loss": 0.7738,
+      "step": 75035
+    },
+    {
+      "epoch": 5.098518820491915,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8134257371925532e-05,
+      "loss": 0.8592,
+      "step": 75040
+    },
+    {
+      "epoch": 5.098858540562577,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.81321341214839e-05,
+      "loss": 0.9009,
+      "step": 75045
+    },
+    {
+      "epoch": 5.099198260633238,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8130010871042264e-05,
+      "loss": 0.7684,
+      "step": 75050
+    },
+    {
+      "epoch": 5.0995379807039,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8127887620600624e-05,
+      "loss": 0.809,
+      "step": 75055
+    },
+    {
+      "epoch": 5.099877700774562,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8125764370158992e-05,
+      "loss": 0.8687,
+      "step": 75060
+    },
+    {
+      "epoch": 5.100217420845223,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8123641119717356e-05,
+      "loss": 0.804,
+      "step": 75065
+    },
+    {
+      "epoch": 5.100557140915885,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8121517869275716e-05,
+      "loss": 0.9402,
+      "step": 75070
+    },
+    {
+      "epoch": 5.1008968609865475,
+      "grad_norm": 8.375,
+      "learning_rate": 1.811939461883408e-05,
+      "loss": 0.7831,
+      "step": 75075
+    },
+    {
+      "epoch": 5.101236581057209,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8117271368392448e-05,
+      "loss": 0.9081,
+      "step": 75080
+    },
+    {
+      "epoch": 5.101576301127871,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.811514811795081e-05,
+      "loss": 0.8355,
+      "step": 75085
+    },
+    {
+      "epoch": 5.101916021198533,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8113024867509172e-05,
+      "loss": 0.8925,
+      "step": 75090
+    },
+    {
+      "epoch": 5.102255741269194,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.811090161706754e-05,
+      "loss": 0.8124,
+      "step": 75095
+    },
+    {
+      "epoch": 5.102595461339856,
+      "grad_norm": 9.0,
+      "learning_rate": 1.81087783666259e-05,
+      "loss": 0.9122,
+      "step": 75100
+    },
+    {
+      "epoch": 5.102935181410518,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8106655116184264e-05,
+      "loss": 0.7841,
+      "step": 75105
+    },
+    {
+      "epoch": 5.103274901481179,
+      "grad_norm": 7.125,
+      "learning_rate": 1.810453186574263e-05,
+      "loss": 0.846,
+      "step": 75110
+    },
+    {
+      "epoch": 5.103614621551841,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.8102408615300992e-05,
+      "loss": 0.821,
+      "step": 75115
+    },
+    {
+      "epoch": 5.103954341622503,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8100285364859356e-05,
+      "loss": 0.8293,
+      "step": 75120
+    },
+    {
+      "epoch": 5.104294061693165,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.809816211441772e-05,
+      "loss": 0.846,
+      "step": 75125
+    },
+    {
+      "epoch": 5.104633781763827,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8096038863976084e-05,
+      "loss": 0.7416,
+      "step": 75130
+    },
+    {
+      "epoch": 5.104973501834488,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.809391561353445e-05,
+      "loss": 0.7727,
+      "step": 75135
+    },
+    {
+      "epoch": 5.10531322190515,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8091792363092812e-05,
+      "loss": 0.8182,
+      "step": 75140
+    },
+    {
+      "epoch": 5.105652941975812,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8089669112651176e-05,
+      "loss": 0.8062,
+      "step": 75145
+    },
+    {
+      "epoch": 5.105992662046473,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.808754586220954e-05,
+      "loss": 0.9035,
+      "step": 75150
+    },
+    {
+      "epoch": 5.106332382117135,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.8085422611767904e-05,
+      "loss": 0.9782,
+      "step": 75155
+    },
+    {
+      "epoch": 5.106672102187797,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.808329936132627e-05,
+      "loss": 0.7844,
+      "step": 75160
+    },
+    {
+      "epoch": 5.107011822258459,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8081176110884632e-05,
+      "loss": 0.9087,
+      "step": 75165
+    },
+    {
+      "epoch": 5.107351542329121,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.8079052860442996e-05,
+      "loss": 0.8119,
+      "step": 75170
+    },
+    {
+      "epoch": 5.107691262399783,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8076929610001357e-05,
+      "loss": 0.7889,
+      "step": 75175
+    },
+    {
+      "epoch": 5.108030982470444,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.8074806359559724e-05,
+      "loss": 0.8009,
+      "step": 75180
+    },
+    {
+      "epoch": 5.108370702541106,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.807268310911809e-05,
+      "loss": 0.897,
+      "step": 75185
+    },
+    {
+      "epoch": 5.108710422611768,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.807055985867645e-05,
+      "loss": 0.8329,
+      "step": 75190
+    },
+    {
+      "epoch": 5.109050142682429,
+      "grad_norm": 7.75,
+      "learning_rate": 1.8068436608234816e-05,
+      "loss": 0.9039,
+      "step": 75195
+    },
+    {
+      "epoch": 5.109389862753091,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.806631335779318e-05,
+      "loss": 0.8526,
+      "step": 75200
+    },
+    {
+      "epoch": 5.1097295828237534,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.806419010735154e-05,
+      "loss": 0.8527,
+      "step": 75205
+    },
+    {
+      "epoch": 5.110069302894415,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8062066856909905e-05,
+      "loss": 0.8129,
+      "step": 75210
+    },
+    {
+      "epoch": 5.110409022965077,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.8059943606468273e-05,
+      "loss": 0.8471,
+      "step": 75215
+    },
+    {
+      "epoch": 5.110748743035739,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.8057820356026637e-05,
+      "loss": 0.8857,
+      "step": 75220
+    },
+    {
+      "epoch": 5.1110884631064,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8055697105584997e-05,
+      "loss": 0.9342,
+      "step": 75225
+    },
+    {
+      "epoch": 5.111428183177062,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8053573855143365e-05,
+      "loss": 0.8928,
+      "step": 75230
+    },
+    {
+      "epoch": 5.111767903247724,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.805145060470173e-05,
+      "loss": 0.8451,
+      "step": 75235
+    },
+    {
+      "epoch": 5.112107623318385,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.804932735426009e-05,
+      "loss": 0.8464,
+      "step": 75240
+    },
+    {
+      "epoch": 5.112447343389047,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.8047204103818453e-05,
+      "loss": 0.8569,
+      "step": 75245
+    },
+    {
+      "epoch": 5.1127870634597095,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.804508085337682e-05,
+      "loss": 0.8406,
+      "step": 75250
+    },
+    {
+      "epoch": 5.113126783530371,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.804295760293518e-05,
+      "loss": 0.8483,
+      "step": 75255
+    },
+    {
+      "epoch": 5.113466503601033,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.8040834352493545e-05,
+      "loss": 0.7847,
+      "step": 75260
+    },
+    {
+      "epoch": 5.113806223671695,
+      "grad_norm": 8.5,
+      "learning_rate": 1.8038711102051913e-05,
+      "loss": 0.8596,
+      "step": 75265
+    },
+    {
+      "epoch": 5.114145943742356,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8036587851610273e-05,
+      "loss": 0.8743,
+      "step": 75270
+    },
+    {
+      "epoch": 5.114485663813018,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8034464601168637e-05,
+      "loss": 0.7719,
+      "step": 75275
+    },
+    {
+      "epoch": 5.11482538388368,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8032341350727e-05,
+      "loss": 0.8143,
+      "step": 75280
+    },
+    {
+      "epoch": 5.115165103954341,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8030218100285365e-05,
+      "loss": 0.9055,
+      "step": 75285
+    },
+    {
+      "epoch": 5.115504824025003,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.802809484984373e-05,
+      "loss": 0.8194,
+      "step": 75290
+    },
+    {
+      "epoch": 5.1158445440956655,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8025971599402093e-05,
+      "loss": 0.9071,
+      "step": 75295
+    },
+    {
+      "epoch": 5.116184264166327,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.8023848348960457e-05,
+      "loss": 0.8264,
+      "step": 75300
+    },
+    {
+      "epoch": 5.116523984236989,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.802172509851882e-05,
+      "loss": 0.7748,
+      "step": 75305
+    },
+    {
+      "epoch": 5.116863704307651,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.8019601848077185e-05,
+      "loss": 0.9064,
+      "step": 75310
+    },
+    {
+      "epoch": 5.117203424378312,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.801747859763555e-05,
+      "loss": 0.839,
+      "step": 75315
+    },
+    {
+      "epoch": 5.117543144448974,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8015355347193913e-05,
+      "loss": 0.9177,
+      "step": 75320
+    },
+    {
+      "epoch": 5.117882864519636,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8013232096752277e-05,
+      "loss": 0.9853,
+      "step": 75325
+    },
+    {
+      "epoch": 5.118222584590297,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.801110884631064e-05,
+      "loss": 0.8819,
+      "step": 75330
+    },
+    {
+      "epoch": 5.118562304660959,
+      "grad_norm": 7.875,
+      "learning_rate": 1.8008985595869005e-05,
+      "loss": 0.8031,
+      "step": 75335
+    },
+    {
+      "epoch": 5.1189020247316215,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.800686234542737e-05,
+      "loss": 0.8227,
+      "step": 75340
+    },
+    {
+      "epoch": 5.119241744802283,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.800473909498573e-05,
+      "loss": 0.827,
+      "step": 75345
+    },
+    {
+      "epoch": 5.119581464872945,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.8002615844544097e-05,
+      "loss": 0.9411,
+      "step": 75350
+    },
+    {
+      "epoch": 5.119921184943607,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.800049259410246e-05,
+      "loss": 0.8487,
+      "step": 75355
+    },
+    {
+      "epoch": 5.120260905014268,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7998369343660822e-05,
+      "loss": 0.8545,
+      "step": 75360
+    },
+    {
+      "epoch": 5.12060062508493,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.799624609321919e-05,
+      "loss": 0.9424,
+      "step": 75365
+    },
+    {
+      "epoch": 5.120940345155592,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.7994122842777553e-05,
+      "loss": 0.8572,
+      "step": 75370
+    },
+    {
+      "epoch": 5.121280065226253,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7991999592335914e-05,
+      "loss": 0.8384,
+      "step": 75375
+    },
+    {
+      "epoch": 5.121619785296915,
+      "grad_norm": 6.75,
+      "learning_rate": 1.7989876341894278e-05,
+      "loss": 0.8552,
+      "step": 75380
+    },
+    {
+      "epoch": 5.1219595053675775,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.7987753091452645e-05,
+      "loss": 0.8047,
+      "step": 75385
+    },
+    {
+      "epoch": 5.122299225438239,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.798562984101101e-05,
+      "loss": 0.8456,
+      "step": 75390
+    },
+    {
+      "epoch": 5.122638945508901,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.798350659056937e-05,
+      "loss": 0.7985,
+      "step": 75395
+    },
+    {
+      "epoch": 5.122978665579563,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7981383340127737e-05,
+      "loss": 0.8381,
+      "step": 75400
+    },
+    {
+      "epoch": 5.123318385650224,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.79792600896861e-05,
+      "loss": 0.8544,
+      "step": 75405
+    },
+    {
+      "epoch": 5.123658105720886,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.7977136839244462e-05,
+      "loss": 0.8723,
+      "step": 75410
+    },
+    {
+      "epoch": 5.123997825791548,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.797501358880283e-05,
+      "loss": 0.8685,
+      "step": 75415
+    },
+    {
+      "epoch": 5.124337545862209,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7972890338361193e-05,
+      "loss": 0.8326,
+      "step": 75420
+    },
+    {
+      "epoch": 5.124677265932871,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7970767087919554e-05,
+      "loss": 0.8679,
+      "step": 75425
+    },
+    {
+      "epoch": 5.1250169860035335,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7968643837477918e-05,
+      "loss": 0.8839,
+      "step": 75430
+    },
+    {
+      "epoch": 5.125356706074195,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7966520587036285e-05,
+      "loss": 0.8571,
+      "step": 75435
+    },
+    {
+      "epoch": 5.125696426144857,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7964397336594646e-05,
+      "loss": 0.8728,
+      "step": 75440
+    },
+    {
+      "epoch": 5.126036146215519,
+      "grad_norm": 7.875,
+      "learning_rate": 1.796227408615301e-05,
+      "loss": 0.9128,
+      "step": 75445
+    },
+    {
+      "epoch": 5.12637586628618,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.7960150835711377e-05,
+      "loss": 0.787,
+      "step": 75450
+    },
+    {
+      "epoch": 5.126715586356842,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7958027585269738e-05,
+      "loss": 0.9386,
+      "step": 75455
+    },
+    {
+      "epoch": 5.127055306427504,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7955904334828102e-05,
+      "loss": 0.8405,
+      "step": 75460
+    },
+    {
+      "epoch": 5.127395026498165,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7953781084386466e-05,
+      "loss": 0.7901,
+      "step": 75465
+    },
+    {
+      "epoch": 5.127734746568827,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.795165783394483e-05,
+      "loss": 0.8531,
+      "step": 75470
+    },
+    {
+      "epoch": 5.1280744666394895,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7949534583503194e-05,
+      "loss": 0.8674,
+      "step": 75475
+    },
+    {
+      "epoch": 5.128414186710151,
+      "grad_norm": 7.125,
+      "learning_rate": 1.7947411333061558e-05,
+      "loss": 0.8799,
+      "step": 75480
+    },
+    {
+      "epoch": 5.128753906780813,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.7945288082619922e-05,
+      "loss": 0.8526,
+      "step": 75485
+    },
+    {
+      "epoch": 5.129093626851474,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.7943164832178286e-05,
+      "loss": 0.8477,
+      "step": 75490
+    },
+    {
+      "epoch": 5.129433346922136,
+      "grad_norm": 8.0,
+      "learning_rate": 1.794104158173665e-05,
+      "loss": 0.9258,
+      "step": 75495
+    },
+    {
+      "epoch": 5.129773066992798,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7938918331295014e-05,
+      "loss": 0.8961,
+      "step": 75500
+    },
+    {
+      "epoch": 5.130112787063459,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.7936795080853378e-05,
+      "loss": 0.867,
+      "step": 75505
+    },
+    {
+      "epoch": 5.130452507134121,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7934671830411742e-05,
+      "loss": 0.8561,
+      "step": 75510
+    },
+    {
+      "epoch": 5.1307922272047835,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.7932548579970106e-05,
+      "loss": 0.9161,
+      "step": 75515
+    },
+    {
+      "epoch": 5.131131947275445,
+      "grad_norm": 7.0,
+      "learning_rate": 1.793042532952847e-05,
+      "loss": 0.8205,
+      "step": 75520
+    },
+    {
+      "epoch": 5.131471667346107,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7928302079086834e-05,
+      "loss": 0.8452,
+      "step": 75525
+    },
+    {
+      "epoch": 5.131811387416769,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7926178828645195e-05,
+      "loss": 0.8644,
+      "step": 75530
+    },
+    {
+      "epoch": 5.13215110748743,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.7924055578203562e-05,
+      "loss": 0.9007,
+      "step": 75535
+    },
+    {
+      "epoch": 5.132490827558092,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.7921932327761926e-05,
+      "loss": 0.828,
+      "step": 75540
+    },
+    {
+      "epoch": 5.132830547628754,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7919809077320287e-05,
+      "loss": 0.9974,
+      "step": 75545
+    },
+    {
+      "epoch": 5.133170267699415,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.7917685826878654e-05,
+      "loss": 0.7755,
+      "step": 75550
+    },
+    {
+      "epoch": 5.133509987770077,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7915562576437018e-05,
+      "loss": 0.8994,
+      "step": 75555
+    },
+    {
+      "epoch": 5.1338497078407395,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.7913439325995382e-05,
+      "loss": 0.8251,
+      "step": 75560
+    },
+    {
+      "epoch": 5.134189427911401,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7911316075553743e-05,
+      "loss": 0.8677,
+      "step": 75565
+    },
+    {
+      "epoch": 5.134529147982063,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.790919282511211e-05,
+      "loss": 0.8513,
+      "step": 75570
+    },
+    {
+      "epoch": 5.134868868052725,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7907069574670474e-05,
+      "loss": 0.826,
+      "step": 75575
+    },
+    {
+      "epoch": 5.135208588123386,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7904946324228835e-05,
+      "loss": 0.9258,
+      "step": 75580
+    },
+    {
+      "epoch": 5.135548308194048,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.7902823073787202e-05,
+      "loss": 0.9187,
+      "step": 75585
+    },
+    {
+      "epoch": 5.13588802826471,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7900699823345566e-05,
+      "loss": 0.8609,
+      "step": 75590
+    },
+    {
+      "epoch": 5.136227748335371,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7898576572903927e-05,
+      "loss": 0.9117,
+      "step": 75595
+    },
+    {
+      "epoch": 5.136567468406033,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.789645332246229e-05,
+      "loss": 0.8656,
+      "step": 75600
+    },
+    {
+      "epoch": 5.1369071884766955,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.7894330072020658e-05,
+      "loss": 0.7819,
+      "step": 75605
+    },
+    {
+      "epoch": 5.137246908547357,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.789220682157902e-05,
+      "loss": 0.8572,
+      "step": 75610
+    },
+    {
+      "epoch": 5.137586628618019,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7890083571137383e-05,
+      "loss": 0.8799,
+      "step": 75615
+    },
+    {
+      "epoch": 5.137926348688681,
+      "grad_norm": 8.25,
+      "learning_rate": 1.788796032069575e-05,
+      "loss": 0.8901,
+      "step": 75620
+    },
+    {
+      "epoch": 5.138266068759342,
+      "grad_norm": 8.375,
+      "learning_rate": 1.788583707025411e-05,
+      "loss": 0.809,
+      "step": 75625
+    },
+    {
+      "epoch": 5.138605788830004,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.7883713819812475e-05,
+      "loss": 0.8221,
+      "step": 75630
+    },
+    {
+      "epoch": 5.138945508900666,
+      "grad_norm": 7.5,
+      "learning_rate": 1.788159056937084e-05,
+      "loss": 0.81,
+      "step": 75635
+    },
+    {
+      "epoch": 5.139285228971327,
+      "grad_norm": 7.125,
+      "learning_rate": 1.7879467318929203e-05,
+      "loss": 0.8584,
+      "step": 75640
+    },
+    {
+      "epoch": 5.139624949041989,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7877344068487567e-05,
+      "loss": 0.7964,
+      "step": 75645
+    },
+    {
+      "epoch": 5.1399646691126515,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.787522081804593e-05,
+      "loss": 0.8207,
+      "step": 75650
+    },
+    {
+      "epoch": 5.140304389183313,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.7873097567604295e-05,
+      "loss": 0.8147,
+      "step": 75655
+    },
+    {
+      "epoch": 5.140644109253975,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.787097431716266e-05,
+      "loss": 0.858,
+      "step": 75660
+    },
+    {
+      "epoch": 5.140983829324637,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.7868851066721023e-05,
+      "loss": 0.8219,
+      "step": 75665
+    },
+    {
+      "epoch": 5.141323549395298,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.7866727816279387e-05,
+      "loss": 0.9744,
+      "step": 75670
+    },
+    {
+      "epoch": 5.14166326946596,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.786460456583775e-05,
+      "loss": 0.9403,
+      "step": 75675
+    },
+    {
+      "epoch": 5.142002989536622,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7862481315396115e-05,
+      "loss": 0.8617,
+      "step": 75680
+    },
+    {
+      "epoch": 5.142342709607283,
+      "grad_norm": 7.375,
+      "learning_rate": 1.786035806495448e-05,
+      "loss": 0.8643,
+      "step": 75685
+    },
+    {
+      "epoch": 5.142682429677945,
+      "grad_norm": 7.25,
+      "learning_rate": 1.7858234814512843e-05,
+      "loss": 0.8442,
+      "step": 75690
+    },
+    {
+      "epoch": 5.1430221497486075,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7856111564071207e-05,
+      "loss": 0.7897,
+      "step": 75695
+    },
+    {
+      "epoch": 5.143361869819269,
+      "grad_norm": 9.25,
+      "learning_rate": 1.7853988313629567e-05,
+      "loss": 0.8984,
+      "step": 75700
+    },
+    {
+      "epoch": 5.143701589889931,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.7851865063187935e-05,
+      "loss": 0.8877,
+      "step": 75705
+    },
+    {
+      "epoch": 5.144041309960593,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.78497418127463e-05,
+      "loss": 0.8266,
+      "step": 75710
+    },
+    {
+      "epoch": 5.144381030031254,
+      "grad_norm": 8.125,
+      "learning_rate": 1.784761856230466e-05,
+      "loss": 0.8532,
+      "step": 75715
+    },
+    {
+      "epoch": 5.144720750101916,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7845495311863027e-05,
+      "loss": 0.8905,
+      "step": 75720
+    },
+    {
+      "epoch": 5.145060470172578,
+      "grad_norm": 8.125,
+      "learning_rate": 1.784337206142139e-05,
+      "loss": 0.9039,
+      "step": 75725
+    },
+    {
+      "epoch": 5.145400190243239,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7841248810979755e-05,
+      "loss": 0.8471,
+      "step": 75730
+    },
+    {
+      "epoch": 5.145739910313901,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.7839125560538115e-05,
+      "loss": 0.8729,
+      "step": 75735
+    },
+    {
+      "epoch": 5.1460796303845635,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7837002310096483e-05,
+      "loss": 0.8119,
+      "step": 75740
+    },
+    {
+      "epoch": 5.146419350455225,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7834879059654847e-05,
+      "loss": 0.8101,
+      "step": 75745
+    },
+    {
+      "epoch": 5.146759070525887,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.7832755809213207e-05,
+      "loss": 0.9455,
+      "step": 75750
+    },
+    {
+      "epoch": 5.147098790596549,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.7830632558771575e-05,
+      "loss": 0.7849,
+      "step": 75755
+    },
+    {
+      "epoch": 5.14743851066721,
+      "grad_norm": 8.125,
+      "learning_rate": 1.782850930832994e-05,
+      "loss": 0.8687,
+      "step": 75760
+    },
+    {
+      "epoch": 5.147778230737872,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.78263860578883e-05,
+      "loss": 0.8229,
+      "step": 75765
+    },
+    {
+      "epoch": 5.148117950808534,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7824262807446663e-05,
+      "loss": 0.8574,
+      "step": 75770
+    },
+    {
+      "epoch": 5.148457670879195,
+      "grad_norm": 7.5,
+      "learning_rate": 1.782213955700503e-05,
+      "loss": 0.871,
+      "step": 75775
+    },
+    {
+      "epoch": 5.1487973909498574,
+      "grad_norm": 7.75,
+      "learning_rate": 1.782001630656339e-05,
+      "loss": 0.8164,
+      "step": 75780
+    },
+    {
+      "epoch": 5.1491371110205195,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7817893056121755e-05,
+      "loss": 0.8278,
+      "step": 75785
+    },
+    {
+      "epoch": 5.149476831091181,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7815769805680123e-05,
+      "loss": 0.8813,
+      "step": 75790
+    },
+    {
+      "epoch": 5.149816551161843,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7813646555238483e-05,
+      "loss": 0.7883,
+      "step": 75795
+    },
+    {
+      "epoch": 5.150156271232504,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7811523304796847e-05,
+      "loss": 0.8712,
+      "step": 75800
+    },
+    {
+      "epoch": 5.150495991303166,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.780940005435521e-05,
+      "loss": 0.9391,
+      "step": 75805
+    },
+    {
+      "epoch": 5.150835711373828,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7807276803913575e-05,
+      "loss": 0.9439,
+      "step": 75810
+    },
+    {
+      "epoch": 5.151175431444489,
+      "grad_norm": 8.75,
+      "learning_rate": 1.780515355347194e-05,
+      "loss": 0.882,
+      "step": 75815
+    },
+    {
+      "epoch": 5.151515151515151,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7803030303030303e-05,
+      "loss": 0.8798,
+      "step": 75820
+    },
+    {
+      "epoch": 5.1518548715858135,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7800907052588667e-05,
+      "loss": 0.822,
+      "step": 75825
+    },
+    {
+      "epoch": 5.152194591656475,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.779878380214703e-05,
+      "loss": 0.8652,
+      "step": 75830
+    },
+    {
+      "epoch": 5.152534311727137,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7796660551705395e-05,
+      "loss": 0.9089,
+      "step": 75835
+    },
+    {
+      "epoch": 5.152874031797799,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.779453730126376e-05,
+      "loss": 0.9364,
+      "step": 75840
+    },
+    {
+      "epoch": 5.15321375186846,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.7792414050822123e-05,
+      "loss": 0.8229,
+      "step": 75845
+    },
+    {
+      "epoch": 5.153553471939122,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7790290800380487e-05,
+      "loss": 0.8076,
+      "step": 75850
+    },
+    {
+      "epoch": 5.153893192009784,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.778816754993885e-05,
+      "loss": 0.8887,
+      "step": 75855
+    },
+    {
+      "epoch": 5.154232912080445,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7786044299497215e-05,
+      "loss": 0.8724,
+      "step": 75860
+    },
+    {
+      "epoch": 5.154572632151107,
+      "grad_norm": 7.75,
+      "learning_rate": 1.778392104905558e-05,
+      "loss": 0.9068,
+      "step": 75865
+    },
+    {
+      "epoch": 5.1549123522217695,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.778179779861394e-05,
+      "loss": 0.98,
+      "step": 75870
+    },
+    {
+      "epoch": 5.155252072292431,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7779674548172307e-05,
+      "loss": 0.798,
+      "step": 75875
+    },
+    {
+      "epoch": 5.155591792363093,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.777755129773067e-05,
+      "loss": 0.8802,
+      "step": 75880
+    },
+    {
+      "epoch": 5.155931512433755,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7775428047289032e-05,
+      "loss": 0.8653,
+      "step": 75885
+    },
+    {
+      "epoch": 5.156271232504416,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.77733047968474e-05,
+      "loss": 0.8909,
+      "step": 75890
+    },
+    {
+      "epoch": 5.156610952575078,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7771181546405763e-05,
+      "loss": 0.8783,
+      "step": 75895
+    },
+    {
+      "epoch": 5.15695067264574,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7769058295964127e-05,
+      "loss": 0.8487,
+      "step": 75900
+    },
+    {
+      "epoch": 5.157290392716401,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7766935045522488e-05,
+      "loss": 0.8277,
+      "step": 75905
+    },
+    {
+      "epoch": 5.157630112787063,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7764811795080855e-05,
+      "loss": 0.8392,
+      "step": 75910
+    },
+    {
+      "epoch": 5.1579698328577255,
+      "grad_norm": 7.625,
+      "learning_rate": 1.776268854463922e-05,
+      "loss": 0.7469,
+      "step": 75915
+    },
+    {
+      "epoch": 5.158309552928387,
+      "grad_norm": 8.25,
+      "learning_rate": 1.776056529419758e-05,
+      "loss": 0.8276,
+      "step": 75920
+    },
+    {
+      "epoch": 5.158649272999049,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.7758442043755948e-05,
+      "loss": 0.872,
+      "step": 75925
+    },
+    {
+      "epoch": 5.158988993069711,
+      "grad_norm": 8.625,
+      "learning_rate": 1.775631879331431e-05,
+      "loss": 0.8512,
+      "step": 75930
+    },
+    {
+      "epoch": 5.159328713140372,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7754195542872672e-05,
+      "loss": 0.7856,
+      "step": 75935
+    },
+    {
+      "epoch": 5.159668433211034,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.775207229243104e-05,
+      "loss": 0.8238,
+      "step": 75940
+    },
+    {
+      "epoch": 5.160008153281696,
+      "grad_norm": 7.125,
+      "learning_rate": 1.7749949041989404e-05,
+      "loss": 0.865,
+      "step": 75945
+    },
+    {
+      "epoch": 5.160347873352357,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7747825791547764e-05,
+      "loss": 0.8188,
+      "step": 75950
+    },
+    {
+      "epoch": 5.160687593423019,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7745702541106128e-05,
+      "loss": 0.8106,
+      "step": 75955
+    },
+    {
+      "epoch": 5.1610273134936815,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7743579290664496e-05,
+      "loss": 0.8596,
+      "step": 75960
+    },
+    {
+      "epoch": 5.161367033564343,
+      "grad_norm": 7.0,
+      "learning_rate": 1.7741456040222856e-05,
+      "loss": 0.7982,
+      "step": 75965
+    },
+    {
+      "epoch": 5.161706753635005,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.773933278978122e-05,
+      "loss": 0.9594,
+      "step": 75970
+    },
+    {
+      "epoch": 5.162046473705667,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7737209539339588e-05,
+      "loss": 0.8281,
+      "step": 75975
+    },
+    {
+      "epoch": 5.162386193776328,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7735086288897948e-05,
+      "loss": 0.8371,
+      "step": 75980
+    },
+    {
+      "epoch": 5.16272591384699,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.7732963038456312e-05,
+      "loss": 0.9205,
+      "step": 75985
+    },
+    {
+      "epoch": 5.163065633917652,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.7730839788014676e-05,
+      "loss": 0.8343,
+      "step": 75990
+    },
+    {
+      "epoch": 5.163405353988313,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.772871653757304e-05,
+      "loss": 0.8938,
+      "step": 75995
+    },
+    {
+      "epoch": 5.163745074058975,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7726593287131404e-05,
+      "loss": 0.958,
+      "step": 76000
+    },
+    {
+      "epoch": 5.1640847941296375,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7724470036689768e-05,
+      "loss": 0.9818,
+      "step": 76005
+    },
+    {
+      "epoch": 5.164424514200299,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7722346786248132e-05,
+      "loss": 0.9146,
+      "step": 76010
+    },
+    {
+      "epoch": 5.164764234270961,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7720223535806496e-05,
+      "loss": 0.8237,
+      "step": 76015
+    },
+    {
+      "epoch": 5.165103954341623,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.771810028536486e-05,
+      "loss": 0.9318,
+      "step": 76020
+    },
+    {
+      "epoch": 5.165443674412284,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7715977034923224e-05,
+      "loss": 0.8526,
+      "step": 76025
+    },
+    {
+      "epoch": 5.165783394482946,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.7713853784481588e-05,
+      "loss": 0.8109,
+      "step": 76030
+    },
+    {
+      "epoch": 5.166123114553608,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7711730534039952e-05,
+      "loss": 0.8079,
+      "step": 76035
+    },
+    {
+      "epoch": 5.166462834624269,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7709607283598316e-05,
+      "loss": 0.9137,
+      "step": 76040
+    },
+    {
+      "epoch": 5.166802554694931,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.770748403315668e-05,
+      "loss": 0.9112,
+      "step": 76045
+    },
+    {
+      "epoch": 5.1671422747655935,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7705360782715044e-05,
+      "loss": 0.8808,
+      "step": 76050
+    },
+    {
+      "epoch": 5.167481994836255,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7703237532273405e-05,
+      "loss": 0.8543,
+      "step": 76055
+    },
+    {
+      "epoch": 5.167821714906917,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7701114281831772e-05,
+      "loss": 0.7216,
+      "step": 76060
+    },
+    {
+      "epoch": 5.168161434977579,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.7698991031390136e-05,
+      "loss": 0.7968,
+      "step": 76065
+    },
+    {
+      "epoch": 5.16850115504824,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.76968677809485e-05,
+      "loss": 0.8243,
+      "step": 76070
+    },
+    {
+      "epoch": 5.168840875118902,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7694744530506864e-05,
+      "loss": 0.872,
+      "step": 76075
+    },
+    {
+      "epoch": 5.169180595189564,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7692621280065228e-05,
+      "loss": 0.9211,
+      "step": 76080
+    },
+    {
+      "epoch": 5.169520315260225,
+      "grad_norm": 9.125,
+      "learning_rate": 1.7690498029623592e-05,
+      "loss": 0.8916,
+      "step": 76085
+    },
+    {
+      "epoch": 5.1698600353308874,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7688374779181953e-05,
+      "loss": 0.9038,
+      "step": 76090
+    },
+    {
+      "epoch": 5.1701997554015495,
+      "grad_norm": 7.75,
+      "learning_rate": 1.768625152874032e-05,
+      "loss": 0.8237,
+      "step": 76095
+    },
+    {
+      "epoch": 5.170539475472211,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.7684128278298684e-05,
+      "loss": 0.8614,
+      "step": 76100
+    },
+    {
+      "epoch": 5.170879195542873,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7682005027857045e-05,
+      "loss": 0.7582,
+      "step": 76105
+    },
+    {
+      "epoch": 5.171218915613535,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7679881777415412e-05,
+      "loss": 0.775,
+      "step": 76110
+    },
+    {
+      "epoch": 5.171558635684196,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7677758526973776e-05,
+      "loss": 0.8799,
+      "step": 76115
+    },
+    {
+      "epoch": 5.171898355754858,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7675635276532137e-05,
+      "loss": 0.8751,
+      "step": 76120
+    },
+    {
+      "epoch": 5.17223807582552,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.76735120260905e-05,
+      "loss": 0.755,
+      "step": 76125
+    },
+    {
+      "epoch": 5.172577795896181,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7671388775648868e-05,
+      "loss": 0.801,
+      "step": 76130
+    },
+    {
+      "epoch": 5.1729175159668435,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.766926552520723e-05,
+      "loss": 0.8888,
+      "step": 76135
+    },
+    {
+      "epoch": 5.1732572360375055,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7667142274765593e-05,
+      "loss": 0.875,
+      "step": 76140
+    },
+    {
+      "epoch": 5.173596956108167,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.766501902432396e-05,
+      "loss": 0.7491,
+      "step": 76145
+    },
+    {
+      "epoch": 5.173936676178829,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.766289577388232e-05,
+      "loss": 0.8918,
+      "step": 76150
+    },
+    {
+      "epoch": 5.174276396249491,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7660772523440685e-05,
+      "loss": 0.8702,
+      "step": 76155
+    },
+    {
+      "epoch": 5.174616116320152,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.765864927299905e-05,
+      "loss": 0.8401,
+      "step": 76160
+    },
+    {
+      "epoch": 5.174955836390814,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7656526022557413e-05,
+      "loss": 0.9634,
+      "step": 76165
+    },
+    {
+      "epoch": 5.175295556461475,
+      "grad_norm": 7.25,
+      "learning_rate": 1.7654402772115777e-05,
+      "loss": 0.8501,
+      "step": 76170
+    },
+    {
+      "epoch": 5.175635276532137,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.765227952167414e-05,
+      "loss": 0.8571,
+      "step": 76175
+    },
+    {
+      "epoch": 5.1759749966027995,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.7650156271232505e-05,
+      "loss": 0.7482,
+      "step": 76180
+    },
+    {
+      "epoch": 5.176314716673461,
+      "grad_norm": 7.375,
+      "learning_rate": 1.764803302079087e-05,
+      "loss": 0.925,
+      "step": 76185
+    },
+    {
+      "epoch": 5.176654436744123,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7645909770349233e-05,
+      "loss": 0.7898,
+      "step": 76190
+    },
+    {
+      "epoch": 5.176994156814785,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7643786519907597e-05,
+      "loss": 0.7843,
+      "step": 76195
+    },
+    {
+      "epoch": 5.177333876885446,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.764166326946596e-05,
+      "loss": 0.8855,
+      "step": 76200
+    },
+    {
+      "epoch": 5.177673596956108,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7639540019024325e-05,
+      "loss": 0.8151,
+      "step": 76205
+    },
+    {
+      "epoch": 5.17801331702677,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.763741676858269e-05,
+      "loss": 0.8349,
+      "step": 76210
+    },
+    {
+      "epoch": 5.178353037097431,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7635293518141053e-05,
+      "loss": 0.9193,
+      "step": 76215
+    },
+    {
+      "epoch": 5.178692757168093,
+      "grad_norm": 7.25,
+      "learning_rate": 1.7633170267699417e-05,
+      "loss": 0.9734,
+      "step": 76220
+    },
+    {
+      "epoch": 5.1790324772387555,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7631047017257778e-05,
+      "loss": 0.8732,
+      "step": 76225
+    },
+    {
+      "epoch": 5.179372197309417,
+      "grad_norm": 7.875,
+      "learning_rate": 1.7628923766816145e-05,
+      "loss": 0.8515,
+      "step": 76230
+    },
+    {
+      "epoch": 5.179711917380079,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.762680051637451e-05,
+      "loss": 0.762,
+      "step": 76235
+    },
+    {
+      "epoch": 5.180051637450741,
+      "grad_norm": 6.75,
+      "learning_rate": 1.7624677265932873e-05,
+      "loss": 0.921,
+      "step": 76240
+    },
+    {
+      "epoch": 5.180391357521402,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.7622554015491237e-05,
+      "loss": 0.801,
+      "step": 76245
+    },
+    {
+      "epoch": 5.180731077592064,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.76204307650496e-05,
+      "loss": 0.8794,
+      "step": 76250
+    },
+    {
+      "epoch": 5.181070797662726,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7618307514607965e-05,
+      "loss": 0.9224,
+      "step": 76255
+    },
+    {
+      "epoch": 5.181410517733387,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.7616184264166326e-05,
+      "loss": 0.8287,
+      "step": 76260
+    },
+    {
+      "epoch": 5.181750237804049,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7614061013724693e-05,
+      "loss": 0.8828,
+      "step": 76265
+    },
+    {
+      "epoch": 5.1820899578747115,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.7611937763283057e-05,
+      "loss": 0.7843,
+      "step": 76270
+    },
+    {
+      "epoch": 5.182429677945373,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7609814512841418e-05,
+      "loss": 0.8929,
+      "step": 76275
+    },
+    {
+      "epoch": 5.182769398016035,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7607691262399785e-05,
+      "loss": 0.8105,
+      "step": 76280
+    },
+    {
+      "epoch": 5.183109118086697,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.760556801195815e-05,
+      "loss": 0.8296,
+      "step": 76285
+    },
+    {
+      "epoch": 5.183448838157358,
+      "grad_norm": 7.75,
+      "learning_rate": 1.760344476151651e-05,
+      "loss": 0.8742,
+      "step": 76290
+    },
+    {
+      "epoch": 5.18378855822802,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7601321511074874e-05,
+      "loss": 0.7665,
+      "step": 76295
+    },
+    {
+      "epoch": 5.184128278298682,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.759919826063324e-05,
+      "loss": 0.9064,
+      "step": 76300
+    },
+    {
+      "epoch": 5.184467998369343,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.75970750101916e-05,
+      "loss": 0.8774,
+      "step": 76305
+    },
+    {
+      "epoch": 5.184807718440005,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7594951759749966e-05,
+      "loss": 0.879,
+      "step": 76310
+    },
+    {
+      "epoch": 5.1851474385106675,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7592828509308333e-05,
+      "loss": 0.7881,
+      "step": 76315
+    },
+    {
+      "epoch": 5.185487158581329,
+      "grad_norm": 7.25,
+      "learning_rate": 1.7590705258866694e-05,
+      "loss": 0.8659,
+      "step": 76320
+    },
+    {
+      "epoch": 5.185826878651991,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7588582008425058e-05,
+      "loss": 0.8743,
+      "step": 76325
+    },
+    {
+      "epoch": 5.186166598722653,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.7586458757983425e-05,
+      "loss": 0.8802,
+      "step": 76330
+    },
+    {
+      "epoch": 5.186506318793314,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.7584335507541786e-05,
+      "loss": 0.9348,
+      "step": 76335
+    },
+    {
+      "epoch": 5.186846038863976,
+      "grad_norm": 7.75,
+      "learning_rate": 1.758221225710015e-05,
+      "loss": 0.818,
+      "step": 76340
+    },
+    {
+      "epoch": 5.187185758934638,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7580089006658514e-05,
+      "loss": 0.7647,
+      "step": 76345
+    },
+    {
+      "epoch": 5.187525479005299,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7577965756216878e-05,
+      "loss": 0.8033,
+      "step": 76350
+    },
+    {
+      "epoch": 5.187865199075961,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.757584250577524e-05,
+      "loss": 0.8887,
+      "step": 76355
+    },
+    {
+      "epoch": 5.1882049191466235,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.7573719255333606e-05,
+      "loss": 0.8294,
+      "step": 76360
+    },
+    {
+      "epoch": 5.188544639217285,
+      "grad_norm": 8.25,
+      "learning_rate": 1.757159600489197e-05,
+      "loss": 0.8383,
+      "step": 76365
+    },
+    {
+      "epoch": 5.188884359287947,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.7569472754450334e-05,
+      "loss": 0.9229,
+      "step": 76370
+    },
+    {
+      "epoch": 5.189224079358609,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7567349504008698e-05,
+      "loss": 0.8094,
+      "step": 76375
+    },
+    {
+      "epoch": 5.18956379942927,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.756522625356706e-05,
+      "loss": 0.9251,
+      "step": 76380
+    },
+    {
+      "epoch": 5.189903519499932,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7563103003125426e-05,
+      "loss": 0.931,
+      "step": 76385
+    },
+    {
+      "epoch": 5.190243239570594,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.756097975268379e-05,
+      "loss": 0.9691,
+      "step": 76390
+    },
+    {
+      "epoch": 5.190582959641255,
+      "grad_norm": 7.625,
+      "learning_rate": 1.755885650224215e-05,
+      "loss": 0.7907,
+      "step": 76395
+    },
+    {
+      "epoch": 5.1909226797119175,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7556733251800518e-05,
+      "loss": 0.8325,
+      "step": 76400
+    },
+    {
+      "epoch": 5.1912623997825795,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7554610001358882e-05,
+      "loss": 0.9008,
+      "step": 76405
+    },
+    {
+      "epoch": 5.191602119853241,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7552486750917246e-05,
+      "loss": 0.8042,
+      "step": 76410
+    },
+    {
+      "epoch": 5.191941839923903,
+      "grad_norm": 7.875,
+      "learning_rate": 1.755036350047561e-05,
+      "loss": 0.79,
+      "step": 76415
+    },
+    {
+      "epoch": 5.192281559994565,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7548240250033974e-05,
+      "loss": 0.7885,
+      "step": 76420
+    },
+    {
+      "epoch": 5.192621280065226,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7546116999592338e-05,
+      "loss": 0.8722,
+      "step": 76425
+    },
+    {
+      "epoch": 5.192961000135888,
+      "grad_norm": 9.0,
+      "learning_rate": 1.7543993749150702e-05,
+      "loss": 0.8801,
+      "step": 76430
+    },
+    {
+      "epoch": 5.19330072020655,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7541870498709066e-05,
+      "loss": 0.7512,
+      "step": 76435
+    },
+    {
+      "epoch": 5.193640440277211,
+      "grad_norm": 8.5,
+      "learning_rate": 1.753974724826743e-05,
+      "loss": 0.7666,
+      "step": 76440
+    },
+    {
+      "epoch": 5.1939801603478735,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.753762399782579e-05,
+      "loss": 0.8886,
+      "step": 76445
+    },
+    {
+      "epoch": 5.1943198804185355,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7535500747384158e-05,
+      "loss": 0.8619,
+      "step": 76450
+    },
+    {
+      "epoch": 5.194659600489197,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7533377496942522e-05,
+      "loss": 0.8294,
+      "step": 76455
+    },
+    {
+      "epoch": 5.194999320559859,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7531254246500882e-05,
+      "loss": 0.9015,
+      "step": 76460
+    },
+    {
+      "epoch": 5.195339040630521,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.752913099605925e-05,
+      "loss": 0.9111,
+      "step": 76465
+    },
+    {
+      "epoch": 5.195678760701182,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7527007745617614e-05,
+      "loss": 0.8116,
+      "step": 76470
+    },
+    {
+      "epoch": 5.196018480771844,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7524884495175974e-05,
+      "loss": 0.8292,
+      "step": 76475
+    },
+    {
+      "epoch": 5.196358200842505,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.752276124473434e-05,
+      "loss": 0.8403,
+      "step": 76480
+    },
+    {
+      "epoch": 5.196697920913167,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7520637994292706e-05,
+      "loss": 0.809,
+      "step": 76485
+    },
+    {
+      "epoch": 5.1970376409838295,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7518514743851066e-05,
+      "loss": 0.8748,
+      "step": 76490
+    },
+    {
+      "epoch": 5.197377361054491,
+      "grad_norm": 8.25,
+      "learning_rate": 1.751639149340943e-05,
+      "loss": 0.8042,
+      "step": 76495
+    },
+    {
+      "epoch": 5.197717081125153,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.7514268242967798e-05,
+      "loss": 0.7995,
+      "step": 76500
+    },
+    {
+      "epoch": 5.198056801195815,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.751214499252616e-05,
+      "loss": 0.898,
+      "step": 76505
+    },
+    {
+      "epoch": 5.198396521266476,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.7510021742084522e-05,
+      "loss": 0.9443,
+      "step": 76510
+    },
+    {
+      "epoch": 5.198736241337138,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.7507898491642886e-05,
+      "loss": 0.8934,
+      "step": 76515
+    },
+    {
+      "epoch": 5.1990759614078,
+      "grad_norm": 8.125,
+      "learning_rate": 1.750577524120125e-05,
+      "loss": 0.8309,
+      "step": 76520
+    },
+    {
+      "epoch": 5.199415681478461,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7503651990759614e-05,
+      "loss": 0.91,
+      "step": 76525
+    },
+    {
+      "epoch": 5.199755401549123,
+      "grad_norm": 8.125,
+      "learning_rate": 1.750152874031798e-05,
+      "loss": 0.8687,
+      "step": 76530
+    },
+    {
+      "epoch": 5.2000951216197855,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.7499405489876342e-05,
+      "loss": 0.8604,
+      "step": 76535
+    },
+    {
+      "epoch": 5.200434841690447,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7497282239434706e-05,
+      "loss": 0.7475,
+      "step": 76540
+    },
+    {
+      "epoch": 5.200774561761109,
+      "grad_norm": 7.75,
+      "learning_rate": 1.749515898899307e-05,
+      "loss": 0.8132,
+      "step": 76545
+    },
+    {
+      "epoch": 5.201114281831771,
+      "grad_norm": 7.125,
+      "learning_rate": 1.7493035738551434e-05,
+      "loss": 0.7632,
+      "step": 76550
+    },
+    {
+      "epoch": 5.201454001902432,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.74909124881098e-05,
+      "loss": 0.8481,
+      "step": 76555
+    },
+    {
+      "epoch": 5.201793721973094,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7488789237668162e-05,
+      "loss": 0.8381,
+      "step": 76560
+    },
+    {
+      "epoch": 5.202133442043756,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.7486665987226526e-05,
+      "loss": 0.7225,
+      "step": 76565
+    },
+    {
+      "epoch": 5.202473162114417,
+      "grad_norm": 9.0,
+      "learning_rate": 1.748454273678489e-05,
+      "loss": 0.7768,
+      "step": 76570
+    },
+    {
+      "epoch": 5.202812882185079,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7482419486343254e-05,
+      "loss": 0.8394,
+      "step": 76575
+    },
+    {
+      "epoch": 5.2031526022557415,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.748029623590162e-05,
+      "loss": 0.8784,
+      "step": 76580
+    },
+    {
+      "epoch": 5.203492322326403,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7478172985459982e-05,
+      "loss": 0.8117,
+      "step": 76585
+    },
+    {
+      "epoch": 5.203832042397065,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7476049735018346e-05,
+      "loss": 0.8729,
+      "step": 76590
+    },
+    {
+      "epoch": 5.204171762467727,
+      "grad_norm": 8.125,
+      "learning_rate": 1.747392648457671e-05,
+      "loss": 0.7824,
+      "step": 76595
+    },
+    {
+      "epoch": 5.204511482538388,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7471803234135074e-05,
+      "loss": 0.8374,
+      "step": 76600
+    },
+    {
+      "epoch": 5.20485120260905,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.746967998369344e-05,
+      "loss": 0.9456,
+      "step": 76605
+    },
+    {
+      "epoch": 5.205190922679712,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7467556733251802e-05,
+      "loss": 0.8745,
+      "step": 76610
+    },
+    {
+      "epoch": 5.205530642750373,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.7465433482810163e-05,
+      "loss": 0.7923,
+      "step": 76615
+    },
+    {
+      "epoch": 5.205870362821035,
+      "grad_norm": 8.875,
+      "learning_rate": 1.746331023236853e-05,
+      "loss": 0.8371,
+      "step": 76620
+    },
+    {
+      "epoch": 5.2062100828916975,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.7461186981926894e-05,
+      "loss": 0.8644,
+      "step": 76625
+    },
+    {
+      "epoch": 5.206549802962359,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7459063731485255e-05,
+      "loss": 0.9151,
+      "step": 76630
+    },
+    {
+      "epoch": 5.206889523033021,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7456940481043622e-05,
+      "loss": 0.849,
+      "step": 76635
+    },
+    {
+      "epoch": 5.207229243103683,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7454817230601987e-05,
+      "loss": 0.7875,
+      "step": 76640
+    },
+    {
+      "epoch": 5.207568963174344,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7452693980160347e-05,
+      "loss": 0.8041,
+      "step": 76645
+    },
+    {
+      "epoch": 5.207908683245006,
+      "grad_norm": 8.25,
+      "learning_rate": 1.745057072971871e-05,
+      "loss": 0.8043,
+      "step": 76650
+    },
+    {
+      "epoch": 5.208248403315668,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.744844747927708e-05,
+      "loss": 0.8764,
+      "step": 76655
+    },
+    {
+      "epoch": 5.208588123386329,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.744632422883544e-05,
+      "loss": 0.8303,
+      "step": 76660
+    },
+    {
+      "epoch": 5.2089278434569914,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7444200978393803e-05,
+      "loss": 0.7697,
+      "step": 76665
+    },
+    {
+      "epoch": 5.2092675635276535,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.744207772795217e-05,
+      "loss": 0.7981,
+      "step": 76670
+    },
+    {
+      "epoch": 5.209607283598315,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.743995447751053e-05,
+      "loss": 0.8339,
+      "step": 76675
+    },
+    {
+      "epoch": 5.209947003668977,
+      "grad_norm": 9.5,
+      "learning_rate": 1.7437831227068895e-05,
+      "loss": 0.8244,
+      "step": 76680
+    },
+    {
+      "epoch": 5.210286723739639,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.743570797662726e-05,
+      "loss": 0.8403,
+      "step": 76685
+    },
+    {
+      "epoch": 5.2106264438103,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7433584726185623e-05,
+      "loss": 0.7892,
+      "step": 76690
+    },
+    {
+      "epoch": 5.210966163880962,
+      "grad_norm": 7.125,
+      "learning_rate": 1.7431461475743987e-05,
+      "loss": 0.8535,
+      "step": 76695
+    },
+    {
+      "epoch": 5.211305883951624,
+      "grad_norm": 8.625,
+      "learning_rate": 1.742933822530235e-05,
+      "loss": 0.9302,
+      "step": 76700
+    },
+    {
+      "epoch": 5.211645604022285,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7427214974860715e-05,
+      "loss": 0.8193,
+      "step": 76705
+    },
+    {
+      "epoch": 5.2119853240929475,
+      "grad_norm": 8.25,
+      "learning_rate": 1.742509172441908e-05,
+      "loss": 0.8448,
+      "step": 76710
+    },
+    {
+      "epoch": 5.2123250441636095,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.7422968473977443e-05,
+      "loss": 0.8429,
+      "step": 76715
+    },
+    {
+      "epoch": 5.212664764234271,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7420845223535807e-05,
+      "loss": 0.8249,
+      "step": 76720
+    },
+    {
+      "epoch": 5.213004484304933,
+      "grad_norm": 8.25,
+      "learning_rate": 1.741872197309417e-05,
+      "loss": 0.8913,
+      "step": 76725
+    },
+    {
+      "epoch": 5.213344204375595,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.7416598722652535e-05,
+      "loss": 0.8432,
+      "step": 76730
+    },
+    {
+      "epoch": 5.213683924446256,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.74144754722109e-05,
+      "loss": 0.9405,
+      "step": 76735
+    },
+    {
+      "epoch": 5.214023644516918,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7412352221769263e-05,
+      "loss": 0.7369,
+      "step": 76740
+    },
+    {
+      "epoch": 5.21436336458758,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7410228971327627e-05,
+      "loss": 0.7819,
+      "step": 76745
+    },
+    {
+      "epoch": 5.214703084658241,
+      "grad_norm": 10.75,
+      "learning_rate": 1.740810572088599e-05,
+      "loss": 0.8167,
+      "step": 76750
+    },
+    {
+      "epoch": 5.2150428047289035,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.7405982470444355e-05,
+      "loss": 0.7989,
+      "step": 76755
+    },
+    {
+      "epoch": 5.2153825247995655,
+      "grad_norm": 7.625,
+      "learning_rate": 1.740385922000272e-05,
+      "loss": 0.8773,
+      "step": 76760
+    },
+    {
+      "epoch": 5.215722244870227,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7401735969561083e-05,
+      "loss": 0.8734,
+      "step": 76765
+    },
+    {
+      "epoch": 5.216061964940889,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7399612719119447e-05,
+      "loss": 0.7432,
+      "step": 76770
+    },
+    {
+      "epoch": 5.216401685011551,
+      "grad_norm": 7.0,
+      "learning_rate": 1.739748946867781e-05,
+      "loss": 0.8478,
+      "step": 76775
+    },
+    {
+      "epoch": 5.216741405082212,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7395366218236175e-05,
+      "loss": 0.8481,
+      "step": 76780
+    },
+    {
+      "epoch": 5.217081125152874,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7393242967794536e-05,
+      "loss": 0.8913,
+      "step": 76785
+    },
+    {
+      "epoch": 5.217420845223536,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7391119717352903e-05,
+      "loss": 0.8396,
+      "step": 76790
+    },
+    {
+      "epoch": 5.217760565294197,
+      "grad_norm": 6.75,
+      "learning_rate": 1.7388996466911267e-05,
+      "loss": 0.8645,
+      "step": 76795
+    },
+    {
+      "epoch": 5.2181002853648595,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.7386873216469628e-05,
+      "loss": 0.8356,
+      "step": 76800
+    },
+    {
+      "epoch": 5.2184400054355216,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7384749966027995e-05,
+      "loss": 0.928,
+      "step": 76805
+    },
+    {
+      "epoch": 5.218779725506183,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.738262671558636e-05,
+      "loss": 0.9481,
+      "step": 76810
+    },
+    {
+      "epoch": 5.219119445576845,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.738050346514472e-05,
+      "loss": 0.9306,
+      "step": 76815
+    },
+    {
+      "epoch": 5.219459165647507,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7378380214703084e-05,
+      "loss": 0.861,
+      "step": 76820
+    },
+    {
+      "epoch": 5.219798885718168,
+      "grad_norm": 8.5,
+      "learning_rate": 1.737625696426145e-05,
+      "loss": 0.8547,
+      "step": 76825
+    },
+    {
+      "epoch": 5.22013860578883,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7374133713819812e-05,
+      "loss": 0.9125,
+      "step": 76830
+    },
+    {
+      "epoch": 5.220478325859492,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7372010463378176e-05,
+      "loss": 0.8608,
+      "step": 76835
+    },
+    {
+      "epoch": 5.220818045930153,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.7369887212936543e-05,
+      "loss": 0.77,
+      "step": 76840
+    },
+    {
+      "epoch": 5.2211577660008155,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7367763962494904e-05,
+      "loss": 0.8375,
+      "step": 76845
+    },
+    {
+      "epoch": 5.221497486071477,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7365640712053268e-05,
+      "loss": 0.8664,
+      "step": 76850
+    },
+    {
+      "epoch": 5.221837206142139,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7363517461611635e-05,
+      "loss": 0.8662,
+      "step": 76855
+    },
+    {
+      "epoch": 5.222176926212801,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7361394211169996e-05,
+      "loss": 0.8282,
+      "step": 76860
+    },
+    {
+      "epoch": 5.222516646283462,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.735927096072836e-05,
+      "loss": 0.876,
+      "step": 76865
+    },
+    {
+      "epoch": 5.222856366354124,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7357147710286724e-05,
+      "loss": 0.9552,
+      "step": 76870
+    },
+    {
+      "epoch": 5.223196086424786,
+      "grad_norm": 8.625,
+      "learning_rate": 1.7355024459845088e-05,
+      "loss": 0.8846,
+      "step": 76875
+    },
+    {
+      "epoch": 5.223535806495447,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7352901209403452e-05,
+      "loss": 0.7637,
+      "step": 76880
+    },
+    {
+      "epoch": 5.223875526566109,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7350777958961816e-05,
+      "loss": 0.8301,
+      "step": 76885
+    },
+    {
+      "epoch": 5.2242152466367715,
+      "grad_norm": 8.125,
+      "learning_rate": 1.734865470852018e-05,
+      "loss": 0.8447,
+      "step": 76890
+    },
+    {
+      "epoch": 5.224554966707433,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7346531458078544e-05,
+      "loss": 0.7639,
+      "step": 76895
+    },
+    {
+      "epoch": 5.224894686778095,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7344408207636908e-05,
+      "loss": 0.8626,
+      "step": 76900
+    },
+    {
+      "epoch": 5.225234406848757,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7342284957195272e-05,
+      "loss": 0.8279,
+      "step": 76905
+    },
+    {
+      "epoch": 5.225574126919418,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7340161706753636e-05,
+      "loss": 0.8571,
+      "step": 76910
+    },
+    {
+      "epoch": 5.22591384699008,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7338038456312e-05,
+      "loss": 0.8106,
+      "step": 76915
+    },
+    {
+      "epoch": 5.226253567060742,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.7335915205870364e-05,
+      "loss": 0.9153,
+      "step": 76920
+    },
+    {
+      "epoch": 5.226593287131403,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.7333791955428728e-05,
+      "loss": 0.7711,
+      "step": 76925
+    },
+    {
+      "epoch": 5.226933007202065,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7331668704987092e-05,
+      "loss": 0.9368,
+      "step": 76930
+    },
+    {
+      "epoch": 5.2272727272727275,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7329545454545456e-05,
+      "loss": 0.823,
+      "step": 76935
+    },
+    {
+      "epoch": 5.227612447343389,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.732742220410382e-05,
+      "loss": 0.9639,
+      "step": 76940
+    },
+    {
+      "epoch": 5.227952167414051,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7325298953662184e-05,
+      "loss": 0.8764,
+      "step": 76945
+    },
+    {
+      "epoch": 5.228291887484713,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7323175703220548e-05,
+      "loss": 0.9202,
+      "step": 76950
+    },
+    {
+      "epoch": 5.228631607555374,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7321052452778912e-05,
+      "loss": 0.8154,
+      "step": 76955
+    },
+    {
+      "epoch": 5.228971327626036,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7318929202337276e-05,
+      "loss": 0.8419,
+      "step": 76960
+    },
+    {
+      "epoch": 5.229311047696698,
+      "grad_norm": 8.0,
+      "learning_rate": 1.731680595189564e-05,
+      "loss": 0.9434,
+      "step": 76965
+    },
+    {
+      "epoch": 5.229650767767359,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.7314682701454e-05,
+      "loss": 0.8756,
+      "step": 76970
+    },
+    {
+      "epoch": 5.2299904878380215,
+      "grad_norm": 7.125,
+      "learning_rate": 1.7312559451012368e-05,
+      "loss": 0.8741,
+      "step": 76975
+    },
+    {
+      "epoch": 5.2303302079086835,
+      "grad_norm": 7.25,
+      "learning_rate": 1.7310436200570732e-05,
+      "loss": 0.8885,
+      "step": 76980
+    },
+    {
+      "epoch": 5.230669927979345,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7308312950129093e-05,
+      "loss": 0.8031,
+      "step": 76985
+    },
+    {
+      "epoch": 5.231009648050007,
+      "grad_norm": 7.25,
+      "learning_rate": 1.730618969968746e-05,
+      "loss": 0.864,
+      "step": 76990
+    },
+    {
+      "epoch": 5.231349368120669,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.7304066449245824e-05,
+      "loss": 0.8482,
+      "step": 76995
+    },
+    {
+      "epoch": 5.23168908819133,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7301943198804185e-05,
+      "loss": 0.7808,
+      "step": 77000
+    },
+    {
+      "epoch": 5.232028808261992,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.729981994836255e-05,
+      "loss": 0.8459,
+      "step": 77005
+    },
+    {
+      "epoch": 5.232368528332654,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.7297696697920916e-05,
+      "loss": 0.8419,
+      "step": 77010
+    },
+    {
+      "epoch": 5.232708248403315,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7295573447479277e-05,
+      "loss": 0.8398,
+      "step": 77015
+    },
+    {
+      "epoch": 5.2330479684739775,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.729345019703764e-05,
+      "loss": 0.7669,
+      "step": 77020
+    },
+    {
+      "epoch": 5.2333876885446395,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.7291326946596008e-05,
+      "loss": 0.8999,
+      "step": 77025
+    },
+    {
+      "epoch": 5.233727408615301,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.728920369615437e-05,
+      "loss": 0.88,
+      "step": 77030
+    },
+    {
+      "epoch": 5.234067128685963,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.7287080445712733e-05,
+      "loss": 0.8788,
+      "step": 77035
+    },
+    {
+      "epoch": 5.234406848756625,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7284957195271097e-05,
+      "loss": 0.8193,
+      "step": 77040
+    },
+    {
+      "epoch": 5.234746568827286,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.728283394482946e-05,
+      "loss": 0.7928,
+      "step": 77045
+    },
+    {
+      "epoch": 5.235086288897948,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7280710694387825e-05,
+      "loss": 0.8701,
+      "step": 77050
+    },
+    {
+      "epoch": 5.23542600896861,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.727858744394619e-05,
+      "loss": 0.7957,
+      "step": 77055
+    },
+    {
+      "epoch": 5.235765729039271,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7276464193504553e-05,
+      "loss": 0.8391,
+      "step": 77060
+    },
+    {
+      "epoch": 5.2361054491099335,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7274340943062917e-05,
+      "loss": 0.8616,
+      "step": 77065
+    },
+    {
+      "epoch": 5.2364451691805955,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.727221769262128e-05,
+      "loss": 0.7806,
+      "step": 77070
+    },
+    {
+      "epoch": 5.236784889251257,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7270094442179645e-05,
+      "loss": 0.8609,
+      "step": 77075
+    },
+    {
+      "epoch": 5.237124609321919,
+      "grad_norm": 8.5,
+      "learning_rate": 1.726797119173801e-05,
+      "loss": 0.9065,
+      "step": 77080
+    },
+    {
+      "epoch": 5.237464329392581,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7265847941296373e-05,
+      "loss": 0.7957,
+      "step": 77085
+    },
+    {
+      "epoch": 5.237804049463242,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7263724690854737e-05,
+      "loss": 0.9352,
+      "step": 77090
+    },
+    {
+      "epoch": 5.238143769533904,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.72616014404131e-05,
+      "loss": 0.8032,
+      "step": 77095
+    },
+    {
+      "epoch": 5.238483489604566,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.7259478189971465e-05,
+      "loss": 0.8761,
+      "step": 77100
+    },
+    {
+      "epoch": 5.238823209675227,
+      "grad_norm": 7.5,
+      "learning_rate": 1.725735493952983e-05,
+      "loss": 0.7075,
+      "step": 77105
+    },
+    {
+      "epoch": 5.2391629297458895,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7255231689088193e-05,
+      "loss": 0.8293,
+      "step": 77110
+    },
+    {
+      "epoch": 5.239502649816552,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7253108438646557e-05,
+      "loss": 0.7965,
+      "step": 77115
+    },
+    {
+      "epoch": 5.239842369887213,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.725098518820492e-05,
+      "loss": 0.822,
+      "step": 77120
+    },
+    {
+      "epoch": 5.240182089957875,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7248861937763285e-05,
+      "loss": 0.7908,
+      "step": 77125
+    },
+    {
+      "epoch": 5.240521810028537,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.724673868732165e-05,
+      "loss": 0.8828,
+      "step": 77130
+    },
+    {
+      "epoch": 5.240861530099198,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.7244615436880013e-05,
+      "loss": 0.889,
+      "step": 77135
+    },
+    {
+      "epoch": 5.24120125016986,
+      "grad_norm": 8.875,
+      "learning_rate": 1.7242492186438373e-05,
+      "loss": 0.8584,
+      "step": 77140
+    },
+    {
+      "epoch": 5.241540970240522,
+      "grad_norm": 8.0,
+      "learning_rate": 1.724036893599674e-05,
+      "loss": 0.9386,
+      "step": 77145
+    },
+    {
+      "epoch": 5.241880690311183,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7238245685555105e-05,
+      "loss": 0.8927,
+      "step": 77150
+    },
+    {
+      "epoch": 5.2422204103818455,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.7236122435113465e-05,
+      "loss": 0.8394,
+      "step": 77155
+    },
+    {
+      "epoch": 5.242560130452507,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7233999184671833e-05,
+      "loss": 0.8629,
+      "step": 77160
+    },
+    {
+      "epoch": 5.242899850523169,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7231875934230197e-05,
+      "loss": 0.836,
+      "step": 77165
+    },
+    {
+      "epoch": 5.243239570593831,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7229752683788557e-05,
+      "loss": 0.9519,
+      "step": 77170
+    },
+    {
+      "epoch": 5.243579290664492,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.722762943334692e-05,
+      "loss": 0.84,
+      "step": 77175
+    },
+    {
+      "epoch": 5.243919010735154,
+      "grad_norm": 8.125,
+      "learning_rate": 1.722550618290529e-05,
+      "loss": 0.8088,
+      "step": 77180
+    },
+    {
+      "epoch": 5.244258730805816,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.722338293246365e-05,
+      "loss": 0.8048,
+      "step": 77185
+    },
+    {
+      "epoch": 5.244598450876477,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7221259682022013e-05,
+      "loss": 0.8949,
+      "step": 77190
+    },
+    {
+      "epoch": 5.244938170947139,
+      "grad_norm": 8.0,
+      "learning_rate": 1.721913643158038e-05,
+      "loss": 0.9586,
+      "step": 77195
+    },
+    {
+      "epoch": 5.2452778910178015,
+      "grad_norm": 7.875,
+      "learning_rate": 1.721701318113874e-05,
+      "loss": 0.8236,
+      "step": 77200
+    },
+    {
+      "epoch": 5.245617611088463,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.7214889930697105e-05,
+      "loss": 0.8318,
+      "step": 77205
+    },
+    {
+      "epoch": 5.245957331159125,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.721276668025547e-05,
+      "loss": 0.7869,
+      "step": 77210
+    },
+    {
+      "epoch": 5.246297051229787,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7210643429813833e-05,
+      "loss": 0.9163,
+      "step": 77215
+    },
+    {
+      "epoch": 5.246636771300448,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.7208520179372197e-05,
+      "loss": 0.9377,
+      "step": 77220
+    },
+    {
+      "epoch": 5.24697649137111,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.720639692893056e-05,
+      "loss": 0.7533,
+      "step": 77225
+    },
+    {
+      "epoch": 5.247316211441772,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7204273678488925e-05,
+      "loss": 0.7947,
+      "step": 77230
+    },
+    {
+      "epoch": 5.247655931512433,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.720215042804729e-05,
+      "loss": 0.8761,
+      "step": 77235
+    },
+    {
+      "epoch": 5.2479956515830954,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7200027177605653e-05,
+      "loss": 0.8164,
+      "step": 77240
+    },
+    {
+      "epoch": 5.2483353716537575,
+      "grad_norm": 7.0,
+      "learning_rate": 1.7197903927164017e-05,
+      "loss": 0.8411,
+      "step": 77245
+    },
+    {
+      "epoch": 5.248675091724419,
+      "grad_norm": 8.875,
+      "learning_rate": 1.719578067672238e-05,
+      "loss": 0.7719,
+      "step": 77250
+    },
+    {
+      "epoch": 5.249014811795081,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.7193657426280745e-05,
+      "loss": 0.792,
+      "step": 77255
+    },
+    {
+      "epoch": 5.249354531865743,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.719153417583911e-05,
+      "loss": 0.8019,
+      "step": 77260
+    },
+    {
+      "epoch": 5.249694251936404,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7189410925397473e-05,
+      "loss": 0.811,
+      "step": 77265
+    },
+    {
+      "epoch": 5.250033972007066,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7187287674955837e-05,
+      "loss": 0.8162,
+      "step": 77270
+    },
+    {
+      "epoch": 5.250373692077728,
+      "grad_norm": 9.0,
+      "learning_rate": 1.71851644245142e-05,
+      "loss": 0.7715,
+      "step": 77275
+    },
+    {
+      "epoch": 5.250713412148389,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.7183041174072565e-05,
+      "loss": 0.7854,
+      "step": 77280
+    },
+    {
+      "epoch": 5.2510531322190515,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.718091792363093e-05,
+      "loss": 0.8888,
+      "step": 77285
+    },
+    {
+      "epoch": 5.2513928522897135,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.7178794673189293e-05,
+      "loss": 0.8215,
+      "step": 77290
+    },
+    {
+      "epoch": 5.251732572360375,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.7176671422747657e-05,
+      "loss": 0.7692,
+      "step": 77295
+    },
+    {
+      "epoch": 5.252072292431037,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.717454817230602e-05,
+      "loss": 0.7949,
+      "step": 77300
+    },
+    {
+      "epoch": 5.252412012501699,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.7172424921864385e-05,
+      "loss": 0.9902,
+      "step": 77305
+    },
+    {
+      "epoch": 5.25275173257236,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7170301671422746e-05,
+      "loss": 0.8706,
+      "step": 77310
+    },
+    {
+      "epoch": 5.253091452643022,
+      "grad_norm": 7.875,
+      "learning_rate": 1.7168178420981113e-05,
+      "loss": 0.8828,
+      "step": 77315
+    },
+    {
+      "epoch": 5.253431172713684,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7166055170539477e-05,
+      "loss": 0.9221,
+      "step": 77320
+    },
+    {
+      "epoch": 5.253770892784345,
+      "grad_norm": 8.625,
+      "learning_rate": 1.7163931920097838e-05,
+      "loss": 0.8664,
+      "step": 77325
+    },
+    {
+      "epoch": 5.2541106128550075,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.7161808669656205e-05,
+      "loss": 0.8289,
+      "step": 77330
+    },
+    {
+      "epoch": 5.2544503329256695,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.715968541921457e-05,
+      "loss": 0.8191,
+      "step": 77335
+    },
+    {
+      "epoch": 5.254790052996331,
+      "grad_norm": 7.25,
+      "learning_rate": 1.715756216877293e-05,
+      "loss": 0.9078,
+      "step": 77340
+    },
+    {
+      "epoch": 5.255129773066993,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7155438918331294e-05,
+      "loss": 0.9268,
+      "step": 77345
+    },
+    {
+      "epoch": 5.255469493137655,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.715331566788966e-05,
+      "loss": 0.8583,
+      "step": 77350
+    },
+    {
+      "epoch": 5.255809213208316,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.7151192417448022e-05,
+      "loss": 0.9112,
+      "step": 77355
+    },
+    {
+      "epoch": 5.256148933278978,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7149069167006386e-05,
+      "loss": 0.7689,
+      "step": 77360
+    },
+    {
+      "epoch": 5.25648865334964,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.7146945916564754e-05,
+      "loss": 0.8925,
+      "step": 77365
+    },
+    {
+      "epoch": 5.256828373420301,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7144822666123114e-05,
+      "loss": 0.8378,
+      "step": 77370
+    },
+    {
+      "epoch": 5.2571680934909635,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7142699415681478e-05,
+      "loss": 0.9697,
+      "step": 77375
+    },
+    {
+      "epoch": 5.2575078135616256,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.7140576165239846e-05,
+      "loss": 0.8142,
+      "step": 77380
+    },
+    {
+      "epoch": 5.257847533632287,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7138452914798206e-05,
+      "loss": 1.037,
+      "step": 77385
+    },
+    {
+      "epoch": 5.258187253702949,
+      "grad_norm": 7.875,
+      "learning_rate": 1.713632966435657e-05,
+      "loss": 0.8596,
+      "step": 77390
+    },
+    {
+      "epoch": 5.258526973773611,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.7134206413914934e-05,
+      "loss": 0.8772,
+      "step": 77395
+    },
+    {
+      "epoch": 5.258866693844272,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7132083163473298e-05,
+      "loss": 0.8169,
+      "step": 77400
+    },
+    {
+      "epoch": 5.259206413914934,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7129959913031662e-05,
+      "loss": 1.0084,
+      "step": 77405
+    },
+    {
+      "epoch": 5.259546133985596,
+      "grad_norm": 7.875,
+      "learning_rate": 1.7127836662590026e-05,
+      "loss": 0.814,
+      "step": 77410
+    },
+    {
+      "epoch": 5.259885854056257,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.712571341214839e-05,
+      "loss": 0.8645,
+      "step": 77415
+    },
+    {
+      "epoch": 5.2602255741269195,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7123590161706754e-05,
+      "loss": 0.8645,
+      "step": 77420
+    },
+    {
+      "epoch": 5.260565294197582,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.7121466911265118e-05,
+      "loss": 0.8365,
+      "step": 77425
+    },
+    {
+      "epoch": 5.260905014268243,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.7119343660823482e-05,
+      "loss": 0.9062,
+      "step": 77430
+    },
+    {
+      "epoch": 5.261244734338905,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7117220410381846e-05,
+      "loss": 0.7933,
+      "step": 77435
+    },
+    {
+      "epoch": 5.261584454409567,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.711509715994021e-05,
+      "loss": 0.8664,
+      "step": 77440
+    },
+    {
+      "epoch": 5.261924174480228,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7112973909498574e-05,
+      "loss": 0.8958,
+      "step": 77445
+    },
+    {
+      "epoch": 5.26226389455089,
+      "grad_norm": 7.875,
+      "learning_rate": 1.7110850659056938e-05,
+      "loss": 0.8814,
+      "step": 77450
+    },
+    {
+      "epoch": 5.262603614621552,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.7108727408615302e-05,
+      "loss": 0.7913,
+      "step": 77455
+    },
+    {
+      "epoch": 5.262943334692213,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7106604158173666e-05,
+      "loss": 0.8574,
+      "step": 77460
+    },
+    {
+      "epoch": 5.2632830547628755,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.710448090773203e-05,
+      "loss": 0.975,
+      "step": 77465
+    },
+    {
+      "epoch": 5.263622774833538,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7102357657290394e-05,
+      "loss": 0.8765,
+      "step": 77470
+    },
+    {
+      "epoch": 5.263962494904199,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7100234406848758e-05,
+      "loss": 0.9921,
+      "step": 77475
+    },
+    {
+      "epoch": 5.264302214974861,
+      "grad_norm": 8.875,
+      "learning_rate": 1.7098111156407122e-05,
+      "loss": 0.7497,
+      "step": 77480
+    },
+    {
+      "epoch": 5.264641935045523,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.7095987905965486e-05,
+      "loss": 0.8815,
+      "step": 77485
+    },
+    {
+      "epoch": 5.264981655116184,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.709386465552385e-05,
+      "loss": 0.9495,
+      "step": 77490
+    },
+    {
+      "epoch": 5.265321375186846,
+      "grad_norm": 9.0,
+      "learning_rate": 1.709174140508221e-05,
+      "loss": 0.8285,
+      "step": 77495
+    },
+    {
+      "epoch": 5.265661095257508,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7089618154640578e-05,
+      "loss": 0.8655,
+      "step": 77500
+    },
+    {
+      "epoch": 5.266000815328169,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7087494904198942e-05,
+      "loss": 0.834,
+      "step": 77505
+    },
+    {
+      "epoch": 5.2663405353988315,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7085371653757303e-05,
+      "loss": 0.8465,
+      "step": 77510
+    },
+    {
+      "epoch": 5.266680255469494,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.708324840331567e-05,
+      "loss": 0.8944,
+      "step": 77515
+    },
+    {
+      "epoch": 5.267019975540155,
+      "grad_norm": 6.5,
+      "learning_rate": 1.7081125152874034e-05,
+      "loss": 0.789,
+      "step": 77520
+    },
+    {
+      "epoch": 5.267359695610817,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7079001902432395e-05,
+      "loss": 0.905,
+      "step": 77525
+    },
+    {
+      "epoch": 5.267699415681479,
+      "grad_norm": 10.0,
+      "learning_rate": 1.707687865199076e-05,
+      "loss": 0.7896,
+      "step": 77530
+    },
+    {
+      "epoch": 5.26803913575214,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.7074755401549126e-05,
+      "loss": 0.8335,
+      "step": 77535
+    },
+    {
+      "epoch": 5.268378855822802,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7072632151107487e-05,
+      "loss": 0.8277,
+      "step": 77540
+    },
+    {
+      "epoch": 5.268718575893463,
+      "grad_norm": 8.625,
+      "learning_rate": 1.707050890066585e-05,
+      "loss": 0.8802,
+      "step": 77545
+    },
+    {
+      "epoch": 5.2690582959641254,
+      "grad_norm": 7.0,
+      "learning_rate": 1.7068385650224218e-05,
+      "loss": 0.8753,
+      "step": 77550
+    },
+    {
+      "epoch": 5.2693980160347875,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.706626239978258e-05,
+      "loss": 0.838,
+      "step": 77555
+    },
+    {
+      "epoch": 5.269737736105449,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7064139149340943e-05,
+      "loss": 0.8682,
+      "step": 77560
+    },
+    {
+      "epoch": 5.270077456176111,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7062015898899307e-05,
+      "loss": 0.8944,
+      "step": 77565
+    },
+    {
+      "epoch": 5.270417176246773,
+      "grad_norm": 8.25,
+      "learning_rate": 1.705989264845767e-05,
+      "loss": 0.911,
+      "step": 77570
+    },
+    {
+      "epoch": 5.270756896317434,
+      "grad_norm": 7.75,
+      "learning_rate": 1.7057769398016035e-05,
+      "loss": 0.7775,
+      "step": 77575
+    },
+    {
+      "epoch": 5.271096616388096,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.70556461475744e-05,
+      "loss": 0.842,
+      "step": 77580
+    },
+    {
+      "epoch": 5.271436336458758,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7053522897132763e-05,
+      "loss": 0.7859,
+      "step": 77585
+    },
+    {
+      "epoch": 5.271776056529419,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7051399646691127e-05,
+      "loss": 0.7949,
+      "step": 77590
+    },
+    {
+      "epoch": 5.2721157766000815,
+      "grad_norm": 7.25,
+      "learning_rate": 1.704927639624949e-05,
+      "loss": 0.8545,
+      "step": 77595
+    },
+    {
+      "epoch": 5.2724554966707435,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.7047153145807855e-05,
+      "loss": 0.8522,
+      "step": 77600
+    },
+    {
+      "epoch": 5.272795216741405,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.704502989536622e-05,
+      "loss": 0.8954,
+      "step": 77605
+    },
+    {
+      "epoch": 5.273134936812067,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7042906644924583e-05,
+      "loss": 0.9294,
+      "step": 77610
+    },
+    {
+      "epoch": 5.273474656882729,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7040783394482947e-05,
+      "loss": 0.8093,
+      "step": 77615
+    },
+    {
+      "epoch": 5.27381437695339,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.703866014404131e-05,
+      "loss": 0.8249,
+      "step": 77620
+    },
+    {
+      "epoch": 5.274154097024052,
+      "grad_norm": 8.125,
+      "learning_rate": 1.7036536893599675e-05,
+      "loss": 0.7926,
+      "step": 77625
+    },
+    {
+      "epoch": 5.274493817094714,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.703441364315804e-05,
+      "loss": 0.9194,
+      "step": 77630
+    },
+    {
+      "epoch": 5.274833537165375,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.7032290392716403e-05,
+      "loss": 0.9019,
+      "step": 77635
+    },
+    {
+      "epoch": 5.2751732572360375,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7030167142274767e-05,
+      "loss": 0.8219,
+      "step": 77640
+    },
+    {
+      "epoch": 5.2755129773066995,
+      "grad_norm": 8.0,
+      "learning_rate": 1.702804389183313e-05,
+      "loss": 0.8204,
+      "step": 77645
+    },
+    {
+      "epoch": 5.275852697377361,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.7025920641391495e-05,
+      "loss": 0.7486,
+      "step": 77650
+    },
+    {
+      "epoch": 5.276192417448023,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.702379739094986e-05,
+      "loss": 0.861,
+      "step": 77655
+    },
+    {
+      "epoch": 5.276532137518685,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.7021674140508223e-05,
+      "loss": 0.8961,
+      "step": 77660
+    },
+    {
+      "epoch": 5.276871857589346,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7019550890066584e-05,
+      "loss": 0.8864,
+      "step": 77665
+    },
+    {
+      "epoch": 5.277211577660008,
+      "grad_norm": 8.5,
+      "learning_rate": 1.701742763962495e-05,
+      "loss": 0.8626,
+      "step": 77670
+    },
+    {
+      "epoch": 5.27755129773067,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7015304389183315e-05,
+      "loss": 0.8186,
+      "step": 77675
+    },
+    {
+      "epoch": 5.277891017801331,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7013181138741676e-05,
+      "loss": 0.7862,
+      "step": 77680
+    },
+    {
+      "epoch": 5.2782307378719935,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.7011057888300043e-05,
+      "loss": 0.8599,
+      "step": 77685
+    },
+    {
+      "epoch": 5.278570457942656,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7008934637858407e-05,
+      "loss": 0.8497,
+      "step": 77690
+    },
+    {
+      "epoch": 5.278910178013317,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.7006811387416768e-05,
+      "loss": 0.9426,
+      "step": 77695
+    },
+    {
+      "epoch": 5.279249898083979,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.700468813697513e-05,
+      "loss": 0.972,
+      "step": 77700
+    },
+    {
+      "epoch": 5.279589618154641,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.70025648865335e-05,
+      "loss": 0.9539,
+      "step": 77705
+    },
+    {
+      "epoch": 5.279929338225302,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.700044163609186e-05,
+      "loss": 0.8349,
+      "step": 77710
+    },
+    {
+      "epoch": 5.280269058295964,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6998318385650224e-05,
+      "loss": 0.7968,
+      "step": 77715
+    },
+    {
+      "epoch": 5.280608778366626,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.699619513520859e-05,
+      "loss": 0.7589,
+      "step": 77720
+    },
+    {
+      "epoch": 5.280948498437287,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.699407188476695e-05,
+      "loss": 0.8531,
+      "step": 77725
+    },
+    {
+      "epoch": 5.2812882185079495,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.6991948634325316e-05,
+      "loss": 0.806,
+      "step": 77730
+    },
+    {
+      "epoch": 5.281627938578612,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.698982538388368e-05,
+      "loss": 0.8873,
+      "step": 77735
+    },
+    {
+      "epoch": 5.281967658649273,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6987702133442044e-05,
+      "loss": 0.8554,
+      "step": 77740
+    },
+    {
+      "epoch": 5.282307378719935,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6985578883000408e-05,
+      "loss": 0.843,
+      "step": 77745
+    },
+    {
+      "epoch": 5.282647098790597,
+      "grad_norm": 7.75,
+      "learning_rate": 1.698345563255877e-05,
+      "loss": 0.8197,
+      "step": 77750
+    },
+    {
+      "epoch": 5.282986818861258,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6981332382117136e-05,
+      "loss": 0.8877,
+      "step": 77755
+    },
+    {
+      "epoch": 5.28332653893192,
+      "grad_norm": 8.25,
+      "learning_rate": 1.69792091316755e-05,
+      "loss": 0.8512,
+      "step": 77760
+    },
+    {
+      "epoch": 5.283666259002582,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6977085881233864e-05,
+      "loss": 0.795,
+      "step": 77765
+    },
+    {
+      "epoch": 5.284005979073243,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.697496263079223e-05,
+      "loss": 0.8776,
+      "step": 77770
+    },
+    {
+      "epoch": 5.2843456991439055,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.697283938035059e-05,
+      "loss": 0.924,
+      "step": 77775
+    },
+    {
+      "epoch": 5.284685419214568,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6970716129908956e-05,
+      "loss": 0.8649,
+      "step": 77780
+    },
+    {
+      "epoch": 5.285025139285229,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.696859287946732e-05,
+      "loss": 0.8858,
+      "step": 77785
+    },
+    {
+      "epoch": 5.285364859355891,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6966469629025684e-05,
+      "loss": 0.9778,
+      "step": 77790
+    },
+    {
+      "epoch": 5.285704579426553,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6964346378584048e-05,
+      "loss": 0.8452,
+      "step": 77795
+    },
+    {
+      "epoch": 5.286044299497214,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.696222312814241e-05,
+      "loss": 0.8107,
+      "step": 77800
+    },
+    {
+      "epoch": 5.286384019567876,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6960099877700776e-05,
+      "loss": 0.8282,
+      "step": 77805
+    },
+    {
+      "epoch": 5.286723739638538,
+      "grad_norm": 7.875,
+      "learning_rate": 1.695797662725914e-05,
+      "loss": 0.912,
+      "step": 77810
+    },
+    {
+      "epoch": 5.287063459709199,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6955853376817504e-05,
+      "loss": 0.7589,
+      "step": 77815
+    },
+    {
+      "epoch": 5.2874031797798615,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6953730126375868e-05,
+      "loss": 0.8054,
+      "step": 77820
+    },
+    {
+      "epoch": 5.287742899850523,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6951606875934232e-05,
+      "loss": 0.8468,
+      "step": 77825
+    },
+    {
+      "epoch": 5.288082619921185,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6949483625492596e-05,
+      "loss": 0.8729,
+      "step": 77830
+    },
+    {
+      "epoch": 5.288422339991847,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6947360375050956e-05,
+      "loss": 0.8511,
+      "step": 77835
+    },
+    {
+      "epoch": 5.288762060062508,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.6945237124609324e-05,
+      "loss": 0.9216,
+      "step": 77840
+    },
+    {
+      "epoch": 5.28910178013317,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6943113874167688e-05,
+      "loss": 0.8322,
+      "step": 77845
+    },
+    {
+      "epoch": 5.289441500203832,
+      "grad_norm": 7.5,
+      "learning_rate": 1.694099062372605e-05,
+      "loss": 0.8593,
+      "step": 77850
+    },
+    {
+      "epoch": 5.289781220274493,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.6938867373284416e-05,
+      "loss": 0.8365,
+      "step": 77855
+    },
+    {
+      "epoch": 5.2901209403451555,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.693674412284278e-05,
+      "loss": 0.758,
+      "step": 77860
+    },
+    {
+      "epoch": 5.2904606604158175,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.693462087240114e-05,
+      "loss": 0.8451,
+      "step": 77865
+    },
+    {
+      "epoch": 5.290800380486479,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.6932497621959508e-05,
+      "loss": 0.8677,
+      "step": 77870
+    },
+    {
+      "epoch": 5.291140100557141,
+      "grad_norm": 9.5,
+      "learning_rate": 1.6930374371517872e-05,
+      "loss": 0.932,
+      "step": 77875
+    },
+    {
+      "epoch": 5.291479820627803,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.6928251121076232e-05,
+      "loss": 0.7961,
+      "step": 77880
+    },
+    {
+      "epoch": 5.291819540698464,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6926127870634596e-05,
+      "loss": 0.8707,
+      "step": 77885
+    },
+    {
+      "epoch": 5.292159260769126,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6924004620192964e-05,
+      "loss": 0.932,
+      "step": 77890
+    },
+    {
+      "epoch": 5.292498980839788,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6921881369751324e-05,
+      "loss": 0.8639,
+      "step": 77895
+    },
+    {
+      "epoch": 5.292838700910449,
+      "grad_norm": 8.5,
+      "learning_rate": 1.691975811930969e-05,
+      "loss": 0.8486,
+      "step": 77900
+    },
+    {
+      "epoch": 5.2931784209811115,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6917634868868056e-05,
+      "loss": 0.8874,
+      "step": 77905
+    },
+    {
+      "epoch": 5.2935181410517735,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.6915511618426416e-05,
+      "loss": 0.8225,
+      "step": 77910
+    },
+    {
+      "epoch": 5.293857861122435,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.691338836798478e-05,
+      "loss": 0.8396,
+      "step": 77915
+    },
+    {
+      "epoch": 5.294197581193097,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.6911265117543144e-05,
+      "loss": 0.7548,
+      "step": 77920
+    },
+    {
+      "epoch": 5.294537301263759,
+      "grad_norm": 8.0,
+      "learning_rate": 1.690914186710151e-05,
+      "loss": 0.883,
+      "step": 77925
+    },
+    {
+      "epoch": 5.29487702133442,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6907018616659872e-05,
+      "loss": 0.7774,
+      "step": 77930
+    },
+    {
+      "epoch": 5.295216741405082,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6904895366218236e-05,
+      "loss": 0.9357,
+      "step": 77935
+    },
+    {
+      "epoch": 5.295556461475744,
+      "grad_norm": 7.75,
+      "learning_rate": 1.6902772115776604e-05,
+      "loss": 0.8295,
+      "step": 77940
+    },
+    {
+      "epoch": 5.295896181546405,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.6900648865334964e-05,
+      "loss": 0.8209,
+      "step": 77945
+    },
+    {
+      "epoch": 5.2962359016170675,
+      "grad_norm": 7.375,
+      "learning_rate": 1.689852561489333e-05,
+      "loss": 0.9109,
+      "step": 77950
+    },
+    {
+      "epoch": 5.2965756216877296,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6896402364451692e-05,
+      "loss": 0.844,
+      "step": 77955
+    },
+    {
+      "epoch": 5.296915341758391,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6894279114010056e-05,
+      "loss": 0.8759,
+      "step": 77960
+    },
+    {
+      "epoch": 5.297255061829053,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.689215586356842e-05,
+      "loss": 0.8125,
+      "step": 77965
+    },
+    {
+      "epoch": 5.297594781899715,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6890032613126784e-05,
+      "loss": 0.8885,
+      "step": 77970
+    },
+    {
+      "epoch": 5.297934501970376,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.688790936268515e-05,
+      "loss": 0.8348,
+      "step": 77975
+    },
+    {
+      "epoch": 5.298274222041038,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6885786112243512e-05,
+      "loss": 0.9455,
+      "step": 77980
+    },
+    {
+      "epoch": 5.2986139421117,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.6883662861801876e-05,
+      "loss": 0.8153,
+      "step": 77985
+    },
+    {
+      "epoch": 5.298953662182361,
+      "grad_norm": 8.25,
+      "learning_rate": 1.688153961136024e-05,
+      "loss": 0.9049,
+      "step": 77990
+    },
+    {
+      "epoch": 5.2992933822530235,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6879416360918604e-05,
+      "loss": 0.8218,
+      "step": 77995
+    },
+    {
+      "epoch": 5.299633102323686,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.687729311047697e-05,
+      "loss": 0.7597,
+      "step": 78000
+    },
+    {
+      "epoch": 5.299972822394347,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6875169860035332e-05,
+      "loss": 0.7902,
+      "step": 78005
+    },
+    {
+      "epoch": 5.300312542465009,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.6873046609593696e-05,
+      "loss": 0.8088,
+      "step": 78010
+    },
+    {
+      "epoch": 5.300652262535671,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.687092335915206e-05,
+      "loss": 0.8902,
+      "step": 78015
+    },
+    {
+      "epoch": 5.300991982606332,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.686880010871042e-05,
+      "loss": 0.7911,
+      "step": 78020
+    },
+    {
+      "epoch": 5.301331702676994,
+      "grad_norm": 8.375,
+      "learning_rate": 1.686667685826879e-05,
+      "loss": 0.791,
+      "step": 78025
+    },
+    {
+      "epoch": 5.301671422747656,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6864553607827152e-05,
+      "loss": 0.8412,
+      "step": 78030
+    },
+    {
+      "epoch": 5.302011142818317,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.6862430357385513e-05,
+      "loss": 0.9113,
+      "step": 78035
+    },
+    {
+      "epoch": 5.3023508628889795,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.686030710694388e-05,
+      "loss": 0.8365,
+      "step": 78040
+    },
+    {
+      "epoch": 5.302690582959642,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6858183856502244e-05,
+      "loss": 0.9334,
+      "step": 78045
+    },
+    {
+      "epoch": 5.303030303030303,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6856060606060605e-05,
+      "loss": 0.8421,
+      "step": 78050
+    },
+    {
+      "epoch": 5.303370023100965,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.685393735561897e-05,
+      "loss": 0.7986,
+      "step": 78055
+    },
+    {
+      "epoch": 5.303709743171627,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6851814105177336e-05,
+      "loss": 0.8738,
+      "step": 78060
+    },
+    {
+      "epoch": 5.304049463242288,
+      "grad_norm": 7.5,
+      "learning_rate": 1.6849690854735697e-05,
+      "loss": 0.8023,
+      "step": 78065
+    },
+    {
+      "epoch": 5.30438918331295,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.684756760429406e-05,
+      "loss": 0.836,
+      "step": 78070
+    },
+    {
+      "epoch": 5.304728903383612,
+      "grad_norm": 8.25,
+      "learning_rate": 1.684544435385243e-05,
+      "loss": 0.8757,
+      "step": 78075
+    },
+    {
+      "epoch": 5.305068623454273,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.684332110341079e-05,
+      "loss": 0.9352,
+      "step": 78080
+    },
+    {
+      "epoch": 5.3054083435249355,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6841197852969153e-05,
+      "loss": 0.8545,
+      "step": 78085
+    },
+    {
+      "epoch": 5.305748063595598,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6839074602527517e-05,
+      "loss": 0.804,
+      "step": 78090
+    },
+    {
+      "epoch": 5.306087783666259,
+      "grad_norm": 7.125,
+      "learning_rate": 1.683695135208588e-05,
+      "loss": 0.7769,
+      "step": 78095
+    },
+    {
+      "epoch": 5.306427503736921,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6834828101644245e-05,
+      "loss": 0.8042,
+      "step": 78100
+    },
+    {
+      "epoch": 5.306767223807583,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.683270485120261e-05,
+      "loss": 0.8802,
+      "step": 78105
+    },
+    {
+      "epoch": 5.307106943878244,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6830581600760977e-05,
+      "loss": 0.8606,
+      "step": 78110
+    },
+    {
+      "epoch": 5.307446663948906,
+      "grad_norm": 8.75,
+      "learning_rate": 1.6828458350319337e-05,
+      "loss": 0.8777,
+      "step": 78115
+    },
+    {
+      "epoch": 5.307786384019568,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.68263350998777e-05,
+      "loss": 0.8682,
+      "step": 78120
+    },
+    {
+      "epoch": 5.3081261040902294,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.6824211849436065e-05,
+      "loss": 0.8362,
+      "step": 78125
+    },
+    {
+      "epoch": 5.3084658241608915,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.682208859899443e-05,
+      "loss": 0.861,
+      "step": 78130
+    },
+    {
+      "epoch": 5.308805544231554,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6819965348552793e-05,
+      "loss": 0.7443,
+      "step": 78135
+    },
+    {
+      "epoch": 5.309145264302215,
+      "grad_norm": 7.75,
+      "learning_rate": 1.6817842098111157e-05,
+      "loss": 0.8029,
+      "step": 78140
+    },
+    {
+      "epoch": 5.309484984372877,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.681571884766952e-05,
+      "loss": 0.8417,
+      "step": 78145
+    },
+    {
+      "epoch": 5.309824704443539,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6813595597227885e-05,
+      "loss": 0.8722,
+      "step": 78150
+    },
+    {
+      "epoch": 5.3101644245142,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.681147234678625e-05,
+      "loss": 0.8415,
+      "step": 78155
+    },
+    {
+      "epoch": 5.310504144584862,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6809349096344613e-05,
+      "loss": 0.925,
+      "step": 78160
+    },
+    {
+      "epoch": 5.310843864655524,
+      "grad_norm": 7.0,
+      "learning_rate": 1.6807225845902977e-05,
+      "loss": 0.9565,
+      "step": 78165
+    },
+    {
+      "epoch": 5.3111835847261855,
+      "grad_norm": 8.125,
+      "learning_rate": 1.680510259546134e-05,
+      "loss": 0.8303,
+      "step": 78170
+    },
+    {
+      "epoch": 5.3115233047968475,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6802979345019705e-05,
+      "loss": 0.8524,
+      "step": 78175
+    },
+    {
+      "epoch": 5.31186302486751,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.680085609457807e-05,
+      "loss": 0.9101,
+      "step": 78180
+    },
+    {
+      "epoch": 5.312202744938171,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6798732844136433e-05,
+      "loss": 0.9303,
+      "step": 78185
+    },
+    {
+      "epoch": 5.312542465008833,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6796609593694794e-05,
+      "loss": 0.7943,
+      "step": 78190
+    },
+    {
+      "epoch": 5.312882185079495,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.679448634325316e-05,
+      "loss": 0.7895,
+      "step": 78195
+    },
+    {
+      "epoch": 5.313221905150156,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6792363092811525e-05,
+      "loss": 0.8175,
+      "step": 78200
+    },
+    {
+      "epoch": 5.313561625220818,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6790239842369886e-05,
+      "loss": 0.7558,
+      "step": 78205
+    },
+    {
+      "epoch": 5.31390134529148,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6788116591928253e-05,
+      "loss": 0.9426,
+      "step": 78210
+    },
+    {
+      "epoch": 5.3142410653621415,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.6785993341486617e-05,
+      "loss": 0.7783,
+      "step": 78215
+    },
+    {
+      "epoch": 5.3145807854328035,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.6783870091044978e-05,
+      "loss": 0.8302,
+      "step": 78220
+    },
+    {
+      "epoch": 5.314920505503465,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6781746840603342e-05,
+      "loss": 0.8491,
+      "step": 78225
+    },
+    {
+      "epoch": 5.315260225574127,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.677962359016171e-05,
+      "loss": 0.8303,
+      "step": 78230
+    },
+    {
+      "epoch": 5.315599945644789,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.677750033972007e-05,
+      "loss": 0.8,
+      "step": 78235
+    },
+    {
+      "epoch": 5.31593966571545,
+      "grad_norm": 7.0,
+      "learning_rate": 1.6775377089278434e-05,
+      "loss": 0.7903,
+      "step": 78240
+    },
+    {
+      "epoch": 5.316279385786112,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.67732538388368e-05,
+      "loss": 0.8779,
+      "step": 78245
+    },
+    {
+      "epoch": 5.316619105856774,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6771130588395162e-05,
+      "loss": 0.7426,
+      "step": 78250
+    },
+    {
+      "epoch": 5.316958825927435,
+      "grad_norm": 7.625,
+      "learning_rate": 1.6769007337953526e-05,
+      "loss": 0.8278,
+      "step": 78255
+    },
+    {
+      "epoch": 5.3172985459980975,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.676688408751189e-05,
+      "loss": 0.9352,
+      "step": 78260
+    },
+    {
+      "epoch": 5.31763826606876,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6764760837070254e-05,
+      "loss": 0.8864,
+      "step": 78265
+    },
+    {
+      "epoch": 5.317977986139421,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6762637586628618e-05,
+      "loss": 0.8603,
+      "step": 78270
+    },
+    {
+      "epoch": 5.318317706210083,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.6760514336186982e-05,
+      "loss": 0.8005,
+      "step": 78275
+    },
+    {
+      "epoch": 5.318657426280745,
+      "grad_norm": 8.25,
+      "learning_rate": 1.675839108574535e-05,
+      "loss": 0.8271,
+      "step": 78280
+    },
+    {
+      "epoch": 5.318997146351406,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.675626783530371e-05,
+      "loss": 0.8714,
+      "step": 78285
+    },
+    {
+      "epoch": 5.319336866422068,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6754144584862074e-05,
+      "loss": 0.8597,
+      "step": 78290
+    },
+    {
+      "epoch": 5.31967658649273,
+      "grad_norm": 8.0,
+      "learning_rate": 1.675202133442044e-05,
+      "loss": 0.7962,
+      "step": 78295
+    },
+    {
+      "epoch": 5.320016306563391,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6749898083978802e-05,
+      "loss": 0.8238,
+      "step": 78300
+    },
+    {
+      "epoch": 5.3203560266340535,
+      "grad_norm": 7.75,
+      "learning_rate": 1.6747774833537166e-05,
+      "loss": 0.815,
+      "step": 78305
+    },
+    {
+      "epoch": 5.320695746704716,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.674565158309553e-05,
+      "loss": 0.7933,
+      "step": 78310
+    },
+    {
+      "epoch": 5.321035466775377,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.6743528332653894e-05,
+      "loss": 0.8306,
+      "step": 78315
+    },
+    {
+      "epoch": 5.321375186846039,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6741405082212258e-05,
+      "loss": 0.8121,
+      "step": 78320
+    },
+    {
+      "epoch": 5.321714906916701,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6739281831770622e-05,
+      "loss": 0.8535,
+      "step": 78325
+    },
+    {
+      "epoch": 5.322054626987362,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6737158581328986e-05,
+      "loss": 0.8216,
+      "step": 78330
+    },
+    {
+      "epoch": 5.322394347058024,
+      "grad_norm": 8.5,
+      "learning_rate": 1.673503533088735e-05,
+      "loss": 0.9089,
+      "step": 78335
+    },
+    {
+      "epoch": 5.322734067128686,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.6732912080445714e-05,
+      "loss": 0.9651,
+      "step": 78340
+    },
+    {
+      "epoch": 5.323073787199347,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6730788830004078e-05,
+      "loss": 0.8726,
+      "step": 78345
+    },
+    {
+      "epoch": 5.3234135072700095,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6728665579562442e-05,
+      "loss": 0.8604,
+      "step": 78350
+    },
+    {
+      "epoch": 5.323753227340672,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6726542329120806e-05,
+      "loss": 0.8898,
+      "step": 78355
+    },
+    {
+      "epoch": 5.324092947411333,
+      "grad_norm": 7.625,
+      "learning_rate": 1.6724419078679167e-05,
+      "loss": 0.7937,
+      "step": 78360
+    },
+    {
+      "epoch": 5.324432667481995,
+      "grad_norm": 7.125,
+      "learning_rate": 1.6722295828237534e-05,
+      "loss": 0.8856,
+      "step": 78365
+    },
+    {
+      "epoch": 5.324772387552657,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6720172577795898e-05,
+      "loss": 0.8032,
+      "step": 78370
+    },
+    {
+      "epoch": 5.325112107623318,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.671804932735426e-05,
+      "loss": 0.8045,
+      "step": 78375
+    },
+    {
+      "epoch": 5.32545182769398,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6715926076912626e-05,
+      "loss": 0.8098,
+      "step": 78380
+    },
+    {
+      "epoch": 5.325791547764642,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.671380282647099e-05,
+      "loss": 0.8694,
+      "step": 78385
+    },
+    {
+      "epoch": 5.326131267835303,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.671167957602935e-05,
+      "loss": 0.8165,
+      "step": 78390
+    },
+    {
+      "epoch": 5.3264709879059655,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.6709556325587718e-05,
+      "loss": 0.8836,
+      "step": 78395
+    },
+    {
+      "epoch": 5.326810707976628,
+      "grad_norm": 9.5,
+      "learning_rate": 1.6707433075146082e-05,
+      "loss": 0.8526,
+      "step": 78400
+    },
+    {
+      "epoch": 5.327150428047289,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6705309824704443e-05,
+      "loss": 0.8762,
+      "step": 78405
+    },
+    {
+      "epoch": 5.327490148117951,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6703186574262807e-05,
+      "loss": 0.8532,
+      "step": 78410
+    },
+    {
+      "epoch": 5.327829868188613,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.6701063323821174e-05,
+      "loss": 0.7873,
+      "step": 78415
+    },
+    {
+      "epoch": 5.328169588259274,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.6698940073379535e-05,
+      "loss": 0.7123,
+      "step": 78420
+    },
+    {
+      "epoch": 5.328509308329936,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.66968168229379e-05,
+      "loss": 0.8688,
+      "step": 78425
+    },
+    {
+      "epoch": 5.328849028400598,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6694693572496266e-05,
+      "loss": 0.9098,
+      "step": 78430
+    },
+    {
+      "epoch": 5.3291887484712595,
+      "grad_norm": 8.875,
+      "learning_rate": 1.6692570322054627e-05,
+      "loss": 0.8159,
+      "step": 78435
+    },
+    {
+      "epoch": 5.3295284685419215,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.669044707161299e-05,
+      "loss": 0.8371,
+      "step": 78440
+    },
+    {
+      "epoch": 5.329868188612584,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6688323821171355e-05,
+      "loss": 0.8447,
+      "step": 78445
+    },
+    {
+      "epoch": 5.330207908683245,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6686200570729722e-05,
+      "loss": 0.8324,
+      "step": 78450
+    },
+    {
+      "epoch": 5.330547628753907,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.6684077320288083e-05,
+      "loss": 0.9074,
+      "step": 78455
+    },
+    {
+      "epoch": 5.330887348824569,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6681954069846447e-05,
+      "loss": 0.917,
+      "step": 78460
+    },
+    {
+      "epoch": 5.33122706889523,
+      "grad_norm": 8.875,
+      "learning_rate": 1.6679830819404814e-05,
+      "loss": 0.8968,
+      "step": 78465
+    },
+    {
+      "epoch": 5.331566788965892,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6677707568963175e-05,
+      "loss": 0.8065,
+      "step": 78470
+    },
+    {
+      "epoch": 5.331906509036554,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.667558431852154e-05,
+      "loss": 0.8158,
+      "step": 78475
+    },
+    {
+      "epoch": 5.3322462291072155,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6673461068079903e-05,
+      "loss": 0.8865,
+      "step": 78480
+    },
+    {
+      "epoch": 5.3325859491778775,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6671337817638267e-05,
+      "loss": 0.8048,
+      "step": 78485
+    },
+    {
+      "epoch": 5.33292566924854,
+      "grad_norm": 7.25,
+      "learning_rate": 1.666921456719663e-05,
+      "loss": 0.8897,
+      "step": 78490
+    },
+    {
+      "epoch": 5.333265389319201,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.6667091316754995e-05,
+      "loss": 0.8332,
+      "step": 78495
+    },
+    {
+      "epoch": 5.333605109389863,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.666496806631336e-05,
+      "loss": 0.8319,
+      "step": 78500
+    },
+    {
+      "epoch": 5.333944829460524,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6662844815871723e-05,
+      "loss": 0.9285,
+      "step": 78505
+    },
+    {
+      "epoch": 5.334284549531186,
+      "grad_norm": 7.625,
+      "learning_rate": 1.6660721565430087e-05,
+      "loss": 0.9097,
+      "step": 78510
+    },
+    {
+      "epoch": 5.334624269601848,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.665859831498845e-05,
+      "loss": 0.8585,
+      "step": 78515
+    },
+    {
+      "epoch": 5.334963989672509,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6656475064546815e-05,
+      "loss": 0.88,
+      "step": 78520
+    },
+    {
+      "epoch": 5.3353037097431715,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.665435181410518e-05,
+      "loss": 0.907,
+      "step": 78525
+    },
+    {
+      "epoch": 5.3356434298138335,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6652228563663543e-05,
+      "loss": 0.909,
+      "step": 78530
+    },
+    {
+      "epoch": 5.335983149884495,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6650105313221907e-05,
+      "loss": 0.8803,
+      "step": 78535
+    },
+    {
+      "epoch": 5.336322869955157,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.664798206278027e-05,
+      "loss": 0.8829,
+      "step": 78540
+    },
+    {
+      "epoch": 5.336662590025819,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.664585881233863e-05,
+      "loss": 0.8029,
+      "step": 78545
+    },
+    {
+      "epoch": 5.33700231009648,
+      "grad_norm": 8.875,
+      "learning_rate": 1.6643735561897e-05,
+      "loss": 0.7927,
+      "step": 78550
+    },
+    {
+      "epoch": 5.337342030167142,
+      "grad_norm": 7.375,
+      "learning_rate": 1.6641612311455363e-05,
+      "loss": 0.8034,
+      "step": 78555
+    },
+    {
+      "epoch": 5.337681750237804,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6639489061013723e-05,
+      "loss": 0.8772,
+      "step": 78560
+    },
+    {
+      "epoch": 5.338021470308465,
+      "grad_norm": 7.75,
+      "learning_rate": 1.663736581057209e-05,
+      "loss": 0.8198,
+      "step": 78565
+    },
+    {
+      "epoch": 5.3383611903791275,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6635242560130455e-05,
+      "loss": 0.8648,
+      "step": 78570
+    },
+    {
+      "epoch": 5.33870091044979,
+      "grad_norm": 9.375,
+      "learning_rate": 1.6633119309688815e-05,
+      "loss": 0.8249,
+      "step": 78575
+    },
+    {
+      "epoch": 5.339040630520451,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.663099605924718e-05,
+      "loss": 0.8919,
+      "step": 78580
+    },
+    {
+      "epoch": 5.339380350591113,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6628872808805547e-05,
+      "loss": 0.8108,
+      "step": 78585
+    },
+    {
+      "epoch": 5.339720070661775,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6626749558363907e-05,
+      "loss": 0.8269,
+      "step": 78590
+    },
+    {
+      "epoch": 5.340059790732436,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.662462630792227e-05,
+      "loss": 0.8367,
+      "step": 78595
+    },
+    {
+      "epoch": 5.340399510803098,
+      "grad_norm": 7.375,
+      "learning_rate": 1.662250305748064e-05,
+      "loss": 0.8555,
+      "step": 78600
+    },
+    {
+      "epoch": 5.34073923087376,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6620379807039e-05,
+      "loss": 0.9179,
+      "step": 78605
+    },
+    {
+      "epoch": 5.341078950944421,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6618256556597363e-05,
+      "loss": 0.8264,
+      "step": 78610
+    },
+    {
+      "epoch": 5.3414186710150835,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6616133306155727e-05,
+      "loss": 0.8279,
+      "step": 78615
+    },
+    {
+      "epoch": 5.341758391085746,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6614010055714095e-05,
+      "loss": 0.8703,
+      "step": 78620
+    },
+    {
+      "epoch": 5.342098111156407,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6611886805272455e-05,
+      "loss": 0.7804,
+      "step": 78625
+    },
+    {
+      "epoch": 5.342437831227069,
+      "grad_norm": 7.875,
+      "learning_rate": 1.660976355483082e-05,
+      "loss": 0.8014,
+      "step": 78630
+    },
+    {
+      "epoch": 5.342777551297731,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6607640304389187e-05,
+      "loss": 0.7945,
+      "step": 78635
+    },
+    {
+      "epoch": 5.343117271368392,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6605517053947547e-05,
+      "loss": 0.7872,
+      "step": 78640
+    },
+    {
+      "epoch": 5.343456991439054,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.660339380350591e-05,
+      "loss": 0.945,
+      "step": 78645
+    },
+    {
+      "epoch": 5.343796711509716,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.6601270553064275e-05,
+      "loss": 0.764,
+      "step": 78650
+    },
+    {
+      "epoch": 5.344136431580377,
+      "grad_norm": 8.625,
+      "learning_rate": 1.659914730262264e-05,
+      "loss": 0.9034,
+      "step": 78655
+    },
+    {
+      "epoch": 5.3444761516510395,
+      "grad_norm": 8.75,
+      "learning_rate": 1.6597024052181003e-05,
+      "loss": 0.9046,
+      "step": 78660
+    },
+    {
+      "epoch": 5.344815871721702,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6594900801739367e-05,
+      "loss": 0.8914,
+      "step": 78665
+    },
+    {
+      "epoch": 5.345155591792363,
+      "grad_norm": 8.125,
+      "learning_rate": 1.659277755129773e-05,
+      "loss": 0.8772,
+      "step": 78670
+    },
+    {
+      "epoch": 5.345495311863025,
+      "grad_norm": 7.375,
+      "learning_rate": 1.6590654300856095e-05,
+      "loss": 0.8118,
+      "step": 78675
+    },
+    {
+      "epoch": 5.345835031933687,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.658853105041446e-05,
+      "loss": 0.7757,
+      "step": 78680
+    },
+    {
+      "epoch": 5.346174752004348,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.6586407799972823e-05,
+      "loss": 0.8447,
+      "step": 78685
+    },
+    {
+      "epoch": 5.34651447207501,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.6584284549531187e-05,
+      "loss": 0.8474,
+      "step": 78690
+    },
+    {
+      "epoch": 5.346854192145672,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.658216129908955e-05,
+      "loss": 0.9871,
+      "step": 78695
+    },
+    {
+      "epoch": 5.3471939122163334,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6580038048647915e-05,
+      "loss": 0.8552,
+      "step": 78700
+    },
+    {
+      "epoch": 5.3475336322869955,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.657791479820628e-05,
+      "loss": 0.8729,
+      "step": 78705
+    },
+    {
+      "epoch": 5.347873352357658,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6575791547764643e-05,
+      "loss": 0.8003,
+      "step": 78710
+    },
+    {
+      "epoch": 5.348213072428319,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6573668297323004e-05,
+      "loss": 0.8328,
+      "step": 78715
+    },
+    {
+      "epoch": 5.348552792498981,
+      "grad_norm": 7.5,
+      "learning_rate": 1.657154504688137e-05,
+      "loss": 0.7399,
+      "step": 78720
+    },
+    {
+      "epoch": 5.348892512569643,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6569421796439735e-05,
+      "loss": 0.858,
+      "step": 78725
+    },
+    {
+      "epoch": 5.349232232640304,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6567298545998096e-05,
+      "loss": 0.9161,
+      "step": 78730
+    },
+    {
+      "epoch": 5.349571952710966,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6565175295556463e-05,
+      "loss": 0.8047,
+      "step": 78735
+    },
+    {
+      "epoch": 5.349911672781628,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6563052045114827e-05,
+      "loss": 0.8172,
+      "step": 78740
+    },
+    {
+      "epoch": 5.3502513928522895,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6560928794673188e-05,
+      "loss": 0.8694,
+      "step": 78745
+    },
+    {
+      "epoch": 5.3505911129229515,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.6558805544231552e-05,
+      "loss": 0.8961,
+      "step": 78750
+    },
+    {
+      "epoch": 5.350930832993614,
+      "grad_norm": 8.875,
+      "learning_rate": 1.655668229378992e-05,
+      "loss": 0.6999,
+      "step": 78755
+    },
+    {
+      "epoch": 5.351270553064275,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.655455904334828e-05,
+      "loss": 0.8804,
+      "step": 78760
+    },
+    {
+      "epoch": 5.351610273134937,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.6552435792906644e-05,
+      "loss": 0.8563,
+      "step": 78765
+    },
+    {
+      "epoch": 5.351949993205599,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.655031254246501e-05,
+      "loss": 0.8703,
+      "step": 78770
+    },
+    {
+      "epoch": 5.35228971327626,
+      "grad_norm": 7.375,
+      "learning_rate": 1.6548189292023372e-05,
+      "loss": 0.8029,
+      "step": 78775
+    },
+    {
+      "epoch": 5.352629433346922,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6546066041581736e-05,
+      "loss": 0.7898,
+      "step": 78780
+    },
+    {
+      "epoch": 5.352969153417584,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.6543942791140104e-05,
+      "loss": 0.82,
+      "step": 78785
+    },
+    {
+      "epoch": 5.3533088734882455,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6541819540698468e-05,
+      "loss": 0.8862,
+      "step": 78790
+    },
+    {
+      "epoch": 5.3536485935589075,
+      "grad_norm": 7.5,
+      "learning_rate": 1.6539696290256828e-05,
+      "loss": 0.9517,
+      "step": 78795
+    },
+    {
+      "epoch": 5.35398831362957,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6537573039815192e-05,
+      "loss": 0.8485,
+      "step": 78800
+    },
+    {
+      "epoch": 5.354328033700231,
+      "grad_norm": 8.375,
+      "learning_rate": 1.653544978937356e-05,
+      "loss": 0.8271,
+      "step": 78805
+    },
+    {
+      "epoch": 5.354667753770893,
+      "grad_norm": 8.25,
+      "learning_rate": 1.653332653893192e-05,
+      "loss": 0.8197,
+      "step": 78810
+    },
+    {
+      "epoch": 5.355007473841555,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6531203288490284e-05,
+      "loss": 0.8091,
+      "step": 78815
+    },
+    {
+      "epoch": 5.355347193912216,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.652908003804865e-05,
+      "loss": 0.8337,
+      "step": 78820
+    },
+    {
+      "epoch": 5.355686913982878,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6526956787607012e-05,
+      "loss": 0.9074,
+      "step": 78825
+    },
+    {
+      "epoch": 5.35602663405354,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6524833537165376e-05,
+      "loss": 0.8462,
+      "step": 78830
+    },
+    {
+      "epoch": 5.3563663541242015,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.652271028672374e-05,
+      "loss": 0.8296,
+      "step": 78835
+    },
+    {
+      "epoch": 5.3567060741948636,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6520587036282104e-05,
+      "loss": 0.9518,
+      "step": 78840
+    },
+    {
+      "epoch": 5.357045794265526,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6518463785840468e-05,
+      "loss": 0.7966,
+      "step": 78845
+    },
+    {
+      "epoch": 5.357385514336187,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6516340535398832e-05,
+      "loss": 0.7807,
+      "step": 78850
+    },
+    {
+      "epoch": 5.357725234406849,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6514217284957196e-05,
+      "loss": 0.8475,
+      "step": 78855
+    },
+    {
+      "epoch": 5.358064954477511,
+      "grad_norm": 8.125,
+      "learning_rate": 1.651209403451556e-05,
+      "loss": 0.8431,
+      "step": 78860
+    },
+    {
+      "epoch": 5.358404674548172,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6509970784073924e-05,
+      "loss": 0.8174,
+      "step": 78865
+    },
+    {
+      "epoch": 5.358744394618834,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6507847533632288e-05,
+      "loss": 0.913,
+      "step": 78870
+    },
+    {
+      "epoch": 5.359084114689496,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6505724283190652e-05,
+      "loss": 0.7862,
+      "step": 78875
+    },
+    {
+      "epoch": 5.3594238347601575,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6503601032749016e-05,
+      "loss": 0.9347,
+      "step": 78880
+    },
+    {
+      "epoch": 5.35976355483082,
+      "grad_norm": 7.75,
+      "learning_rate": 1.650147778230738e-05,
+      "loss": 0.9323,
+      "step": 78885
+    },
+    {
+      "epoch": 5.360103274901482,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6499354531865744e-05,
+      "loss": 0.8703,
+      "step": 78890
+    },
+    {
+      "epoch": 5.360442994972143,
+      "grad_norm": 6.625,
+      "learning_rate": 1.6497231281424108e-05,
+      "loss": 0.8272,
+      "step": 78895
+    },
+    {
+      "epoch": 5.360782715042805,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.649510803098247e-05,
+      "loss": 0.8225,
+      "step": 78900
+    },
+    {
+      "epoch": 5.361122435113466,
+      "grad_norm": 7.25,
+      "learning_rate": 1.6492984780540836e-05,
+      "loss": 0.809,
+      "step": 78905
+    },
+    {
+      "epoch": 5.361462155184128,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.64908615300992e-05,
+      "loss": 0.8161,
+      "step": 78910
+    },
+    {
+      "epoch": 5.36180187525479,
+      "grad_norm": 8.75,
+      "learning_rate": 1.648873827965756e-05,
+      "loss": 0.791,
+      "step": 78915
+    },
+    {
+      "epoch": 5.362141595325451,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.6486615029215928e-05,
+      "loss": 0.8082,
+      "step": 78920
+    },
+    {
+      "epoch": 5.3624813153961135,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.6484491778774292e-05,
+      "loss": 0.805,
+      "step": 78925
+    },
+    {
+      "epoch": 5.362821035466776,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6482368528332653e-05,
+      "loss": 0.7903,
+      "step": 78930
+    },
+    {
+      "epoch": 5.363160755537437,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.6480245277891017e-05,
+      "loss": 0.8921,
+      "step": 78935
+    },
+    {
+      "epoch": 5.363500475608099,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.6478122027449384e-05,
+      "loss": 0.8242,
+      "step": 78940
+    },
+    {
+      "epoch": 5.363840195678761,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6475998777007745e-05,
+      "loss": 0.7887,
+      "step": 78945
+    },
+    {
+      "epoch": 5.364179915749422,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.647387552656611e-05,
+      "loss": 0.8077,
+      "step": 78950
+    },
+    {
+      "epoch": 5.364519635820084,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6471752276124476e-05,
+      "loss": 0.8444,
+      "step": 78955
+    },
+    {
+      "epoch": 5.364859355890746,
+      "grad_norm": 7.875,
+      "learning_rate": 1.646962902568284e-05,
+      "loss": 0.8011,
+      "step": 78960
+    },
+    {
+      "epoch": 5.365199075961407,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.64675057752412e-05,
+      "loss": 0.7941,
+      "step": 78965
+    },
+    {
+      "epoch": 5.3655387960320695,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.6465382524799565e-05,
+      "loss": 0.842,
+      "step": 78970
+    },
+    {
+      "epoch": 5.365878516102732,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.6463259274357932e-05,
+      "loss": 0.8834,
+      "step": 78975
+    },
+    {
+      "epoch": 5.366218236173393,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6461136023916293e-05,
+      "loss": 0.8633,
+      "step": 78980
+    },
+    {
+      "epoch": 5.366557956244055,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.6459012773474657e-05,
+      "loss": 0.7394,
+      "step": 78985
+    },
+    {
+      "epoch": 5.366897676314717,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6456889523033024e-05,
+      "loss": 0.7962,
+      "step": 78990
+    },
+    {
+      "epoch": 5.367237396385378,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6454766272591385e-05,
+      "loss": 0.7949,
+      "step": 78995
+    },
+    {
+      "epoch": 5.36757711645604,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.645264302214975e-05,
+      "loss": 0.8153,
+      "step": 79000
+    },
+    {
+      "epoch": 5.367916836526702,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.6450519771708113e-05,
+      "loss": 0.769,
+      "step": 79005
+    },
+    {
+      "epoch": 5.3682565565973634,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.6448396521266477e-05,
+      "loss": 0.9237,
+      "step": 79010
+    },
+    {
+      "epoch": 5.3685962766680255,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.644627327082484e-05,
+      "loss": 0.8591,
+      "step": 79015
+    },
+    {
+      "epoch": 5.368935996738688,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.6444150020383205e-05,
+      "loss": 0.9092,
+      "step": 79020
+    },
+    {
+      "epoch": 5.369275716809349,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.644202676994157e-05,
+      "loss": 0.7845,
+      "step": 79025
+    },
+    {
+      "epoch": 5.369615436880011,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6439903519499933e-05,
+      "loss": 0.8878,
+      "step": 79030
+    },
+    {
+      "epoch": 5.369955156950673,
+      "grad_norm": 7.25,
+      "learning_rate": 1.6437780269058297e-05,
+      "loss": 0.869,
+      "step": 79035
+    },
+    {
+      "epoch": 5.370294877021334,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.643565701861666e-05,
+      "loss": 0.7619,
+      "step": 79040
+    },
+    {
+      "epoch": 5.370634597091996,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.6433533768175025e-05,
+      "loss": 0.8771,
+      "step": 79045
+    },
+    {
+      "epoch": 5.370974317162658,
+      "grad_norm": 8.125,
+      "learning_rate": 1.643141051773339e-05,
+      "loss": 0.782,
+      "step": 79050
+    },
+    {
+      "epoch": 5.3713140372333195,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.6429287267291753e-05,
+      "loss": 0.8021,
+      "step": 79055
+    },
+    {
+      "epoch": 5.3716537573039815,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6427164016850117e-05,
+      "loss": 0.8277,
+      "step": 79060
+    },
+    {
+      "epoch": 5.371993477374644,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.642504076640848e-05,
+      "loss": 0.8797,
+      "step": 79065
+    },
+    {
+      "epoch": 5.372333197445305,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.642291751596684e-05,
+      "loss": 0.8702,
+      "step": 79070
+    },
+    {
+      "epoch": 5.372672917515967,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.642079426552521e-05,
+      "loss": 0.91,
+      "step": 79075
+    },
+    {
+      "epoch": 5.373012637586629,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6418671015083573e-05,
+      "loss": 0.8332,
+      "step": 79080
+    },
+    {
+      "epoch": 5.37335235765729,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6416547764641934e-05,
+      "loss": 0.9085,
+      "step": 79085
+    },
+    {
+      "epoch": 5.373692077727952,
+      "grad_norm": 8.125,
+      "learning_rate": 1.64144245142003e-05,
+      "loss": 0.7943,
+      "step": 79090
+    },
+    {
+      "epoch": 5.374031797798614,
+      "grad_norm": 7.125,
+      "learning_rate": 1.6412301263758665e-05,
+      "loss": 0.8625,
+      "step": 79095
+    },
+    {
+      "epoch": 5.3743715178692755,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6410178013317026e-05,
+      "loss": 0.8226,
+      "step": 79100
+    },
+    {
+      "epoch": 5.3747112379399375,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.640805476287539e-05,
+      "loss": 0.8418,
+      "step": 79105
+    },
+    {
+      "epoch": 5.3750509580106,
+      "grad_norm": 7.5,
+      "learning_rate": 1.6405931512433757e-05,
+      "loss": 0.7672,
+      "step": 79110
+    },
+    {
+      "epoch": 5.375390678081261,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6403808261992118e-05,
+      "loss": 0.8703,
+      "step": 79115
+    },
+    {
+      "epoch": 5.375730398151923,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.640168501155048e-05,
+      "loss": 0.891,
+      "step": 79120
+    },
+    {
+      "epoch": 5.376070118222585,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.639956176110885e-05,
+      "loss": 0.8822,
+      "step": 79125
+    },
+    {
+      "epoch": 5.376409838293246,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.6397438510667213e-05,
+      "loss": 0.8893,
+      "step": 79130
+    },
+    {
+      "epoch": 5.376749558363908,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6395315260225574e-05,
+      "loss": 0.8495,
+      "step": 79135
+    },
+    {
+      "epoch": 5.37708927843457,
+      "grad_norm": 9.25,
+      "learning_rate": 1.6393192009783938e-05,
+      "loss": 0.8098,
+      "step": 79140
+    },
+    {
+      "epoch": 5.3774289985052315,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6391068759342305e-05,
+      "loss": 0.8465,
+      "step": 79145
+    },
+    {
+      "epoch": 5.377768718575894,
+      "grad_norm": 7.25,
+      "learning_rate": 1.6388945508900666e-05,
+      "loss": 0.8589,
+      "step": 79150
+    },
+    {
+      "epoch": 5.378108438646556,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.638682225845903e-05,
+      "loss": 0.7749,
+      "step": 79155
+    },
+    {
+      "epoch": 5.378448158717217,
+      "grad_norm": 7.0,
+      "learning_rate": 1.6384699008017397e-05,
+      "loss": 0.7973,
+      "step": 79160
+    },
+    {
+      "epoch": 5.378787878787879,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6382575757575758e-05,
+      "loss": 0.8595,
+      "step": 79165
+    },
+    {
+      "epoch": 5.379127598858541,
+      "grad_norm": 8.625,
+      "learning_rate": 1.638045250713412e-05,
+      "loss": 0.937,
+      "step": 79170
+    },
+    {
+      "epoch": 5.379467318929202,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6378329256692486e-05,
+      "loss": 0.7773,
+      "step": 79175
+    },
+    {
+      "epoch": 5.379807038999864,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.637620600625085e-05,
+      "loss": 0.8809,
+      "step": 79180
+    },
+    {
+      "epoch": 5.380146759070525,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6374082755809214e-05,
+      "loss": 0.8387,
+      "step": 79185
+    },
+    {
+      "epoch": 5.3804864791411875,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6371959505367578e-05,
+      "loss": 0.8964,
+      "step": 79190
+    },
+    {
+      "epoch": 5.38082619921185,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.636983625492594e-05,
+      "loss": 0.8339,
+      "step": 79195
+    },
+    {
+      "epoch": 5.381165919282511,
+      "grad_norm": 8.875,
+      "learning_rate": 1.6367713004484306e-05,
+      "loss": 0.8654,
+      "step": 79200
+    },
+    {
+      "epoch": 5.381505639353173,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.636558975404267e-05,
+      "loss": 0.8461,
+      "step": 79205
+    },
+    {
+      "epoch": 5.381845359423835,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.6363466503601034e-05,
+      "loss": 0.8254,
+      "step": 79210
+    },
+    {
+      "epoch": 5.382185079494496,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6361343253159398e-05,
+      "loss": 0.8525,
+      "step": 79215
+    },
+    {
+      "epoch": 5.382524799565158,
+      "grad_norm": 8.5,
+      "learning_rate": 1.635922000271776e-05,
+      "loss": 0.8614,
+      "step": 79220
+    },
+    {
+      "epoch": 5.38286451963582,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6357096752276126e-05,
+      "loss": 0.9354,
+      "step": 79225
+    },
+    {
+      "epoch": 5.383204239706481,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.635497350183449e-05,
+      "loss": 0.8894,
+      "step": 79230
+    },
+    {
+      "epoch": 5.3835439597771435,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6352850251392854e-05,
+      "loss": 0.8958,
+      "step": 79235
+    },
+    {
+      "epoch": 5.383883679847806,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.6350727000951214e-05,
+      "loss": 0.8906,
+      "step": 79240
+    },
+    {
+      "epoch": 5.384223399918467,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.634860375050958e-05,
+      "loss": 0.7682,
+      "step": 79245
+    },
+    {
+      "epoch": 5.384563119989129,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6346480500067946e-05,
+      "loss": 0.8688,
+      "step": 79250
+    },
+    {
+      "epoch": 5.384902840059791,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6344357249626306e-05,
+      "loss": 0.7652,
+      "step": 79255
+    },
+    {
+      "epoch": 5.385242560130452,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6342233999184674e-05,
+      "loss": 0.8817,
+      "step": 79260
+    },
+    {
+      "epoch": 5.385582280201114,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6340110748743038e-05,
+      "loss": 0.8493,
+      "step": 79265
+    },
+    {
+      "epoch": 5.385922000271776,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.63379874983014e-05,
+      "loss": 0.8758,
+      "step": 79270
+    },
+    {
+      "epoch": 5.386261720342437,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6335864247859762e-05,
+      "loss": 0.8585,
+      "step": 79275
+    },
+    {
+      "epoch": 5.3866014404130995,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.633374099741813e-05,
+      "loss": 0.8081,
+      "step": 79280
+    },
+    {
+      "epoch": 5.386941160483762,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.633161774697649e-05,
+      "loss": 0.8361,
+      "step": 79285
+    },
+    {
+      "epoch": 5.387280880554423,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6329494496534854e-05,
+      "loss": 0.8724,
+      "step": 79290
+    },
+    {
+      "epoch": 5.387620600625085,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6327371246093222e-05,
+      "loss": 0.8526,
+      "step": 79295
+    },
+    {
+      "epoch": 5.387960320695747,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.6325247995651586e-05,
+      "loss": 0.8704,
+      "step": 79300
+    },
+    {
+      "epoch": 5.388300040766408,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6323124745209946e-05,
+      "loss": 0.8046,
+      "step": 79305
+    },
+    {
+      "epoch": 5.38863976083707,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.6321001494768314e-05,
+      "loss": 0.8771,
+      "step": 79310
+    },
+    {
+      "epoch": 5.388979480907732,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6318878244326678e-05,
+      "loss": 0.8199,
+      "step": 79315
+    },
+    {
+      "epoch": 5.3893192009783935,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.631675499388504e-05,
+      "loss": 0.8551,
+      "step": 79320
+    },
+    {
+      "epoch": 5.3896589210490555,
+      "grad_norm": 9.25,
+      "learning_rate": 1.6314631743443402e-05,
+      "loss": 0.8374,
+      "step": 79325
+    },
+    {
+      "epoch": 5.389998641119718,
+      "grad_norm": 8.5,
+      "learning_rate": 1.631250849300177e-05,
+      "loss": 0.8724,
+      "step": 79330
+    },
+    {
+      "epoch": 5.390338361190379,
+      "grad_norm": 7.75,
+      "learning_rate": 1.631038524256013e-05,
+      "loss": 0.8586,
+      "step": 79335
+    },
+    {
+      "epoch": 5.390678081261041,
+      "grad_norm": 9.125,
+      "learning_rate": 1.6308261992118494e-05,
+      "loss": 0.9117,
+      "step": 79340
+    },
+    {
+      "epoch": 5.391017801331703,
+      "grad_norm": 6.875,
+      "learning_rate": 1.6306138741676862e-05,
+      "loss": 0.7753,
+      "step": 79345
+    },
+    {
+      "epoch": 5.391357521402364,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6304015491235222e-05,
+      "loss": 0.8057,
+      "step": 79350
+    },
+    {
+      "epoch": 5.391697241473026,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6301892240793586e-05,
+      "loss": 0.8699,
+      "step": 79355
+    },
+    {
+      "epoch": 5.392036961543688,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.629976899035195e-05,
+      "loss": 0.871,
+      "step": 79360
+    },
+    {
+      "epoch": 5.3923766816143495,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6297645739910314e-05,
+      "loss": 0.9198,
+      "step": 79365
+    },
+    {
+      "epoch": 5.3927164016850115,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.629552248946868e-05,
+      "loss": 0.8214,
+      "step": 79370
+    },
+    {
+      "epoch": 5.393056121755674,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.6293399239027042e-05,
+      "loss": 0.8826,
+      "step": 79375
+    },
+    {
+      "epoch": 5.393395841826335,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6291275988585406e-05,
+      "loss": 0.831,
+      "step": 79380
+    },
+    {
+      "epoch": 5.393735561896997,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.628915273814377e-05,
+      "loss": 0.8934,
+      "step": 79385
+    },
+    {
+      "epoch": 5.394075281967659,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6287029487702134e-05,
+      "loss": 0.8204,
+      "step": 79390
+    },
+    {
+      "epoch": 5.39441500203832,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.62849062372605e-05,
+      "loss": 0.8697,
+      "step": 79395
+    },
+    {
+      "epoch": 5.394754722108982,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.6282782986818862e-05,
+      "loss": 0.8972,
+      "step": 79400
+    },
+    {
+      "epoch": 5.395094442179644,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6280659736377226e-05,
+      "loss": 0.8714,
+      "step": 79405
+    },
+    {
+      "epoch": 5.3954341622503055,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.627853648593559e-05,
+      "loss": 0.874,
+      "step": 79410
+    },
+    {
+      "epoch": 5.3957738823209676,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.6276413235493954e-05,
+      "loss": 0.7862,
+      "step": 79415
+    },
+    {
+      "epoch": 5.39611360239163,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.627428998505232e-05,
+      "loss": 0.8299,
+      "step": 79420
+    },
+    {
+      "epoch": 5.396453322462291,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.627216673461068e-05,
+      "loss": 0.9183,
+      "step": 79425
+    },
+    {
+      "epoch": 5.396793042532953,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6270043484169046e-05,
+      "loss": 0.8519,
+      "step": 79430
+    },
+    {
+      "epoch": 5.397132762603615,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.626792023372741e-05,
+      "loss": 0.8684,
+      "step": 79435
+    },
+    {
+      "epoch": 5.397472482674276,
+      "grad_norm": 8.5,
+      "learning_rate": 1.626579698328577e-05,
+      "loss": 0.9583,
+      "step": 79440
+    },
+    {
+      "epoch": 5.397812202744938,
+      "grad_norm": 7.5,
+      "learning_rate": 1.626367373284414e-05,
+      "loss": 0.8662,
+      "step": 79445
+    },
+    {
+      "epoch": 5.3981519228156,
+      "grad_norm": 8.75,
+      "learning_rate": 1.6261550482402502e-05,
+      "loss": 0.8339,
+      "step": 79450
+    },
+    {
+      "epoch": 5.3984916428862615,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.6259427231960863e-05,
+      "loss": 0.9196,
+      "step": 79455
+    },
+    {
+      "epoch": 5.398831362956924,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6257303981519227e-05,
+      "loss": 0.8267,
+      "step": 79460
+    },
+    {
+      "epoch": 5.399171083027586,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.6255180731077594e-05,
+      "loss": 0.7795,
+      "step": 79465
+    },
+    {
+      "epoch": 5.399510803098247,
+      "grad_norm": 8.125,
+      "learning_rate": 1.625305748063596e-05,
+      "loss": 0.8441,
+      "step": 79470
+    },
+    {
+      "epoch": 5.399850523168909,
+      "grad_norm": 8.25,
+      "learning_rate": 1.625093423019432e-05,
+      "loss": 0.9115,
+      "step": 79475
+    },
+    {
+      "epoch": 5.400190243239571,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6248810979752686e-05,
+      "loss": 0.9677,
+      "step": 79480
+    },
+    {
+      "epoch": 5.400529963310232,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.624668772931105e-05,
+      "loss": 0.8878,
+      "step": 79485
+    },
+    {
+      "epoch": 5.400869683380894,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.624456447886941e-05,
+      "loss": 0.909,
+      "step": 79490
+    },
+    {
+      "epoch": 5.401209403451556,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.6242441228427775e-05,
+      "loss": 0.9181,
+      "step": 79495
+    },
+    {
+      "epoch": 5.4015491235222175,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.6240317977986142e-05,
+      "loss": 0.8473,
+      "step": 79500
+    },
+    {
+      "epoch": 5.40188884359288,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6238194727544503e-05,
+      "loss": 0.8853,
+      "step": 79505
+    },
+    {
+      "epoch": 5.402228563663542,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6236071477102867e-05,
+      "loss": 0.8088,
+      "step": 79510
+    },
+    {
+      "epoch": 5.402568283734203,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6233948226661235e-05,
+      "loss": 0.8271,
+      "step": 79515
+    },
+    {
+      "epoch": 5.402908003804865,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6231824976219595e-05,
+      "loss": 0.8621,
+      "step": 79520
+    },
+    {
+      "epoch": 5.403247723875527,
+      "grad_norm": 8.0,
+      "learning_rate": 1.622970172577796e-05,
+      "loss": 0.9029,
+      "step": 79525
+    },
+    {
+      "epoch": 5.403587443946188,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6227578475336323e-05,
+      "loss": 0.8437,
+      "step": 79530
+    },
+    {
+      "epoch": 5.40392716401685,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6225455224894687e-05,
+      "loss": 0.8791,
+      "step": 79535
+    },
+    {
+      "epoch": 5.404266884087512,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.622333197445305e-05,
+      "loss": 0.8398,
+      "step": 79540
+    },
+    {
+      "epoch": 5.4046066041581735,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6221208724011415e-05,
+      "loss": 0.9305,
+      "step": 79545
+    },
+    {
+      "epoch": 5.404946324228836,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.621908547356978e-05,
+      "loss": 0.8714,
+      "step": 79550
+    },
+    {
+      "epoch": 5.405286044299498,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.6216962223128143e-05,
+      "loss": 0.9087,
+      "step": 79555
+    },
+    {
+      "epoch": 5.405625764370159,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6214838972686507e-05,
+      "loss": 0.8022,
+      "step": 79560
+    },
+    {
+      "epoch": 5.405965484440821,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.621271572224487e-05,
+      "loss": 0.8,
+      "step": 79565
+    },
+    {
+      "epoch": 5.406305204511483,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6210592471803235e-05,
+      "loss": 0.8284,
+      "step": 79570
+    },
+    {
+      "epoch": 5.406644924582144,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.62084692213616e-05,
+      "loss": 0.8615,
+      "step": 79575
+    },
+    {
+      "epoch": 5.406984644652806,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.6206345970919963e-05,
+      "loss": 0.9173,
+      "step": 79580
+    },
+    {
+      "epoch": 5.4073243647234674,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6204222720478327e-05,
+      "loss": 0.7823,
+      "step": 79585
+    },
+    {
+      "epoch": 5.4076640847941295,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.620209947003669e-05,
+      "loss": 0.8173,
+      "step": 79590
+    },
+    {
+      "epoch": 5.408003804864792,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6199976219595052e-05,
+      "loss": 0.8121,
+      "step": 79595
+    },
+    {
+      "epoch": 5.408343524935453,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.619785296915342e-05,
+      "loss": 0.9628,
+      "step": 79600
+    },
+    {
+      "epoch": 5.408683245006115,
+      "grad_norm": 6.75,
+      "learning_rate": 1.6195729718711783e-05,
+      "loss": 0.8492,
+      "step": 79605
+    },
+    {
+      "epoch": 5.409022965076777,
+      "grad_norm": 6.25,
+      "learning_rate": 1.6193606468270144e-05,
+      "loss": 0.8032,
+      "step": 79610
+    },
+    {
+      "epoch": 5.409362685147438,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.619148321782851e-05,
+      "loss": 0.7936,
+      "step": 79615
+    },
+    {
+      "epoch": 5.4097024052181,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.6189359967386875e-05,
+      "loss": 0.8218,
+      "step": 79620
+    },
+    {
+      "epoch": 5.410042125288762,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.6187236716945236e-05,
+      "loss": 0.9732,
+      "step": 79625
+    },
+    {
+      "epoch": 5.4103818453594235,
+      "grad_norm": 7.875,
+      "learning_rate": 1.61851134665036e-05,
+      "loss": 0.9203,
+      "step": 79630
+    },
+    {
+      "epoch": 5.4107215654300855,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6182990216061967e-05,
+      "loss": 0.9041,
+      "step": 79635
+    },
+    {
+      "epoch": 5.411061285500748,
+      "grad_norm": 5.8125,
+      "learning_rate": 1.618086696562033e-05,
+      "loss": 0.9393,
+      "step": 79640
+    },
+    {
+      "epoch": 5.411401005571409,
+      "grad_norm": 6.875,
+      "learning_rate": 1.6178743715178692e-05,
+      "loss": 0.8926,
+      "step": 79645
+    },
+    {
+      "epoch": 5.411740725642071,
+      "grad_norm": 8.875,
+      "learning_rate": 1.617662046473706e-05,
+      "loss": 0.7882,
+      "step": 79650
+    },
+    {
+      "epoch": 5.412080445712733,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.6174497214295423e-05,
+      "loss": 0.9406,
+      "step": 79655
+    },
+    {
+      "epoch": 5.412420165783394,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6172373963853784e-05,
+      "loss": 0.9049,
+      "step": 79660
+    },
+    {
+      "epoch": 5.412759885854056,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.6170250713412148e-05,
+      "loss": 0.8655,
+      "step": 79665
+    },
+    {
+      "epoch": 5.413099605924718,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6168127462970515e-05,
+      "loss": 0.8967,
+      "step": 79670
+    },
+    {
+      "epoch": 5.4134393259953795,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6166004212528876e-05,
+      "loss": 0.8926,
+      "step": 79675
+    },
+    {
+      "epoch": 5.4137790460660415,
+      "grad_norm": 8.5,
+      "learning_rate": 1.616388096208724e-05,
+      "loss": 0.7752,
+      "step": 79680
+    },
+    {
+      "epoch": 5.414118766136704,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6161757711645607e-05,
+      "loss": 0.7929,
+      "step": 79685
+    },
+    {
+      "epoch": 5.414458486207365,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6159634461203968e-05,
+      "loss": 0.8396,
+      "step": 79690
+    },
+    {
+      "epoch": 5.414798206278027,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6157511210762332e-05,
+      "loss": 0.9007,
+      "step": 79695
+    },
+    {
+      "epoch": 5.415137926348689,
+      "grad_norm": 9.375,
+      "learning_rate": 1.6155387960320696e-05,
+      "loss": 0.915,
+      "step": 79700
+    },
+    {
+      "epoch": 5.41547764641935,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.615326470987906e-05,
+      "loss": 0.8285,
+      "step": 79705
+    },
+    {
+      "epoch": 5.415817366490012,
+      "grad_norm": 7.5,
+      "learning_rate": 1.6151141459437424e-05,
+      "loss": 0.7153,
+      "step": 79710
+    },
+    {
+      "epoch": 5.416157086560674,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6149018208995788e-05,
+      "loss": 0.8404,
+      "step": 79715
+    },
+    {
+      "epoch": 5.4164968066313355,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6146894958554152e-05,
+      "loss": 0.8417,
+      "step": 79720
+    },
+    {
+      "epoch": 5.416836526701998,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.6144771708112516e-05,
+      "loss": 0.9056,
+      "step": 79725
+    },
+    {
+      "epoch": 5.41717624677266,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.614264845767088e-05,
+      "loss": 0.8817,
+      "step": 79730
+    },
+    {
+      "epoch": 5.417515966843321,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.6140525207229244e-05,
+      "loss": 0.8607,
+      "step": 79735
+    },
+    {
+      "epoch": 5.417855686913983,
+      "grad_norm": 7.75,
+      "learning_rate": 1.6138401956787608e-05,
+      "loss": 0.7989,
+      "step": 79740
+    },
+    {
+      "epoch": 5.418195406984645,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6136278706345972e-05,
+      "loss": 0.8051,
+      "step": 79745
+    },
+    {
+      "epoch": 5.418535127055306,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.6134155455904336e-05,
+      "loss": 0.8493,
+      "step": 79750
+    },
+    {
+      "epoch": 5.418874847125968,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.61320322054627e-05,
+      "loss": 0.8801,
+      "step": 79755
+    },
+    {
+      "epoch": 5.41921456719663,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6129908955021064e-05,
+      "loss": 0.86,
+      "step": 79760
+    },
+    {
+      "epoch": 5.4195542872672915,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6127785704579425e-05,
+      "loss": 0.8868,
+      "step": 79765
+    },
+    {
+      "epoch": 5.419894007337954,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.6125662454137792e-05,
+      "loss": 0.782,
+      "step": 79770
+    },
+    {
+      "epoch": 5.420233727408616,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6123539203696156e-05,
+      "loss": 0.9035,
+      "step": 79775
+    },
+    {
+      "epoch": 5.420573447479277,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6121415953254517e-05,
+      "loss": 0.915,
+      "step": 79780
+    },
+    {
+      "epoch": 5.420913167549939,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.6119292702812884e-05,
+      "loss": 0.9613,
+      "step": 79785
+    },
+    {
+      "epoch": 5.421252887620601,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6117169452371248e-05,
+      "loss": 0.8963,
+      "step": 79790
+    },
+    {
+      "epoch": 5.421592607691262,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.611504620192961e-05,
+      "loss": 0.8162,
+      "step": 79795
+    },
+    {
+      "epoch": 5.421932327761924,
+      "grad_norm": 7.625,
+      "learning_rate": 1.6112922951487973e-05,
+      "loss": 0.8906,
+      "step": 79800
+    },
+    {
+      "epoch": 5.422272047832586,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.611079970104634e-05,
+      "loss": 0.8693,
+      "step": 79805
+    },
+    {
+      "epoch": 5.4226117679032475,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.6108676450604704e-05,
+      "loss": 0.7914,
+      "step": 79810
+    },
+    {
+      "epoch": 5.42295148797391,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6106553200163065e-05,
+      "loss": 0.7681,
+      "step": 79815
+    },
+    {
+      "epoch": 5.423291208044572,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.6104429949721432e-05,
+      "loss": 0.7928,
+      "step": 79820
+    },
+    {
+      "epoch": 5.423630928115233,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6102306699279796e-05,
+      "loss": 0.82,
+      "step": 79825
+    },
+    {
+      "epoch": 5.423970648185895,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6100183448838157e-05,
+      "loss": 0.8803,
+      "step": 79830
+    },
+    {
+      "epoch": 5.424310368256557,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6098060198396524e-05,
+      "loss": 0.8581,
+      "step": 79835
+    },
+    {
+      "epoch": 5.424650088327218,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.6095936947954888e-05,
+      "loss": 0.8668,
+      "step": 79840
+    },
+    {
+      "epoch": 5.42498980839788,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.609381369751325e-05,
+      "loss": 0.799,
+      "step": 79845
+    },
+    {
+      "epoch": 5.425329528468542,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.6091690447071613e-05,
+      "loss": 0.8988,
+      "step": 79850
+    },
+    {
+      "epoch": 5.4256692485392035,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.608956719662998e-05,
+      "loss": 0.853,
+      "step": 79855
+    },
+    {
+      "epoch": 5.426008968609866,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.608744394618834e-05,
+      "loss": 0.7911,
+      "step": 79860
+    },
+    {
+      "epoch": 5.426348688680527,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6085320695746705e-05,
+      "loss": 0.8168,
+      "step": 79865
+    },
+    {
+      "epoch": 5.426688408751189,
+      "grad_norm": 7.375,
+      "learning_rate": 1.6083197445305072e-05,
+      "loss": 0.806,
+      "step": 79870
+    },
+    {
+      "epoch": 5.427028128821851,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.6081074194863433e-05,
+      "loss": 0.8627,
+      "step": 79875
+    },
+    {
+      "epoch": 5.427367848892512,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6078950944421797e-05,
+      "loss": 0.8775,
+      "step": 79880
+    },
+    {
+      "epoch": 5.427707568963174,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.607682769398016e-05,
+      "loss": 0.8087,
+      "step": 79885
+    },
+    {
+      "epoch": 5.428047289033836,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.6074704443538525e-05,
+      "loss": 0.8817,
+      "step": 79890
+    },
+    {
+      "epoch": 5.4283870091044975,
+      "grad_norm": 8.25,
+      "learning_rate": 1.607258119309689e-05,
+      "loss": 0.9597,
+      "step": 79895
+    },
+    {
+      "epoch": 5.4287267291751595,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.6070457942655253e-05,
+      "loss": 0.7875,
+      "step": 79900
+    },
+    {
+      "epoch": 5.429066449245822,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6068334692213617e-05,
+      "loss": 0.8082,
+      "step": 79905
+    },
+    {
+      "epoch": 5.429406169316483,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.606621144177198e-05,
+      "loss": 0.7957,
+      "step": 79910
+    },
+    {
+      "epoch": 5.429745889387145,
+      "grad_norm": 7.5,
+      "learning_rate": 1.6064088191330345e-05,
+      "loss": 0.7979,
+      "step": 79915
+    },
+    {
+      "epoch": 5.430085609457807,
+      "grad_norm": 8.125,
+      "learning_rate": 1.606196494088871e-05,
+      "loss": 0.7979,
+      "step": 79920
+    },
+    {
+      "epoch": 5.430425329528468,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6059841690447073e-05,
+      "loss": 0.8895,
+      "step": 79925
+    },
+    {
+      "epoch": 5.43076504959913,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.6057718440005437e-05,
+      "loss": 0.7594,
+      "step": 79930
+    },
+    {
+      "epoch": 5.431104769669792,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.60555951895638e-05,
+      "loss": 0.7908,
+      "step": 79935
+    },
+    {
+      "epoch": 5.4314444897404535,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.6053471939122165e-05,
+      "loss": 0.8662,
+      "step": 79940
+    },
+    {
+      "epoch": 5.4317842098111155,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.605134868868053e-05,
+      "loss": 0.8345,
+      "step": 79945
+    },
+    {
+      "epoch": 5.432123929881778,
+      "grad_norm": 7.625,
+      "learning_rate": 1.604922543823889e-05,
+      "loss": 0.8568,
+      "step": 79950
+    },
+    {
+      "epoch": 5.432463649952439,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6047102187797257e-05,
+      "loss": 0.8383,
+      "step": 79955
+    },
+    {
+      "epoch": 5.432803370023101,
+      "grad_norm": 7.625,
+      "learning_rate": 1.604497893735562e-05,
+      "loss": 0.8143,
+      "step": 79960
+    },
+    {
+      "epoch": 5.433143090093763,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.604285568691398e-05,
+      "loss": 0.8673,
+      "step": 79965
+    },
+    {
+      "epoch": 5.433482810164424,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.604073243647235e-05,
+      "loss": 0.8004,
+      "step": 79970
+    },
+    {
+      "epoch": 5.433822530235086,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6038609186030713e-05,
+      "loss": 0.7687,
+      "step": 79975
+    },
+    {
+      "epoch": 5.434162250305748,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6036485935589077e-05,
+      "loss": 0.8758,
+      "step": 79980
+    },
+    {
+      "epoch": 5.4345019703764095,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6034362685147437e-05,
+      "loss": 0.7764,
+      "step": 79985
+    },
+    {
+      "epoch": 5.4348416904470715,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6032239434705805e-05,
+      "loss": 0.8111,
+      "step": 79990
+    },
+    {
+      "epoch": 5.435181410517734,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.603011618426417e-05,
+      "loss": 0.7457,
+      "step": 79995
+    },
+    {
+      "epoch": 5.435521130588395,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.602799293382253e-05,
+      "loss": 0.8823,
+      "step": 80000
+    },
+    {
+      "epoch": 5.435860850659057,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.6025869683380897e-05,
+      "loss": 0.8603,
+      "step": 80005
+    },
+    {
+      "epoch": 5.436200570729719,
+      "grad_norm": 8.375,
+      "learning_rate": 1.602374643293926e-05,
+      "loss": 0.9029,
+      "step": 80010
+    },
+    {
+      "epoch": 5.43654029080038,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.602162318249762e-05,
+      "loss": 0.7838,
+      "step": 80015
+    },
+    {
+      "epoch": 5.436880010871042,
+      "grad_norm": 7.125,
+      "learning_rate": 1.6019499932055985e-05,
+      "loss": 0.8378,
+      "step": 80020
+    },
+    {
+      "epoch": 5.437219730941704,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.6017376681614353e-05,
+      "loss": 0.8029,
+      "step": 80025
+    },
+    {
+      "epoch": 5.4375594510123655,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6015253431172713e-05,
+      "loss": 0.926,
+      "step": 80030
+    },
+    {
+      "epoch": 5.437899171083028,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6013130180731077e-05,
+      "loss": 0.8649,
+      "step": 80035
+    },
+    {
+      "epoch": 5.43823889115369,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6011006930289445e-05,
+      "loss": 0.7956,
+      "step": 80040
+    },
+    {
+      "epoch": 5.438578611224351,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6008883679847805e-05,
+      "loss": 0.7927,
+      "step": 80045
+    },
+    {
+      "epoch": 5.438918331295013,
+      "grad_norm": 8.625,
+      "learning_rate": 1.600676042940617e-05,
+      "loss": 0.8495,
+      "step": 80050
+    },
+    {
+      "epoch": 5.439258051365675,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.6004637178964533e-05,
+      "loss": 1.0238,
+      "step": 80055
+    },
+    {
+      "epoch": 5.439597771436336,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6002513928522897e-05,
+      "loss": 0.9669,
+      "step": 80060
+    },
+    {
+      "epoch": 5.439937491506998,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.600039067808126e-05,
+      "loss": 0.854,
+      "step": 80065
+    },
+    {
+      "epoch": 5.44027721157766,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.5998267427639625e-05,
+      "loss": 0.8373,
+      "step": 80070
+    },
+    {
+      "epoch": 5.4406169316483215,
+      "grad_norm": 7.375,
+      "learning_rate": 1.599614417719799e-05,
+      "loss": 0.9692,
+      "step": 80075
+    },
+    {
+      "epoch": 5.440956651718984,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5994020926756353e-05,
+      "loss": 0.8007,
+      "step": 80080
+    },
+    {
+      "epoch": 5.441296371789646,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5991897676314717e-05,
+      "loss": 0.909,
+      "step": 80085
+    },
+    {
+      "epoch": 5.441636091860307,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.598977442587308e-05,
+      "loss": 0.8883,
+      "step": 80090
+    },
+    {
+      "epoch": 5.441975811930969,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5987651175431445e-05,
+      "loss": 0.785,
+      "step": 80095
+    },
+    {
+      "epoch": 5.442315532001631,
+      "grad_norm": 8.25,
+      "learning_rate": 1.598552792498981e-05,
+      "loss": 0.7853,
+      "step": 80100
+    },
+    {
+      "epoch": 5.442655252072292,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5983404674548173e-05,
+      "loss": 0.8707,
+      "step": 80105
+    },
+    {
+      "epoch": 5.442994972142954,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5981281424106537e-05,
+      "loss": 0.8493,
+      "step": 80110
+    },
+    {
+      "epoch": 5.443334692213616,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.59791581736649e-05,
+      "loss": 0.8735,
+      "step": 80115
+    },
+    {
+      "epoch": 5.4436744122842775,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5977034923223262e-05,
+      "loss": 0.9122,
+      "step": 80120
+    },
+    {
+      "epoch": 5.44401413235494,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.597491167278163e-05,
+      "loss": 0.8704,
+      "step": 80125
+    },
+    {
+      "epoch": 5.444353852425602,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.5972788422339993e-05,
+      "loss": 0.8623,
+      "step": 80130
+    },
+    {
+      "epoch": 5.444693572496263,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5970665171898354e-05,
+      "loss": 0.8888,
+      "step": 80135
+    },
+    {
+      "epoch": 5.445033292566925,
+      "grad_norm": 8.375,
+      "learning_rate": 1.596854192145672e-05,
+      "loss": 0.8149,
+      "step": 80140
+    },
+    {
+      "epoch": 5.445373012637587,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5966418671015085e-05,
+      "loss": 0.8446,
+      "step": 80145
+    },
+    {
+      "epoch": 5.445712732708248,
+      "grad_norm": 6.375,
+      "learning_rate": 1.596429542057345e-05,
+      "loss": 0.832,
+      "step": 80150
+    },
+    {
+      "epoch": 5.44605245277891,
+      "grad_norm": 7.875,
+      "learning_rate": 1.596217217013181e-05,
+      "loss": 0.7592,
+      "step": 80155
+    },
+    {
+      "epoch": 5.446392172849572,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5960048919690177e-05,
+      "loss": 0.9213,
+      "step": 80160
+    },
+    {
+      "epoch": 5.4467318929202335,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.595792566924854e-05,
+      "loss": 0.8183,
+      "step": 80165
+    },
+    {
+      "epoch": 5.447071612990896,
+      "grad_norm": 8.75,
+      "learning_rate": 1.5955802418806902e-05,
+      "loss": 0.9753,
+      "step": 80170
+    },
+    {
+      "epoch": 5.447411333061558,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.595367916836527e-05,
+      "loss": 0.9091,
+      "step": 80175
+    },
+    {
+      "epoch": 5.447751053132219,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5951555917923633e-05,
+      "loss": 0.8396,
+      "step": 80180
+    },
+    {
+      "epoch": 5.448090773202881,
+      "grad_norm": 8.75,
+      "learning_rate": 1.5949432667481994e-05,
+      "loss": 0.8612,
+      "step": 80185
+    },
+    {
+      "epoch": 5.448430493273543,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5947309417040358e-05,
+      "loss": 0.9503,
+      "step": 80190
+    },
+    {
+      "epoch": 5.448770213344204,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5945186166598725e-05,
+      "loss": 0.9039,
+      "step": 80195
+    },
+    {
+      "epoch": 5.449109933414866,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5943062916157086e-05,
+      "loss": 0.8329,
+      "step": 80200
+    },
+    {
+      "epoch": 5.449449653485528,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.594093966571545e-05,
+      "loss": 0.9124,
+      "step": 80205
+    },
+    {
+      "epoch": 5.4497893735561895,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5938816415273817e-05,
+      "loss": 0.8881,
+      "step": 80210
+    },
+    {
+      "epoch": 5.450129093626852,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5936693164832178e-05,
+      "loss": 1.0147,
+      "step": 80215
+    },
+    {
+      "epoch": 5.450468813697514,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5934569914390542e-05,
+      "loss": 0.8586,
+      "step": 80220
+    },
+    {
+      "epoch": 5.450808533768175,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.593244666394891e-05,
+      "loss": 0.8438,
+      "step": 80225
+    },
+    {
+      "epoch": 5.451148253838837,
+      "grad_norm": 8.625,
+      "learning_rate": 1.593032341350727e-05,
+      "loss": 0.8841,
+      "step": 80230
+    },
+    {
+      "epoch": 5.451487973909499,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.5928200163065634e-05,
+      "loss": 0.9237,
+      "step": 80235
+    },
+    {
+      "epoch": 5.45182769398016,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5926076912623998e-05,
+      "loss": 0.8286,
+      "step": 80240
+    },
+    {
+      "epoch": 5.452167414050822,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5923953662182362e-05,
+      "loss": 0.8028,
+      "step": 80245
+    },
+    {
+      "epoch": 5.452507134121484,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5921830411740726e-05,
+      "loss": 0.8749,
+      "step": 80250
+    },
+    {
+      "epoch": 5.4528468541921455,
+      "grad_norm": 8.125,
+      "learning_rate": 1.591970716129909e-05,
+      "loss": 0.9431,
+      "step": 80255
+    },
+    {
+      "epoch": 5.453186574262808,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.5917583910857454e-05,
+      "loss": 0.8126,
+      "step": 80260
+    },
+    {
+      "epoch": 5.45352629433347,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.5915460660415818e-05,
+      "loss": 0.9312,
+      "step": 80265
+    },
+    {
+      "epoch": 5.453866014404131,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5913337409974182e-05,
+      "loss": 0.9929,
+      "step": 80270
+    },
+    {
+      "epoch": 5.454205734474793,
+      "grad_norm": 9.125,
+      "learning_rate": 1.5911214159532546e-05,
+      "loss": 0.8594,
+      "step": 80275
+    },
+    {
+      "epoch": 5.454545454545454,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.590909090909091e-05,
+      "loss": 0.8312,
+      "step": 80280
+    },
+    {
+      "epoch": 5.454885174616116,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5906967658649274e-05,
+      "loss": 0.8782,
+      "step": 80285
+    },
+    {
+      "epoch": 5.455224894686778,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5904844408207635e-05,
+      "loss": 0.77,
+      "step": 80290
+    },
+    {
+      "epoch": 5.4555646147574395,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5902721157766002e-05,
+      "loss": 0.8895,
+      "step": 80295
+    },
+    {
+      "epoch": 5.4559043348281016,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5900597907324366e-05,
+      "loss": 0.8824,
+      "step": 80300
+    },
+    {
+      "epoch": 5.456244054898764,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5898474656882727e-05,
+      "loss": 0.8385,
+      "step": 80305
+    },
+    {
+      "epoch": 5.456583774969425,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5896351406441094e-05,
+      "loss": 0.8742,
+      "step": 80310
+    },
+    {
+      "epoch": 5.456923495040087,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5894228155999458e-05,
+      "loss": 0.7763,
+      "step": 80315
+    },
+    {
+      "epoch": 5.457263215110749,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5892104905557822e-05,
+      "loss": 0.8273,
+      "step": 80320
+    },
+    {
+      "epoch": 5.45760293518141,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5889981655116186e-05,
+      "loss": 0.8622,
+      "step": 80325
+    },
+    {
+      "epoch": 5.457942655252072,
+      "grad_norm": 8.25,
+      "learning_rate": 1.588785840467455e-05,
+      "loss": 0.8878,
+      "step": 80330
+    },
+    {
+      "epoch": 5.458282375322734,
+      "grad_norm": 7.25,
+      "learning_rate": 1.5885735154232914e-05,
+      "loss": 0.8906,
+      "step": 80335
+    },
+    {
+      "epoch": 5.4586220953933955,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.5883611903791275e-05,
+      "loss": 0.8177,
+      "step": 80340
+    },
+    {
+      "epoch": 5.458961815464058,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.5881488653349642e-05,
+      "loss": 0.8442,
+      "step": 80345
+    },
+    {
+      "epoch": 5.45930153553472,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5879365402908006e-05,
+      "loss": 0.8859,
+      "step": 80350
+    },
+    {
+      "epoch": 5.459641255605381,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5877242152466367e-05,
+      "loss": 0.8828,
+      "step": 80355
+    },
+    {
+      "epoch": 5.459980975676043,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5875118902024734e-05,
+      "loss": 0.9051,
+      "step": 80360
+    },
+    {
+      "epoch": 5.460320695746705,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5872995651583098e-05,
+      "loss": 0.912,
+      "step": 80365
+    },
+    {
+      "epoch": 5.460660415817366,
+      "grad_norm": 7.25,
+      "learning_rate": 1.587087240114146e-05,
+      "loss": 0.913,
+      "step": 80370
+    },
+    {
+      "epoch": 5.461000135888028,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.5868749150699823e-05,
+      "loss": 0.8469,
+      "step": 80375
+    },
+    {
+      "epoch": 5.46133985595869,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.586662590025819e-05,
+      "loss": 0.8635,
+      "step": 80380
+    },
+    {
+      "epoch": 5.4616795760293515,
+      "grad_norm": 7.25,
+      "learning_rate": 1.586450264981655e-05,
+      "loss": 0.8638,
+      "step": 80385
+    },
+    {
+      "epoch": 5.462019296100014,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.5862379399374915e-05,
+      "loss": 0.8712,
+      "step": 80390
+    },
+    {
+      "epoch": 5.462359016170676,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5860256148933282e-05,
+      "loss": 0.8941,
+      "step": 80395
+    },
+    {
+      "epoch": 5.462698736241337,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5858132898491643e-05,
+      "loss": 0.8734,
+      "step": 80400
+    },
+    {
+      "epoch": 5.463038456311999,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5856009648050007e-05,
+      "loss": 0.9456,
+      "step": 80405
+    },
+    {
+      "epoch": 5.463378176382661,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.585388639760837e-05,
+      "loss": 0.9671,
+      "step": 80410
+    },
+    {
+      "epoch": 5.463717896453322,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5851763147166735e-05,
+      "loss": 0.8688,
+      "step": 80415
+    },
+    {
+      "epoch": 5.464057616523984,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.58496398967251e-05,
+      "loss": 0.8054,
+      "step": 80420
+    },
+    {
+      "epoch": 5.464397336594646,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5847516646283463e-05,
+      "loss": 0.7561,
+      "step": 80425
+    },
+    {
+      "epoch": 5.4647370566653075,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5845393395841827e-05,
+      "loss": 0.8559,
+      "step": 80430
+    },
+    {
+      "epoch": 5.46507677673597,
+      "grad_norm": 8.125,
+      "learning_rate": 1.584327014540019e-05,
+      "loss": 0.8458,
+      "step": 80435
+    },
+    {
+      "epoch": 5.465416496806632,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5841146894958555e-05,
+      "loss": 0.8879,
+      "step": 80440
+    },
+    {
+      "epoch": 5.465756216877293,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.583902364451692e-05,
+      "loss": 0.8283,
+      "step": 80445
+    },
+    {
+      "epoch": 5.466095936947955,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5836900394075283e-05,
+      "loss": 0.8916,
+      "step": 80450
+    },
+    {
+      "epoch": 5.466435657018617,
+      "grad_norm": 5.375,
+      "learning_rate": 1.5834777143633647e-05,
+      "loss": 0.912,
+      "step": 80455
+    },
+    {
+      "epoch": 5.466775377089278,
+      "grad_norm": 8.25,
+      "learning_rate": 1.583265389319201e-05,
+      "loss": 0.8882,
+      "step": 80460
+    },
+    {
+      "epoch": 5.46711509715994,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5830530642750375e-05,
+      "loss": 0.8541,
+      "step": 80465
+    },
+    {
+      "epoch": 5.467454817230602,
+      "grad_norm": 7.875,
+      "learning_rate": 1.582840739230874e-05,
+      "loss": 0.7917,
+      "step": 80470
+    },
+    {
+      "epoch": 5.4677945373012635,
+      "grad_norm": 7.875,
+      "learning_rate": 1.58262841418671e-05,
+      "loss": 0.8639,
+      "step": 80475
+    },
+    {
+      "epoch": 5.468134257371926,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.5824160891425467e-05,
+      "loss": 0.8862,
+      "step": 80480
+    },
+    {
+      "epoch": 5.468473977442588,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.582203764098383e-05,
+      "loss": 0.8249,
+      "step": 80485
+    },
+    {
+      "epoch": 5.468813697513249,
+      "grad_norm": 7.375,
+      "learning_rate": 1.5819914390542195e-05,
+      "loss": 0.7412,
+      "step": 80490
+    },
+    {
+      "epoch": 5.469153417583911,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.581779114010056e-05,
+      "loss": 0.7598,
+      "step": 80495
+    },
+    {
+      "epoch": 5.469493137654573,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5815667889658923e-05,
+      "loss": 0.8795,
+      "step": 80500
+    },
+    {
+      "epoch": 5.469832857725234,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5813544639217287e-05,
+      "loss": 0.918,
+      "step": 80505
+    },
+    {
+      "epoch": 5.470172577795896,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.5811421388775648e-05,
+      "loss": 0.8615,
+      "step": 80510
+    },
+    {
+      "epoch": 5.470512297866558,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.5809298138334015e-05,
+      "loss": 0.7808,
+      "step": 80515
+    },
+    {
+      "epoch": 5.4708520179372195,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.580717488789238e-05,
+      "loss": 0.9165,
+      "step": 80520
+    },
+    {
+      "epoch": 5.471191738007882,
+      "grad_norm": 8.375,
+      "learning_rate": 1.580505163745074e-05,
+      "loss": 0.9287,
+      "step": 80525
+    },
+    {
+      "epoch": 5.471531458078544,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5802928387009107e-05,
+      "loss": 0.9293,
+      "step": 80530
+    },
+    {
+      "epoch": 5.471871178149205,
+      "grad_norm": 8.0,
+      "learning_rate": 1.580080513656747e-05,
+      "loss": 0.7943,
+      "step": 80535
+    },
+    {
+      "epoch": 5.472210898219867,
+      "grad_norm": 7.25,
+      "learning_rate": 1.579868188612583e-05,
+      "loss": 0.7819,
+      "step": 80540
+    },
+    {
+      "epoch": 5.472550618290528,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.5796558635684196e-05,
+      "loss": 0.839,
+      "step": 80545
+    },
+    {
+      "epoch": 5.47289033836119,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5794435385242563e-05,
+      "loss": 0.8458,
+      "step": 80550
+    },
+    {
+      "epoch": 5.473230058431852,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5792312134800924e-05,
+      "loss": 0.8256,
+      "step": 80555
+    },
+    {
+      "epoch": 5.4735697785025135,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5790188884359288e-05,
+      "loss": 0.8239,
+      "step": 80560
+    },
+    {
+      "epoch": 5.4739094985731755,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5788065633917655e-05,
+      "loss": 0.7857,
+      "step": 80565
+    },
+    {
+      "epoch": 5.474249218643838,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5785942383476016e-05,
+      "loss": 0.843,
+      "step": 80570
+    },
+    {
+      "epoch": 5.474588938714499,
+      "grad_norm": 6.875,
+      "learning_rate": 1.578381913303438e-05,
+      "loss": 0.8795,
+      "step": 80575
+    },
+    {
+      "epoch": 5.474928658785161,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.5781695882592744e-05,
+      "loss": 0.9006,
+      "step": 80580
+    },
+    {
+      "epoch": 5.475268378855823,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5779572632151108e-05,
+      "loss": 0.8427,
+      "step": 80585
+    },
+    {
+      "epoch": 5.475608098926484,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.577744938170947e-05,
+      "loss": 0.9075,
+      "step": 80590
+    },
+    {
+      "epoch": 5.475947818997146,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5775326131267836e-05,
+      "loss": 0.9325,
+      "step": 80595
+    },
+    {
+      "epoch": 5.476287539067808,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.57732028808262e-05,
+      "loss": 0.9055,
+      "step": 80600
+    },
+    {
+      "epoch": 5.4766272591384695,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.5771079630384564e-05,
+      "loss": 0.7737,
+      "step": 80605
+    },
+    {
+      "epoch": 5.476966979209132,
+      "grad_norm": 8.875,
+      "learning_rate": 1.5768956379942928e-05,
+      "loss": 0.8219,
+      "step": 80610
+    },
+    {
+      "epoch": 5.477306699279794,
+      "grad_norm": 8.125,
+      "learning_rate": 1.576683312950129e-05,
+      "loss": 0.8283,
+      "step": 80615
+    },
+    {
+      "epoch": 5.477646419350455,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5764709879059656e-05,
+      "loss": 0.8305,
+      "step": 80620
+    },
+    {
+      "epoch": 5.477986139421117,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.576258662861802e-05,
+      "loss": 0.8763,
+      "step": 80625
+    },
+    {
+      "epoch": 5.478325859491779,
+      "grad_norm": 7.25,
+      "learning_rate": 1.5760463378176384e-05,
+      "loss": 0.8017,
+      "step": 80630
+    },
+    {
+      "epoch": 5.47866557956244,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5758340127734748e-05,
+      "loss": 0.9224,
+      "step": 80635
+    },
+    {
+      "epoch": 5.479005299633102,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.575621687729311e-05,
+      "loss": 0.8964,
+      "step": 80640
+    },
+    {
+      "epoch": 5.479345019703764,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5754093626851472e-05,
+      "loss": 0.8653,
+      "step": 80645
+    },
+    {
+      "epoch": 5.4796847397744255,
+      "grad_norm": 8.0,
+      "learning_rate": 1.575197037640984e-05,
+      "loss": 0.8488,
+      "step": 80650
+    },
+    {
+      "epoch": 5.480024459845088,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5749847125968204e-05,
+      "loss": 1.0159,
+      "step": 80655
+    },
+    {
+      "epoch": 5.48036417991575,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.5747723875526568e-05,
+      "loss": 0.7516,
+      "step": 80660
+    },
+    {
+      "epoch": 5.480703899986411,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.574560062508493e-05,
+      "loss": 0.8597,
+      "step": 80665
+    },
+    {
+      "epoch": 5.481043620057073,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.5743477374643296e-05,
+      "loss": 0.9107,
+      "step": 80670
+    },
+    {
+      "epoch": 5.481383340127735,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.574135412420166e-05,
+      "loss": 0.7751,
+      "step": 80675
+    },
+    {
+      "epoch": 5.481723060198396,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.573923087376002e-05,
+      "loss": 0.8507,
+      "step": 80680
+    },
+    {
+      "epoch": 5.482062780269058,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5737107623318388e-05,
+      "loss": 0.8417,
+      "step": 80685
+    },
+    {
+      "epoch": 5.48240250033972,
+      "grad_norm": 7.375,
+      "learning_rate": 1.5734984372876752e-05,
+      "loss": 0.8378,
+      "step": 80690
+    },
+    {
+      "epoch": 5.4827422204103815,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5732861122435112e-05,
+      "loss": 0.8805,
+      "step": 80695
+    },
+    {
+      "epoch": 5.483081940481044,
+      "grad_norm": 7.875,
+      "learning_rate": 1.573073787199348e-05,
+      "loss": 0.811,
+      "step": 80700
+    },
+    {
+      "epoch": 5.483421660551706,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5728614621551844e-05,
+      "loss": 0.8235,
+      "step": 80705
+    },
+    {
+      "epoch": 5.483761380622367,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.5726491371110204e-05,
+      "loss": 0.8829,
+      "step": 80710
+    },
+    {
+      "epoch": 5.484101100693029,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.572436812066857e-05,
+      "loss": 0.8482,
+      "step": 80715
+    },
+    {
+      "epoch": 5.484440820763691,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5722244870226936e-05,
+      "loss": 0.8486,
+      "step": 80720
+    },
+    {
+      "epoch": 5.484780540834352,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5720121619785296e-05,
+      "loss": 0.843,
+      "step": 80725
+    },
+    {
+      "epoch": 5.485120260905014,
+      "grad_norm": 8.375,
+      "learning_rate": 1.571799836934366e-05,
+      "loss": 0.9027,
+      "step": 80730
+    },
+    {
+      "epoch": 5.485459980975676,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5715875118902028e-05,
+      "loss": 0.8322,
+      "step": 80735
+    },
+    {
+      "epoch": 5.4857997010463375,
+      "grad_norm": 8.5,
+      "learning_rate": 1.571375186846039e-05,
+      "loss": 0.7751,
+      "step": 80740
+    },
+    {
+      "epoch": 5.486139421117,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5711628618018752e-05,
+      "loss": 0.8114,
+      "step": 80745
+    },
+    {
+      "epoch": 5.486479141187662,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.570950536757712e-05,
+      "loss": 0.8888,
+      "step": 80750
+    },
+    {
+      "epoch": 5.486818861258323,
+      "grad_norm": 8.125,
+      "learning_rate": 1.570738211713548e-05,
+      "loss": 0.948,
+      "step": 80755
+    },
+    {
+      "epoch": 5.487158581328985,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5705258866693844e-05,
+      "loss": 0.8532,
+      "step": 80760
+    },
+    {
+      "epoch": 5.487498301399647,
+      "grad_norm": 8.375,
+      "learning_rate": 1.570313561625221e-05,
+      "loss": 0.7937,
+      "step": 80765
+    },
+    {
+      "epoch": 5.487838021470308,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5701012365810572e-05,
+      "loss": 0.852,
+      "step": 80770
+    },
+    {
+      "epoch": 5.48817774154097,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.5698889115368936e-05,
+      "loss": 0.8233,
+      "step": 80775
+    },
+    {
+      "epoch": 5.488517461611632,
+      "grad_norm": 8.5,
+      "learning_rate": 1.56967658649273e-05,
+      "loss": 0.9094,
+      "step": 80780
+    },
+    {
+      "epoch": 5.4888571816822935,
+      "grad_norm": 7.25,
+      "learning_rate": 1.5694642614485664e-05,
+      "loss": 0.8079,
+      "step": 80785
+    },
+    {
+      "epoch": 5.489196901752956,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.569251936404403e-05,
+      "loss": 0.8965,
+      "step": 80790
+    },
+    {
+      "epoch": 5.489536621823618,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.5690396113602392e-05,
+      "loss": 0.9735,
+      "step": 80795
+    },
+    {
+      "epoch": 5.489876341894279,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5688272863160756e-05,
+      "loss": 0.8413,
+      "step": 80800
+    },
+    {
+      "epoch": 5.490216061964941,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.568614961271912e-05,
+      "loss": 0.9067,
+      "step": 80805
+    },
+    {
+      "epoch": 5.490555782035603,
+      "grad_norm": 9.8125,
+      "learning_rate": 1.5684026362277484e-05,
+      "loss": 0.8985,
+      "step": 80810
+    },
+    {
+      "epoch": 5.490895502106264,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5681903111835845e-05,
+      "loss": 0.8862,
+      "step": 80815
+    },
+    {
+      "epoch": 5.491235222176926,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5679779861394212e-05,
+      "loss": 0.7778,
+      "step": 80820
+    },
+    {
+      "epoch": 5.491574942247588,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5677656610952576e-05,
+      "loss": 0.907,
+      "step": 80825
+    },
+    {
+      "epoch": 5.4919146623182495,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.567553336051094e-05,
+      "loss": 0.8659,
+      "step": 80830
+    },
+    {
+      "epoch": 5.492254382388912,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.5673410110069304e-05,
+      "loss": 0.7738,
+      "step": 80835
+    },
+    {
+      "epoch": 5.492594102459574,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.567128685962767e-05,
+      "loss": 0.9288,
+      "step": 80840
+    },
+    {
+      "epoch": 5.492933822530235,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5669163609186032e-05,
+      "loss": 0.8074,
+      "step": 80845
+    },
+    {
+      "epoch": 5.493273542600897,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5667040358744396e-05,
+      "loss": 0.8335,
+      "step": 80850
+    },
+    {
+      "epoch": 5.493613262671559,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.566491710830276e-05,
+      "loss": 0.7962,
+      "step": 80855
+    },
+    {
+      "epoch": 5.49395298274222,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5662793857861124e-05,
+      "loss": 0.9258,
+      "step": 80860
+    },
+    {
+      "epoch": 5.494292702812882,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.5660670607419485e-05,
+      "loss": 0.8197,
+      "step": 80865
+    },
+    {
+      "epoch": 5.494632422883544,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5658547356977852e-05,
+      "loss": 0.8408,
+      "step": 80870
+    },
+    {
+      "epoch": 5.4949721429542056,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5656424106536216e-05,
+      "loss": 0.9055,
+      "step": 80875
+    },
+    {
+      "epoch": 5.495311863024868,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5654300856094577e-05,
+      "loss": 0.829,
+      "step": 80880
+    },
+    {
+      "epoch": 5.49565158309553,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.5652177605652944e-05,
+      "loss": 0.8466,
+      "step": 80885
+    },
+    {
+      "epoch": 5.495991303166191,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.565005435521131e-05,
+      "loss": 0.8953,
+      "step": 80890
+    },
+    {
+      "epoch": 5.496331023236853,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.564793110476967e-05,
+      "loss": 0.8704,
+      "step": 80895
+    },
+    {
+      "epoch": 5.496670743307515,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5645807854328033e-05,
+      "loss": 0.8511,
+      "step": 80900
+    },
+    {
+      "epoch": 5.497010463378176,
+      "grad_norm": 8.75,
+      "learning_rate": 1.56436846038864e-05,
+      "loss": 0.9374,
+      "step": 80905
+    },
+    {
+      "epoch": 5.497350183448838,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.564156135344476e-05,
+      "loss": 0.813,
+      "step": 80910
+    },
+    {
+      "epoch": 5.4976899035195,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5639438103003125e-05,
+      "loss": 0.8917,
+      "step": 80915
+    },
+    {
+      "epoch": 5.498029623590162,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5637314852561492e-05,
+      "loss": 0.8122,
+      "step": 80920
+    },
+    {
+      "epoch": 5.498369343660824,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5635191602119853e-05,
+      "loss": 0.7887,
+      "step": 80925
+    },
+    {
+      "epoch": 5.498709063731486,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5633068351678217e-05,
+      "loss": 0.9145,
+      "step": 80930
+    },
+    {
+      "epoch": 5.499048783802147,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.563094510123658e-05,
+      "loss": 0.8495,
+      "step": 80935
+    },
+    {
+      "epoch": 5.499388503872809,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5628821850794945e-05,
+      "loss": 0.9393,
+      "step": 80940
+    },
+    {
+      "epoch": 5.499728223943471,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.562669860035331e-05,
+      "loss": 0.8742,
+      "step": 80945
+    },
+    {
+      "epoch": 5.500067944014132,
+      "grad_norm": 9.0,
+      "learning_rate": 1.5624575349911673e-05,
+      "loss": 0.7539,
+      "step": 80950
+    },
+    {
+      "epoch": 5.500407664084794,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.5622452099470037e-05,
+      "loss": 0.9224,
+      "step": 80955
+    },
+    {
+      "epoch": 5.500747384155456,
+      "grad_norm": 7.75,
+      "learning_rate": 1.56203288490284e-05,
+      "loss": 0.898,
+      "step": 80960
+    },
+    {
+      "epoch": 5.501087104226118,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5618205598586765e-05,
+      "loss": 0.8038,
+      "step": 80965
+    },
+    {
+      "epoch": 5.50142682429678,
+      "grad_norm": 8.125,
+      "learning_rate": 1.561608234814513e-05,
+      "loss": 1.0163,
+      "step": 80970
+    },
+    {
+      "epoch": 5.501766544367442,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5613959097703493e-05,
+      "loss": 0.8127,
+      "step": 80975
+    },
+    {
+      "epoch": 5.502106264438103,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5611835847261857e-05,
+      "loss": 0.8576,
+      "step": 80980
+    },
+    {
+      "epoch": 5.502445984508765,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.560971259682022e-05,
+      "loss": 0.8724,
+      "step": 80985
+    },
+    {
+      "epoch": 5.502785704579426,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5607589346378585e-05,
+      "loss": 0.8261,
+      "step": 80990
+    },
+    {
+      "epoch": 5.503125424650088,
+      "grad_norm": 8.0,
+      "learning_rate": 1.560546609593695e-05,
+      "loss": 0.8891,
+      "step": 80995
+    },
+    {
+      "epoch": 5.50346514472075,
+      "grad_norm": 8.5,
+      "learning_rate": 1.560334284549531e-05,
+      "loss": 0.9502,
+      "step": 81000
+    },
+    {
+      "epoch": 5.5038048647914115,
+      "grad_norm": 8.75,
+      "learning_rate": 1.5601219595053677e-05,
+      "loss": 0.8832,
+      "step": 81005
+    },
+    {
+      "epoch": 5.504144584862074,
+      "grad_norm": 7.875,
+      "learning_rate": 1.559909634461204e-05,
+      "loss": 0.8153,
+      "step": 81010
+    },
+    {
+      "epoch": 5.504484304932736,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5596973094170405e-05,
+      "loss": 0.8717,
+      "step": 81015
+    },
+    {
+      "epoch": 5.504824025003397,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.559484984372877e-05,
+      "loss": 0.8598,
+      "step": 81020
+    },
+    {
+      "epoch": 5.505163745074059,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5592726593287133e-05,
+      "loss": 0.8586,
+      "step": 81025
+    },
+    {
+      "epoch": 5.505503465144721,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5590603342845497e-05,
+      "loss": 0.8736,
+      "step": 81030
+    },
+    {
+      "epoch": 5.505843185215382,
+      "grad_norm": 9.5,
+      "learning_rate": 1.5588480092403858e-05,
+      "loss": 0.8765,
+      "step": 81035
+    },
+    {
+      "epoch": 5.506182905286044,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5586356841962225e-05,
+      "loss": 0.8357,
+      "step": 81040
+    },
+    {
+      "epoch": 5.506522625356706,
+      "grad_norm": 7.25,
+      "learning_rate": 1.558423359152059e-05,
+      "loss": 0.8537,
+      "step": 81045
+    },
+    {
+      "epoch": 5.5068623454273675,
+      "grad_norm": 8.25,
+      "learning_rate": 1.558211034107895e-05,
+      "loss": 0.9771,
+      "step": 81050
+    },
+    {
+      "epoch": 5.50720206549803,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5579987090637317e-05,
+      "loss": 0.7784,
+      "step": 81055
+    },
+    {
+      "epoch": 5.507541785568692,
+      "grad_norm": 7.75,
+      "learning_rate": 1.557786384019568e-05,
+      "loss": 0.8423,
+      "step": 81060
+    },
+    {
+      "epoch": 5.507881505639353,
+      "grad_norm": 6.625,
+      "learning_rate": 1.5575740589754042e-05,
+      "loss": 0.7803,
+      "step": 81065
+    },
+    {
+      "epoch": 5.508221225710015,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5573617339312406e-05,
+      "loss": 0.9312,
+      "step": 81070
+    },
+    {
+      "epoch": 5.508560945780677,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5571494088870773e-05,
+      "loss": 0.7682,
+      "step": 81075
+    },
+    {
+      "epoch": 5.508900665851338,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5569370838429134e-05,
+      "loss": 0.8334,
+      "step": 81080
+    },
+    {
+      "epoch": 5.509240385922,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.5567247587987498e-05,
+      "loss": 0.8525,
+      "step": 81085
+    },
+    {
+      "epoch": 5.509580105992662,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5565124337545865e-05,
+      "loss": 0.9449,
+      "step": 81090
+    },
+    {
+      "epoch": 5.5099198260633235,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5563001087104226e-05,
+      "loss": 0.8156,
+      "step": 81095
+    },
+    {
+      "epoch": 5.510259546133986,
+      "grad_norm": 9.375,
+      "learning_rate": 1.556087783666259e-05,
+      "loss": 0.9234,
+      "step": 81100
+    },
+    {
+      "epoch": 5.510599266204648,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5558754586220954e-05,
+      "loss": 0.9002,
+      "step": 81105
+    },
+    {
+      "epoch": 5.510938986275309,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.5556631335779318e-05,
+      "loss": 0.9176,
+      "step": 81110
+    },
+    {
+      "epoch": 5.511278706345971,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5554508085337682e-05,
+      "loss": 0.8462,
+      "step": 81115
+    },
+    {
+      "epoch": 5.511618426416633,
+      "grad_norm": 8.75,
+      "learning_rate": 1.5552384834896046e-05,
+      "loss": 0.8463,
+      "step": 81120
+    },
+    {
+      "epoch": 5.511958146487294,
+      "grad_norm": 7.5,
+      "learning_rate": 1.555026158445441e-05,
+      "loss": 0.8714,
+      "step": 81125
+    },
+    {
+      "epoch": 5.512297866557956,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.5548138334012774e-05,
+      "loss": 0.8851,
+      "step": 81130
+    },
+    {
+      "epoch": 5.512637586628618,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5546015083571138e-05,
+      "loss": 0.832,
+      "step": 81135
+    },
+    {
+      "epoch": 5.5129773066992795,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5543891833129502e-05,
+      "loss": 0.7485,
+      "step": 81140
+    },
+    {
+      "epoch": 5.513317026769942,
+      "grad_norm": 9.625,
+      "learning_rate": 1.5541768582687866e-05,
+      "loss": 0.8873,
+      "step": 81145
+    },
+    {
+      "epoch": 5.513656746840604,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.553964533224623e-05,
+      "loss": 0.8301,
+      "step": 81150
+    },
+    {
+      "epoch": 5.513996466911265,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.5537522081804594e-05,
+      "loss": 0.8532,
+      "step": 81155
+    },
+    {
+      "epoch": 5.514336186981927,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5535398831362958e-05,
+      "loss": 0.9971,
+      "step": 81160
+    },
+    {
+      "epoch": 5.514675907052589,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5533275580921322e-05,
+      "loss": 0.8605,
+      "step": 81165
+    },
+    {
+      "epoch": 5.51501562712325,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5531152330479683e-05,
+      "loss": 0.8395,
+      "step": 81170
+    },
+    {
+      "epoch": 5.515355347193912,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.552902908003805e-05,
+      "loss": 0.8662,
+      "step": 81175
+    },
+    {
+      "epoch": 5.515695067264574,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5526905829596414e-05,
+      "loss": 0.9029,
+      "step": 81180
+    },
+    {
+      "epoch": 5.516034787335236,
+      "grad_norm": 8.875,
+      "learning_rate": 1.5524782579154778e-05,
+      "loss": 0.8634,
+      "step": 81185
+    },
+    {
+      "epoch": 5.516374507405898,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5522659328713142e-05,
+      "loss": 0.8919,
+      "step": 81190
+    },
+    {
+      "epoch": 5.516714227476559,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.5520536078271506e-05,
+      "loss": 0.8194,
+      "step": 81195
+    },
+    {
+      "epoch": 5.517053947547221,
+      "grad_norm": 8.625,
+      "learning_rate": 1.551841282782987e-05,
+      "loss": 0.9131,
+      "step": 81200
+    },
+    {
+      "epoch": 5.517393667617883,
+      "grad_norm": 7.25,
+      "learning_rate": 1.551628957738823e-05,
+      "loss": 0.8244,
+      "step": 81205
+    },
+    {
+      "epoch": 5.517733387688544,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.5514166326946598e-05,
+      "loss": 0.8727,
+      "step": 81210
+    },
+    {
+      "epoch": 5.518073107759206,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.5512043076504962e-05,
+      "loss": 0.8793,
+      "step": 81215
+    },
+    {
+      "epoch": 5.518412827829868,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.5509919826063323e-05,
+      "loss": 0.8422,
+      "step": 81220
+    },
+    {
+      "epoch": 5.5187525479005295,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.550779657562169e-05,
+      "loss": 0.8293,
+      "step": 81225
+    },
+    {
+      "epoch": 5.519092267971192,
+      "grad_norm": 9.0,
+      "learning_rate": 1.5505673325180054e-05,
+      "loss": 0.8853,
+      "step": 81230
+    },
+    {
+      "epoch": 5.519431988041854,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.5503550074738415e-05,
+      "loss": 0.8408,
+      "step": 81235
+    },
+    {
+      "epoch": 5.519771708112515,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5501426824296782e-05,
+      "loss": 0.8608,
+      "step": 81240
+    },
+    {
+      "epoch": 5.520111428183177,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5499303573855146e-05,
+      "loss": 0.9297,
+      "step": 81245
+    },
+    {
+      "epoch": 5.520451148253839,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5497180323413507e-05,
+      "loss": 0.8551,
+      "step": 81250
+    },
+    {
+      "epoch": 5.5207908683245,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.549505707297187e-05,
+      "loss": 0.7803,
+      "step": 81255
+    },
+    {
+      "epoch": 5.521130588395162,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5492933822530238e-05,
+      "loss": 0.9757,
+      "step": 81260
+    },
+    {
+      "epoch": 5.521470308465824,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.54908105720886e-05,
+      "loss": 0.8008,
+      "step": 81265
+    },
+    {
+      "epoch": 5.5218100285364855,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5488687321646963e-05,
+      "loss": 0.8748,
+      "step": 81270
+    },
+    {
+      "epoch": 5.522149748607148,
+      "grad_norm": 9.0,
+      "learning_rate": 1.548656407120533e-05,
+      "loss": 0.9103,
+      "step": 81275
+    },
+    {
+      "epoch": 5.52248946867781,
+      "grad_norm": 8.375,
+      "learning_rate": 1.548444082076369e-05,
+      "loss": 0.9087,
+      "step": 81280
+    },
+    {
+      "epoch": 5.522829188748471,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.5482317570322055e-05,
+      "loss": 0.8063,
+      "step": 81285
+    },
+    {
+      "epoch": 5.523168908819133,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.548019431988042e-05,
+      "loss": 0.831,
+      "step": 81290
+    },
+    {
+      "epoch": 5.523508628889795,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5478071069438783e-05,
+      "loss": 0.8851,
+      "step": 81295
+    },
+    {
+      "epoch": 5.523848348960456,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5475947818997147e-05,
+      "loss": 0.8366,
+      "step": 81300
+    },
+    {
+      "epoch": 5.524188069031118,
+      "grad_norm": 6.1875,
+      "learning_rate": 1.547382456855551e-05,
+      "loss": 0.8547,
+      "step": 81305
+    },
+    {
+      "epoch": 5.52452778910178,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.5471701318113875e-05,
+      "loss": 0.8779,
+      "step": 81310
+    },
+    {
+      "epoch": 5.5248675091724415,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.546957806767224e-05,
+      "loss": 0.8824,
+      "step": 81315
+    },
+    {
+      "epoch": 5.525207229243104,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5467454817230603e-05,
+      "loss": 0.753,
+      "step": 81320
+    },
+    {
+      "epoch": 5.525546949313766,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5465331566788967e-05,
+      "loss": 0.8321,
+      "step": 81325
+    },
+    {
+      "epoch": 5.525886669384427,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.546320831634733e-05,
+      "loss": 0.8492,
+      "step": 81330
+    },
+    {
+      "epoch": 5.526226389455089,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5461085065905695e-05,
+      "loss": 0.8548,
+      "step": 81335
+    },
+    {
+      "epoch": 5.526566109525751,
+      "grad_norm": 8.25,
+      "learning_rate": 1.545896181546406e-05,
+      "loss": 0.886,
+      "step": 81340
+    },
+    {
+      "epoch": 5.526905829596412,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.5456838565022423e-05,
+      "loss": 0.8337,
+      "step": 81345
+    },
+    {
+      "epoch": 5.527245549667074,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5454715314580787e-05,
+      "loss": 0.8068,
+      "step": 81350
+    },
+    {
+      "epoch": 5.527585269737736,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.545259206413915e-05,
+      "loss": 0.8572,
+      "step": 81355
+    },
+    {
+      "epoch": 5.5279249898083975,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5450468813697515e-05,
+      "loss": 0.7479,
+      "step": 81360
+    },
+    {
+      "epoch": 5.52826470987906,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.544834556325588e-05,
+      "loss": 0.8316,
+      "step": 81365
+    },
+    {
+      "epoch": 5.528604429949722,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5446222312814243e-05,
+      "loss": 0.8482,
+      "step": 81370
+    },
+    {
+      "epoch": 5.528944150020383,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.5444099062372607e-05,
+      "loss": 0.9149,
+      "step": 81375
+    },
+    {
+      "epoch": 5.529283870091045,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.544197581193097e-05,
+      "loss": 0.8009,
+      "step": 81380
+    },
+    {
+      "epoch": 5.529623590161707,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5439852561489335e-05,
+      "loss": 0.8419,
+      "step": 81385
+    },
+    {
+      "epoch": 5.529963310232368,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.5437729311047695e-05,
+      "loss": 0.9386,
+      "step": 81390
+    },
+    {
+      "epoch": 5.53030303030303,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5435606060606063e-05,
+      "loss": 0.7697,
+      "step": 81395
+    },
+    {
+      "epoch": 5.530642750373692,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.5433482810164427e-05,
+      "loss": 0.8328,
+      "step": 81400
+    },
+    {
+      "epoch": 5.5309824704443535,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.5431359559722787e-05,
+      "loss": 0.8588,
+      "step": 81405
+    },
+    {
+      "epoch": 5.531322190515016,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.5429236309281155e-05,
+      "loss": 0.7545,
+      "step": 81410
+    },
+    {
+      "epoch": 5.531661910585678,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.542711305883952e-05,
+      "loss": 0.9403,
+      "step": 81415
+    },
+    {
+      "epoch": 5.532001630656339,
+      "grad_norm": 9.125,
+      "learning_rate": 1.542498980839788e-05,
+      "loss": 0.8161,
+      "step": 81420
+    },
+    {
+      "epoch": 5.532341350727001,
+      "grad_norm": 7.625,
+      "learning_rate": 1.5422866557956243e-05,
+      "loss": 0.7838,
+      "step": 81425
+    },
+    {
+      "epoch": 5.532681070797663,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.542074330751461e-05,
+      "loss": 0.809,
+      "step": 81430
+    },
+    {
+      "epoch": 5.533020790868324,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.541862005707297e-05,
+      "loss": 0.8448,
+      "step": 81435
+    },
+    {
+      "epoch": 5.533360510938986,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5416496806631335e-05,
+      "loss": 0.9239,
+      "step": 81440
+    },
+    {
+      "epoch": 5.533700231009648,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5414373556189703e-05,
+      "loss": 0.8504,
+      "step": 81445
+    },
+    {
+      "epoch": 5.5340399510803095,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.5412250305748063e-05,
+      "loss": 0.7633,
+      "step": 81450
+    },
+    {
+      "epoch": 5.534379671150972,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5410127055306427e-05,
+      "loss": 0.8169,
+      "step": 81455
+    },
+    {
+      "epoch": 5.534719391221634,
+      "grad_norm": 8.25,
+      "learning_rate": 1.540800380486479e-05,
+      "loss": 0.8828,
+      "step": 81460
+    },
+    {
+      "epoch": 5.535059111292295,
+      "grad_norm": 8.875,
+      "learning_rate": 1.5405880554423155e-05,
+      "loss": 0.8683,
+      "step": 81465
+    },
+    {
+      "epoch": 5.535398831362957,
+      "grad_norm": 7.75,
+      "learning_rate": 1.540375730398152e-05,
+      "loss": 0.8447,
+      "step": 81470
+    },
+    {
+      "epoch": 5.535738551433619,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5401634053539883e-05,
+      "loss": 0.9478,
+      "step": 81475
+    },
+    {
+      "epoch": 5.53607827150428,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5399510803098247e-05,
+      "loss": 0.7944,
+      "step": 81480
+    },
+    {
+      "epoch": 5.536417991574942,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.539738755265661e-05,
+      "loss": 0.8082,
+      "step": 81485
+    },
+    {
+      "epoch": 5.536757711645604,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.5395264302214975e-05,
+      "loss": 0.8164,
+      "step": 81490
+    },
+    {
+      "epoch": 5.537097431716266,
+      "grad_norm": 8.875,
+      "learning_rate": 1.539314105177334e-05,
+      "loss": 0.8814,
+      "step": 81495
+    },
+    {
+      "epoch": 5.537437151786928,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5391017801331703e-05,
+      "loss": 0.7861,
+      "step": 81500
+    },
+    {
+      "epoch": 5.53777687185759,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5388894550890067e-05,
+      "loss": 0.9826,
+      "step": 81505
+    },
+    {
+      "epoch": 5.538116591928251,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.538677130044843e-05,
+      "loss": 0.7678,
+      "step": 81510
+    },
+    {
+      "epoch": 5.538456311998913,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.5384648050006795e-05,
+      "loss": 0.8229,
+      "step": 81515
+    },
+    {
+      "epoch": 5.538796032069575,
+      "grad_norm": 8.25,
+      "learning_rate": 1.538252479956516e-05,
+      "loss": 0.8903,
+      "step": 81520
+    },
+    {
+      "epoch": 5.539135752140236,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5380401549123523e-05,
+      "loss": 0.927,
+      "step": 81525
+    },
+    {
+      "epoch": 5.539475472210898,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5378278298681887e-05,
+      "loss": 0.8532,
+      "step": 81530
+    },
+    {
+      "epoch": 5.53981519228156,
+      "grad_norm": 8.375,
+      "learning_rate": 1.537615504824025e-05,
+      "loss": 0.8723,
+      "step": 81535
+    },
+    {
+      "epoch": 5.540154912352222,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.5374031797798615e-05,
+      "loss": 0.8416,
+      "step": 81540
+    },
+    {
+      "epoch": 5.540494632422884,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.537190854735698e-05,
+      "loss": 0.8517,
+      "step": 81545
+    },
+    {
+      "epoch": 5.540834352493546,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5369785296915343e-05,
+      "loss": 0.8772,
+      "step": 81550
+    },
+    {
+      "epoch": 5.541174072564207,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.5367662046473707e-05,
+      "loss": 0.8979,
+      "step": 81555
+    },
+    {
+      "epoch": 5.541513792634869,
+      "grad_norm": 7.0,
+      "learning_rate": 1.5365538796032068e-05,
+      "loss": 0.8449,
+      "step": 81560
+    },
+    {
+      "epoch": 5.541853512705531,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5363415545590435e-05,
+      "loss": 0.7496,
+      "step": 81565
+    },
+    {
+      "epoch": 5.542193232776192,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.53612922951488e-05,
+      "loss": 0.7971,
+      "step": 81570
+    },
+    {
+      "epoch": 5.542532952846854,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.535916904470716e-05,
+      "loss": 0.7999,
+      "step": 81575
+    },
+    {
+      "epoch": 5.542872672917516,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5357045794265527e-05,
+      "loss": 0.8146,
+      "step": 81580
+    },
+    {
+      "epoch": 5.543212392988178,
+      "grad_norm": 8.25,
+      "learning_rate": 1.535492254382389e-05,
+      "loss": 0.9399,
+      "step": 81585
+    },
+    {
+      "epoch": 5.54355211305884,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5352799293382252e-05,
+      "loss": 0.819,
+      "step": 81590
+    },
+    {
+      "epoch": 5.543891833129502,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5350676042940616e-05,
+      "loss": 0.9236,
+      "step": 81595
+    },
+    {
+      "epoch": 5.544231553200163,
+      "grad_norm": 7.625,
+      "learning_rate": 1.5348552792498983e-05,
+      "loss": 0.8497,
+      "step": 81600
+    },
+    {
+      "epoch": 5.544571273270825,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5346429542057344e-05,
+      "loss": 0.8976,
+      "step": 81605
+    },
+    {
+      "epoch": 5.544910993341487,
+      "grad_norm": 7.125,
+      "learning_rate": 1.5344306291615708e-05,
+      "loss": 0.923,
+      "step": 81610
+    },
+    {
+      "epoch": 5.545250713412148,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5342183041174075e-05,
+      "loss": 0.8653,
+      "step": 81615
+    },
+    {
+      "epoch": 5.54559043348281,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5340059790732436e-05,
+      "loss": 0.9273,
+      "step": 81620
+    },
+    {
+      "epoch": 5.545930153553472,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.53379365402908e-05,
+      "loss": 0.8738,
+      "step": 81625
+    },
+    {
+      "epoch": 5.546269873624134,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.5335813289849164e-05,
+      "loss": 0.8269,
+      "step": 81630
+    },
+    {
+      "epoch": 5.546609593694796,
+      "grad_norm": 7.125,
+      "learning_rate": 1.5333690039407528e-05,
+      "loss": 0.846,
+      "step": 81635
+    },
+    {
+      "epoch": 5.546949313765458,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5331566788965892e-05,
+      "loss": 0.8264,
+      "step": 81640
+    },
+    {
+      "epoch": 5.547289033836119,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5329443538524256e-05,
+      "loss": 0.9118,
+      "step": 81645
+    },
+    {
+      "epoch": 5.547628753906781,
+      "grad_norm": 8.0,
+      "learning_rate": 1.532732028808262e-05,
+      "loss": 0.9013,
+      "step": 81650
+    },
+    {
+      "epoch": 5.547968473977443,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.5325197037640984e-05,
+      "loss": 0.7633,
+      "step": 81655
+    },
+    {
+      "epoch": 5.548308194048104,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5323073787199348e-05,
+      "loss": 0.8469,
+      "step": 81660
+    },
+    {
+      "epoch": 5.548647914118766,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.5320950536757712e-05,
+      "loss": 0.8174,
+      "step": 81665
+    },
+    {
+      "epoch": 5.5489876341894275,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.5318827286316076e-05,
+      "loss": 0.8248,
+      "step": 81670
+    },
+    {
+      "epoch": 5.54932735426009,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.531670403587444e-05,
+      "loss": 0.789,
+      "step": 81675
+    },
+    {
+      "epoch": 5.549667074330752,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5314580785432804e-05,
+      "loss": 0.8023,
+      "step": 81680
+    },
+    {
+      "epoch": 5.550006794401413,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5312457534991168e-05,
+      "loss": 0.7985,
+      "step": 81685
+    },
+    {
+      "epoch": 5.550346514472075,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5310334284549532e-05,
+      "loss": 0.8656,
+      "step": 81690
+    },
+    {
+      "epoch": 5.550686234542737,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5308211034107896e-05,
+      "loss": 0.7825,
+      "step": 81695
+    },
+    {
+      "epoch": 5.551025954613398,
+      "grad_norm": 8.5,
+      "learning_rate": 1.530608778366626e-05,
+      "loss": 0.9153,
+      "step": 81700
+    },
+    {
+      "epoch": 5.55136567468406,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5303964533224624e-05,
+      "loss": 0.8616,
+      "step": 81705
+    },
+    {
+      "epoch": 5.551705394754722,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5301841282782988e-05,
+      "loss": 0.7784,
+      "step": 81710
+    },
+    {
+      "epoch": 5.5520451148253835,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.5299718032341352e-05,
+      "loss": 0.8871,
+      "step": 81715
+    },
+    {
+      "epoch": 5.552384834896046,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5297594781899716e-05,
+      "loss": 0.8394,
+      "step": 81720
+    },
+    {
+      "epoch": 5.552724554966708,
+      "grad_norm": 8.0,
+      "learning_rate": 1.529547153145808e-05,
+      "loss": 0.7963,
+      "step": 81725
+    },
+    {
+      "epoch": 5.553064275037369,
+      "grad_norm": 8.125,
+      "learning_rate": 1.529334828101644e-05,
+      "loss": 0.8245,
+      "step": 81730
+    },
+    {
+      "epoch": 5.553403995108031,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5291225030574808e-05,
+      "loss": 0.8517,
+      "step": 81735
+    },
+    {
+      "epoch": 5.553743715178693,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5289101780133172e-05,
+      "loss": 0.8739,
+      "step": 81740
+    },
+    {
+      "epoch": 5.554083435249354,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5286978529691533e-05,
+      "loss": 0.8719,
+      "step": 81745
+    },
+    {
+      "epoch": 5.554423155320016,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.52848552792499e-05,
+      "loss": 0.8021,
+      "step": 81750
+    },
+    {
+      "epoch": 5.554762875390678,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5282732028808264e-05,
+      "loss": 0.8884,
+      "step": 81755
+    },
+    {
+      "epoch": 5.5551025954613396,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5280608778366625e-05,
+      "loss": 0.8183,
+      "step": 81760
+    },
+    {
+      "epoch": 5.555442315532002,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5278485527924992e-05,
+      "loss": 0.7501,
+      "step": 81765
+    },
+    {
+      "epoch": 5.555782035602664,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5276362277483356e-05,
+      "loss": 0.9008,
+      "step": 81770
+    },
+    {
+      "epoch": 5.556121755673325,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5274239027041717e-05,
+      "loss": 0.8381,
+      "step": 81775
+    },
+    {
+      "epoch": 5.556461475743987,
+      "grad_norm": 8.375,
+      "learning_rate": 1.527211577660008e-05,
+      "loss": 0.8332,
+      "step": 81780
+    },
+    {
+      "epoch": 5.556801195814649,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.5269992526158448e-05,
+      "loss": 0.8863,
+      "step": 81785
+    },
+    {
+      "epoch": 5.55714091588531,
+      "grad_norm": 7.25,
+      "learning_rate": 1.526786927571681e-05,
+      "loss": 0.8015,
+      "step": 81790
+    },
+    {
+      "epoch": 5.557480635955972,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.5265746025275173e-05,
+      "loss": 0.8862,
+      "step": 81795
+    },
+    {
+      "epoch": 5.557820356026634,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.526362277483354e-05,
+      "loss": 0.8596,
+      "step": 81800
+    },
+    {
+      "epoch": 5.558160076097296,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.52614995243919e-05,
+      "loss": 0.9118,
+      "step": 81805
+    },
+    {
+      "epoch": 5.558499796167958,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5259376273950265e-05,
+      "loss": 0.8774,
+      "step": 81810
+    },
+    {
+      "epoch": 5.55883951623862,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.525725302350863e-05,
+      "loss": 0.8919,
+      "step": 81815
+    },
+    {
+      "epoch": 5.559179236309281,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.5255129773066993e-05,
+      "loss": 0.8066,
+      "step": 81820
+    },
+    {
+      "epoch": 5.559518956379943,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5253006522625357e-05,
+      "loss": 0.8861,
+      "step": 81825
+    },
+    {
+      "epoch": 5.559858676450605,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.5250883272183723e-05,
+      "loss": 0.895,
+      "step": 81830
+    },
+    {
+      "epoch": 5.560198396521266,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5248760021742083e-05,
+      "loss": 0.8782,
+      "step": 81835
+    },
+    {
+      "epoch": 5.560538116591928,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5246636771300449e-05,
+      "loss": 0.8942,
+      "step": 81840
+    },
+    {
+      "epoch": 5.56087783666259,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.5244513520858813e-05,
+      "loss": 0.9238,
+      "step": 81845
+    },
+    {
+      "epoch": 5.561217556733252,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5242390270417175e-05,
+      "loss": 0.8458,
+      "step": 81850
+    },
+    {
+      "epoch": 5.561557276803914,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5240267019975541e-05,
+      "loss": 0.8049,
+      "step": 81855
+    },
+    {
+      "epoch": 5.561896996874576,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5238143769533905e-05,
+      "loss": 0.8333,
+      "step": 81860
+    },
+    {
+      "epoch": 5.562236716945237,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.523602051909227e-05,
+      "loss": 0.8072,
+      "step": 81865
+    },
+    {
+      "epoch": 5.562576437015899,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5233897268650631e-05,
+      "loss": 0.814,
+      "step": 81870
+    },
+    {
+      "epoch": 5.56291615708656,
+      "grad_norm": 9.25,
+      "learning_rate": 1.5231774018208997e-05,
+      "loss": 0.8271,
+      "step": 81875
+    },
+    {
+      "epoch": 5.563255877157222,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5229650767767361e-05,
+      "loss": 0.7645,
+      "step": 81880
+    },
+    {
+      "epoch": 5.563595597227884,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5227527517325723e-05,
+      "loss": 0.8478,
+      "step": 81885
+    },
+    {
+      "epoch": 5.5639353172985455,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5225404266884089e-05,
+      "loss": 0.8698,
+      "step": 81890
+    },
+    {
+      "epoch": 5.564275037369208,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.5223281016442453e-05,
+      "loss": 0.8113,
+      "step": 81895
+    },
+    {
+      "epoch": 5.56461475743987,
+      "grad_norm": 6.875,
+      "learning_rate": 1.5221157766000815e-05,
+      "loss": 0.8187,
+      "step": 81900
+    },
+    {
+      "epoch": 5.564954477510531,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.5219034515559181e-05,
+      "loss": 0.7935,
+      "step": 81905
+    },
+    {
+      "epoch": 5.565294197581193,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5216911265117545e-05,
+      "loss": 0.8907,
+      "step": 81910
+    },
+    {
+      "epoch": 5.565633917651855,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5214788014675907e-05,
+      "loss": 0.868,
+      "step": 81915
+    },
+    {
+      "epoch": 5.565973637722516,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5212664764234271e-05,
+      "loss": 0.8624,
+      "step": 81920
+    },
+    {
+      "epoch": 5.566313357793178,
+      "grad_norm": 9.375,
+      "learning_rate": 1.5210541513792637e-05,
+      "loss": 0.8994,
+      "step": 81925
+    },
+    {
+      "epoch": 5.56665307786384,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5208418263351e-05,
+      "loss": 0.8379,
+      "step": 81930
+    },
+    {
+      "epoch": 5.5669927979345015,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5206295012909363e-05,
+      "loss": 0.8904,
+      "step": 81935
+    },
+    {
+      "epoch": 5.567332518005164,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.5204171762467729e-05,
+      "loss": 0.9021,
+      "step": 81940
+    },
+    {
+      "epoch": 5.567672238075826,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.520204851202609e-05,
+      "loss": 0.8437,
+      "step": 81945
+    },
+    {
+      "epoch": 5.568011958146487,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5199925261584455e-05,
+      "loss": 0.8789,
+      "step": 81950
+    },
+    {
+      "epoch": 5.568351678217149,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.519780201114282e-05,
+      "loss": 0.8881,
+      "step": 81955
+    },
+    {
+      "epoch": 5.568691398287811,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5195678760701182e-05,
+      "loss": 0.8507,
+      "step": 81960
+    },
+    {
+      "epoch": 5.569031118358472,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.5193555510259547e-05,
+      "loss": 0.8428,
+      "step": 81965
+    },
+    {
+      "epoch": 5.569370838429134,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5191432259817911e-05,
+      "loss": 0.8554,
+      "step": 81970
+    },
+    {
+      "epoch": 5.569710558499796,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5189309009376274e-05,
+      "loss": 0.9469,
+      "step": 81975
+    },
+    {
+      "epoch": 5.5700502785704575,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.5187185758934638e-05,
+      "loss": 0.855,
+      "step": 81980
+    },
+    {
+      "epoch": 5.57038999864112,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5185062508493003e-05,
+      "loss": 0.881,
+      "step": 81985
+    },
+    {
+      "epoch": 5.570729718711782,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.5182939258051366e-05,
+      "loss": 0.8486,
+      "step": 81990
+    },
+    {
+      "epoch": 5.571069438782443,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.518081600760973e-05,
+      "loss": 0.778,
+      "step": 81995
+    },
+    {
+      "epoch": 5.571409158853105,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5178692757168095e-05,
+      "loss": 0.8773,
+      "step": 82000
+    },
+    {
+      "epoch": 5.571748878923767,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5176569506726458e-05,
+      "loss": 0.8078,
+      "step": 82005
+    },
+    {
+      "epoch": 5.572088598994428,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5174446256284822e-05,
+      "loss": 0.8425,
+      "step": 82010
+    },
+    {
+      "epoch": 5.57242831906509,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5172323005843186e-05,
+      "loss": 0.8342,
+      "step": 82015
+    },
+    {
+      "epoch": 5.572768039135752,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5170199755401548e-05,
+      "loss": 0.8428,
+      "step": 82020
+    },
+    {
+      "epoch": 5.5731077592064135,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5168076504959914e-05,
+      "loss": 0.9338,
+      "step": 82025
+    },
+    {
+      "epoch": 5.573447479277076,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5165953254518278e-05,
+      "loss": 0.8754,
+      "step": 82030
+    },
+    {
+      "epoch": 5.573787199347738,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5163830004076643e-05,
+      "loss": 0.9084,
+      "step": 82035
+    },
+    {
+      "epoch": 5.574126919418399,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5161706753635006e-05,
+      "loss": 0.8781,
+      "step": 82040
+    },
+    {
+      "epoch": 5.574466639489061,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.515958350319337e-05,
+      "loss": 0.86,
+      "step": 82045
+    },
+    {
+      "epoch": 5.574806359559723,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5157460252751735e-05,
+      "loss": 0.8934,
+      "step": 82050
+    },
+    {
+      "epoch": 5.575146079630384,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.5155337002310096e-05,
+      "loss": 0.8538,
+      "step": 82055
+    },
+    {
+      "epoch": 5.575485799701046,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5153213751868462e-05,
+      "loss": 0.8502,
+      "step": 82060
+    },
+    {
+      "epoch": 5.575825519771708,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5151090501426826e-05,
+      "loss": 0.8364,
+      "step": 82065
+    },
+    {
+      "epoch": 5.57616523984237,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5148967250985188e-05,
+      "loss": 0.8857,
+      "step": 82070
+    },
+    {
+      "epoch": 5.576504959913032,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5146844000543554e-05,
+      "loss": 0.8539,
+      "step": 82075
+    },
+    {
+      "epoch": 5.576844679983694,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5144720750101918e-05,
+      "loss": 0.9439,
+      "step": 82080
+    },
+    {
+      "epoch": 5.577184400054355,
+      "grad_norm": 8.75,
+      "learning_rate": 1.514259749966028e-05,
+      "loss": 0.9324,
+      "step": 82085
+    },
+    {
+      "epoch": 5.577524120125017,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5140474249218644e-05,
+      "loss": 0.903,
+      "step": 82090
+    },
+    {
+      "epoch": 5.577863840195679,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.513835099877701e-05,
+      "loss": 0.7245,
+      "step": 82095
+    },
+    {
+      "epoch": 5.57820356026634,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5136227748335372e-05,
+      "loss": 0.8787,
+      "step": 82100
+    },
+    {
+      "epoch": 5.578543280337002,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5134104497893736e-05,
+      "loss": 0.8372,
+      "step": 82105
+    },
+    {
+      "epoch": 5.578883000407664,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.5131981247452102e-05,
+      "loss": 0.8044,
+      "step": 82110
+    },
+    {
+      "epoch": 5.579222720478326,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.5129857997010462e-05,
+      "loss": 0.824,
+      "step": 82115
+    },
+    {
+      "epoch": 5.579562440548988,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5127734746568828e-05,
+      "loss": 0.8452,
+      "step": 82120
+    },
+    {
+      "epoch": 5.57990216061965,
+      "grad_norm": 7.0,
+      "learning_rate": 1.5125611496127192e-05,
+      "loss": 0.7615,
+      "step": 82125
+    },
+    {
+      "epoch": 5.580241880690311,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5123488245685554e-05,
+      "loss": 0.8365,
+      "step": 82130
+    },
+    {
+      "epoch": 5.580581600760973,
+      "grad_norm": 9.0,
+      "learning_rate": 1.512136499524392e-05,
+      "loss": 0.8478,
+      "step": 82135
+    },
+    {
+      "epoch": 5.580921320831635,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5119241744802284e-05,
+      "loss": 0.9425,
+      "step": 82140
+    },
+    {
+      "epoch": 5.581261040902296,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.5117118494360646e-05,
+      "loss": 0.8729,
+      "step": 82145
+    },
+    {
+      "epoch": 5.581600760972958,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5114995243919012e-05,
+      "loss": 0.9212,
+      "step": 82150
+    },
+    {
+      "epoch": 5.58194048104362,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5112871993477376e-05,
+      "loss": 0.8886,
+      "step": 82155
+    },
+    {
+      "epoch": 5.582280201114282,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5110748743035738e-05,
+      "loss": 0.9255,
+      "step": 82160
+    },
+    {
+      "epoch": 5.582619921184944,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.5108625492594102e-05,
+      "loss": 0.8866,
+      "step": 82165
+    },
+    {
+      "epoch": 5.582959641255606,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5106502242152468e-05,
+      "loss": 0.8351,
+      "step": 82170
+    },
+    {
+      "epoch": 5.583299361326267,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.510437899171083e-05,
+      "loss": 0.9029,
+      "step": 82175
+    },
+    {
+      "epoch": 5.583639081396929,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5102255741269194e-05,
+      "loss": 0.7778,
+      "step": 82180
+    },
+    {
+      "epoch": 5.583978801467591,
+      "grad_norm": 8.875,
+      "learning_rate": 1.510013249082756e-05,
+      "loss": 0.7387,
+      "step": 82185
+    },
+    {
+      "epoch": 5.584318521538252,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.509800924038592e-05,
+      "loss": 0.8335,
+      "step": 82190
+    },
+    {
+      "epoch": 5.584658241608914,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5095885989944286e-05,
+      "loss": 0.8001,
+      "step": 82195
+    },
+    {
+      "epoch": 5.584997961679576,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.509376273950265e-05,
+      "loss": 0.9159,
+      "step": 82200
+    },
+    {
+      "epoch": 5.585337681750238,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.5091639489061016e-05,
+      "loss": 0.8251,
+      "step": 82205
+    },
+    {
+      "epoch": 5.5856774018209,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5089516238619378e-05,
+      "loss": 0.8049,
+      "step": 82210
+    },
+    {
+      "epoch": 5.586017121891562,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5087392988177742e-05,
+      "loss": 0.9144,
+      "step": 82215
+    },
+    {
+      "epoch": 5.586356841962223,
+      "grad_norm": 7.625,
+      "learning_rate": 1.5085269737736108e-05,
+      "loss": 0.8146,
+      "step": 82220
+    },
+    {
+      "epoch": 5.586696562032885,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5083146487294469e-05,
+      "loss": 0.8323,
+      "step": 82225
+    },
+    {
+      "epoch": 5.587036282103547,
+      "grad_norm": 8.625,
+      "learning_rate": 1.5081023236852834e-05,
+      "loss": 0.8199,
+      "step": 82230
+    },
+    {
+      "epoch": 5.587376002174208,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5078899986411198e-05,
+      "loss": 0.8406,
+      "step": 82235
+    },
+    {
+      "epoch": 5.58771572224487,
+      "grad_norm": 7.875,
+      "learning_rate": 1.507677673596956e-05,
+      "loss": 0.8654,
+      "step": 82240
+    },
+    {
+      "epoch": 5.588055442315532,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.5074653485527926e-05,
+      "loss": 0.8694,
+      "step": 82245
+    },
+    {
+      "epoch": 5.588395162386194,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.507253023508629e-05,
+      "loss": 0.846,
+      "step": 82250
+    },
+    {
+      "epoch": 5.588734882456856,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5070406984644653e-05,
+      "loss": 0.9083,
+      "step": 82255
+    },
+    {
+      "epoch": 5.589074602527518,
+      "grad_norm": 7.5,
+      "learning_rate": 1.5068283734203017e-05,
+      "loss": 0.8881,
+      "step": 82260
+    },
+    {
+      "epoch": 5.589414322598179,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.5066160483761382e-05,
+      "loss": 0.7755,
+      "step": 82265
+    },
+    {
+      "epoch": 5.589754042668841,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.5064037233319745e-05,
+      "loss": 0.8685,
+      "step": 82270
+    },
+    {
+      "epoch": 5.590093762739503,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.5061913982878109e-05,
+      "loss": 0.8125,
+      "step": 82275
+    },
+    {
+      "epoch": 5.590433482810164,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.5059790732436474e-05,
+      "loss": 0.8437,
+      "step": 82280
+    },
+    {
+      "epoch": 5.590773202880826,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5057667481994837e-05,
+      "loss": 0.9124,
+      "step": 82285
+    },
+    {
+      "epoch": 5.591112922951488,
+      "grad_norm": 8.0,
+      "learning_rate": 1.50555442315532e-05,
+      "loss": 0.8802,
+      "step": 82290
+    },
+    {
+      "epoch": 5.59145264302215,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5053420981111566e-05,
+      "loss": 0.9209,
+      "step": 82295
+    },
+    {
+      "epoch": 5.591792363092812,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5051297730669927e-05,
+      "loss": 0.8637,
+      "step": 82300
+    },
+    {
+      "epoch": 5.592132083163474,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.5049174480228293e-05,
+      "loss": 0.9362,
+      "step": 82305
+    },
+    {
+      "epoch": 5.592471803234135,
+      "grad_norm": 7.375,
+      "learning_rate": 1.5047051229786657e-05,
+      "loss": 0.8412,
+      "step": 82310
+    },
+    {
+      "epoch": 5.592811523304797,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5044927979345019e-05,
+      "loss": 0.8212,
+      "step": 82315
+    },
+    {
+      "epoch": 5.593151243375459,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5042804728903385e-05,
+      "loss": 0.8774,
+      "step": 82320
+    },
+    {
+      "epoch": 5.59349096344612,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.5040681478461749e-05,
+      "loss": 0.7003,
+      "step": 82325
+    },
+    {
+      "epoch": 5.593830683516782,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5038558228020111e-05,
+      "loss": 0.9032,
+      "step": 82330
+    },
+    {
+      "epoch": 5.594170403587444,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5036434977578475e-05,
+      "loss": 0.8885,
+      "step": 82335
+    },
+    {
+      "epoch": 5.594510123658106,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.503431172713684e-05,
+      "loss": 0.8587,
+      "step": 82340
+    },
+    {
+      "epoch": 5.594849843728768,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.5032188476695203e-05,
+      "loss": 0.8264,
+      "step": 82345
+    },
+    {
+      "epoch": 5.595189563799429,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5030065226253567e-05,
+      "loss": 0.7576,
+      "step": 82350
+    },
+    {
+      "epoch": 5.595529283870091,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5027941975811933e-05,
+      "loss": 0.816,
+      "step": 82355
+    },
+    {
+      "epoch": 5.595869003940753,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.5025818725370293e-05,
+      "loss": 0.9742,
+      "step": 82360
+    },
+    {
+      "epoch": 5.596208724011414,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5023695474928659e-05,
+      "loss": 0.9107,
+      "step": 82365
+    },
+    {
+      "epoch": 5.596548444082076,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5021572224487023e-05,
+      "loss": 0.8994,
+      "step": 82370
+    },
+    {
+      "epoch": 5.596888164152738,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.5019448974045389e-05,
+      "loss": 0.7944,
+      "step": 82375
+    },
+    {
+      "epoch": 5.5972278842234,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.5017325723603751e-05,
+      "loss": 0.9352,
+      "step": 82380
+    },
+    {
+      "epoch": 5.597567604294062,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.5015202473162115e-05,
+      "loss": 0.7953,
+      "step": 82385
+    },
+    {
+      "epoch": 5.597907324364724,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.501307922272048e-05,
+      "loss": 0.7444,
+      "step": 82390
+    },
+    {
+      "epoch": 5.598247044435385,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.5010955972278843e-05,
+      "loss": 0.8536,
+      "step": 82395
+    },
+    {
+      "epoch": 5.598586764506047,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.5008832721837207e-05,
+      "loss": 0.9128,
+      "step": 82400
+    },
+    {
+      "epoch": 5.598926484576709,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.5006709471395571e-05,
+      "loss": 0.7671,
+      "step": 82405
+    },
+    {
+      "epoch": 5.59926620464737,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.5004586220953933e-05,
+      "loss": 0.8714,
+      "step": 82410
+    },
+    {
+      "epoch": 5.599605924718032,
+      "grad_norm": 9.5,
+      "learning_rate": 1.5002462970512299e-05,
+      "loss": 0.8184,
+      "step": 82415
+    },
+    {
+      "epoch": 5.599945644788694,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5000339720070663e-05,
+      "loss": 0.8215,
+      "step": 82420
+    },
+    {
+      "epoch": 5.600285364859356,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4998216469629025e-05,
+      "loss": 0.8898,
+      "step": 82425
+    },
+    {
+      "epoch": 5.600625084930018,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.4996093219187391e-05,
+      "loss": 0.8341,
+      "step": 82430
+    },
+    {
+      "epoch": 5.60096480500068,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4993969968745755e-05,
+      "loss": 0.817,
+      "step": 82435
+    },
+    {
+      "epoch": 5.601304525071341,
+      "grad_norm": 16.125,
+      "learning_rate": 1.4991846718304117e-05,
+      "loss": 0.8789,
+      "step": 82440
+    },
+    {
+      "epoch": 5.601644245142003,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4989723467862481e-05,
+      "loss": 0.8766,
+      "step": 82445
+    },
+    {
+      "epoch": 5.601983965212665,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4987600217420847e-05,
+      "loss": 0.83,
+      "step": 82450
+    },
+    {
+      "epoch": 5.602323685283326,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.498547696697921e-05,
+      "loss": 0.9058,
+      "step": 82455
+    },
+    {
+      "epoch": 5.602663405353988,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.4983353716537573e-05,
+      "loss": 0.8365,
+      "step": 82460
+    },
+    {
+      "epoch": 5.60300312542465,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.498123046609594e-05,
+      "loss": 0.8759,
+      "step": 82465
+    },
+    {
+      "epoch": 5.603342845495312,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.49791072156543e-05,
+      "loss": 0.8766,
+      "step": 82470
+    },
+    {
+      "epoch": 5.603682565565974,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4976983965212666e-05,
+      "loss": 0.9449,
+      "step": 82475
+    },
+    {
+      "epoch": 5.604022285636636,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.497486071477103e-05,
+      "loss": 0.861,
+      "step": 82480
+    },
+    {
+      "epoch": 5.604362005707297,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4972737464329392e-05,
+      "loss": 0.9255,
+      "step": 82485
+    },
+    {
+      "epoch": 5.604701725777959,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4970614213887758e-05,
+      "loss": 0.9181,
+      "step": 82490
+    },
+    {
+      "epoch": 5.605041445848621,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.4968490963446122e-05,
+      "loss": 0.7601,
+      "step": 82495
+    },
+    {
+      "epoch": 5.605381165919282,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4966367713004484e-05,
+      "loss": 0.9285,
+      "step": 82500
+    },
+    {
+      "epoch": 5.605720885989944,
+      "grad_norm": 8.625,
+      "learning_rate": 1.4964244462562848e-05,
+      "loss": 0.9115,
+      "step": 82505
+    },
+    {
+      "epoch": 5.606060606060606,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4962121212121214e-05,
+      "loss": 0.8264,
+      "step": 82510
+    },
+    {
+      "epoch": 5.606400326131268,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4959997961679576e-05,
+      "loss": 0.8344,
+      "step": 82515
+    },
+    {
+      "epoch": 5.60674004620193,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.495787471123794e-05,
+      "loss": 0.9074,
+      "step": 82520
+    },
+    {
+      "epoch": 5.607079766272592,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4955751460796306e-05,
+      "loss": 0.8377,
+      "step": 82525
+    },
+    {
+      "epoch": 5.607419486343253,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4953628210354668e-05,
+      "loss": 0.9209,
+      "step": 82530
+    },
+    {
+      "epoch": 5.607759206413915,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4951504959913032e-05,
+      "loss": 0.9236,
+      "step": 82535
+    },
+    {
+      "epoch": 5.608098926484577,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4949381709471398e-05,
+      "loss": 0.8656,
+      "step": 82540
+    },
+    {
+      "epoch": 5.608438646555238,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4947258459029762e-05,
+      "loss": 0.8705,
+      "step": 82545
+    },
+    {
+      "epoch": 5.6087783666259,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4945135208588124e-05,
+      "loss": 0.8866,
+      "step": 82550
+    },
+    {
+      "epoch": 5.6091180866965615,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4943011958146488e-05,
+      "loss": 0.7906,
+      "step": 82555
+    },
+    {
+      "epoch": 5.609457806767224,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4940888707704854e-05,
+      "loss": 0.9256,
+      "step": 82560
+    },
+    {
+      "epoch": 5.609797526837886,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4938765457263216e-05,
+      "loss": 0.8821,
+      "step": 82565
+    },
+    {
+      "epoch": 5.610137246908547,
+      "grad_norm": 8.375,
+      "learning_rate": 1.493664220682158e-05,
+      "loss": 0.8717,
+      "step": 82570
+    },
+    {
+      "epoch": 5.610476966979209,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4934518956379946e-05,
+      "loss": 0.7847,
+      "step": 82575
+    },
+    {
+      "epoch": 5.610816687049871,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4932395705938306e-05,
+      "loss": 0.88,
+      "step": 82580
+    },
+    {
+      "epoch": 5.611156407120532,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4930272455496672e-05,
+      "loss": 0.999,
+      "step": 82585
+    },
+    {
+      "epoch": 5.611496127191194,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4928149205055036e-05,
+      "loss": 0.8408,
+      "step": 82590
+    },
+    {
+      "epoch": 5.611835847261856,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4926025954613398e-05,
+      "loss": 0.8896,
+      "step": 82595
+    },
+    {
+      "epoch": 5.6121755673325175,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4923902704171764e-05,
+      "loss": 0.834,
+      "step": 82600
+    },
+    {
+      "epoch": 5.61251528740318,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4921779453730128e-05,
+      "loss": 0.8559,
+      "step": 82605
+    },
+    {
+      "epoch": 5.612855007473842,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.491965620328849e-05,
+      "loss": 0.9043,
+      "step": 82610
+    },
+    {
+      "epoch": 5.613194727544503,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4917532952846854e-05,
+      "loss": 0.8525,
+      "step": 82615
+    },
+    {
+      "epoch": 5.613534447615165,
+      "grad_norm": 8.5,
+      "learning_rate": 1.491540970240522e-05,
+      "loss": 0.7751,
+      "step": 82620
+    },
+    {
+      "epoch": 5.613874167685827,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4913286451963582e-05,
+      "loss": 0.8524,
+      "step": 82625
+    },
+    {
+      "epoch": 5.614213887756488,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4911163201521946e-05,
+      "loss": 1.0028,
+      "step": 82630
+    },
+    {
+      "epoch": 5.61455360782715,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4909039951080312e-05,
+      "loss": 1.0409,
+      "step": 82635
+    },
+    {
+      "epoch": 5.614893327897812,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4906916700638674e-05,
+      "loss": 0.8678,
+      "step": 82640
+    },
+    {
+      "epoch": 5.615233047968474,
+      "grad_norm": 7.0,
+      "learning_rate": 1.4904793450197038e-05,
+      "loss": 0.8861,
+      "step": 82645
+    },
+    {
+      "epoch": 5.615572768039136,
+      "grad_norm": 6.875,
+      "learning_rate": 1.4902670199755402e-05,
+      "loss": 0.8892,
+      "step": 82650
+    },
+    {
+      "epoch": 5.615912488109798,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4900546949313765e-05,
+      "loss": 0.8048,
+      "step": 82655
+    },
+    {
+      "epoch": 5.616252208180459,
+      "grad_norm": 8.75,
+      "learning_rate": 1.489842369887213e-05,
+      "loss": 0.8907,
+      "step": 82660
+    },
+    {
+      "epoch": 5.616591928251121,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4896300448430494e-05,
+      "loss": 0.8716,
+      "step": 82665
+    },
+    {
+      "epoch": 5.616931648321783,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4894177197988857e-05,
+      "loss": 0.8439,
+      "step": 82670
+    },
+    {
+      "epoch": 5.617271368392444,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4892053947547222e-05,
+      "loss": 0.8842,
+      "step": 82675
+    },
+    {
+      "epoch": 5.617611088463106,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4889930697105586e-05,
+      "loss": 0.8079,
+      "step": 82680
+    },
+    {
+      "epoch": 5.617950808533768,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4887807446663949e-05,
+      "loss": 0.927,
+      "step": 82685
+    },
+    {
+      "epoch": 5.61829052860443,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.4885684196222313e-05,
+      "loss": 0.8114,
+      "step": 82690
+    },
+    {
+      "epoch": 5.618630248675092,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4883560945780678e-05,
+      "loss": 0.8634,
+      "step": 82695
+    },
+    {
+      "epoch": 5.618969968745754,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.488143769533904e-05,
+      "loss": 0.8371,
+      "step": 82700
+    },
+    {
+      "epoch": 5.619309688816415,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.4879314444897405e-05,
+      "loss": 0.8941,
+      "step": 82705
+    },
+    {
+      "epoch": 5.619649408887077,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.487719119445577e-05,
+      "loss": 0.8241,
+      "step": 82710
+    },
+    {
+      "epoch": 5.619989128957739,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4875067944014134e-05,
+      "loss": 0.8732,
+      "step": 82715
+    },
+    {
+      "epoch": 5.6203288490284,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4872944693572497e-05,
+      "loss": 0.8008,
+      "step": 82720
+    },
+    {
+      "epoch": 5.620668569099062,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.487082144313086e-05,
+      "loss": 0.8697,
+      "step": 82725
+    },
+    {
+      "epoch": 5.621008289169724,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4868698192689226e-05,
+      "loss": 0.7811,
+      "step": 82730
+    },
+    {
+      "epoch": 5.621348009240386,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4866574942247589e-05,
+      "loss": 0.9688,
+      "step": 82735
+    },
+    {
+      "epoch": 5.621687729311048,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4864451691805953e-05,
+      "loss": 0.8709,
+      "step": 82740
+    },
+    {
+      "epoch": 5.62202744938171,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4862328441364318e-05,
+      "loss": 1.0986,
+      "step": 82745
+    },
+    {
+      "epoch": 5.622367169452371,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4860205190922679e-05,
+      "loss": 0.8202,
+      "step": 82750
+    },
+    {
+      "epoch": 5.622706889523033,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4858081940481045e-05,
+      "loss": 0.823,
+      "step": 82755
+    },
+    {
+      "epoch": 5.623046609593695,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4855958690039409e-05,
+      "loss": 0.8567,
+      "step": 82760
+    },
+    {
+      "epoch": 5.623386329664356,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4853835439597771e-05,
+      "loss": 0.8695,
+      "step": 82765
+    },
+    {
+      "epoch": 5.623726049735018,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4851712189156137e-05,
+      "loss": 0.9176,
+      "step": 82770
+    },
+    {
+      "epoch": 5.62406576980568,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.48495889387145e-05,
+      "loss": 0.7915,
+      "step": 82775
+    },
+    {
+      "epoch": 5.624405489876342,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4847465688272863e-05,
+      "loss": 0.8248,
+      "step": 82780
+    },
+    {
+      "epoch": 5.624745209947004,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4845342437831227e-05,
+      "loss": 0.8663,
+      "step": 82785
+    },
+    {
+      "epoch": 5.625084930017666,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4843219187389593e-05,
+      "loss": 0.8462,
+      "step": 82790
+    },
+    {
+      "epoch": 5.625424650088327,
+      "grad_norm": 9.75,
+      "learning_rate": 1.4841095936947955e-05,
+      "loss": 0.8119,
+      "step": 82795
+    },
+    {
+      "epoch": 5.625764370158989,
+      "grad_norm": 6.75,
+      "learning_rate": 1.4838972686506319e-05,
+      "loss": 0.947,
+      "step": 82800
+    },
+    {
+      "epoch": 5.626104090229651,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4836849436064685e-05,
+      "loss": 0.9718,
+      "step": 82805
+    },
+    {
+      "epoch": 5.626443810300312,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4834726185623047e-05,
+      "loss": 0.862,
+      "step": 82810
+    },
+    {
+      "epoch": 5.626783530370974,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4832602935181411e-05,
+      "loss": 0.8062,
+      "step": 82815
+    },
+    {
+      "epoch": 5.627123250441636,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4830479684739777e-05,
+      "loss": 0.8652,
+      "step": 82820
+    },
+    {
+      "epoch": 5.627462970512298,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4828356434298137e-05,
+      "loss": 0.8254,
+      "step": 82825
+    },
+    {
+      "epoch": 5.62780269058296,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4826233183856503e-05,
+      "loss": 0.8808,
+      "step": 82830
+    },
+    {
+      "epoch": 5.628142410653622,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.4824109933414867e-05,
+      "loss": 0.9151,
+      "step": 82835
+    },
+    {
+      "epoch": 5.628482130724283,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.482198668297323e-05,
+      "loss": 0.8662,
+      "step": 82840
+    },
+    {
+      "epoch": 5.628821850794945,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4819863432531595e-05,
+      "loss": 0.812,
+      "step": 82845
+    },
+    {
+      "epoch": 5.629161570865607,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4817740182089959e-05,
+      "loss": 0.8418,
+      "step": 82850
+    },
+    {
+      "epoch": 5.629501290936268,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.4815616931648321e-05,
+      "loss": 0.8432,
+      "step": 82855
+    },
+    {
+      "epoch": 5.62984101100693,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4813493681206685e-05,
+      "loss": 0.9659,
+      "step": 82860
+    },
+    {
+      "epoch": 5.630180731077592,
+      "grad_norm": 9.25,
+      "learning_rate": 1.4811370430765051e-05,
+      "loss": 0.9349,
+      "step": 82865
+    },
+    {
+      "epoch": 5.630520451148254,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4809247180323413e-05,
+      "loss": 0.8363,
+      "step": 82870
+    },
+    {
+      "epoch": 5.630860171218916,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4807123929881777e-05,
+      "loss": 0.83,
+      "step": 82875
+    },
+    {
+      "epoch": 5.631199891289578,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4805000679440143e-05,
+      "loss": 0.8219,
+      "step": 82880
+    },
+    {
+      "epoch": 5.631539611360239,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4802877428998507e-05,
+      "loss": 0.7965,
+      "step": 82885
+    },
+    {
+      "epoch": 5.631879331430901,
+      "grad_norm": 8.25,
+      "learning_rate": 1.480075417855687e-05,
+      "loss": 0.9101,
+      "step": 82890
+    },
+    {
+      "epoch": 5.632219051501563,
+      "grad_norm": 9.375,
+      "learning_rate": 1.4798630928115233e-05,
+      "loss": 0.8195,
+      "step": 82895
+    },
+    {
+      "epoch": 5.632558771572224,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4796507677673599e-05,
+      "loss": 0.8596,
+      "step": 82900
+    },
+    {
+      "epoch": 5.632898491642886,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4794384427231961e-05,
+      "loss": 0.9772,
+      "step": 82905
+    },
+    {
+      "epoch": 5.633238211713548,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4792261176790325e-05,
+      "loss": 0.9544,
+      "step": 82910
+    },
+    {
+      "epoch": 5.63357793178421,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4790137926348691e-05,
+      "loss": 0.8364,
+      "step": 82915
+    },
+    {
+      "epoch": 5.633917651854872,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.4788014675907053e-05,
+      "loss": 0.8578,
+      "step": 82920
+    },
+    {
+      "epoch": 5.634257371925534,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.4785891425465417e-05,
+      "loss": 0.7957,
+      "step": 82925
+    },
+    {
+      "epoch": 5.634597091996195,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.4783768175023781e-05,
+      "loss": 0.9027,
+      "step": 82930
+    },
+    {
+      "epoch": 5.634936812066857,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4781644924582144e-05,
+      "loss": 0.8094,
+      "step": 82935
+    },
+    {
+      "epoch": 5.635276532137519,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.477952167414051e-05,
+      "loss": 0.7732,
+      "step": 82940
+    },
+    {
+      "epoch": 5.63561625220818,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.4777398423698873e-05,
+      "loss": 0.7805,
+      "step": 82945
+    },
+    {
+      "epoch": 5.635955972278842,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4775275173257236e-05,
+      "loss": 0.9016,
+      "step": 82950
+    },
+    {
+      "epoch": 5.6362956923495044,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4773151922815601e-05,
+      "loss": 0.8933,
+      "step": 82955
+    },
+    {
+      "epoch": 5.636635412420166,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4771028672373965e-05,
+      "loss": 0.8722,
+      "step": 82960
+    },
+    {
+      "epoch": 5.636975132490828,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4768905421932328e-05,
+      "loss": 0.8336,
+      "step": 82965
+    },
+    {
+      "epoch": 5.63731485256149,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4766782171490692e-05,
+      "loss": 0.8247,
+      "step": 82970
+    },
+    {
+      "epoch": 5.637654572632151,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4764658921049057e-05,
+      "loss": 0.8785,
+      "step": 82975
+    },
+    {
+      "epoch": 5.637994292702813,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.476253567060742e-05,
+      "loss": 0.8648,
+      "step": 82980
+    },
+    {
+      "epoch": 5.638334012773475,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4760412420165784e-05,
+      "loss": 0.8919,
+      "step": 82985
+    },
+    {
+      "epoch": 5.638673732844136,
+      "grad_norm": 8.25,
+      "learning_rate": 1.475828916972415e-05,
+      "loss": 0.8929,
+      "step": 82990
+    },
+    {
+      "epoch": 5.639013452914798,
+      "grad_norm": 7.75,
+      "learning_rate": 1.475616591928251e-05,
+      "loss": 0.8548,
+      "step": 82995
+    },
+    {
+      "epoch": 5.6393531729854605,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4754042668840876e-05,
+      "loss": 0.8658,
+      "step": 83000
+    },
+    {
+      "epoch": 5.639692893056122,
+      "grad_norm": 8.125,
+      "learning_rate": 1.475191941839924e-05,
+      "loss": 0.8579,
+      "step": 83005
+    },
+    {
+      "epoch": 5.640032613126784,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4749796167957602e-05,
+      "loss": 0.8024,
+      "step": 83010
+    },
+    {
+      "epoch": 5.640372333197446,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4747672917515968e-05,
+      "loss": 0.7318,
+      "step": 83015
+    },
+    {
+      "epoch": 5.640712053268107,
+      "grad_norm": 7.5,
+      "learning_rate": 1.4745549667074332e-05,
+      "loss": 0.7841,
+      "step": 83020
+    },
+    {
+      "epoch": 5.641051773338769,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4743426416632694e-05,
+      "loss": 0.8317,
+      "step": 83025
+    },
+    {
+      "epoch": 5.64139149340943,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4741303166191058e-05,
+      "loss": 0.8944,
+      "step": 83030
+    },
+    {
+      "epoch": 5.641731213480092,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4739179915749424e-05,
+      "loss": 0.8861,
+      "step": 83035
+    },
+    {
+      "epoch": 5.642070933550754,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4737056665307786e-05,
+      "loss": 0.8714,
+      "step": 83040
+    },
+    {
+      "epoch": 5.642410653621416,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.473493341486615e-05,
+      "loss": 0.9005,
+      "step": 83045
+    },
+    {
+      "epoch": 5.642750373692078,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.4732810164424516e-05,
+      "loss": 0.9546,
+      "step": 83050
+    },
+    {
+      "epoch": 5.64309009376274,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.473068691398288e-05,
+      "loss": 0.8529,
+      "step": 83055
+    },
+    {
+      "epoch": 5.643429813833401,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4728563663541242e-05,
+      "loss": 0.8534,
+      "step": 83060
+    },
+    {
+      "epoch": 5.643769533904063,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4726440413099608e-05,
+      "loss": 0.863,
+      "step": 83065
+    },
+    {
+      "epoch": 5.644109253974725,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4724317162657972e-05,
+      "loss": 0.8605,
+      "step": 83070
+    },
+    {
+      "epoch": 5.644448974045386,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4722193912216334e-05,
+      "loss": 0.8695,
+      "step": 83075
+    },
+    {
+      "epoch": 5.644788694116048,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4720070661774698e-05,
+      "loss": 0.8123,
+      "step": 83080
+    },
+    {
+      "epoch": 5.64512841418671,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4717947411333064e-05,
+      "loss": 0.7864,
+      "step": 83085
+    },
+    {
+      "epoch": 5.645468134257372,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4715824160891426e-05,
+      "loss": 0.9494,
+      "step": 83090
+    },
+    {
+      "epoch": 5.645807854328034,
+      "grad_norm": 7.625,
+      "learning_rate": 1.471370091044979e-05,
+      "loss": 0.836,
+      "step": 83095
+    },
+    {
+      "epoch": 5.646147574398696,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4711577660008156e-05,
+      "loss": 0.9133,
+      "step": 83100
+    },
+    {
+      "epoch": 5.646487294469357,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4709454409566516e-05,
+      "loss": 0.7706,
+      "step": 83105
+    },
+    {
+      "epoch": 5.646827014540019,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.4707331159124882e-05,
+      "loss": 0.8507,
+      "step": 83110
+    },
+    {
+      "epoch": 5.647166734610681,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4705207908683246e-05,
+      "loss": 0.881,
+      "step": 83115
+    },
+    {
+      "epoch": 5.647506454681342,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4703084658241608e-05,
+      "loss": 0.7881,
+      "step": 83120
+    },
+    {
+      "epoch": 5.647846174752004,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.4700961407799974e-05,
+      "loss": 0.8142,
+      "step": 83125
+    },
+    {
+      "epoch": 5.648185894822666,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4698838157358338e-05,
+      "loss": 0.8519,
+      "step": 83130
+    },
+    {
+      "epoch": 5.648525614893328,
+      "grad_norm": 7.75,
+      "learning_rate": 1.46967149069167e-05,
+      "loss": 0.7833,
+      "step": 83135
+    },
+    {
+      "epoch": 5.64886533496399,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4694591656475064e-05,
+      "loss": 0.872,
+      "step": 83140
+    },
+    {
+      "epoch": 5.649205055034652,
+      "grad_norm": 8.75,
+      "learning_rate": 1.469246840603343e-05,
+      "loss": 0.8991,
+      "step": 83145
+    },
+    {
+      "epoch": 5.649544775105313,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4690345155591792e-05,
+      "loss": 0.8583,
+      "step": 83150
+    },
+    {
+      "epoch": 5.649884495175975,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4688221905150156e-05,
+      "loss": 0.8997,
+      "step": 83155
+    },
+    {
+      "epoch": 5.650224215246637,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4686098654708522e-05,
+      "loss": 0.8355,
+      "step": 83160
+    },
+    {
+      "epoch": 5.650563935317298,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4683975404266884e-05,
+      "loss": 0.8759,
+      "step": 83165
+    },
+    {
+      "epoch": 5.65090365538796,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4681852153825248e-05,
+      "loss": 0.9165,
+      "step": 83170
+    },
+    {
+      "epoch": 5.651243375458622,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4679728903383612e-05,
+      "loss": 0.9529,
+      "step": 83175
+    },
+    {
+      "epoch": 5.651583095529284,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4677605652941975e-05,
+      "loss": 0.7703,
+      "step": 83180
+    },
+    {
+      "epoch": 5.651922815599946,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.467548240250034e-05,
+      "loss": 0.8142,
+      "step": 83185
+    },
+    {
+      "epoch": 5.652262535670608,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4673359152058705e-05,
+      "loss": 0.7871,
+      "step": 83190
+    },
+    {
+      "epoch": 5.652602255741269,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.4671235901617067e-05,
+      "loss": 0.7421,
+      "step": 83195
+    },
+    {
+      "epoch": 5.652941975811931,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4669112651175433e-05,
+      "loss": 0.9189,
+      "step": 83200
+    },
+    {
+      "epoch": 5.653281695882593,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4666989400733797e-05,
+      "loss": 0.8485,
+      "step": 83205
+    },
+    {
+      "epoch": 5.653621415953254,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4664866150292159e-05,
+      "loss": 0.9083,
+      "step": 83210
+    },
+    {
+      "epoch": 5.653961136023916,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.4662742899850523e-05,
+      "loss": 0.9067,
+      "step": 83215
+    },
+    {
+      "epoch": 5.654300856094578,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4660619649408889e-05,
+      "loss": 0.9236,
+      "step": 83220
+    },
+    {
+      "epoch": 5.65464057616524,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4658496398967253e-05,
+      "loss": 0.889,
+      "step": 83225
+    },
+    {
+      "epoch": 5.654980296235902,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4656373148525615e-05,
+      "loss": 0.8364,
+      "step": 83230
+    },
+    {
+      "epoch": 5.655320016306564,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.465424989808398e-05,
+      "loss": 0.8488,
+      "step": 83235
+    },
+    {
+      "epoch": 5.655659736377225,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4652126647642345e-05,
+      "loss": 0.9185,
+      "step": 83240
+    },
+    {
+      "epoch": 5.655999456447887,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4650003397200707e-05,
+      "loss": 0.9012,
+      "step": 83245
+    },
+    {
+      "epoch": 5.656339176518548,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4647880146759071e-05,
+      "loss": 0.813,
+      "step": 83250
+    },
+    {
+      "epoch": 5.65667889658921,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4645756896317437e-05,
+      "loss": 0.8454,
+      "step": 83255
+    },
+    {
+      "epoch": 5.657018616659872,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4643633645875799e-05,
+      "loss": 0.9053,
+      "step": 83260
+    },
+    {
+      "epoch": 5.657358336730534,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4641510395434163e-05,
+      "loss": 0.896,
+      "step": 83265
+    },
+    {
+      "epoch": 5.657698056801196,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4639387144992529e-05,
+      "loss": 0.8513,
+      "step": 83270
+    },
+    {
+      "epoch": 5.658037776871858,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.463726389455089e-05,
+      "loss": 0.7913,
+      "step": 83275
+    },
+    {
+      "epoch": 5.658377496942519,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4635140644109255e-05,
+      "loss": 0.9345,
+      "step": 83280
+    },
+    {
+      "epoch": 5.658717217013181,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4633017393667619e-05,
+      "loss": 0.7946,
+      "step": 83285
+    },
+    {
+      "epoch": 5.659056937083843,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.4630894143225981e-05,
+      "loss": 0.8044,
+      "step": 83290
+    },
+    {
+      "epoch": 5.659396657154504,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4628770892784347e-05,
+      "loss": 0.7912,
+      "step": 83295
+    },
+    {
+      "epoch": 5.659736377225166,
+      "grad_norm": 9.0,
+      "learning_rate": 1.4626647642342711e-05,
+      "loss": 0.8737,
+      "step": 83300
+    },
+    {
+      "epoch": 5.660076097295828,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4624524391901073e-05,
+      "loss": 0.7266,
+      "step": 83305
+    },
+    {
+      "epoch": 5.66041581736649,
+      "grad_norm": 7.125,
+      "learning_rate": 1.4622401141459439e-05,
+      "loss": 0.8037,
+      "step": 83310
+    },
+    {
+      "epoch": 5.660755537437152,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.4620277891017803e-05,
+      "loss": 0.9087,
+      "step": 83315
+    },
+    {
+      "epoch": 5.661095257507814,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4618154640576165e-05,
+      "loss": 0.8141,
+      "step": 83320
+    },
+    {
+      "epoch": 5.661434977578475,
+      "grad_norm": 8.25,
+      "learning_rate": 1.461603139013453e-05,
+      "loss": 0.8503,
+      "step": 83325
+    },
+    {
+      "epoch": 5.661774697649137,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4613908139692895e-05,
+      "loss": 0.9076,
+      "step": 83330
+    },
+    {
+      "epoch": 5.662114417719799,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4611784889251257e-05,
+      "loss": 0.8386,
+      "step": 83335
+    },
+    {
+      "epoch": 5.66245413779046,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4609661638809621e-05,
+      "loss": 0.906,
+      "step": 83340
+    },
+    {
+      "epoch": 5.662793857861122,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4607538388367987e-05,
+      "loss": 0.8511,
+      "step": 83345
+    },
+    {
+      "epoch": 5.663133577931784,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.4605415137926348e-05,
+      "loss": 0.8401,
+      "step": 83350
+    },
+    {
+      "epoch": 5.663473298002446,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4603291887484713e-05,
+      "loss": 0.9074,
+      "step": 83355
+    },
+    {
+      "epoch": 5.663813018073108,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.4601168637043077e-05,
+      "loss": 0.826,
+      "step": 83360
+    },
+    {
+      "epoch": 5.66415273814377,
+      "grad_norm": 8.0,
+      "learning_rate": 1.459904538660144e-05,
+      "loss": 0.754,
+      "step": 83365
+    },
+    {
+      "epoch": 5.664492458214431,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4596922136159805e-05,
+      "loss": 0.8269,
+      "step": 83370
+    },
+    {
+      "epoch": 5.664832178285093,
+      "grad_norm": 8.125,
+      "learning_rate": 1.459479888571817e-05,
+      "loss": 0.8577,
+      "step": 83375
+    },
+    {
+      "epoch": 5.665171898355755,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4592675635276532e-05,
+      "loss": 0.9123,
+      "step": 83380
+    },
+    {
+      "epoch": 5.665511618426416,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4590552384834896e-05,
+      "loss": 0.7644,
+      "step": 83385
+    },
+    {
+      "epoch": 5.665851338497078,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4588429134393261e-05,
+      "loss": 0.8191,
+      "step": 83390
+    },
+    {
+      "epoch": 5.66619105856774,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4586305883951625e-05,
+      "loss": 0.8383,
+      "step": 83395
+    },
+    {
+      "epoch": 5.666530778638402,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4584182633509988e-05,
+      "loss": 0.6826,
+      "step": 83400
+    },
+    {
+      "epoch": 5.666870498709064,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.4582059383068353e-05,
+      "loss": 0.7839,
+      "step": 83405
+    },
+    {
+      "epoch": 5.667210218779726,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4579936132626717e-05,
+      "loss": 0.8798,
+      "step": 83410
+    },
+    {
+      "epoch": 5.667549938850387,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.457781288218508e-05,
+      "loss": 0.8806,
+      "step": 83415
+    },
+    {
+      "epoch": 5.667889658921049,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4575689631743444e-05,
+      "loss": 0.7704,
+      "step": 83420
+    },
+    {
+      "epoch": 5.668229378991711,
+      "grad_norm": 8.75,
+      "learning_rate": 1.457356638130181e-05,
+      "loss": 0.7984,
+      "step": 83425
+    },
+    {
+      "epoch": 5.668569099062372,
+      "grad_norm": 8.75,
+      "learning_rate": 1.4571443130860172e-05,
+      "loss": 0.9576,
+      "step": 83430
+    },
+    {
+      "epoch": 5.668908819133034,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4569319880418536e-05,
+      "loss": 0.9927,
+      "step": 83435
+    },
+    {
+      "epoch": 5.669248539203696,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4567196629976901e-05,
+      "loss": 0.7986,
+      "step": 83440
+    },
+    {
+      "epoch": 5.669588259274358,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4565073379535264e-05,
+      "loss": 0.79,
+      "step": 83445
+    },
+    {
+      "epoch": 5.66992797934502,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4562950129093628e-05,
+      "loss": 0.8209,
+      "step": 83450
+    },
+    {
+      "epoch": 5.670267699415682,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4560826878651992e-05,
+      "loss": 0.8396,
+      "step": 83455
+    },
+    {
+      "epoch": 5.670607419486343,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4558703628210354e-05,
+      "loss": 0.8954,
+      "step": 83460
+    },
+    {
+      "epoch": 5.670947139557005,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.455658037776872e-05,
+      "loss": 0.9234,
+      "step": 83465
+    },
+    {
+      "epoch": 5.671286859627667,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4554457127327084e-05,
+      "loss": 0.8752,
+      "step": 83470
+    },
+    {
+      "epoch": 5.671626579698328,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4552333876885446e-05,
+      "loss": 0.8504,
+      "step": 83475
+    },
+    {
+      "epoch": 5.67196629976899,
+      "grad_norm": 7.375,
+      "learning_rate": 1.4550210626443812e-05,
+      "loss": 0.8562,
+      "step": 83480
+    },
+    {
+      "epoch": 5.672306019839652,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4548087376002176e-05,
+      "loss": 0.8027,
+      "step": 83485
+    },
+    {
+      "epoch": 5.672645739910314,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4545964125560538e-05,
+      "loss": 0.8819,
+      "step": 83490
+    },
+    {
+      "epoch": 5.672985459980976,
+      "grad_norm": 7.125,
+      "learning_rate": 1.4543840875118902e-05,
+      "loss": 0.7572,
+      "step": 83495
+    },
+    {
+      "epoch": 5.673325180051638,
+      "grad_norm": 7.375,
+      "learning_rate": 1.4541717624677268e-05,
+      "loss": 0.9028,
+      "step": 83500
+    },
+    {
+      "epoch": 5.673664900122299,
+      "grad_norm": 8.25,
+      "learning_rate": 1.453959437423563e-05,
+      "loss": 0.8473,
+      "step": 83505
+    },
+    {
+      "epoch": 5.674004620192961,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4537471123793994e-05,
+      "loss": 0.8078,
+      "step": 83510
+    },
+    {
+      "epoch": 5.674344340263623,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.453534787335236e-05,
+      "loss": 0.8359,
+      "step": 83515
+    },
+    {
+      "epoch": 5.674684060334284,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.453322462291072e-05,
+      "loss": 0.8013,
+      "step": 83520
+    },
+    {
+      "epoch": 5.675023780404946,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4531101372469086e-05,
+      "loss": 0.816,
+      "step": 83525
+    },
+    {
+      "epoch": 5.6753635004756084,
+      "grad_norm": 8.375,
+      "learning_rate": 1.452897812202745e-05,
+      "loss": 0.8277,
+      "step": 83530
+    },
+    {
+      "epoch": 5.67570322054627,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.4526854871585812e-05,
+      "loss": 0.864,
+      "step": 83535
+    },
+    {
+      "epoch": 5.676042940616932,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4524731621144178e-05,
+      "loss": 0.8216,
+      "step": 83540
+    },
+    {
+      "epoch": 5.676382660687594,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4522608370702542e-05,
+      "loss": 0.7733,
+      "step": 83545
+    },
+    {
+      "epoch": 5.676722380758255,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4520485120260904e-05,
+      "loss": 0.7724,
+      "step": 83550
+    },
+    {
+      "epoch": 5.677062100828917,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4518361869819268e-05,
+      "loss": 0.9302,
+      "step": 83555
+    },
+    {
+      "epoch": 5.677401820899579,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4516238619377634e-05,
+      "loss": 0.8483,
+      "step": 83560
+    },
+    {
+      "epoch": 5.67774154097024,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4514115368935998e-05,
+      "loss": 0.866,
+      "step": 83565
+    },
+    {
+      "epoch": 5.678081261040902,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.451199211849436e-05,
+      "loss": 0.804,
+      "step": 83570
+    },
+    {
+      "epoch": 5.6784209811115645,
+      "grad_norm": 9.25,
+      "learning_rate": 1.4509868868052726e-05,
+      "loss": 0.8509,
+      "step": 83575
+    },
+    {
+      "epoch": 5.678760701182226,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.450774561761109e-05,
+      "loss": 0.858,
+      "step": 83580
+    },
+    {
+      "epoch": 5.679100421252888,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4505622367169452e-05,
+      "loss": 0.8539,
+      "step": 83585
+    },
+    {
+      "epoch": 5.67944014132355,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4503499116727818e-05,
+      "loss": 0.8179,
+      "step": 83590
+    },
+    {
+      "epoch": 5.679779861394211,
+      "grad_norm": 6.125,
+      "learning_rate": 1.4501375866286182e-05,
+      "loss": 0.7882,
+      "step": 83595
+    },
+    {
+      "epoch": 5.680119581464873,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4499252615844544e-05,
+      "loss": 0.8287,
+      "step": 83600
+    },
+    {
+      "epoch": 5.680459301535535,
+      "grad_norm": 8.75,
+      "learning_rate": 1.4497129365402908e-05,
+      "loss": 0.8235,
+      "step": 83605
+    },
+    {
+      "epoch": 5.680799021606196,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4495006114961274e-05,
+      "loss": 0.8376,
+      "step": 83610
+    },
+    {
+      "epoch": 5.681138741676858,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4492882864519636e-05,
+      "loss": 0.904,
+      "step": 83615
+    },
+    {
+      "epoch": 5.6814784617475205,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4490759614078e-05,
+      "loss": 0.8645,
+      "step": 83620
+    },
+    {
+      "epoch": 5.681818181818182,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.4488636363636366e-05,
+      "loss": 0.8032,
+      "step": 83625
+    },
+    {
+      "epoch": 5.682157901888844,
+      "grad_norm": 9.125,
+      "learning_rate": 1.4486513113194727e-05,
+      "loss": 0.7989,
+      "step": 83630
+    },
+    {
+      "epoch": 5.682497621959506,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.4484389862753092e-05,
+      "loss": 0.8973,
+      "step": 83635
+    },
+    {
+      "epoch": 5.682837342030167,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4482266612311456e-05,
+      "loss": 0.8815,
+      "step": 83640
+    },
+    {
+      "epoch": 5.683177062100829,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4480143361869819e-05,
+      "loss": 0.9068,
+      "step": 83645
+    },
+    {
+      "epoch": 5.683516782171491,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4478020111428184e-05,
+      "loss": 0.8168,
+      "step": 83650
+    },
+    {
+      "epoch": 5.683856502242152,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4475896860986548e-05,
+      "loss": 0.8655,
+      "step": 83655
+    },
+    {
+      "epoch": 5.684196222312814,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.447377361054491e-05,
+      "loss": 0.8897,
+      "step": 83660
+    },
+    {
+      "epoch": 5.6845359423834765,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4471650360103275e-05,
+      "loss": 0.8035,
+      "step": 83665
+    },
+    {
+      "epoch": 5.684875662454138,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.446952710966164e-05,
+      "loss": 0.8309,
+      "step": 83670
+    },
+    {
+      "epoch": 5.6852153825248,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4467403859220003e-05,
+      "loss": 0.9491,
+      "step": 83675
+    },
+    {
+      "epoch": 5.685555102595462,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4465280608778367e-05,
+      "loss": 0.9651,
+      "step": 83680
+    },
+    {
+      "epoch": 5.685894822666123,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.4463157358336732e-05,
+      "loss": 0.7444,
+      "step": 83685
+    },
+    {
+      "epoch": 5.686234542736785,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4461034107895095e-05,
+      "loss": 0.726,
+      "step": 83690
+    },
+    {
+      "epoch": 5.686574262807447,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4458910857453459e-05,
+      "loss": 0.8877,
+      "step": 83695
+    },
+    {
+      "epoch": 5.686913982878108,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4456787607011823e-05,
+      "loss": 0.8101,
+      "step": 83700
+    },
+    {
+      "epoch": 5.68725370294877,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4454664356570185e-05,
+      "loss": 0.8226,
+      "step": 83705
+    },
+    {
+      "epoch": 5.687593423019432,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.445254110612855e-05,
+      "loss": 0.8144,
+      "step": 83710
+    },
+    {
+      "epoch": 5.687933143090094,
+      "grad_norm": 7.25,
+      "learning_rate": 1.4450417855686915e-05,
+      "loss": 0.9169,
+      "step": 83715
+    },
+    {
+      "epoch": 5.688272863160756,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4448294605245277e-05,
+      "loss": 0.8942,
+      "step": 83720
+    },
+    {
+      "epoch": 5.688612583231417,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.4446171354803643e-05,
+      "loss": 0.8443,
+      "step": 83725
+    },
+    {
+      "epoch": 5.688952303302079,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4444048104362007e-05,
+      "loss": 0.8453,
+      "step": 83730
+    },
+    {
+      "epoch": 5.689292023372741,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4441924853920372e-05,
+      "loss": 0.8025,
+      "step": 83735
+    },
+    {
+      "epoch": 5.689631743443402,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.4439801603478733e-05,
+      "loss": 0.8281,
+      "step": 83740
+    },
+    {
+      "epoch": 5.689971463514064,
+      "grad_norm": 8.875,
+      "learning_rate": 1.4437678353037099e-05,
+      "loss": 0.8341,
+      "step": 83745
+    },
+    {
+      "epoch": 5.690311183584726,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4435555102595463e-05,
+      "loss": 0.8621,
+      "step": 83750
+    },
+    {
+      "epoch": 5.690650903655388,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4433431852153825e-05,
+      "loss": 0.7948,
+      "step": 83755
+    },
+    {
+      "epoch": 5.69099062372605,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.443130860171219e-05,
+      "loss": 0.8506,
+      "step": 83760
+    },
+    {
+      "epoch": 5.691330343796712,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4429185351270555e-05,
+      "loss": 0.8462,
+      "step": 83765
+    },
+    {
+      "epoch": 5.691670063867373,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.4427062100828917e-05,
+      "loss": 0.8628,
+      "step": 83770
+    },
+    {
+      "epoch": 5.692009783938035,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4424938850387281e-05,
+      "loss": 0.8413,
+      "step": 83775
+    },
+    {
+      "epoch": 5.692349504008697,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4422815599945647e-05,
+      "loss": 0.8194,
+      "step": 83780
+    },
+    {
+      "epoch": 5.692689224079358,
+      "grad_norm": 7.5,
+      "learning_rate": 1.4420692349504009e-05,
+      "loss": 0.9029,
+      "step": 83785
+    },
+    {
+      "epoch": 5.69302894415002,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4418569099062373e-05,
+      "loss": 0.7764,
+      "step": 83790
+    },
+    {
+      "epoch": 5.693368664220682,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4416445848620739e-05,
+      "loss": 0.8874,
+      "step": 83795
+    },
+    {
+      "epoch": 5.693708384291344,
+      "grad_norm": 6.875,
+      "learning_rate": 1.44143225981791e-05,
+      "loss": 0.8103,
+      "step": 83800
+    },
+    {
+      "epoch": 5.694048104362006,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4412199347737465e-05,
+      "loss": 0.8179,
+      "step": 83805
+    },
+    {
+      "epoch": 5.694387824432668,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.4410076097295829e-05,
+      "loss": 0.8439,
+      "step": 83810
+    },
+    {
+      "epoch": 5.694727544503329,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4407952846854191e-05,
+      "loss": 0.9105,
+      "step": 83815
+    },
+    {
+      "epoch": 5.695067264573991,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4405829596412557e-05,
+      "loss": 0.8404,
+      "step": 83820
+    },
+    {
+      "epoch": 5.695406984644653,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4403706345970921e-05,
+      "loss": 0.8866,
+      "step": 83825
+    },
+    {
+      "epoch": 5.695746704715314,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4401583095529283e-05,
+      "loss": 0.8277,
+      "step": 83830
+    },
+    {
+      "epoch": 5.696086424785976,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4399459845087649e-05,
+      "loss": 0.8658,
+      "step": 83835
+    },
+    {
+      "epoch": 5.6964261448566385,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4397336594646013e-05,
+      "loss": 0.8229,
+      "step": 83840
+    },
+    {
+      "epoch": 5.6967658649273,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4395213344204375e-05,
+      "loss": 0.7936,
+      "step": 83845
+    },
+    {
+      "epoch": 5.697105584997962,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.439309009376274e-05,
+      "loss": 0.8476,
+      "step": 83850
+    },
+    {
+      "epoch": 5.697445305068624,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4390966843321105e-05,
+      "loss": 0.8738,
+      "step": 83855
+    },
+    {
+      "epoch": 5.697785025139285,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.4388843592879467e-05,
+      "loss": 0.847,
+      "step": 83860
+    },
+    {
+      "epoch": 5.698124745209947,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.4386720342437831e-05,
+      "loss": 0.7728,
+      "step": 83865
+    },
+    {
+      "epoch": 5.698464465280609,
+      "grad_norm": 6.875,
+      "learning_rate": 1.4384597091996197e-05,
+      "loss": 0.8716,
+      "step": 83870
+    },
+    {
+      "epoch": 5.69880418535127,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4382473841554558e-05,
+      "loss": 0.8538,
+      "step": 83875
+    },
+    {
+      "epoch": 5.699143905421932,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4380350591112923e-05,
+      "loss": 0.8792,
+      "step": 83880
+    },
+    {
+      "epoch": 5.6994836254925945,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4378227340671287e-05,
+      "loss": 0.8117,
+      "step": 83885
+    },
+    {
+      "epoch": 5.699823345563256,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.437610409022965e-05,
+      "loss": 0.8442,
+      "step": 83890
+    },
+    {
+      "epoch": 5.700163065633918,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.4373980839788015e-05,
+      "loss": 0.9032,
+      "step": 83895
+    },
+    {
+      "epoch": 5.70050278570458,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.437185758934638e-05,
+      "loss": 0.8851,
+      "step": 83900
+    },
+    {
+      "epoch": 5.700842505775241,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4369734338904745e-05,
+      "loss": 0.7501,
+      "step": 83905
+    },
+    {
+      "epoch": 5.701182225845903,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4367611088463106e-05,
+      "loss": 0.8482,
+      "step": 83910
+    },
+    {
+      "epoch": 5.701521945916565,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4365487838021472e-05,
+      "loss": 0.8767,
+      "step": 83915
+    },
+    {
+      "epoch": 5.701861665987226,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4363364587579836e-05,
+      "loss": 0.7494,
+      "step": 83920
+    },
+    {
+      "epoch": 5.702201386057888,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4361241337138198e-05,
+      "loss": 0.9303,
+      "step": 83925
+    },
+    {
+      "epoch": 5.70254110612855,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4359118086696564e-05,
+      "loss": 0.7932,
+      "step": 83930
+    },
+    {
+      "epoch": 5.702880826199212,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4356994836254928e-05,
+      "loss": 0.8506,
+      "step": 83935
+    },
+    {
+      "epoch": 5.703220546269874,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.435487158581329e-05,
+      "loss": 0.8463,
+      "step": 83940
+    },
+    {
+      "epoch": 5.703560266340535,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4352748335371654e-05,
+      "loss": 0.8752,
+      "step": 83945
+    },
+    {
+      "epoch": 5.703899986411197,
+      "grad_norm": 8.25,
+      "learning_rate": 1.435062508493002e-05,
+      "loss": 0.8002,
+      "step": 83950
+    },
+    {
+      "epoch": 5.704239706481859,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4348501834488382e-05,
+      "loss": 0.8396,
+      "step": 83955
+    },
+    {
+      "epoch": 5.70457942655252,
+      "grad_norm": 8.625,
+      "learning_rate": 1.4346378584046746e-05,
+      "loss": 0.8271,
+      "step": 83960
+    },
+    {
+      "epoch": 5.704919146623182,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4344255333605112e-05,
+      "loss": 0.8164,
+      "step": 83965
+    },
+    {
+      "epoch": 5.705258866693844,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4342132083163474e-05,
+      "loss": 0.8052,
+      "step": 83970
+    },
+    {
+      "epoch": 5.705598586764506,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4340008832721838e-05,
+      "loss": 0.9371,
+      "step": 83975
+    },
+    {
+      "epoch": 5.705938306835168,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4337885582280204e-05,
+      "loss": 0.8851,
+      "step": 83980
+    },
+    {
+      "epoch": 5.70627802690583,
+      "grad_norm": 8.875,
+      "learning_rate": 1.4335762331838564e-05,
+      "loss": 0.8198,
+      "step": 83985
+    },
+    {
+      "epoch": 5.706617746976491,
+      "grad_norm": 6.53125,
+      "learning_rate": 1.433363908139693e-05,
+      "loss": 0.8563,
+      "step": 83990
+    },
+    {
+      "epoch": 5.706957467047153,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4331515830955294e-05,
+      "loss": 0.8112,
+      "step": 83995
+    },
+    {
+      "epoch": 5.707297187117815,
+      "grad_norm": 6.625,
+      "learning_rate": 1.4329392580513656e-05,
+      "loss": 0.7809,
+      "step": 84000
+    },
+    {
+      "epoch": 5.707636907188476,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4327269330072022e-05,
+      "loss": 0.8295,
+      "step": 84005
+    },
+    {
+      "epoch": 5.707976627259138,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4325146079630386e-05,
+      "loss": 0.8512,
+      "step": 84010
+    },
+    {
+      "epoch": 5.7083163473298,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4323022829188748e-05,
+      "loss": 0.8485,
+      "step": 84015
+    },
+    {
+      "epoch": 5.708656067400462,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4320899578747112e-05,
+      "loss": 0.8447,
+      "step": 84020
+    },
+    {
+      "epoch": 5.708995787471124,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.4318776328305478e-05,
+      "loss": 0.8103,
+      "step": 84025
+    },
+    {
+      "epoch": 5.709335507541786,
+      "grad_norm": 8.375,
+      "learning_rate": 1.431665307786384e-05,
+      "loss": 0.8779,
+      "step": 84030
+    },
+    {
+      "epoch": 5.709675227612447,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4314529827422204e-05,
+      "loss": 0.9409,
+      "step": 84035
+    },
+    {
+      "epoch": 5.710014947683109,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.431240657698057e-05,
+      "loss": 0.8669,
+      "step": 84040
+    },
+    {
+      "epoch": 5.710354667753771,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.431028332653893e-05,
+      "loss": 0.8434,
+      "step": 84045
+    },
+    {
+      "epoch": 5.710694387824432,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4308160076097296e-05,
+      "loss": 0.982,
+      "step": 84050
+    },
+    {
+      "epoch": 5.711034107895094,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.430603682565566e-05,
+      "loss": 0.9355,
+      "step": 84055
+    },
+    {
+      "epoch": 5.711373827965756,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4303913575214023e-05,
+      "loss": 0.8435,
+      "step": 84060
+    },
+    {
+      "epoch": 5.711713548036418,
+      "grad_norm": 9.0,
+      "learning_rate": 1.4301790324772388e-05,
+      "loss": 0.8778,
+      "step": 84065
+    },
+    {
+      "epoch": 5.71205326810708,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4299667074330752e-05,
+      "loss": 0.8338,
+      "step": 84070
+    },
+    {
+      "epoch": 5.712392988177742,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4297543823889118e-05,
+      "loss": 0.7751,
+      "step": 84075
+    },
+    {
+      "epoch": 5.712732708248403,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.429542057344748e-05,
+      "loss": 0.9819,
+      "step": 84080
+    },
+    {
+      "epoch": 5.713072428319065,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4293297323005844e-05,
+      "loss": 0.812,
+      "step": 84085
+    },
+    {
+      "epoch": 5.713412148389727,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4291174072564208e-05,
+      "loss": 0.9389,
+      "step": 84090
+    },
+    {
+      "epoch": 5.713751868460388,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.428905082212257e-05,
+      "loss": 0.907,
+      "step": 84095
+    },
+    {
+      "epoch": 5.71409158853105,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4286927571680936e-05,
+      "loss": 0.8391,
+      "step": 84100
+    },
+    {
+      "epoch": 5.7144313086017124,
+      "grad_norm": 7.875,
+      "learning_rate": 1.42848043212393e-05,
+      "loss": 0.8997,
+      "step": 84105
+    },
+    {
+      "epoch": 5.714771028672374,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4282681070797663e-05,
+      "loss": 0.8053,
+      "step": 84110
+    },
+    {
+      "epoch": 5.715110748743036,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.4280557820356028e-05,
+      "loss": 0.8353,
+      "step": 84115
+    },
+    {
+      "epoch": 5.715450468813698,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4278434569914392e-05,
+      "loss": 0.8262,
+      "step": 84120
+    },
+    {
+      "epoch": 5.715790188884359,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4276311319472755e-05,
+      "loss": 0.8513,
+      "step": 84125
+    },
+    {
+      "epoch": 5.716129908955021,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4274188069031119e-05,
+      "loss": 0.7657,
+      "step": 84130
+    },
+    {
+      "epoch": 5.716469629025683,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4272064818589484e-05,
+      "loss": 0.9575,
+      "step": 84135
+    },
+    {
+      "epoch": 5.716809349096344,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4269941568147847e-05,
+      "loss": 0.9201,
+      "step": 84140
+    },
+    {
+      "epoch": 5.717149069167006,
+      "grad_norm": 9.0,
+      "learning_rate": 1.426781831770621e-05,
+      "loss": 0.8546,
+      "step": 84145
+    },
+    {
+      "epoch": 5.7174887892376685,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4265695067264576e-05,
+      "loss": 0.9066,
+      "step": 84150
+    },
+    {
+      "epoch": 5.71782850930833,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4263571816822937e-05,
+      "loss": 0.9701,
+      "step": 84155
+    },
+    {
+      "epoch": 5.718168229378992,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4261448566381303e-05,
+      "loss": 0.9132,
+      "step": 84160
+    },
+    {
+      "epoch": 5.718507949449654,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.4259325315939667e-05,
+      "loss": 0.9668,
+      "step": 84165
+    },
+    {
+      "epoch": 5.718847669520315,
+      "grad_norm": 8.625,
+      "learning_rate": 1.4257202065498029e-05,
+      "loss": 0.8319,
+      "step": 84170
+    },
+    {
+      "epoch": 5.719187389590977,
+      "grad_norm": 6.75,
+      "learning_rate": 1.4255078815056395e-05,
+      "loss": 0.678,
+      "step": 84175
+    },
+    {
+      "epoch": 5.719527109661639,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4252955564614759e-05,
+      "loss": 0.8333,
+      "step": 84180
+    },
+    {
+      "epoch": 5.7198668297323,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4250832314173121e-05,
+      "loss": 0.8419,
+      "step": 84185
+    },
+    {
+      "epoch": 5.720206549802962,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.4248709063731485e-05,
+      "loss": 0.847,
+      "step": 84190
+    },
+    {
+      "epoch": 5.7205462698736245,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.424658581328985e-05,
+      "loss": 0.7381,
+      "step": 84195
+    },
+    {
+      "epoch": 5.720885989944286,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.4244462562848213e-05,
+      "loss": 0.7894,
+      "step": 84200
+    },
+    {
+      "epoch": 5.721225710014948,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4242339312406577e-05,
+      "loss": 0.8228,
+      "step": 84205
+    },
+    {
+      "epoch": 5.72156543008561,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4240216061964943e-05,
+      "loss": 0.7716,
+      "step": 84210
+    },
+    {
+      "epoch": 5.721905150156271,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4238092811523305e-05,
+      "loss": 0.8207,
+      "step": 84215
+    },
+    {
+      "epoch": 5.722244870226933,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4235969561081669e-05,
+      "loss": 0.9161,
+      "step": 84220
+    },
+    {
+      "epoch": 5.722584590297595,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4233846310640035e-05,
+      "loss": 0.886,
+      "step": 84225
+    },
+    {
+      "epoch": 5.722924310368256,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4231723060198395e-05,
+      "loss": 0.9471,
+      "step": 84230
+    },
+    {
+      "epoch": 5.723264030438918,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4229599809756761e-05,
+      "loss": 0.8524,
+      "step": 84235
+    },
+    {
+      "epoch": 5.7236037505095805,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4227476559315125e-05,
+      "loss": 0.8689,
+      "step": 84240
+    },
+    {
+      "epoch": 5.723943470580242,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.422535330887349e-05,
+      "loss": 0.9101,
+      "step": 84245
+    },
+    {
+      "epoch": 5.724283190650904,
+      "grad_norm": 8.75,
+      "learning_rate": 1.4223230058431853e-05,
+      "loss": 0.7394,
+      "step": 84250
+    },
+    {
+      "epoch": 5.724622910721566,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.4221106807990217e-05,
+      "loss": 0.7699,
+      "step": 84255
+    },
+    {
+      "epoch": 5.724962630792227,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4218983557548583e-05,
+      "loss": 0.8593,
+      "step": 84260
+    },
+    {
+      "epoch": 5.725302350862889,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.4216860307106943e-05,
+      "loss": 0.817,
+      "step": 84265
+    },
+    {
+      "epoch": 5.725642070933551,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4214737056665309e-05,
+      "loss": 0.7778,
+      "step": 84270
+    },
+    {
+      "epoch": 5.725981791004212,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4212613806223673e-05,
+      "loss": 0.8814,
+      "step": 84275
+    },
+    {
+      "epoch": 5.726321511074874,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4210490555782035e-05,
+      "loss": 0.8556,
+      "step": 84280
+    },
+    {
+      "epoch": 5.7266612311455365,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4208367305340401e-05,
+      "loss": 0.7958,
+      "step": 84285
+    },
+    {
+      "epoch": 5.727000951216198,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4206244054898765e-05,
+      "loss": 0.9144,
+      "step": 84290
+    },
+    {
+      "epoch": 5.72734067128686,
+      "grad_norm": 7.125,
+      "learning_rate": 1.4204120804457127e-05,
+      "loss": 0.7889,
+      "step": 84295
+    },
+    {
+      "epoch": 5.727680391357522,
+      "grad_norm": 8.75,
+      "learning_rate": 1.4201997554015491e-05,
+      "loss": 0.8812,
+      "step": 84300
+    },
+    {
+      "epoch": 5.728020111428183,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4199874303573857e-05,
+      "loss": 0.8505,
+      "step": 84305
+    },
+    {
+      "epoch": 5.728359831498845,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.419775105313222e-05,
+      "loss": 0.8935,
+      "step": 84310
+    },
+    {
+      "epoch": 5.728699551569507,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4195627802690583e-05,
+      "loss": 0.9147,
+      "step": 84315
+    },
+    {
+      "epoch": 5.729039271640168,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4193504552248949e-05,
+      "loss": 0.8119,
+      "step": 84320
+    },
+    {
+      "epoch": 5.72937899171083,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4191381301807311e-05,
+      "loss": 0.8896,
+      "step": 84325
+    },
+    {
+      "epoch": 5.7297187117814925,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4189258051365675e-05,
+      "loss": 0.8312,
+      "step": 84330
+    },
+    {
+      "epoch": 5.730058431852154,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.418713480092404e-05,
+      "loss": 0.8194,
+      "step": 84335
+    },
+    {
+      "epoch": 5.730398151922816,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4185011550482402e-05,
+      "loss": 0.8987,
+      "step": 84340
+    },
+    {
+      "epoch": 5.730737871993478,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4182888300040767e-05,
+      "loss": 0.9526,
+      "step": 84345
+    },
+    {
+      "epoch": 5.731077592064139,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.4180765049599131e-05,
+      "loss": 0.9241,
+      "step": 84350
+    },
+    {
+      "epoch": 5.731417312134801,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4178641799157494e-05,
+      "loss": 0.8791,
+      "step": 84355
+    },
+    {
+      "epoch": 5.731757032205463,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.417651854871586e-05,
+      "loss": 0.8547,
+      "step": 84360
+    },
+    {
+      "epoch": 5.732096752276124,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4174395298274223e-05,
+      "loss": 0.8279,
+      "step": 84365
+    },
+    {
+      "epoch": 5.732436472346786,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4172272047832586e-05,
+      "loss": 0.8044,
+      "step": 84370
+    },
+    {
+      "epoch": 5.7327761924174485,
+      "grad_norm": 8.125,
+      "learning_rate": 1.417014879739095e-05,
+      "loss": 0.844,
+      "step": 84375
+    },
+    {
+      "epoch": 5.73311591248811,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4168025546949315e-05,
+      "loss": 0.8762,
+      "step": 84380
+    },
+    {
+      "epoch": 5.733455632558772,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4165902296507678e-05,
+      "loss": 0.843,
+      "step": 84385
+    },
+    {
+      "epoch": 5.733795352629433,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4163779046066042e-05,
+      "loss": 0.8387,
+      "step": 84390
+    },
+    {
+      "epoch": 5.734135072700095,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.4161655795624407e-05,
+      "loss": 0.8302,
+      "step": 84395
+    },
+    {
+      "epoch": 5.734474792770757,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4159532545182768e-05,
+      "loss": 0.8949,
+      "step": 84400
+    },
+    {
+      "epoch": 5.734814512841418,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4157409294741134e-05,
+      "loss": 0.8519,
+      "step": 84405
+    },
+    {
+      "epoch": 5.73515423291208,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4155286044299498e-05,
+      "loss": 0.9789,
+      "step": 84410
+    },
+    {
+      "epoch": 5.7354939529827424,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4153162793857863e-05,
+      "loss": 0.8058,
+      "step": 84415
+    },
+    {
+      "epoch": 5.735833673053404,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4151039543416226e-05,
+      "loss": 0.8439,
+      "step": 84420
+    },
+    {
+      "epoch": 5.736173393124066,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.414891629297459e-05,
+      "loss": 0.978,
+      "step": 84425
+    },
+    {
+      "epoch": 5.736513113194728,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4146793042532955e-05,
+      "loss": 0.8086,
+      "step": 84430
+    },
+    {
+      "epoch": 5.736852833265389,
+      "grad_norm": 10.0,
+      "learning_rate": 1.4144669792091316e-05,
+      "loss": 0.8008,
+      "step": 84435
+    },
+    {
+      "epoch": 5.737192553336051,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4142546541649682e-05,
+      "loss": 0.9548,
+      "step": 84440
+    },
+    {
+      "epoch": 5.737532273406713,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4140423291208046e-05,
+      "loss": 0.8596,
+      "step": 84445
+    },
+    {
+      "epoch": 5.737871993477374,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4138300040766408e-05,
+      "loss": 0.9111,
+      "step": 84450
+    },
+    {
+      "epoch": 5.738211713548036,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4136176790324774e-05,
+      "loss": 0.9095,
+      "step": 84455
+    },
+    {
+      "epoch": 5.7385514336186985,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4134053539883138e-05,
+      "loss": 0.842,
+      "step": 84460
+    },
+    {
+      "epoch": 5.73889115368936,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.41319302894415e-05,
+      "loss": 0.8965,
+      "step": 84465
+    },
+    {
+      "epoch": 5.739230873760022,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4129807038999864e-05,
+      "loss": 0.8468,
+      "step": 84470
+    },
+    {
+      "epoch": 5.739570593830684,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.412768378855823e-05,
+      "loss": 0.9266,
+      "step": 84475
+    },
+    {
+      "epoch": 5.739910313901345,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4125560538116592e-05,
+      "loss": 0.7199,
+      "step": 84480
+    },
+    {
+      "epoch": 5.740250033972007,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4123437287674956e-05,
+      "loss": 0.8504,
+      "step": 84485
+    },
+    {
+      "epoch": 5.740589754042669,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4121314037233322e-05,
+      "loss": 0.9257,
+      "step": 84490
+    },
+    {
+      "epoch": 5.74092947411333,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4119190786791684e-05,
+      "loss": 0.8958,
+      "step": 84495
+    },
+    {
+      "epoch": 5.741269194183992,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4117067536350048e-05,
+      "loss": 0.8423,
+      "step": 84500
+    },
+    {
+      "epoch": 5.7416089142546545,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4114944285908414e-05,
+      "loss": 0.772,
+      "step": 84505
+    },
+    {
+      "epoch": 5.741948634325316,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4112821035466774e-05,
+      "loss": 0.9286,
+      "step": 84510
+    },
+    {
+      "epoch": 5.742288354395978,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.411069778502514e-05,
+      "loss": 0.8437,
+      "step": 84515
+    },
+    {
+      "epoch": 5.74262807446664,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4108574534583504e-05,
+      "loss": 0.8386,
+      "step": 84520
+    },
+    {
+      "epoch": 5.742967794537301,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4106451284141866e-05,
+      "loss": 0.8816,
+      "step": 84525
+    },
+    {
+      "epoch": 5.743307514607963,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4104328033700232e-05,
+      "loss": 0.8825,
+      "step": 84530
+    },
+    {
+      "epoch": 5.743647234678625,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.4102204783258596e-05,
+      "loss": 0.8662,
+      "step": 84535
+    },
+    {
+      "epoch": 5.743986954749286,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.4100081532816958e-05,
+      "loss": 0.945,
+      "step": 84540
+    },
+    {
+      "epoch": 5.744326674819948,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.4097958282375322e-05,
+      "loss": 0.8966,
+      "step": 84545
+    },
+    {
+      "epoch": 5.7446663948906105,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4095835031933688e-05,
+      "loss": 0.7464,
+      "step": 84550
+    },
+    {
+      "epoch": 5.745006114961272,
+      "grad_norm": 8.125,
+      "learning_rate": 1.409371178149205e-05,
+      "loss": 0.7627,
+      "step": 84555
+    },
+    {
+      "epoch": 5.745345835031934,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4091588531050414e-05,
+      "loss": 0.9386,
+      "step": 84560
+    },
+    {
+      "epoch": 5.745685555102596,
+      "grad_norm": 8.125,
+      "learning_rate": 1.408946528060878e-05,
+      "loss": 0.7555,
+      "step": 84565
+    },
+    {
+      "epoch": 5.746025275173257,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.408734203016714e-05,
+      "loss": 0.8949,
+      "step": 84570
+    },
+    {
+      "epoch": 5.746364995243919,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4085218779725506e-05,
+      "loss": 0.8015,
+      "step": 84575
+    },
+    {
+      "epoch": 5.746704715314581,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.408309552928387e-05,
+      "loss": 0.8653,
+      "step": 84580
+    },
+    {
+      "epoch": 5.747044435385242,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4080972278842236e-05,
+      "loss": 0.8197,
+      "step": 84585
+    },
+    {
+      "epoch": 5.747384155455904,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4078849028400598e-05,
+      "loss": 0.9069,
+      "step": 84590
+    },
+    {
+      "epoch": 5.7477238755265665,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4076725777958962e-05,
+      "loss": 0.8633,
+      "step": 84595
+    },
+    {
+      "epoch": 5.748063595597228,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4074602527517328e-05,
+      "loss": 0.8666,
+      "step": 84600
+    },
+    {
+      "epoch": 5.74840331566789,
+      "grad_norm": 7.25,
+      "learning_rate": 1.407247927707569e-05,
+      "loss": 0.8915,
+      "step": 84605
+    },
+    {
+      "epoch": 5.748743035738551,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4070356026634054e-05,
+      "loss": 0.8816,
+      "step": 84610
+    },
+    {
+      "epoch": 5.749082755809213,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4068232776192418e-05,
+      "loss": 0.8769,
+      "step": 84615
+    },
+    {
+      "epoch": 5.749422475879875,
+      "grad_norm": 7.375,
+      "learning_rate": 1.406610952575078e-05,
+      "loss": 0.7929,
+      "step": 84620
+    },
+    {
+      "epoch": 5.749762195950536,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4063986275309147e-05,
+      "loss": 0.9293,
+      "step": 84625
+    },
+    {
+      "epoch": 5.750101916021198,
+      "grad_norm": 8.75,
+      "learning_rate": 1.406186302486751e-05,
+      "loss": 0.9187,
+      "step": 84630
+    },
+    {
+      "epoch": 5.75044163609186,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4059739774425873e-05,
+      "loss": 0.8813,
+      "step": 84635
+    },
+    {
+      "epoch": 5.750781356162522,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4057616523984239e-05,
+      "loss": 0.9056,
+      "step": 84640
+    },
+    {
+      "epoch": 5.751121076233184,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4055493273542603e-05,
+      "loss": 0.8793,
+      "step": 84645
+    },
+    {
+      "epoch": 5.751460796303846,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4053370023100965e-05,
+      "loss": 0.8231,
+      "step": 84650
+    },
+    {
+      "epoch": 5.751800516374507,
+      "grad_norm": 8.375,
+      "learning_rate": 1.4051246772659329e-05,
+      "loss": 0.8934,
+      "step": 84655
+    },
+    {
+      "epoch": 5.752140236445169,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4049123522217695e-05,
+      "loss": 0.8948,
+      "step": 84660
+    },
+    {
+      "epoch": 5.752479956515831,
+      "grad_norm": 8.75,
+      "learning_rate": 1.4047000271776057e-05,
+      "loss": 0.806,
+      "step": 84665
+    },
+    {
+      "epoch": 5.752819676586492,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.4044877021334421e-05,
+      "loss": 0.941,
+      "step": 84670
+    },
+    {
+      "epoch": 5.753159396657154,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4042753770892787e-05,
+      "loss": 0.856,
+      "step": 84675
+    },
+    {
+      "epoch": 5.753499116727816,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.4040630520451147e-05,
+      "loss": 0.8334,
+      "step": 84680
+    },
+    {
+      "epoch": 5.753838836798478,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.4038507270009513e-05,
+      "loss": 0.7592,
+      "step": 84685
+    },
+    {
+      "epoch": 5.75417855686914,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4036384019567877e-05,
+      "loss": 0.8554,
+      "step": 84690
+    },
+    {
+      "epoch": 5.754518276939802,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.403426076912624e-05,
+      "loss": 0.8406,
+      "step": 84695
+    },
+    {
+      "epoch": 5.754857997010463,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4032137518684605e-05,
+      "loss": 0.8082,
+      "step": 84700
+    },
+    {
+      "epoch": 5.755197717081125,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.4030014268242969e-05,
+      "loss": 0.8353,
+      "step": 84705
+    },
+    {
+      "epoch": 5.755537437151787,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4027891017801331e-05,
+      "loss": 0.7941,
+      "step": 84710
+    },
+    {
+      "epoch": 5.755877157222448,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4025767767359695e-05,
+      "loss": 0.8197,
+      "step": 84715
+    },
+    {
+      "epoch": 5.75621687729311,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4023644516918061e-05,
+      "loss": 0.8068,
+      "step": 84720
+    },
+    {
+      "epoch": 5.7565565973637725,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4021521266476423e-05,
+      "loss": 0.7882,
+      "step": 84725
+    },
+    {
+      "epoch": 5.756896317434434,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4019398016034787e-05,
+      "loss": 0.794,
+      "step": 84730
+    },
+    {
+      "epoch": 5.757236037505096,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4017274765593153e-05,
+      "loss": 0.894,
+      "step": 84735
+    },
+    {
+      "epoch": 5.757575757575758,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4015151515151515e-05,
+      "loss": 0.7983,
+      "step": 84740
+    },
+    {
+      "epoch": 5.757915477646419,
+      "grad_norm": 8.25,
+      "learning_rate": 1.401302826470988e-05,
+      "loss": 0.8898,
+      "step": 84745
+    },
+    {
+      "epoch": 5.758255197717081,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.4010905014268245e-05,
+      "loss": 0.864,
+      "step": 84750
+    },
+    {
+      "epoch": 5.758594917787743,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4008781763826609e-05,
+      "loss": 0.9133,
+      "step": 84755
+    },
+    {
+      "epoch": 5.758934637858404,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.4006658513384971e-05,
+      "loss": 0.8266,
+      "step": 84760
+    },
+    {
+      "epoch": 5.759274357929066,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4004535262943335e-05,
+      "loss": 0.8502,
+      "step": 84765
+    },
+    {
+      "epoch": 5.7596140779997285,
+      "grad_norm": 8.0,
+      "learning_rate": 1.4002412012501701e-05,
+      "loss": 0.8787,
+      "step": 84770
+    },
+    {
+      "epoch": 5.75995379807039,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4000288762060063e-05,
+      "loss": 0.764,
+      "step": 84775
+    },
+    {
+      "epoch": 5.760293518141052,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3998165511618427e-05,
+      "loss": 0.9058,
+      "step": 84780
+    },
+    {
+      "epoch": 5.760633238211714,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3996042261176793e-05,
+      "loss": 0.7814,
+      "step": 84785
+    },
+    {
+      "epoch": 5.760972958282375,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3993919010735154e-05,
+      "loss": 0.7543,
+      "step": 84790
+    },
+    {
+      "epoch": 5.761312678353037,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.399179576029352e-05,
+      "loss": 0.871,
+      "step": 84795
+    },
+    {
+      "epoch": 5.761652398423699,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3989672509851883e-05,
+      "loss": 0.7305,
+      "step": 84800
+    },
+    {
+      "epoch": 5.76199211849436,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3987549259410246e-05,
+      "loss": 0.7743,
+      "step": 84805
+    },
+    {
+      "epoch": 5.762331838565022,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3985426008968611e-05,
+      "loss": 0.7635,
+      "step": 84810
+    },
+    {
+      "epoch": 5.7626715586356845,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3983302758526975e-05,
+      "loss": 0.8157,
+      "step": 84815
+    },
+    {
+      "epoch": 5.763011278706346,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3981179508085338e-05,
+      "loss": 0.9101,
+      "step": 84820
+    },
+    {
+      "epoch": 5.763350998777008,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.3979056257643702e-05,
+      "loss": 0.8496,
+      "step": 84825
+    },
+    {
+      "epoch": 5.76369071884767,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.3976933007202067e-05,
+      "loss": 0.7915,
+      "step": 84830
+    },
+    {
+      "epoch": 5.764030438918331,
+      "grad_norm": 9.0,
+      "learning_rate": 1.397480975676043e-05,
+      "loss": 0.8599,
+      "step": 84835
+    },
+    {
+      "epoch": 5.764370158988993,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3972686506318794e-05,
+      "loss": 0.8609,
+      "step": 84840
+    },
+    {
+      "epoch": 5.764709879059655,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.397056325587716e-05,
+      "loss": 0.8007,
+      "step": 84845
+    },
+    {
+      "epoch": 5.765049599130316,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3968440005435522e-05,
+      "loss": 0.8052,
+      "step": 84850
+    },
+    {
+      "epoch": 5.765389319200978,
+      "grad_norm": 6.28125,
+      "learning_rate": 1.3966316754993886e-05,
+      "loss": 0.8132,
+      "step": 84855
+    },
+    {
+      "epoch": 5.7657290392716405,
+      "grad_norm": 8.25,
+      "learning_rate": 1.396419350455225e-05,
+      "loss": 0.9169,
+      "step": 84860
+    },
+    {
+      "epoch": 5.766068759342302,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3962070254110612e-05,
+      "loss": 0.8738,
+      "step": 84865
+    },
+    {
+      "epoch": 5.766408479412964,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3959947003668978e-05,
+      "loss": 0.9165,
+      "step": 84870
+    },
+    {
+      "epoch": 5.766748199483626,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3957823753227342e-05,
+      "loss": 0.828,
+      "step": 84875
+    },
+    {
+      "epoch": 5.767087919554287,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3955700502785704e-05,
+      "loss": 0.8428,
+      "step": 84880
+    },
+    {
+      "epoch": 5.767427639624949,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.395357725234407e-05,
+      "loss": 0.8468,
+      "step": 84885
+    },
+    {
+      "epoch": 5.767767359695611,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3951454001902434e-05,
+      "loss": 0.8578,
+      "step": 84890
+    },
+    {
+      "epoch": 5.768107079766272,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3949330751460796e-05,
+      "loss": 0.8955,
+      "step": 84895
+    },
+    {
+      "epoch": 5.768446799836934,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.394720750101916e-05,
+      "loss": 0.7667,
+      "step": 84900
+    },
+    {
+      "epoch": 5.7687865199075965,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3945084250577526e-05,
+      "loss": 0.8974,
+      "step": 84905
+    },
+    {
+      "epoch": 5.769126239978258,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3942961000135888e-05,
+      "loss": 0.8841,
+      "step": 84910
+    },
+    {
+      "epoch": 5.76946596004892,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.3940837749694252e-05,
+      "loss": 0.8781,
+      "step": 84915
+    },
+    {
+      "epoch": 5.769805680119582,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3938714499252618e-05,
+      "loss": 0.8199,
+      "step": 84920
+    },
+    {
+      "epoch": 5.770145400190243,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3936591248810982e-05,
+      "loss": 0.8418,
+      "step": 84925
+    },
+    {
+      "epoch": 5.770485120260905,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3934467998369344e-05,
+      "loss": 0.8664,
+      "step": 84930
+    },
+    {
+      "epoch": 5.770824840331567,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3932344747927708e-05,
+      "loss": 0.9838,
+      "step": 84935
+    },
+    {
+      "epoch": 5.771164560402228,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3930221497486074e-05,
+      "loss": 0.8298,
+      "step": 84940
+    },
+    {
+      "epoch": 5.77150428047289,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3928098247044436e-05,
+      "loss": 0.7857,
+      "step": 84945
+    },
+    {
+      "epoch": 5.7718440005435525,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.39259749966028e-05,
+      "loss": 0.8307,
+      "step": 84950
+    },
+    {
+      "epoch": 5.772183720614214,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3923851746161166e-05,
+      "loss": 0.8976,
+      "step": 84955
+    },
+    {
+      "epoch": 5.772523440684876,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3921728495719526e-05,
+      "loss": 0.8747,
+      "step": 84960
+    },
+    {
+      "epoch": 5.772863160755538,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3919605245277892e-05,
+      "loss": 0.8532,
+      "step": 84965
+    },
+    {
+      "epoch": 5.773202880826199,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3917481994836256e-05,
+      "loss": 0.844,
+      "step": 84970
+    },
+    {
+      "epoch": 5.773542600896861,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.3915358744394618e-05,
+      "loss": 0.9026,
+      "step": 84975
+    },
+    {
+      "epoch": 5.773882320967523,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3913235493952984e-05,
+      "loss": 0.8449,
+      "step": 84980
+    },
+    {
+      "epoch": 5.774222041038184,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3911112243511348e-05,
+      "loss": 0.8486,
+      "step": 84985
+    },
+    {
+      "epoch": 5.7745617611088464,
+      "grad_norm": 7.75,
+      "learning_rate": 1.390898899306971e-05,
+      "loss": 0.8536,
+      "step": 84990
+    },
+    {
+      "epoch": 5.7749014811795085,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3906865742628076e-05,
+      "loss": 0.8537,
+      "step": 84995
+    },
+    {
+      "epoch": 5.77524120125017,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.390474249218644e-05,
+      "loss": 0.8154,
+      "step": 85000
+    },
+    {
+      "epoch": 5.775580921320832,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3902619241744802e-05,
+      "loss": 0.8441,
+      "step": 85005
+    },
+    {
+      "epoch": 5.775920641391494,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3900495991303166e-05,
+      "loss": 0.8973,
+      "step": 85010
+    },
+    {
+      "epoch": 5.776260361462155,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3898372740861532e-05,
+      "loss": 0.8323,
+      "step": 85015
+    },
+    {
+      "epoch": 5.776600081532817,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3896249490419894e-05,
+      "loss": 0.8432,
+      "step": 85020
+    },
+    {
+      "epoch": 5.776939801603479,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3894126239978258e-05,
+      "loss": 0.8881,
+      "step": 85025
+    },
+    {
+      "epoch": 5.77727952167414,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.3892002989536624e-05,
+      "loss": 0.7697,
+      "step": 85030
+    },
+    {
+      "epoch": 5.7776192417448025,
+      "grad_norm": 6.875,
+      "learning_rate": 1.3889879739094985e-05,
+      "loss": 0.9904,
+      "step": 85035
+    },
+    {
+      "epoch": 5.7779589618154645,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.388775648865335e-05,
+      "loss": 0.8898,
+      "step": 85040
+    },
+    {
+      "epoch": 5.778298681886126,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3885633238211714e-05,
+      "loss": 0.893,
+      "step": 85045
+    },
+    {
+      "epoch": 5.778638401956788,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3883509987770077e-05,
+      "loss": 0.841,
+      "step": 85050
+    },
+    {
+      "epoch": 5.77897812202745,
+      "grad_norm": 8.75,
+      "learning_rate": 1.3881386737328442e-05,
+      "loss": 0.8598,
+      "step": 85055
+    },
+    {
+      "epoch": 5.779317842098111,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3879263486886806e-05,
+      "loss": 0.817,
+      "step": 85060
+    },
+    {
+      "epoch": 5.779657562168773,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3877140236445169e-05,
+      "loss": 0.8741,
+      "step": 85065
+    },
+    {
+      "epoch": 5.779997282239434,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3875016986003533e-05,
+      "loss": 0.8219,
+      "step": 85070
+    },
+    {
+      "epoch": 5.780337002310096,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3872893735561898e-05,
+      "loss": 0.8478,
+      "step": 85075
+    },
+    {
+      "epoch": 5.7806767223807585,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.387077048512026e-05,
+      "loss": 0.8702,
+      "step": 85080
+    },
+    {
+      "epoch": 5.78101644245142,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3868647234678625e-05,
+      "loss": 0.8456,
+      "step": 85085
+    },
+    {
+      "epoch": 5.781356162522082,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.386652398423699e-05,
+      "loss": 0.911,
+      "step": 85090
+    },
+    {
+      "epoch": 5.781695882592744,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3864400733795354e-05,
+      "loss": 0.9226,
+      "step": 85095
+    },
+    {
+      "epoch": 5.782035602663405,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.3862277483353717e-05,
+      "loss": 0.8092,
+      "step": 85100
+    },
+    {
+      "epoch": 5.782375322734067,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.386015423291208e-05,
+      "loss": 0.8288,
+      "step": 85105
+    },
+    {
+      "epoch": 5.782715042804729,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3858030982470446e-05,
+      "loss": 0.8797,
+      "step": 85110
+    },
+    {
+      "epoch": 5.78305476287539,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3855907732028809e-05,
+      "loss": 0.8685,
+      "step": 85115
+    },
+    {
+      "epoch": 5.783394482946052,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.3853784481587173e-05,
+      "loss": 0.7999,
+      "step": 85120
+    },
+    {
+      "epoch": 5.7837342030167145,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3851661231145538e-05,
+      "loss": 0.7902,
+      "step": 85125
+    },
+    {
+      "epoch": 5.784073923087376,
+      "grad_norm": 7.125,
+      "learning_rate": 1.38495379807039e-05,
+      "loss": 0.8086,
+      "step": 85130
+    },
+    {
+      "epoch": 5.784413643158038,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3847414730262265e-05,
+      "loss": 0.8307,
+      "step": 85135
+    },
+    {
+      "epoch": 5.7847533632287,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3845291479820629e-05,
+      "loss": 0.8158,
+      "step": 85140
+    },
+    {
+      "epoch": 5.785093083299361,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.3843168229378991e-05,
+      "loss": 0.8128,
+      "step": 85145
+    },
+    {
+      "epoch": 5.785432803370023,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.3841044978937357e-05,
+      "loss": 0.8479,
+      "step": 85150
+    },
+    {
+      "epoch": 5.785772523440685,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.383892172849572e-05,
+      "loss": 0.8525,
+      "step": 85155
+    },
+    {
+      "epoch": 5.786112243511346,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3836798478054083e-05,
+      "loss": 0.8514,
+      "step": 85160
+    },
+    {
+      "epoch": 5.786451963582008,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3834675227612449e-05,
+      "loss": 0.7943,
+      "step": 85165
+    },
+    {
+      "epoch": 5.7867916836526705,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3832551977170813e-05,
+      "loss": 0.8848,
+      "step": 85170
+    },
+    {
+      "epoch": 5.787131403723332,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3830428726729175e-05,
+      "loss": 0.8509,
+      "step": 85175
+    },
+    {
+      "epoch": 5.787471123793994,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3828305476287539e-05,
+      "loss": 0.8283,
+      "step": 85180
+    },
+    {
+      "epoch": 5.787810843864656,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.3826182225845905e-05,
+      "loss": 0.6867,
+      "step": 85185
+    },
+    {
+      "epoch": 5.788150563935317,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3824058975404267e-05,
+      "loss": 0.9745,
+      "step": 85190
+    },
+    {
+      "epoch": 5.788490284005979,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3821935724962631e-05,
+      "loss": 0.8336,
+      "step": 85195
+    },
+    {
+      "epoch": 5.788830004076641,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3819812474520997e-05,
+      "loss": 0.779,
+      "step": 85200
+    },
+    {
+      "epoch": 5.789169724147302,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3817689224079357e-05,
+      "loss": 0.9652,
+      "step": 85205
+    },
+    {
+      "epoch": 5.789509444217964,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3815565973637723e-05,
+      "loss": 0.8867,
+      "step": 85210
+    },
+    {
+      "epoch": 5.7898491642886265,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.3813442723196087e-05,
+      "loss": 0.8386,
+      "step": 85215
+    },
+    {
+      "epoch": 5.790188884359288,
+      "grad_norm": 7.375,
+      "learning_rate": 1.381131947275445e-05,
+      "loss": 0.8138,
+      "step": 85220
+    },
+    {
+      "epoch": 5.79052860442995,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3809196222312815e-05,
+      "loss": 0.7658,
+      "step": 85225
+    },
+    {
+      "epoch": 5.790868324500612,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3807072971871179e-05,
+      "loss": 0.9285,
+      "step": 85230
+    },
+    {
+      "epoch": 5.791208044571273,
+      "grad_norm": 7.0,
+      "learning_rate": 1.3804949721429541e-05,
+      "loss": 0.7991,
+      "step": 85235
+    },
+    {
+      "epoch": 5.791547764641935,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.3802826470987905e-05,
+      "loss": 0.8148,
+      "step": 85240
+    },
+    {
+      "epoch": 5.791887484712597,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3800703220546271e-05,
+      "loss": 0.8155,
+      "step": 85245
+    },
+    {
+      "epoch": 5.792227204783258,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3798579970104633e-05,
+      "loss": 0.8825,
+      "step": 85250
+    },
+    {
+      "epoch": 5.79256692485392,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.3796456719662997e-05,
+      "loss": 0.8125,
+      "step": 85255
+    },
+    {
+      "epoch": 5.7929066449245825,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3794333469221363e-05,
+      "loss": 0.8565,
+      "step": 85260
+    },
+    {
+      "epoch": 5.793246364995244,
+      "grad_norm": 9.25,
+      "learning_rate": 1.3792210218779727e-05,
+      "loss": 0.7403,
+      "step": 85265
+    },
+    {
+      "epoch": 5.793586085065906,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.379008696833809e-05,
+      "loss": 0.7931,
+      "step": 85270
+    },
+    {
+      "epoch": 5.793925805136568,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3787963717896455e-05,
+      "loss": 0.8371,
+      "step": 85275
+    },
+    {
+      "epoch": 5.794265525207229,
+      "grad_norm": 7.75,
+      "learning_rate": 1.378584046745482e-05,
+      "loss": 0.9598,
+      "step": 85280
+    },
+    {
+      "epoch": 5.794605245277891,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3783717217013181e-05,
+      "loss": 0.8589,
+      "step": 85285
+    },
+    {
+      "epoch": 5.794944965348552,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3781593966571545e-05,
+      "loss": 0.8387,
+      "step": 85290
+    },
+    {
+      "epoch": 5.795284685419214,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3779470716129911e-05,
+      "loss": 0.835,
+      "step": 85295
+    },
+    {
+      "epoch": 5.7956244054898765,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3777347465688273e-05,
+      "loss": 0.8429,
+      "step": 85300
+    },
+    {
+      "epoch": 5.795964125560538,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3775224215246637e-05,
+      "loss": 0.8443,
+      "step": 85305
+    },
+    {
+      "epoch": 5.7963038456312,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3773100964805003e-05,
+      "loss": 0.8626,
+      "step": 85310
+    },
+    {
+      "epoch": 5.796643565701862,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3770977714363364e-05,
+      "loss": 0.8586,
+      "step": 85315
+    },
+    {
+      "epoch": 5.796983285772523,
+      "grad_norm": 7.0,
+      "learning_rate": 1.376885446392173e-05,
+      "loss": 0.9204,
+      "step": 85320
+    },
+    {
+      "epoch": 5.797323005843185,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3766731213480093e-05,
+      "loss": 0.8784,
+      "step": 85325
+    },
+    {
+      "epoch": 5.797662725913847,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3764607963038456e-05,
+      "loss": 0.9017,
+      "step": 85330
+    },
+    {
+      "epoch": 5.798002445984508,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3762484712596822e-05,
+      "loss": 1.0168,
+      "step": 85335
+    },
+    {
+      "epoch": 5.79834216605517,
+      "grad_norm": 9.125,
+      "learning_rate": 1.3760361462155186e-05,
+      "loss": 0.823,
+      "step": 85340
+    },
+    {
+      "epoch": 5.7986818861258325,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3758238211713548e-05,
+      "loss": 0.8465,
+      "step": 85345
+    },
+    {
+      "epoch": 5.799021606196494,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3756114961271912e-05,
+      "loss": 0.7465,
+      "step": 85350
+    },
+    {
+      "epoch": 5.799361326267156,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3753991710830278e-05,
+      "loss": 0.9111,
+      "step": 85355
+    },
+    {
+      "epoch": 5.799701046337818,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.375186846038864e-05,
+      "loss": 0.8251,
+      "step": 85360
+    },
+    {
+      "epoch": 5.800040766408479,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3749745209947004e-05,
+      "loss": 0.8092,
+      "step": 85365
+    },
+    {
+      "epoch": 5.800380486479141,
+      "grad_norm": 8.375,
+      "learning_rate": 1.374762195950537e-05,
+      "loss": 0.8219,
+      "step": 85370
+    },
+    {
+      "epoch": 5.800720206549803,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.3745498709063732e-05,
+      "loss": 0.8936,
+      "step": 85375
+    },
+    {
+      "epoch": 5.801059926620464,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3743375458622096e-05,
+      "loss": 0.8226,
+      "step": 85380
+    },
+    {
+      "epoch": 5.801399646691126,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.374125220818046e-05,
+      "loss": 0.851,
+      "step": 85385
+    },
+    {
+      "epoch": 5.8017393667617885,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3739128957738822e-05,
+      "loss": 0.842,
+      "step": 85390
+    },
+    {
+      "epoch": 5.80207908683245,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.3737005707297188e-05,
+      "loss": 0.8376,
+      "step": 85395
+    },
+    {
+      "epoch": 5.802418806903112,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3734882456855552e-05,
+      "loss": 0.8051,
+      "step": 85400
+    },
+    {
+      "epoch": 5.802758526973774,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3732759206413914e-05,
+      "loss": 0.8314,
+      "step": 85405
+    },
+    {
+      "epoch": 5.803098247044435,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.373063595597228e-05,
+      "loss": 0.8592,
+      "step": 85410
+    },
+    {
+      "epoch": 5.803437967115097,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3728512705530644e-05,
+      "loss": 0.7225,
+      "step": 85415
+    },
+    {
+      "epoch": 5.803777687185759,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.3726389455089006e-05,
+      "loss": 0.7935,
+      "step": 85420
+    },
+    {
+      "epoch": 5.80411740725642,
+      "grad_norm": 7.25,
+      "learning_rate": 1.372426620464737e-05,
+      "loss": 0.8366,
+      "step": 85425
+    },
+    {
+      "epoch": 5.804457127327082,
+      "grad_norm": 8.75,
+      "learning_rate": 1.3722142954205736e-05,
+      "loss": 0.8395,
+      "step": 85430
+    },
+    {
+      "epoch": 5.8047968473977445,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.37200197037641e-05,
+      "loss": 0.8804,
+      "step": 85435
+    },
+    {
+      "epoch": 5.805136567468406,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3717896453322462e-05,
+      "loss": 0.8614,
+      "step": 85440
+    },
+    {
+      "epoch": 5.805476287539068,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3715773202880828e-05,
+      "loss": 0.8224,
+      "step": 85445
+    },
+    {
+      "epoch": 5.80581600760973,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3713649952439192e-05,
+      "loss": 0.7804,
+      "step": 85450
+    },
+    {
+      "epoch": 5.806155727680391,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3711526701997554e-05,
+      "loss": 0.9394,
+      "step": 85455
+    },
+    {
+      "epoch": 5.806495447751053,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3709403451555918e-05,
+      "loss": 0.8395,
+      "step": 85460
+    },
+    {
+      "epoch": 5.806835167821715,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3707280201114284e-05,
+      "loss": 0.9624,
+      "step": 85465
+    },
+    {
+      "epoch": 5.807174887892376,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3705156950672646e-05,
+      "loss": 0.8426,
+      "step": 85470
+    },
+    {
+      "epoch": 5.807514607963038,
+      "grad_norm": 8.375,
+      "learning_rate": 1.370303370023101e-05,
+      "loss": 0.8799,
+      "step": 85475
+    },
+    {
+      "epoch": 5.8078543280337005,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3700910449789376e-05,
+      "loss": 0.8822,
+      "step": 85480
+    },
+    {
+      "epoch": 5.808194048104362,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3698787199347737e-05,
+      "loss": 0.8899,
+      "step": 85485
+    },
+    {
+      "epoch": 5.808533768175024,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3696663948906102e-05,
+      "loss": 0.7753,
+      "step": 85490
+    },
+    {
+      "epoch": 5.808873488245686,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3694540698464466e-05,
+      "loss": 0.8954,
+      "step": 85495
+    },
+    {
+      "epoch": 5.809213208316347,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.3692417448022829e-05,
+      "loss": 0.8698,
+      "step": 85500
+    },
+    {
+      "epoch": 5.809552928387009,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3690294197581194e-05,
+      "loss": 0.8044,
+      "step": 85505
+    },
+    {
+      "epoch": 5.809892648457671,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3688170947139558e-05,
+      "loss": 0.8679,
+      "step": 85510
+    },
+    {
+      "epoch": 5.810232368528332,
+      "grad_norm": 7.75,
+      "learning_rate": 1.368604769669792e-05,
+      "loss": 0.8874,
+      "step": 85515
+    },
+    {
+      "epoch": 5.810572088598994,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3683924446256286e-05,
+      "loss": 0.7946,
+      "step": 85520
+    },
+    {
+      "epoch": 5.8109118086696565,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.368180119581465e-05,
+      "loss": 0.8646,
+      "step": 85525
+    },
+    {
+      "epoch": 5.811251528740318,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3679677945373013e-05,
+      "loss": 0.8344,
+      "step": 85530
+    },
+    {
+      "epoch": 5.81159124881098,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3677554694931377e-05,
+      "loss": 0.9163,
+      "step": 85535
+    },
+    {
+      "epoch": 5.811930968881642,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3675431444489742e-05,
+      "loss": 0.8262,
+      "step": 85540
+    },
+    {
+      "epoch": 5.812270688952303,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3673308194048105e-05,
+      "loss": 0.8427,
+      "step": 85545
+    },
+    {
+      "epoch": 5.812610409022965,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.3671184943606469e-05,
+      "loss": 0.812,
+      "step": 85550
+    },
+    {
+      "epoch": 5.812950129093627,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3669061693164834e-05,
+      "loss": 0.8757,
+      "step": 85555
+    },
+    {
+      "epoch": 5.813289849164288,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3666938442723195e-05,
+      "loss": 0.8409,
+      "step": 85560
+    },
+    {
+      "epoch": 5.8136295692349504,
+      "grad_norm": 9.125,
+      "learning_rate": 1.366481519228156e-05,
+      "loss": 0.8379,
+      "step": 85565
+    },
+    {
+      "epoch": 5.8139692893056125,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3662691941839925e-05,
+      "loss": 0.8511,
+      "step": 85570
+    },
+    {
+      "epoch": 5.814309009376274,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3660568691398287e-05,
+      "loss": 0.8458,
+      "step": 85575
+    },
+    {
+      "epoch": 5.814648729446936,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3658445440956653e-05,
+      "loss": 0.88,
+      "step": 85580
+    },
+    {
+      "epoch": 5.814988449517598,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3656322190515017e-05,
+      "loss": 0.8722,
+      "step": 85585
+    },
+    {
+      "epoch": 5.815328169588259,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3654198940073379e-05,
+      "loss": 0.865,
+      "step": 85590
+    },
+    {
+      "epoch": 5.815667889658921,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3652075689631743e-05,
+      "loss": 0.7913,
+      "step": 85595
+    },
+    {
+      "epoch": 5.816007609729583,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3649952439190109e-05,
+      "loss": 0.8867,
+      "step": 85600
+    },
+    {
+      "epoch": 5.816347329800244,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3647829188748473e-05,
+      "loss": 0.8508,
+      "step": 85605
+    },
+    {
+      "epoch": 5.8166870498709065,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3645705938306835e-05,
+      "loss": 0.8804,
+      "step": 85610
+    },
+    {
+      "epoch": 5.8170267699415685,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.36435826878652e-05,
+      "loss": 0.8023,
+      "step": 85615
+    },
+    {
+      "epoch": 5.81736649001223,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.3641459437423565e-05,
+      "loss": 0.8481,
+      "step": 85620
+    },
+    {
+      "epoch": 5.817706210082892,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3639336186981927e-05,
+      "loss": 0.7864,
+      "step": 85625
+    },
+    {
+      "epoch": 5.818045930153554,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3637212936540291e-05,
+      "loss": 0.9318,
+      "step": 85630
+    },
+    {
+      "epoch": 5.818385650224215,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3635089686098657e-05,
+      "loss": 0.897,
+      "step": 85635
+    },
+    {
+      "epoch": 5.818725370294877,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3632966435657019e-05,
+      "loss": 0.8602,
+      "step": 85640
+    },
+    {
+      "epoch": 5.819065090365539,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3630843185215383e-05,
+      "loss": 0.8219,
+      "step": 85645
+    },
+    {
+      "epoch": 5.8194048104362,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3628719934773749e-05,
+      "loss": 0.9291,
+      "step": 85650
+    },
+    {
+      "epoch": 5.8197445305068625,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3626596684332111e-05,
+      "loss": 0.9336,
+      "step": 85655
+    },
+    {
+      "epoch": 5.8200842505775245,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3624473433890475e-05,
+      "loss": 0.8141,
+      "step": 85660
+    },
+    {
+      "epoch": 5.820423970648186,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.362235018344884e-05,
+      "loss": 0.9179,
+      "step": 85665
+    },
+    {
+      "epoch": 5.820763690718848,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3620226933007201e-05,
+      "loss": 0.8031,
+      "step": 85670
+    },
+    {
+      "epoch": 5.82110341078951,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3618103682565567e-05,
+      "loss": 0.9955,
+      "step": 85675
+    },
+    {
+      "epoch": 5.821443130860171,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3615980432123931e-05,
+      "loss": 0.9129,
+      "step": 85680
+    },
+    {
+      "epoch": 5.821782850930833,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3613857181682293e-05,
+      "loss": 0.9118,
+      "step": 85685
+    },
+    {
+      "epoch": 5.822122571001495,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3611733931240659e-05,
+      "loss": 0.7894,
+      "step": 85690
+    },
+    {
+      "epoch": 5.822462291072156,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3609610680799023e-05,
+      "loss": 0.8164,
+      "step": 85695
+    },
+    {
+      "epoch": 5.8228020111428185,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3607487430357385e-05,
+      "loss": 0.8281,
+      "step": 85700
+    },
+    {
+      "epoch": 5.8231417312134806,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.360536417991575e-05,
+      "loss": 0.8468,
+      "step": 85705
+    },
+    {
+      "epoch": 5.823481451284142,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3603240929474115e-05,
+      "loss": 0.808,
+      "step": 85710
+    },
+    {
+      "epoch": 5.823821171354804,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3601117679032477e-05,
+      "loss": 0.8392,
+      "step": 85715
+    },
+    {
+      "epoch": 5.824160891425466,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3598994428590841e-05,
+      "loss": 0.7732,
+      "step": 85720
+    },
+    {
+      "epoch": 5.824500611496127,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3596871178149207e-05,
+      "loss": 0.808,
+      "step": 85725
+    },
+    {
+      "epoch": 5.824840331566789,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3594747927707568e-05,
+      "loss": 0.9094,
+      "step": 85730
+    },
+    {
+      "epoch": 5.825180051637451,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.3592624677265933e-05,
+      "loss": 0.8687,
+      "step": 85735
+    },
+    {
+      "epoch": 5.825519771708112,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3590501426824297e-05,
+      "loss": 0.8659,
+      "step": 85740
+    },
+    {
+      "epoch": 5.8258594917787745,
+      "grad_norm": 8.5,
+      "learning_rate": 1.358837817638266e-05,
+      "loss": 0.847,
+      "step": 85745
+    },
+    {
+      "epoch": 5.826199211849436,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3586254925941025e-05,
+      "loss": 0.8947,
+      "step": 85750
+    },
+    {
+      "epoch": 5.826538931920098,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.358413167549939e-05,
+      "loss": 0.625,
+      "step": 85755
+    },
+    {
+      "epoch": 5.82687865199076,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.3582008425057752e-05,
+      "loss": 1.1166,
+      "step": 85760
+    },
+    {
+      "epoch": 5.827218372061421,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3579885174616117e-05,
+      "loss": 0.9215,
+      "step": 85765
+    },
+    {
+      "epoch": 5.827558092132083,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3577761924174481e-05,
+      "loss": 0.9277,
+      "step": 85770
+    },
+    {
+      "epoch": 5.827897812202745,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3575638673732845e-05,
+      "loss": 0.86,
+      "step": 85775
+    },
+    {
+      "epoch": 5.828237532273406,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3573515423291208e-05,
+      "loss": 0.9229,
+      "step": 85780
+    },
+    {
+      "epoch": 5.828577252344068,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3571392172849573e-05,
+      "loss": 0.7785,
+      "step": 85785
+    },
+    {
+      "epoch": 5.8289169724147305,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.3569268922407937e-05,
+      "loss": 0.9446,
+      "step": 85790
+    },
+    {
+      "epoch": 5.829256692485392,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.35671456719663e-05,
+      "loss": 0.9247,
+      "step": 85795
+    },
+    {
+      "epoch": 5.829596412556054,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3565022421524665e-05,
+      "loss": 0.8452,
+      "step": 85800
+    },
+    {
+      "epoch": 5.829936132626716,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.356289917108303e-05,
+      "loss": 0.8042,
+      "step": 85805
+    },
+    {
+      "epoch": 5.830275852697377,
+      "grad_norm": 9.375,
+      "learning_rate": 1.3560775920641392e-05,
+      "loss": 0.9777,
+      "step": 85810
+    },
+    {
+      "epoch": 5.830615572768039,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3558652670199756e-05,
+      "loss": 0.8136,
+      "step": 85815
+    },
+    {
+      "epoch": 5.830955292838701,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3556529419758121e-05,
+      "loss": 0.8227,
+      "step": 85820
+    },
+    {
+      "epoch": 5.831295012909362,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3554406169316484e-05,
+      "loss": 0.9907,
+      "step": 85825
+    },
+    {
+      "epoch": 5.831634732980024,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.3552282918874848e-05,
+      "loss": 0.7961,
+      "step": 85830
+    },
+    {
+      "epoch": 5.8319744530506865,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3550159668433213e-05,
+      "loss": 0.9026,
+      "step": 85835
+    },
+    {
+      "epoch": 5.832314173121348,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3548036417991574e-05,
+      "loss": 0.9612,
+      "step": 85840
+    },
+    {
+      "epoch": 5.83265389319201,
+      "grad_norm": 8.875,
+      "learning_rate": 1.354591316754994e-05,
+      "loss": 0.836,
+      "step": 85845
+    },
+    {
+      "epoch": 5.832993613262672,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3543789917108304e-05,
+      "loss": 0.8252,
+      "step": 85850
+    },
+    {
+      "epoch": 5.833333333333333,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3541666666666666e-05,
+      "loss": 0.9274,
+      "step": 85855
+    },
+    {
+      "epoch": 5.833673053403995,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3539543416225032e-05,
+      "loss": 0.8343,
+      "step": 85860
+    },
+    {
+      "epoch": 5.834012773474657,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.3537420165783396e-05,
+      "loss": 0.9187,
+      "step": 85865
+    },
+    {
+      "epoch": 5.834352493545318,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3535296915341758e-05,
+      "loss": 0.8059,
+      "step": 85870
+    },
+    {
+      "epoch": 5.8346922136159804,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3533173664900122e-05,
+      "loss": 0.838,
+      "step": 85875
+    },
+    {
+      "epoch": 5.8350319336866425,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3531050414458488e-05,
+      "loss": 0.9153,
+      "step": 85880
+    },
+    {
+      "epoch": 5.835371653757304,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.352892716401685e-05,
+      "loss": 0.8357,
+      "step": 85885
+    },
+    {
+      "epoch": 5.835711373827966,
+      "grad_norm": 9.25,
+      "learning_rate": 1.3526803913575214e-05,
+      "loss": 0.8218,
+      "step": 85890
+    },
+    {
+      "epoch": 5.836051093898628,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.352468066313358e-05,
+      "loss": 0.806,
+      "step": 85895
+    },
+    {
+      "epoch": 5.836390813969289,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3522557412691942e-05,
+      "loss": 0.8422,
+      "step": 85900
+    },
+    {
+      "epoch": 5.836730534039951,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3520434162250306e-05,
+      "loss": 0.8807,
+      "step": 85905
+    },
+    {
+      "epoch": 5.837070254110613,
+      "grad_norm": 8.0,
+      "learning_rate": 1.351831091180867e-05,
+      "loss": 0.8362,
+      "step": 85910
+    },
+    {
+      "epoch": 5.837409974181274,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3516187661367032e-05,
+      "loss": 0.8565,
+      "step": 85915
+    },
+    {
+      "epoch": 5.8377496942519365,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3514064410925398e-05,
+      "loss": 0.8467,
+      "step": 85920
+    },
+    {
+      "epoch": 5.8380894143225985,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.3511941160483762e-05,
+      "loss": 1.0998,
+      "step": 85925
+    },
+    {
+      "epoch": 5.83842913439326,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3509817910042124e-05,
+      "loss": 0.9274,
+      "step": 85930
+    },
+    {
+      "epoch": 5.838768854463922,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.350769465960049e-05,
+      "loss": 0.8339,
+      "step": 85935
+    },
+    {
+      "epoch": 5.839108574534584,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3505571409158854e-05,
+      "loss": 0.7316,
+      "step": 85940
+    },
+    {
+      "epoch": 5.839448294605245,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.350344815871722e-05,
+      "loss": 0.9287,
+      "step": 85945
+    },
+    {
+      "epoch": 5.839788014675907,
+      "grad_norm": 8.375,
+      "learning_rate": 1.350132490827558e-05,
+      "loss": 1.0152,
+      "step": 85950
+    },
+    {
+      "epoch": 5.840127734746569,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3499201657833946e-05,
+      "loss": 0.8608,
+      "step": 85955
+    },
+    {
+      "epoch": 5.84046745481723,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.349707840739231e-05,
+      "loss": 0.8476,
+      "step": 85960
+    },
+    {
+      "epoch": 5.8408071748878925,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3494955156950672e-05,
+      "loss": 0.8296,
+      "step": 85965
+    },
+    {
+      "epoch": 5.841146894958554,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.3492831906509038e-05,
+      "loss": 0.7316,
+      "step": 85970
+    },
+    {
+      "epoch": 5.841486615029216,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3490708656067402e-05,
+      "loss": 0.8387,
+      "step": 85975
+    },
+    {
+      "epoch": 5.841826335099878,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3488585405625764e-05,
+      "loss": 0.8785,
+      "step": 85980
+    },
+    {
+      "epoch": 5.842166055170539,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3486462155184128e-05,
+      "loss": 0.792,
+      "step": 85985
+    },
+    {
+      "epoch": 5.842505775241201,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3484338904742494e-05,
+      "loss": 0.861,
+      "step": 85990
+    },
+    {
+      "epoch": 5.842845495311863,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3482215654300856e-05,
+      "loss": 0.9265,
+      "step": 85995
+    },
+    {
+      "epoch": 5.843185215382524,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.348009240385922e-05,
+      "loss": 0.862,
+      "step": 86000
+    },
+    {
+      "epoch": 5.843524935453186,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.3477969153417586e-05,
+      "loss": 0.8304,
+      "step": 86005
+    },
+    {
+      "epoch": 5.8438646555238485,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3475845902975947e-05,
+      "loss": 0.8771,
+      "step": 86010
+    },
+    {
+      "epoch": 5.84420437559451,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3473722652534312e-05,
+      "loss": 0.8405,
+      "step": 86015
+    },
+    {
+      "epoch": 5.844544095665172,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.3471599402092676e-05,
+      "loss": 0.8098,
+      "step": 86020
+    },
+    {
+      "epoch": 5.844883815735834,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3469476151651039e-05,
+      "loss": 0.8454,
+      "step": 86025
+    },
+    {
+      "epoch": 5.845223535806495,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3467352901209404e-05,
+      "loss": 0.8411,
+      "step": 86030
+    },
+    {
+      "epoch": 5.845563255877157,
+      "grad_norm": 6.46875,
+      "learning_rate": 1.3465229650767768e-05,
+      "loss": 0.8161,
+      "step": 86035
+    },
+    {
+      "epoch": 5.845902975947819,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.346310640032613e-05,
+      "loss": 0.9395,
+      "step": 86040
+    },
+    {
+      "epoch": 5.84624269601848,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3460983149884496e-05,
+      "loss": 0.7497,
+      "step": 86045
+    },
+    {
+      "epoch": 5.846582416089142,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.345885989944286e-05,
+      "loss": 0.88,
+      "step": 86050
+    },
+    {
+      "epoch": 5.8469221361598045,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3456736649001223e-05,
+      "loss": 0.8043,
+      "step": 86055
+    },
+    {
+      "epoch": 5.847261856230466,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.3454613398559587e-05,
+      "loss": 0.8391,
+      "step": 86060
+    },
+    {
+      "epoch": 5.847601576301128,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3452490148117953e-05,
+      "loss": 0.8751,
+      "step": 86065
+    },
+    {
+      "epoch": 5.84794129637179,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3450366897676315e-05,
+      "loss": 0.8166,
+      "step": 86070
+    },
+    {
+      "epoch": 5.848281016442451,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3448243647234679e-05,
+      "loss": 0.9528,
+      "step": 86075
+    },
+    {
+      "epoch": 5.848620736513113,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3446120396793045e-05,
+      "loss": 0.8931,
+      "step": 86080
+    },
+    {
+      "epoch": 5.848960456583775,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.3443997146351405e-05,
+      "loss": 0.8323,
+      "step": 86085
+    },
+    {
+      "epoch": 5.849300176654436,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.344187389590977e-05,
+      "loss": 0.8127,
+      "step": 86090
+    },
+    {
+      "epoch": 5.849639896725098,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3439750645468135e-05,
+      "loss": 0.7873,
+      "step": 86095
+    },
+    {
+      "epoch": 5.8499796167957605,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.3437627395026497e-05,
+      "loss": 0.8279,
+      "step": 86100
+    },
+    {
+      "epoch": 5.850319336866422,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3435504144584863e-05,
+      "loss": 0.8269,
+      "step": 86105
+    },
+    {
+      "epoch": 5.850659056937084,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3433380894143227e-05,
+      "loss": 0.8478,
+      "step": 86110
+    },
+    {
+      "epoch": 5.850998777007746,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3431257643701593e-05,
+      "loss": 0.9083,
+      "step": 86115
+    },
+    {
+      "epoch": 5.851338497078407,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3429134393259953e-05,
+      "loss": 0.8162,
+      "step": 86120
+    },
+    {
+      "epoch": 5.851678217149069,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.3427011142818319e-05,
+      "loss": 0.8878,
+      "step": 86125
+    },
+    {
+      "epoch": 5.852017937219731,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3424887892376683e-05,
+      "loss": 0.8429,
+      "step": 86130
+    },
+    {
+      "epoch": 5.852357657290392,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3422764641935045e-05,
+      "loss": 0.8773,
+      "step": 86135
+    },
+    {
+      "epoch": 5.852697377361054,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3420641391493411e-05,
+      "loss": 0.8864,
+      "step": 86140
+    },
+    {
+      "epoch": 5.8530370974317165,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3418518141051775e-05,
+      "loss": 0.8885,
+      "step": 86145
+    },
+    {
+      "epoch": 5.853376817502378,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3416394890610137e-05,
+      "loss": 0.8448,
+      "step": 86150
+    },
+    {
+      "epoch": 5.85371653757304,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3414271640168501e-05,
+      "loss": 0.7973,
+      "step": 86155
+    },
+    {
+      "epoch": 5.854056257643702,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3412148389726867e-05,
+      "loss": 0.7732,
+      "step": 86160
+    },
+    {
+      "epoch": 5.854395977714363,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.341002513928523e-05,
+      "loss": 0.8602,
+      "step": 86165
+    },
+    {
+      "epoch": 5.854735697785025,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.3407901888843593e-05,
+      "loss": 0.8742,
+      "step": 86170
+    },
+    {
+      "epoch": 5.855075417855687,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3405778638401959e-05,
+      "loss": 0.8095,
+      "step": 86175
+    },
+    {
+      "epoch": 5.855415137926348,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3403655387960321e-05,
+      "loss": 0.9088,
+      "step": 86180
+    },
+    {
+      "epoch": 5.8557548579970105,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3401532137518685e-05,
+      "loss": 0.9784,
+      "step": 86185
+    },
+    {
+      "epoch": 5.8560945780676725,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.3399408887077051e-05,
+      "loss": 0.7807,
+      "step": 86190
+    },
+    {
+      "epoch": 5.856434298138334,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3397285636635412e-05,
+      "loss": 0.8902,
+      "step": 86195
+    },
+    {
+      "epoch": 5.856774018208996,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3395162386193777e-05,
+      "loss": 0.8665,
+      "step": 86200
+    },
+    {
+      "epoch": 5.857113738279658,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3393039135752141e-05,
+      "loss": 0.8279,
+      "step": 86205
+    },
+    {
+      "epoch": 5.857453458350319,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3390915885310504e-05,
+      "loss": 0.8268,
+      "step": 86210
+    },
+    {
+      "epoch": 5.857793178420981,
+      "grad_norm": 8.0,
+      "learning_rate": 1.338879263486887e-05,
+      "loss": 0.8373,
+      "step": 86215
+    },
+    {
+      "epoch": 5.858132898491643,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3386669384427233e-05,
+      "loss": 0.8757,
+      "step": 86220
+    },
+    {
+      "epoch": 5.858472618562304,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3384546133985596e-05,
+      "loss": 0.822,
+      "step": 86225
+    },
+    {
+      "epoch": 5.8588123386329665,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.338242288354396e-05,
+      "loss": 0.9097,
+      "step": 86230
+    },
+    {
+      "epoch": 5.8591520587036285,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3380299633102325e-05,
+      "loss": 0.8891,
+      "step": 86235
+    },
+    {
+      "epoch": 5.85949177877429,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.3378176382660688e-05,
+      "loss": 0.7751,
+      "step": 86240
+    },
+    {
+      "epoch": 5.859831498844952,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3376053132219052e-05,
+      "loss": 0.8676,
+      "step": 86245
+    },
+    {
+      "epoch": 5.860171218915614,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3373929881777417e-05,
+      "loss": 0.8299,
+      "step": 86250
+    },
+    {
+      "epoch": 5.860510938986275,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3371806631335778e-05,
+      "loss": 0.8183,
+      "step": 86255
+    },
+    {
+      "epoch": 5.860850659056937,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.3369683380894144e-05,
+      "loss": 0.8502,
+      "step": 86260
+    },
+    {
+      "epoch": 5.861190379127599,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3367560130452508e-05,
+      "loss": 0.7504,
+      "step": 86265
+    },
+    {
+      "epoch": 5.86153009919826,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.336543688001087e-05,
+      "loss": 0.8576,
+      "step": 86270
+    },
+    {
+      "epoch": 5.8618698192689225,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3363313629569236e-05,
+      "loss": 0.8761,
+      "step": 86275
+    },
+    {
+      "epoch": 5.8622095393395846,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.33611903791276e-05,
+      "loss": 0.8725,
+      "step": 86280
+    },
+    {
+      "epoch": 5.862549259410246,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3359067128685965e-05,
+      "loss": 0.8213,
+      "step": 86285
+    },
+    {
+      "epoch": 5.862888979480908,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3356943878244328e-05,
+      "loss": 0.7934,
+      "step": 86290
+    },
+    {
+      "epoch": 5.86322869955157,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.3354820627802692e-05,
+      "loss": 0.8703,
+      "step": 86295
+    },
+    {
+      "epoch": 5.863568419622231,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3352697377361056e-05,
+      "loss": 0.81,
+      "step": 86300
+    },
+    {
+      "epoch": 5.863908139692893,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3350574126919418e-05,
+      "loss": 0.9239,
+      "step": 86305
+    },
+    {
+      "epoch": 5.864247859763555,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3348450876477784e-05,
+      "loss": 0.9342,
+      "step": 86310
+    },
+    {
+      "epoch": 5.864587579834216,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3346327626036148e-05,
+      "loss": 0.9475,
+      "step": 86315
+    },
+    {
+      "epoch": 5.8649272999048785,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.334420437559451e-05,
+      "loss": 0.9477,
+      "step": 86320
+    },
+    {
+      "epoch": 5.865267019975541,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3342081125152876e-05,
+      "loss": 0.8777,
+      "step": 86325
+    },
+    {
+      "epoch": 5.865606740046202,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.333995787471124e-05,
+      "loss": 0.8589,
+      "step": 86330
+    },
+    {
+      "epoch": 5.865946460116864,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3337834624269602e-05,
+      "loss": 0.8994,
+      "step": 86335
+    },
+    {
+      "epoch": 5.866286180187526,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.3335711373827966e-05,
+      "loss": 0.8408,
+      "step": 86340
+    },
+    {
+      "epoch": 5.866625900258187,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3333588123386332e-05,
+      "loss": 0.8465,
+      "step": 86345
+    },
+    {
+      "epoch": 5.866965620328849,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3331464872944694e-05,
+      "loss": 0.9059,
+      "step": 86350
+    },
+    {
+      "epoch": 5.867305340399511,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3329341622503058e-05,
+      "loss": 0.7968,
+      "step": 86355
+    },
+    {
+      "epoch": 5.867645060470172,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3327218372061424e-05,
+      "loss": 0.8249,
+      "step": 86360
+    },
+    {
+      "epoch": 5.8679847805408345,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3325095121619784e-05,
+      "loss": 0.8717,
+      "step": 86365
+    },
+    {
+      "epoch": 5.868324500611497,
+      "grad_norm": 8.0,
+      "learning_rate": 1.332297187117815e-05,
+      "loss": 0.9242,
+      "step": 86370
+    },
+    {
+      "epoch": 5.868664220682158,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.3320848620736514e-05,
+      "loss": 0.9715,
+      "step": 86375
+    },
+    {
+      "epoch": 5.86900394075282,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3318725370294876e-05,
+      "loss": 0.8088,
+      "step": 86380
+    },
+    {
+      "epoch": 5.869343660823482,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3316602119853242e-05,
+      "loss": 0.9669,
+      "step": 86385
+    },
+    {
+      "epoch": 5.869683380894143,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3314478869411606e-05,
+      "loss": 0.8593,
+      "step": 86390
+    },
+    {
+      "epoch": 5.870023100964805,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3312355618969968e-05,
+      "loss": 0.8444,
+      "step": 86395
+    },
+    {
+      "epoch": 5.870362821035467,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.3310232368528332e-05,
+      "loss": 0.8674,
+      "step": 86400
+    },
+    {
+      "epoch": 5.870702541106128,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.3308109118086698e-05,
+      "loss": 0.8625,
+      "step": 86405
+    },
+    {
+      "epoch": 5.8710422611767905,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.330598586764506e-05,
+      "loss": 0.9123,
+      "step": 86410
+    },
+    {
+      "epoch": 5.871381981247453,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3303862617203424e-05,
+      "loss": 0.8762,
+      "step": 86415
+    },
+    {
+      "epoch": 5.871721701318114,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.330173936676179e-05,
+      "loss": 0.976,
+      "step": 86420
+    },
+    {
+      "epoch": 5.872061421388776,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3299616116320152e-05,
+      "loss": 0.6314,
+      "step": 86425
+    },
+    {
+      "epoch": 5.872401141459437,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.3297492865878516e-05,
+      "loss": 0.8901,
+      "step": 86430
+    },
+    {
+      "epoch": 5.872740861530099,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3295369615436882e-05,
+      "loss": 0.8403,
+      "step": 86435
+    },
+    {
+      "epoch": 5.873080581600761,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3293246364995243e-05,
+      "loss": 0.7781,
+      "step": 86440
+    },
+    {
+      "epoch": 5.873420301671422,
+      "grad_norm": 7.25,
+      "learning_rate": 1.3291123114553608e-05,
+      "loss": 0.8366,
+      "step": 86445
+    },
+    {
+      "epoch": 5.8737600217420844,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3288999864111972e-05,
+      "loss": 0.8114,
+      "step": 86450
+    },
+    {
+      "epoch": 5.8740997418127465,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.3286876613670338e-05,
+      "loss": 0.9277,
+      "step": 86455
+    },
+    {
+      "epoch": 5.874439461883408,
+      "grad_norm": 7.875,
+      "learning_rate": 1.32847533632287e-05,
+      "loss": 0.8646,
+      "step": 86460
+    },
+    {
+      "epoch": 5.87477918195407,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3282630112787064e-05,
+      "loss": 0.8625,
+      "step": 86465
+    },
+    {
+      "epoch": 5.875118902024732,
+      "grad_norm": 8.5,
+      "learning_rate": 1.328050686234543e-05,
+      "loss": 0.8574,
+      "step": 86470
+    },
+    {
+      "epoch": 5.875458622095393,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.327838361190379e-05,
+      "loss": 0.9289,
+      "step": 86475
+    },
+    {
+      "epoch": 5.875798342166055,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3276260361462156e-05,
+      "loss": 0.8713,
+      "step": 86480
+    },
+    {
+      "epoch": 5.876138062236717,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.327413711102052e-05,
+      "loss": 0.8508,
+      "step": 86485
+    },
+    {
+      "epoch": 5.876477782307378,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3272013860578883e-05,
+      "loss": 0.8641,
+      "step": 86490
+    },
+    {
+      "epoch": 5.8768175023780405,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3269890610137248e-05,
+      "loss": 0.8909,
+      "step": 86495
+    },
+    {
+      "epoch": 5.8771572224487025,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3267767359695612e-05,
+      "loss": 0.8132,
+      "step": 86500
+    },
+    {
+      "epoch": 5.877496942519364,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3265644109253975e-05,
+      "loss": 0.8639,
+      "step": 86505
+    },
+    {
+      "epoch": 5.877836662590026,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3263520858812339e-05,
+      "loss": 0.7499,
+      "step": 86510
+    },
+    {
+      "epoch": 5.878176382660688,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3261397608370704e-05,
+      "loss": 0.8685,
+      "step": 86515
+    },
+    {
+      "epoch": 5.878516102731349,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3259274357929067e-05,
+      "loss": 0.8969,
+      "step": 86520
+    },
+    {
+      "epoch": 5.878855822802011,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.325715110748743e-05,
+      "loss": 0.9509,
+      "step": 86525
+    },
+    {
+      "epoch": 5.879195542872673,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3255027857045796e-05,
+      "loss": 0.8368,
+      "step": 86530
+    },
+    {
+      "epoch": 5.879535262943334,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.3252904606604159e-05,
+      "loss": 0.8276,
+      "step": 86535
+    },
+    {
+      "epoch": 5.8798749830139965,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3250781356162523e-05,
+      "loss": 0.8845,
+      "step": 86540
+    },
+    {
+      "epoch": 5.8802147030846585,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3248658105720887e-05,
+      "loss": 0.8832,
+      "step": 86545
+    },
+    {
+      "epoch": 5.88055442315532,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3246534855279249e-05,
+      "loss": 0.893,
+      "step": 86550
+    },
+    {
+      "epoch": 5.880894143225982,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3244411604837615e-05,
+      "loss": 0.8145,
+      "step": 86555
+    },
+    {
+      "epoch": 5.881233863296644,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3242288354395979e-05,
+      "loss": 0.9067,
+      "step": 86560
+    },
+    {
+      "epoch": 5.881573583367305,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3240165103954341e-05,
+      "loss": 0.8412,
+      "step": 86565
+    },
+    {
+      "epoch": 5.881913303437967,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3238041853512707e-05,
+      "loss": 0.8776,
+      "step": 86570
+    },
+    {
+      "epoch": 5.882253023508629,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.323591860307107e-05,
+      "loss": 0.7676,
+      "step": 86575
+    },
+    {
+      "epoch": 5.88259274357929,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3233795352629433e-05,
+      "loss": 0.8328,
+      "step": 86580
+    },
+    {
+      "epoch": 5.8829324636499525,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3231672102187797e-05,
+      "loss": 0.8571,
+      "step": 86585
+    },
+    {
+      "epoch": 5.883272183720615,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3229548851746163e-05,
+      "loss": 0.8571,
+      "step": 86590
+    },
+    {
+      "epoch": 5.883611903791276,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3227425601304525e-05,
+      "loss": 1.0713,
+      "step": 86595
+    },
+    {
+      "epoch": 5.883951623861938,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3225302350862889e-05,
+      "loss": 0.819,
+      "step": 86600
+    },
+    {
+      "epoch": 5.8842913439326,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3223179100421255e-05,
+      "loss": 0.8708,
+      "step": 86605
+    },
+    {
+      "epoch": 5.884631064003261,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3221055849979615e-05,
+      "loss": 0.7927,
+      "step": 86610
+    },
+    {
+      "epoch": 5.884970784073923,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.3218932599537981e-05,
+      "loss": 0.8054,
+      "step": 86615
+    },
+    {
+      "epoch": 5.885310504144585,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3216809349096345e-05,
+      "loss": 0.9106,
+      "step": 86620
+    },
+    {
+      "epoch": 5.885650224215246,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.321468609865471e-05,
+      "loss": 0.7511,
+      "step": 86625
+    },
+    {
+      "epoch": 5.8859899442859085,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3212562848213073e-05,
+      "loss": 0.9011,
+      "step": 86630
+    },
+    {
+      "epoch": 5.886329664356571,
+      "grad_norm": 7.5,
+      "learning_rate": 1.3210439597771437e-05,
+      "loss": 0.7929,
+      "step": 86635
+    },
+    {
+      "epoch": 5.886669384427232,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3208316347329803e-05,
+      "loss": 0.8273,
+      "step": 86640
+    },
+    {
+      "epoch": 5.887009104497894,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.3206193096888163e-05,
+      "loss": 0.8751,
+      "step": 86645
+    },
+    {
+      "epoch": 5.887348824568555,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3204069846446529e-05,
+      "loss": 0.9405,
+      "step": 86650
+    },
+    {
+      "epoch": 5.887688544639217,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3201946596004893e-05,
+      "loss": 0.8759,
+      "step": 86655
+    },
+    {
+      "epoch": 5.888028264709879,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3199823345563255e-05,
+      "loss": 0.8542,
+      "step": 86660
+    },
+    {
+      "epoch": 5.88836798478054,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3197700095121621e-05,
+      "loss": 0.9316,
+      "step": 86665
+    },
+    {
+      "epoch": 5.888707704851202,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3195576844679985e-05,
+      "loss": 0.8231,
+      "step": 86670
+    },
+    {
+      "epoch": 5.8890474249218645,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3193453594238347e-05,
+      "loss": 0.8627,
+      "step": 86675
+    },
+    {
+      "epoch": 5.889387144992526,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.3191330343796713e-05,
+      "loss": 0.7792,
+      "step": 86680
+    },
+    {
+      "epoch": 5.889726865063188,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3189207093355077e-05,
+      "loss": 0.8208,
+      "step": 86685
+    },
+    {
+      "epoch": 5.89006658513385,
+      "grad_norm": 7.5,
+      "learning_rate": 1.318708384291344e-05,
+      "loss": 0.8747,
+      "step": 86690
+    },
+    {
+      "epoch": 5.890406305204511,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3184960592471803e-05,
+      "loss": 0.7616,
+      "step": 86695
+    },
+    {
+      "epoch": 5.890746025275173,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3182837342030169e-05,
+      "loss": 0.796,
+      "step": 86700
+    },
+    {
+      "epoch": 5.891085745345835,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3180714091588531e-05,
+      "loss": 0.8449,
+      "step": 86705
+    },
+    {
+      "epoch": 5.891425465416496,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3178590841146895e-05,
+      "loss": 0.8493,
+      "step": 86710
+    },
+    {
+      "epoch": 5.891765185487158,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3176467590705261e-05,
+      "loss": 0.8043,
+      "step": 86715
+    },
+    {
+      "epoch": 5.8921049055578205,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3174344340263622e-05,
+      "loss": 0.8237,
+      "step": 86720
+    },
+    {
+      "epoch": 5.892444625628482,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3172221089821987e-05,
+      "loss": 0.9954,
+      "step": 86725
+    },
+    {
+      "epoch": 5.892784345699144,
+      "grad_norm": 7.375,
+      "learning_rate": 1.3170097839380351e-05,
+      "loss": 0.9687,
+      "step": 86730
+    },
+    {
+      "epoch": 5.893124065769806,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.3167974588938714e-05,
+      "loss": 0.8897,
+      "step": 86735
+    },
+    {
+      "epoch": 5.893463785840467,
+      "grad_norm": 8.0,
+      "learning_rate": 1.316585133849708e-05,
+      "loss": 0.7951,
+      "step": 86740
+    },
+    {
+      "epoch": 5.893803505911129,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3163728088055443e-05,
+      "loss": 0.8171,
+      "step": 86745
+    },
+    {
+      "epoch": 5.894143225981791,
+      "grad_norm": 9.25,
+      "learning_rate": 1.3161604837613806e-05,
+      "loss": 0.7736,
+      "step": 86750
+    },
+    {
+      "epoch": 5.894482946052452,
+      "grad_norm": 8.375,
+      "learning_rate": 1.315948158717217e-05,
+      "loss": 0.8269,
+      "step": 86755
+    },
+    {
+      "epoch": 5.8948226661231145,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3157358336730535e-05,
+      "loss": 0.8774,
+      "step": 86760
+    },
+    {
+      "epoch": 5.8951623861937765,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3155235086288898e-05,
+      "loss": 0.821,
+      "step": 86765
+    },
+    {
+      "epoch": 5.895502106264438,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3153111835847262e-05,
+      "loss": 0.9984,
+      "step": 86770
+    },
+    {
+      "epoch": 5.8958418263351,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3150988585405628e-05,
+      "loss": 0.8617,
+      "step": 86775
+    },
+    {
+      "epoch": 5.896181546405762,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.314886533496399e-05,
+      "loss": 0.8656,
+      "step": 86780
+    },
+    {
+      "epoch": 5.896521266476423,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3146742084522354e-05,
+      "loss": 0.7937,
+      "step": 86785
+    },
+    {
+      "epoch": 5.896860986547085,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3144618834080718e-05,
+      "loss": 0.8029,
+      "step": 86790
+    },
+    {
+      "epoch": 5.897200706617747,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3142495583639084e-05,
+      "loss": 0.8312,
+      "step": 86795
+    },
+    {
+      "epoch": 5.897540426688408,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3140372333197446e-05,
+      "loss": 0.8863,
+      "step": 86800
+    },
+    {
+      "epoch": 5.8978801467590705,
+      "grad_norm": 8.125,
+      "learning_rate": 1.313824908275581e-05,
+      "loss": 0.9211,
+      "step": 86805
+    },
+    {
+      "epoch": 5.8982198668297325,
+      "grad_norm": 7.625,
+      "learning_rate": 1.3136125832314176e-05,
+      "loss": 0.9176,
+      "step": 86810
+    },
+    {
+      "epoch": 5.898559586900394,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3134002581872538e-05,
+      "loss": 0.8596,
+      "step": 86815
+    },
+    {
+      "epoch": 5.898899306971056,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3131879331430902e-05,
+      "loss": 0.9505,
+      "step": 86820
+    },
+    {
+      "epoch": 5.899239027041718,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.3129756080989266e-05,
+      "loss": 0.8849,
+      "step": 86825
+    },
+    {
+      "epoch": 5.899578747112379,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3127632830547628e-05,
+      "loss": 0.809,
+      "step": 86830
+    },
+    {
+      "epoch": 5.899918467183041,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3125509580105994e-05,
+      "loss": 0.8645,
+      "step": 86835
+    },
+    {
+      "epoch": 5.900258187253703,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3123386329664358e-05,
+      "loss": 0.8567,
+      "step": 86840
+    },
+    {
+      "epoch": 5.900597907324364,
+      "grad_norm": 8.125,
+      "learning_rate": 1.312126307922272e-05,
+      "loss": 0.8051,
+      "step": 86845
+    },
+    {
+      "epoch": 5.9009376273950265,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3119139828781086e-05,
+      "loss": 0.952,
+      "step": 86850
+    },
+    {
+      "epoch": 5.9012773474656885,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.311701657833945e-05,
+      "loss": 0.8649,
+      "step": 86855
+    },
+    {
+      "epoch": 5.90161706753635,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3114893327897812e-05,
+      "loss": 0.9118,
+      "step": 86860
+    },
+    {
+      "epoch": 5.901956787607012,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3112770077456176e-05,
+      "loss": 0.818,
+      "step": 86865
+    },
+    {
+      "epoch": 5.902296507677674,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3110646827014542e-05,
+      "loss": 0.902,
+      "step": 86870
+    },
+    {
+      "epoch": 5.902636227748335,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3108523576572904e-05,
+      "loss": 0.7968,
+      "step": 86875
+    },
+    {
+      "epoch": 5.902975947818997,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.3106400326131268e-05,
+      "loss": 0.8683,
+      "step": 86880
+    },
+    {
+      "epoch": 5.903315667889659,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3104277075689634e-05,
+      "loss": 0.9263,
+      "step": 86885
+    },
+    {
+      "epoch": 5.90365538796032,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3102153825247995e-05,
+      "loss": 0.9142,
+      "step": 86890
+    },
+    {
+      "epoch": 5.9039951080309825,
+      "grad_norm": 7.875,
+      "learning_rate": 1.310003057480636e-05,
+      "loss": 0.8841,
+      "step": 86895
+    },
+    {
+      "epoch": 5.904334828101645,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3097907324364724e-05,
+      "loss": 0.942,
+      "step": 86900
+    },
+    {
+      "epoch": 5.904674548172306,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.3095784073923087e-05,
+      "loss": 0.8385,
+      "step": 86905
+    },
+    {
+      "epoch": 5.905014268242968,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3093660823481452e-05,
+      "loss": 0.8072,
+      "step": 86910
+    },
+    {
+      "epoch": 5.90535398831363,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.3091537573039816e-05,
+      "loss": 0.8999,
+      "step": 86915
+    },
+    {
+      "epoch": 5.905693708384291,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3089414322598179e-05,
+      "loss": 0.8806,
+      "step": 86920
+    },
+    {
+      "epoch": 5.906033428454953,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3087291072156543e-05,
+      "loss": 0.7772,
+      "step": 86925
+    },
+    {
+      "epoch": 5.906373148525615,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.3085167821714908e-05,
+      "loss": 0.8166,
+      "step": 86930
+    },
+    {
+      "epoch": 5.906712868596276,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.308304457127327e-05,
+      "loss": 0.9046,
+      "step": 86935
+    },
+    {
+      "epoch": 5.9070525886669385,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3080921320831635e-05,
+      "loss": 0.7751,
+      "step": 86940
+    },
+    {
+      "epoch": 5.907392308737601,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.307879807039e-05,
+      "loss": 0.8823,
+      "step": 86945
+    },
+    {
+      "epoch": 5.907732028808262,
+      "grad_norm": 8.625,
+      "learning_rate": 1.3076674819948363e-05,
+      "loss": 0.831,
+      "step": 86950
+    },
+    {
+      "epoch": 5.908071748878924,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3074551569506727e-05,
+      "loss": 0.8531,
+      "step": 86955
+    },
+    {
+      "epoch": 5.908411468949586,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3072428319065092e-05,
+      "loss": 0.8084,
+      "step": 86960
+    },
+    {
+      "epoch": 5.908751189020247,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3070305068623456e-05,
+      "loss": 0.95,
+      "step": 86965
+    },
+    {
+      "epoch": 5.909090909090909,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3068181818181819e-05,
+      "loss": 0.8738,
+      "step": 86970
+    },
+    {
+      "epoch": 5.909430629161571,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3066058567740183e-05,
+      "loss": 0.9004,
+      "step": 86975
+    },
+    {
+      "epoch": 5.909770349232232,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3063935317298548e-05,
+      "loss": 0.815,
+      "step": 86980
+    },
+    {
+      "epoch": 5.9101100693028945,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.306181206685691e-05,
+      "loss": 0.9517,
+      "step": 86985
+    },
+    {
+      "epoch": 5.910449789373557,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3059688816415275e-05,
+      "loss": 0.8282,
+      "step": 86990
+    },
+    {
+      "epoch": 5.910789509444218,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.305756556597364e-05,
+      "loss": 0.8466,
+      "step": 86995
+    },
+    {
+      "epoch": 5.91112922951488,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.3055442315532001e-05,
+      "loss": 0.8818,
+      "step": 87000
+    },
+    {
+      "epoch": 5.911468949585542,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.3053319065090367e-05,
+      "loss": 0.8893,
+      "step": 87005
+    },
+    {
+      "epoch": 5.911808669656203,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.305119581464873e-05,
+      "loss": 0.8317,
+      "step": 87010
+    },
+    {
+      "epoch": 5.912148389726865,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.3049072564207093e-05,
+      "loss": 0.8315,
+      "step": 87015
+    },
+    {
+      "epoch": 5.912488109797527,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3046949313765459e-05,
+      "loss": 0.909,
+      "step": 87020
+    },
+    {
+      "epoch": 5.9128278298681884,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.3044826063323823e-05,
+      "loss": 0.8923,
+      "step": 87025
+    },
+    {
+      "epoch": 5.9131675499388505,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3042702812882185e-05,
+      "loss": 0.8482,
+      "step": 87030
+    },
+    {
+      "epoch": 5.913507270009513,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3040579562440549e-05,
+      "loss": 0.8549,
+      "step": 87035
+    },
+    {
+      "epoch": 5.913846990080174,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3038456311998915e-05,
+      "loss": 0.7652,
+      "step": 87040
+    },
+    {
+      "epoch": 5.914186710150836,
+      "grad_norm": 7.0,
+      "learning_rate": 1.3036333061557277e-05,
+      "loss": 0.8246,
+      "step": 87045
+    },
+    {
+      "epoch": 5.914526430221498,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3034209811115641e-05,
+      "loss": 0.8271,
+      "step": 87050
+    },
+    {
+      "epoch": 5.914866150292159,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.3032086560674007e-05,
+      "loss": 0.7299,
+      "step": 87055
+    },
+    {
+      "epoch": 5.915205870362821,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3029963310232369e-05,
+      "loss": 0.7974,
+      "step": 87060
+    },
+    {
+      "epoch": 5.915545590433483,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3027840059790733e-05,
+      "loss": 0.8292,
+      "step": 87065
+    },
+    {
+      "epoch": 5.9158853105041445,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3025716809349097e-05,
+      "loss": 0.8651,
+      "step": 87070
+    },
+    {
+      "epoch": 5.9162250305748065,
+      "grad_norm": 8.625,
+      "learning_rate": 1.302359355890746e-05,
+      "loss": 0.9295,
+      "step": 87075
+    },
+    {
+      "epoch": 5.916564750645469,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.3021470308465825e-05,
+      "loss": 0.962,
+      "step": 87080
+    },
+    {
+      "epoch": 5.91690447071613,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3019347058024189e-05,
+      "loss": 0.792,
+      "step": 87085
+    },
+    {
+      "epoch": 5.917244190786792,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3017223807582551e-05,
+      "loss": 0.9497,
+      "step": 87090
+    },
+    {
+      "epoch": 5.917583910857454,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3015100557140917e-05,
+      "loss": 0.9256,
+      "step": 87095
+    },
+    {
+      "epoch": 5.917923630928115,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3012977306699281e-05,
+      "loss": 0.7768,
+      "step": 87100
+    },
+    {
+      "epoch": 5.918263350998777,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.3010854056257643e-05,
+      "loss": 0.7755,
+      "step": 87105
+    },
+    {
+      "epoch": 5.918603071069439,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3008730805816007e-05,
+      "loss": 0.8199,
+      "step": 87110
+    },
+    {
+      "epoch": 5.9189427911401005,
+      "grad_norm": 7.125,
+      "learning_rate": 1.3006607555374373e-05,
+      "loss": 0.8325,
+      "step": 87115
+    },
+    {
+      "epoch": 5.9192825112107625,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3004484304932735e-05,
+      "loss": 0.9093,
+      "step": 87120
+    },
+    {
+      "epoch": 5.919622231281424,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.30023610544911e-05,
+      "loss": 0.8445,
+      "step": 87125
+    },
+    {
+      "epoch": 5.919961951352086,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3000237804049465e-05,
+      "loss": 0.8137,
+      "step": 87130
+    },
+    {
+      "epoch": 5.920301671422748,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2998114553607829e-05,
+      "loss": 0.8348,
+      "step": 87135
+    },
+    {
+      "epoch": 5.920641391493409,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2995991303166191e-05,
+      "loss": 0.863,
+      "step": 87140
+    },
+    {
+      "epoch": 5.920981111564071,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.2993868052724555e-05,
+      "loss": 0.8522,
+      "step": 87145
+    },
+    {
+      "epoch": 5.921320831634733,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2991744802282921e-05,
+      "loss": 0.8622,
+      "step": 87150
+    },
+    {
+      "epoch": 5.921660551705394,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2989621551841283e-05,
+      "loss": 0.8189,
+      "step": 87155
+    },
+    {
+      "epoch": 5.9220002717760565,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2987498301399647e-05,
+      "loss": 0.8589,
+      "step": 87160
+    },
+    {
+      "epoch": 5.9223399918467186,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2985375050958013e-05,
+      "loss": 0.9018,
+      "step": 87165
+    },
+    {
+      "epoch": 5.92267971191738,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2983251800516374e-05,
+      "loss": 0.9172,
+      "step": 87170
+    },
+    {
+      "epoch": 5.923019431988042,
+      "grad_norm": 7.875,
+      "learning_rate": 1.298112855007474e-05,
+      "loss": 0.9311,
+      "step": 87175
+    },
+    {
+      "epoch": 5.923359152058704,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2979005299633103e-05,
+      "loss": 0.8667,
+      "step": 87180
+    },
+    {
+      "epoch": 5.923698872129365,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.2976882049191466e-05,
+      "loss": 0.8174,
+      "step": 87185
+    },
+    {
+      "epoch": 5.924038592200027,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2974758798749831e-05,
+      "loss": 0.9421,
+      "step": 87190
+    },
+    {
+      "epoch": 5.924378312270689,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2972635548308195e-05,
+      "loss": 0.9339,
+      "step": 87195
+    },
+    {
+      "epoch": 5.92471803234135,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.2970512297866558e-05,
+      "loss": 0.8138,
+      "step": 87200
+    },
+    {
+      "epoch": 5.9250577524120125,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2968389047424923e-05,
+      "loss": 0.8211,
+      "step": 87205
+    },
+    {
+      "epoch": 5.925397472482675,
+      "grad_norm": 7.625,
+      "learning_rate": 1.2966265796983287e-05,
+      "loss": 0.8758,
+      "step": 87210
+    },
+    {
+      "epoch": 5.925737192553336,
+      "grad_norm": 7.875,
+      "learning_rate": 1.296414254654165e-05,
+      "loss": 0.8234,
+      "step": 87215
+    },
+    {
+      "epoch": 5.926076912623998,
+      "grad_norm": 7.0,
+      "learning_rate": 1.2962019296100014e-05,
+      "loss": 0.8916,
+      "step": 87220
+    },
+    {
+      "epoch": 5.92641663269466,
+      "grad_norm": 8.125,
+      "learning_rate": 1.295989604565838e-05,
+      "loss": 0.8366,
+      "step": 87225
+    },
+    {
+      "epoch": 5.926756352765321,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2957772795216742e-05,
+      "loss": 0.8966,
+      "step": 87230
+    },
+    {
+      "epoch": 5.927096072835983,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2955649544775106e-05,
+      "loss": 0.8782,
+      "step": 87235
+    },
+    {
+      "epoch": 5.927435792906645,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2953526294333471e-05,
+      "loss": 0.8176,
+      "step": 87240
+    },
+    {
+      "epoch": 5.927775512977306,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2951403043891832e-05,
+      "loss": 0.8158,
+      "step": 87245
+    },
+    {
+      "epoch": 5.9281152330479685,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2949279793450198e-05,
+      "loss": 0.792,
+      "step": 87250
+    },
+    {
+      "epoch": 5.928454953118631,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.2947156543008562e-05,
+      "loss": 0.786,
+      "step": 87255
+    },
+    {
+      "epoch": 5.928794673189292,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2945033292566924e-05,
+      "loss": 0.7935,
+      "step": 87260
+    },
+    {
+      "epoch": 5.929134393259954,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.294291004212529e-05,
+      "loss": 0.8798,
+      "step": 87265
+    },
+    {
+      "epoch": 5.929474113330616,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2940786791683654e-05,
+      "loss": 0.9029,
+      "step": 87270
+    },
+    {
+      "epoch": 5.929813833401277,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2938663541242016e-05,
+      "loss": 0.8346,
+      "step": 87275
+    },
+    {
+      "epoch": 5.930153553471939,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.293654029080038e-05,
+      "loss": 0.7637,
+      "step": 87280
+    },
+    {
+      "epoch": 5.930493273542601,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2934417040358746e-05,
+      "loss": 0.8196,
+      "step": 87285
+    },
+    {
+      "epoch": 5.930832993613262,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.2932293789917108e-05,
+      "loss": 0.8662,
+      "step": 87290
+    },
+    {
+      "epoch": 5.9311727136839245,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.2930170539475472e-05,
+      "loss": 0.839,
+      "step": 87295
+    },
+    {
+      "epoch": 5.931512433754587,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2928047289033838e-05,
+      "loss": 0.8284,
+      "step": 87300
+    },
+    {
+      "epoch": 5.931852153825248,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2925924038592202e-05,
+      "loss": 0.9075,
+      "step": 87305
+    },
+    {
+      "epoch": 5.93219187389591,
+      "grad_norm": 9.25,
+      "learning_rate": 1.2923800788150564e-05,
+      "loss": 0.8576,
+      "step": 87310
+    },
+    {
+      "epoch": 5.932531593966572,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2921677537708928e-05,
+      "loss": 0.9349,
+      "step": 87315
+    },
+    {
+      "epoch": 5.932871314037233,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.2919554287267294e-05,
+      "loss": 0.9276,
+      "step": 87320
+    },
+    {
+      "epoch": 5.933211034107895,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2917431036825656e-05,
+      "loss": 0.8877,
+      "step": 87325
+    },
+    {
+      "epoch": 5.933550754178556,
+      "grad_norm": 8.125,
+      "learning_rate": 1.291530778638402e-05,
+      "loss": 0.9599,
+      "step": 87330
+    },
+    {
+      "epoch": 5.9338904742492184,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2913184535942386e-05,
+      "loss": 0.9485,
+      "step": 87335
+    },
+    {
+      "epoch": 5.9342301943198805,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2911061285500748e-05,
+      "loss": 0.902,
+      "step": 87340
+    },
+    {
+      "epoch": 5.934569914390542,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2908938035059112e-05,
+      "loss": 0.9318,
+      "step": 87345
+    },
+    {
+      "epoch": 5.934909634461204,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2906814784617478e-05,
+      "loss": 0.8895,
+      "step": 87350
+    },
+    {
+      "epoch": 5.935249354531866,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2904691534175838e-05,
+      "loss": 0.7833,
+      "step": 87355
+    },
+    {
+      "epoch": 5.935589074602527,
+      "grad_norm": 9.375,
+      "learning_rate": 1.2902568283734204e-05,
+      "loss": 0.8669,
+      "step": 87360
+    },
+    {
+      "epoch": 5.935928794673189,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2900445033292568e-05,
+      "loss": 0.7969,
+      "step": 87365
+    },
+    {
+      "epoch": 5.936268514743851,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.289832178285093e-05,
+      "loss": 0.8744,
+      "step": 87370
+    },
+    {
+      "epoch": 5.936608234814512,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.2896198532409296e-05,
+      "loss": 0.7884,
+      "step": 87375
+    },
+    {
+      "epoch": 5.9369479548851745,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.289407528196766e-05,
+      "loss": 0.8252,
+      "step": 87380
+    },
+    {
+      "epoch": 5.9372876749558365,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2891952031526022e-05,
+      "loss": 0.8596,
+      "step": 87385
+    },
+    {
+      "epoch": 5.937627395026498,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2889828781084386e-05,
+      "loss": 0.8227,
+      "step": 87390
+    },
+    {
+      "epoch": 5.93796711509716,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.2887705530642752e-05,
+      "loss": 0.8626,
+      "step": 87395
+    },
+    {
+      "epoch": 5.938306835167822,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.2885582280201114e-05,
+      "loss": 0.8603,
+      "step": 87400
+    },
+    {
+      "epoch": 5.938646555238483,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2883459029759478e-05,
+      "loss": 0.886,
+      "step": 87405
+    },
+    {
+      "epoch": 5.938986275309145,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2881335779317844e-05,
+      "loss": 0.8371,
+      "step": 87410
+    },
+    {
+      "epoch": 5.939325995379807,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2879212528876205e-05,
+      "loss": 0.806,
+      "step": 87415
+    },
+    {
+      "epoch": 5.939665715450468,
+      "grad_norm": 7.5,
+      "learning_rate": 1.287708927843457e-05,
+      "loss": 0.756,
+      "step": 87420
+    },
+    {
+      "epoch": 5.9400054355211305,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2874966027992934e-05,
+      "loss": 0.8585,
+      "step": 87425
+    },
+    {
+      "epoch": 5.9403451555917925,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2872842777551297e-05,
+      "loss": 0.9419,
+      "step": 87430
+    },
+    {
+      "epoch": 5.940684875662454,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2870719527109662e-05,
+      "loss": 0.9454,
+      "step": 87435
+    },
+    {
+      "epoch": 5.941024595733116,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2868596276668026e-05,
+      "loss": 0.8637,
+      "step": 87440
+    },
+    {
+      "epoch": 5.941364315803778,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2866473026226389e-05,
+      "loss": 0.8466,
+      "step": 87445
+    },
+    {
+      "epoch": 5.941704035874439,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2864349775784754e-05,
+      "loss": 0.878,
+      "step": 87450
+    },
+    {
+      "epoch": 5.942043755945101,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2862226525343118e-05,
+      "loss": 0.7701,
+      "step": 87455
+    },
+    {
+      "epoch": 5.942383476015763,
+      "grad_norm": 7.375,
+      "learning_rate": 1.286010327490148e-05,
+      "loss": 0.9328,
+      "step": 87460
+    },
+    {
+      "epoch": 5.942723196086424,
+      "grad_norm": 7.0,
+      "learning_rate": 1.2857980024459845e-05,
+      "loss": 0.842,
+      "step": 87465
+    },
+    {
+      "epoch": 5.9430629161570865,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.285585677401821e-05,
+      "loss": 0.7836,
+      "step": 87470
+    },
+    {
+      "epoch": 5.943402636227749,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2853733523576574e-05,
+      "loss": 0.8735,
+      "step": 87475
+    },
+    {
+      "epoch": 5.94374235629841,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2851610273134937e-05,
+      "loss": 0.8,
+      "step": 87480
+    },
+    {
+      "epoch": 5.944082076369072,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.2849487022693303e-05,
+      "loss": 0.883,
+      "step": 87485
+    },
+    {
+      "epoch": 5.944421796439734,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2847363772251667e-05,
+      "loss": 0.8941,
+      "step": 87490
+    },
+    {
+      "epoch": 5.944761516510395,
+      "grad_norm": 7.25,
+      "learning_rate": 1.2845240521810029e-05,
+      "loss": 0.8408,
+      "step": 87495
+    },
+    {
+      "epoch": 5.945101236581057,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2843117271368393e-05,
+      "loss": 0.8594,
+      "step": 87500
+    },
+    {
+      "epoch": 5.945440956651719,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2840994020926759e-05,
+      "loss": 0.8346,
+      "step": 87505
+    },
+    {
+      "epoch": 5.94578067672238,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.283887077048512e-05,
+      "loss": 0.8343,
+      "step": 87510
+    },
+    {
+      "epoch": 5.9461203967930425,
+      "grad_norm": 6.75,
+      "learning_rate": 1.2836747520043485e-05,
+      "loss": 1.0119,
+      "step": 87515
+    },
+    {
+      "epoch": 5.946460116863705,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.283462426960185e-05,
+      "loss": 0.8827,
+      "step": 87520
+    },
+    {
+      "epoch": 5.946799836934366,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2832501019160211e-05,
+      "loss": 0.8559,
+      "step": 87525
+    },
+    {
+      "epoch": 5.947139557005028,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.2830377768718577e-05,
+      "loss": 0.7865,
+      "step": 87530
+    },
+    {
+      "epoch": 5.94747927707569,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2828254518276941e-05,
+      "loss": 0.9262,
+      "step": 87535
+    },
+    {
+      "epoch": 5.947818997146351,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2826131267835303e-05,
+      "loss": 0.819,
+      "step": 87540
+    },
+    {
+      "epoch": 5.948158717217013,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2824008017393669e-05,
+      "loss": 0.8048,
+      "step": 87545
+    },
+    {
+      "epoch": 5.948498437287675,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2821884766952033e-05,
+      "loss": 0.9088,
+      "step": 87550
+    },
+    {
+      "epoch": 5.948838157358336,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.2819761516510395e-05,
+      "loss": 0.9741,
+      "step": 87555
+    },
+    {
+      "epoch": 5.9491778774289985,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.281763826606876e-05,
+      "loss": 0.9238,
+      "step": 87560
+    },
+    {
+      "epoch": 5.949517597499661,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2815515015627125e-05,
+      "loss": 0.8154,
+      "step": 87565
+    },
+    {
+      "epoch": 5.949857317570322,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.2813391765185487e-05,
+      "loss": 0.8086,
+      "step": 87570
+    },
+    {
+      "epoch": 5.950197037640984,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.2811268514743851e-05,
+      "loss": 0.7376,
+      "step": 87575
+    },
+    {
+      "epoch": 5.950536757711646,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2809145264302217e-05,
+      "loss": 0.8452,
+      "step": 87580
+    },
+    {
+      "epoch": 5.950876477782307,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.280702201386058e-05,
+      "loss": 0.8114,
+      "step": 87585
+    },
+    {
+      "epoch": 5.951216197852969,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2804898763418943e-05,
+      "loss": 0.8824,
+      "step": 87590
+    },
+    {
+      "epoch": 5.951555917923631,
+      "grad_norm": 7.25,
+      "learning_rate": 1.2802775512977307e-05,
+      "loss": 0.9211,
+      "step": 87595
+    },
+    {
+      "epoch": 5.951895637994292,
+      "grad_norm": 8.0,
+      "learning_rate": 1.280065226253567e-05,
+      "loss": 0.8902,
+      "step": 87600
+    },
+    {
+      "epoch": 5.9522353580649545,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2798529012094035e-05,
+      "loss": 0.8293,
+      "step": 87605
+    },
+    {
+      "epoch": 5.952575078135617,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.27964057616524e-05,
+      "loss": 0.874,
+      "step": 87610
+    },
+    {
+      "epoch": 5.952914798206278,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2794282511210762e-05,
+      "loss": 0.8787,
+      "step": 87615
+    },
+    {
+      "epoch": 5.95325451827694,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2792159260769127e-05,
+      "loss": 0.8957,
+      "step": 87620
+    },
+    {
+      "epoch": 5.953594238347602,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2790036010327491e-05,
+      "loss": 0.8767,
+      "step": 87625
+    },
+    {
+      "epoch": 5.953933958418263,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2787912759885854e-05,
+      "loss": 0.8887,
+      "step": 87630
+    },
+    {
+      "epoch": 5.954273678488925,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2785789509444218e-05,
+      "loss": 0.9413,
+      "step": 87635
+    },
+    {
+      "epoch": 5.954613398559587,
+      "grad_norm": 6.65625,
+      "learning_rate": 1.2783666259002583e-05,
+      "loss": 0.8244,
+      "step": 87640
+    },
+    {
+      "epoch": 5.9549531186302485,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2781543008560947e-05,
+      "loss": 0.8643,
+      "step": 87645
+    },
+    {
+      "epoch": 5.9552928387009105,
+      "grad_norm": 6.875,
+      "learning_rate": 1.277941975811931e-05,
+      "loss": 0.8332,
+      "step": 87650
+    },
+    {
+      "epoch": 5.955632558771573,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2777296507677675e-05,
+      "loss": 0.7801,
+      "step": 87655
+    },
+    {
+      "epoch": 5.955972278842234,
+      "grad_norm": 8.5,
+      "learning_rate": 1.277517325723604e-05,
+      "loss": 0.8433,
+      "step": 87660
+    },
+    {
+      "epoch": 5.956311998912896,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2773050006794402e-05,
+      "loss": 0.8772,
+      "step": 87665
+    },
+    {
+      "epoch": 5.956651718983558,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2770926756352766e-05,
+      "loss": 0.846,
+      "step": 87670
+    },
+    {
+      "epoch": 5.956991439054219,
+      "grad_norm": 7.875,
+      "learning_rate": 1.2768803505911131e-05,
+      "loss": 0.8702,
+      "step": 87675
+    },
+    {
+      "epoch": 5.957331159124881,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2766680255469494e-05,
+      "loss": 0.9052,
+      "step": 87680
+    },
+    {
+      "epoch": 5.957670879195543,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2764557005027858e-05,
+      "loss": 0.9183,
+      "step": 87685
+    },
+    {
+      "epoch": 5.9580105992662045,
+      "grad_norm": 6.21875,
+      "learning_rate": 1.2762433754586223e-05,
+      "loss": 0.8269,
+      "step": 87690
+    },
+    {
+      "epoch": 5.9583503193368665,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2760310504144584e-05,
+      "loss": 0.8857,
+      "step": 87695
+    },
+    {
+      "epoch": 5.958690039407529,
+      "grad_norm": 8.125,
+      "learning_rate": 1.275818725370295e-05,
+      "loss": 0.849,
+      "step": 87700
+    },
+    {
+      "epoch": 5.95902975947819,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.2756064003261314e-05,
+      "loss": 0.8407,
+      "step": 87705
+    },
+    {
+      "epoch": 5.959369479548852,
+      "grad_norm": 9.125,
+      "learning_rate": 1.2753940752819676e-05,
+      "loss": 0.8449,
+      "step": 87710
+    },
+    {
+      "epoch": 5.959709199619514,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2751817502378042e-05,
+      "loss": 0.8106,
+      "step": 87715
+    },
+    {
+      "epoch": 5.960048919690175,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2749694251936406e-05,
+      "loss": 0.8325,
+      "step": 87720
+    },
+    {
+      "epoch": 5.960388639760837,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2747571001494768e-05,
+      "loss": 0.9447,
+      "step": 87725
+    },
+    {
+      "epoch": 5.960728359831499,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.2745447751053134e-05,
+      "loss": 0.9086,
+      "step": 87730
+    },
+    {
+      "epoch": 5.9610680799021605,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2743324500611498e-05,
+      "loss": 0.8482,
+      "step": 87735
+    },
+    {
+      "epoch": 5.9614077999728226,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.274120125016986e-05,
+      "loss": 0.8484,
+      "step": 87740
+    },
+    {
+      "epoch": 5.961747520043485,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2739077999728224e-05,
+      "loss": 0.7901,
+      "step": 87745
+    },
+    {
+      "epoch": 5.962087240114146,
+      "grad_norm": 8.0,
+      "learning_rate": 1.273695474928659e-05,
+      "loss": 0.9022,
+      "step": 87750
+    },
+    {
+      "epoch": 5.962426960184808,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2734831498844952e-05,
+      "loss": 0.8643,
+      "step": 87755
+    },
+    {
+      "epoch": 5.96276668025547,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2732708248403316e-05,
+      "loss": 0.7759,
+      "step": 87760
+    },
+    {
+      "epoch": 5.963106400326131,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2730584997961682e-05,
+      "loss": 0.9459,
+      "step": 87765
+    },
+    {
+      "epoch": 5.963446120396793,
+      "grad_norm": 7.875,
+      "learning_rate": 1.2728461747520042e-05,
+      "loss": 0.7913,
+      "step": 87770
+    },
+    {
+      "epoch": 5.963785840467455,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2726338497078408e-05,
+      "loss": 0.8648,
+      "step": 87775
+    },
+    {
+      "epoch": 5.9641255605381165,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2724215246636772e-05,
+      "loss": 0.864,
+      "step": 87780
+    },
+    {
+      "epoch": 5.964465280608779,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2722091996195134e-05,
+      "loss": 0.8527,
+      "step": 87785
+    },
+    {
+      "epoch": 5.964805000679441,
+      "grad_norm": 8.125,
+      "learning_rate": 1.27199687457535e-05,
+      "loss": 0.8621,
+      "step": 87790
+    },
+    {
+      "epoch": 5.965144720750102,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.2717845495311864e-05,
+      "loss": 0.8105,
+      "step": 87795
+    },
+    {
+      "epoch": 5.965484440820764,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.2715722244870226e-05,
+      "loss": 0.8344,
+      "step": 87800
+    },
+    {
+      "epoch": 5.965824160891425,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.271359899442859e-05,
+      "loss": 0.77,
+      "step": 87805
+    },
+    {
+      "epoch": 5.966163880962087,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2711475743986956e-05,
+      "loss": 0.869,
+      "step": 87810
+    },
+    {
+      "epoch": 5.966503601032749,
+      "grad_norm": 9.0,
+      "learning_rate": 1.270935249354532e-05,
+      "loss": 1.0286,
+      "step": 87815
+    },
+    {
+      "epoch": 5.96684332110341,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2707229243103682e-05,
+      "loss": 0.9235,
+      "step": 87820
+    },
+    {
+      "epoch": 5.9671830411740725,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.2705105992662048e-05,
+      "loss": 0.8334,
+      "step": 87825
+    },
+    {
+      "epoch": 5.967522761244735,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2702982742220412e-05,
+      "loss": 0.8527,
+      "step": 87830
+    },
+    {
+      "epoch": 5.967862481315396,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.2700859491778774e-05,
+      "loss": 0.8613,
+      "step": 87835
+    },
+    {
+      "epoch": 5.968202201386058,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.2698736241337138e-05,
+      "loss": 0.9336,
+      "step": 87840
+    },
+    {
+      "epoch": 5.96854192145672,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2696612990895504e-05,
+      "loss": 0.951,
+      "step": 87845
+    },
+    {
+      "epoch": 5.968881641527381,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2694489740453866e-05,
+      "loss": 0.8261,
+      "step": 87850
+    },
+    {
+      "epoch": 5.969221361598043,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.269236649001223e-05,
+      "loss": 0.9526,
+      "step": 87855
+    },
+    {
+      "epoch": 5.969561081668705,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.2690243239570596e-05,
+      "loss": 0.8116,
+      "step": 87860
+    },
+    {
+      "epoch": 5.969900801739366,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2688119989128958e-05,
+      "loss": 0.8811,
+      "step": 87865
+    },
+    {
+      "epoch": 5.9702405218100285,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2685996738687322e-05,
+      "loss": 0.8907,
+      "step": 87870
+    },
+    {
+      "epoch": 5.970580241880691,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2683873488245688e-05,
+      "loss": 0.8969,
+      "step": 87875
+    },
+    {
+      "epoch": 5.970919961951352,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2681750237804049e-05,
+      "loss": 0.8908,
+      "step": 87880
+    },
+    {
+      "epoch": 5.971259682022014,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.2679626987362414e-05,
+      "loss": 0.843,
+      "step": 87885
+    },
+    {
+      "epoch": 5.971599402092676,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.2677503736920778e-05,
+      "loss": 0.7315,
+      "step": 87890
+    },
+    {
+      "epoch": 5.971939122163337,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.267538048647914e-05,
+      "loss": 0.8767,
+      "step": 87895
+    },
+    {
+      "epoch": 5.972278842233999,
+      "grad_norm": 7.125,
+      "learning_rate": 1.2673257236037506e-05,
+      "loss": 0.7699,
+      "step": 87900
+    },
+    {
+      "epoch": 5.972618562304661,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.267113398559587e-05,
+      "loss": 0.7847,
+      "step": 87905
+    },
+    {
+      "epoch": 5.9729582823753224,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2669010735154233e-05,
+      "loss": 0.8608,
+      "step": 87910
+    },
+    {
+      "epoch": 5.9732980024459845,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2666887484712597e-05,
+      "loss": 0.8649,
+      "step": 87915
+    },
+    {
+      "epoch": 5.973637722516647,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.2664764234270962e-05,
+      "loss": 0.832,
+      "step": 87920
+    },
+    {
+      "epoch": 5.973977442587308,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2662640983829325e-05,
+      "loss": 0.8548,
+      "step": 87925
+    },
+    {
+      "epoch": 5.97431716265797,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2660517733387689e-05,
+      "loss": 0.8635,
+      "step": 87930
+    },
+    {
+      "epoch": 5.974656882728632,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2658394482946054e-05,
+      "loss": 0.8773,
+      "step": 87935
+    },
+    {
+      "epoch": 5.974996602799293,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2656271232504415e-05,
+      "loss": 0.8442,
+      "step": 87940
+    },
+    {
+      "epoch": 5.975336322869955,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.265414798206278e-05,
+      "loss": 0.926,
+      "step": 87945
+    },
+    {
+      "epoch": 5.975676042940617,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2652024731621145e-05,
+      "loss": 0.7026,
+      "step": 87950
+    },
+    {
+      "epoch": 5.9760157630112785,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2649901481179507e-05,
+      "loss": 0.8119,
+      "step": 87955
+    },
+    {
+      "epoch": 5.9763554830819405,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2647778230737873e-05,
+      "loss": 0.9565,
+      "step": 87960
+    },
+    {
+      "epoch": 5.976695203152603,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.2645654980296237e-05,
+      "loss": 0.7595,
+      "step": 87965
+    },
+    {
+      "epoch": 5.977034923223264,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2643531729854599e-05,
+      "loss": 0.8838,
+      "step": 87970
+    },
+    {
+      "epoch": 5.977374643293926,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2641408479412965e-05,
+      "loss": 0.9058,
+      "step": 87975
+    },
+    {
+      "epoch": 5.977714363364588,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2639285228971329e-05,
+      "loss": 0.8451,
+      "step": 87980
+    },
+    {
+      "epoch": 5.978054083435249,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2637161978529693e-05,
+      "loss": 0.9248,
+      "step": 87985
+    },
+    {
+      "epoch": 5.978393803505911,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2635038728088055e-05,
+      "loss": 0.8812,
+      "step": 87990
+    },
+    {
+      "epoch": 5.978733523576573,
+      "grad_norm": 8.25,
+      "learning_rate": 1.263291547764642e-05,
+      "loss": 0.8658,
+      "step": 87995
+    },
+    {
+      "epoch": 5.9790732436472345,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.2630792227204785e-05,
+      "loss": 0.8225,
+      "step": 88000
+    },
+    {
+      "epoch": 5.9794129637178965,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2628668976763147e-05,
+      "loss": 0.8679,
+      "step": 88005
+    },
+    {
+      "epoch": 5.979752683788558,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2626545726321513e-05,
+      "loss": 0.7567,
+      "step": 88010
+    },
+    {
+      "epoch": 5.98009240385922,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2624422475879877e-05,
+      "loss": 0.8053,
+      "step": 88015
+    },
+    {
+      "epoch": 5.980432123929882,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2622299225438239e-05,
+      "loss": 0.8637,
+      "step": 88020
+    },
+    {
+      "epoch": 5.980771844000543,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.2620175974996603e-05,
+      "loss": 0.8916,
+      "step": 88025
+    },
+    {
+      "epoch": 5.981111564071205,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2618052724554969e-05,
+      "loss": 0.908,
+      "step": 88030
+    },
+    {
+      "epoch": 5.981451284141867,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2615929474113331e-05,
+      "loss": 0.8781,
+      "step": 88035
+    },
+    {
+      "epoch": 5.981791004212528,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2613806223671695e-05,
+      "loss": 0.8388,
+      "step": 88040
+    },
+    {
+      "epoch": 5.9821307242831905,
+      "grad_norm": 7.75,
+      "learning_rate": 1.261168297323006e-05,
+      "loss": 0.8113,
+      "step": 88045
+    },
+    {
+      "epoch": 5.982470444353853,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2609559722788421e-05,
+      "loss": 0.8028,
+      "step": 88050
+    },
+    {
+      "epoch": 5.982810164424514,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2607436472346787e-05,
+      "loss": 0.7927,
+      "step": 88055
+    },
+    {
+      "epoch": 5.983149884495176,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.2605313221905151e-05,
+      "loss": 0.9831,
+      "step": 88060
+    },
+    {
+      "epoch": 5.983489604565838,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.2603189971463513e-05,
+      "loss": 0.8506,
+      "step": 88065
+    },
+    {
+      "epoch": 5.983829324636499,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.2601066721021879e-05,
+      "loss": 0.9316,
+      "step": 88070
+    },
+    {
+      "epoch": 5.984169044707161,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2598943470580243e-05,
+      "loss": 0.9162,
+      "step": 88075
+    },
+    {
+      "epoch": 5.984508764777823,
+      "grad_norm": 7.625,
+      "learning_rate": 1.2596820220138605e-05,
+      "loss": 0.8382,
+      "step": 88080
+    },
+    {
+      "epoch": 5.984848484848484,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.259469696969697e-05,
+      "loss": 0.8463,
+      "step": 88085
+    },
+    {
+      "epoch": 5.9851882049191465,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2592573719255335e-05,
+      "loss": 0.8637,
+      "step": 88090
+    },
+    {
+      "epoch": 5.985527924989809,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2590450468813697e-05,
+      "loss": 0.8666,
+      "step": 88095
+    },
+    {
+      "epoch": 5.98586764506047,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2588327218372061e-05,
+      "loss": 0.8475,
+      "step": 88100
+    },
+    {
+      "epoch": 5.986207365131132,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.2586203967930427e-05,
+      "loss": 0.8717,
+      "step": 88105
+    },
+    {
+      "epoch": 5.986547085201794,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.258408071748879e-05,
+      "loss": 0.9373,
+      "step": 88110
+    },
+    {
+      "epoch": 5.986886805272455,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2581957467047153e-05,
+      "loss": 0.9217,
+      "step": 88115
+    },
+    {
+      "epoch": 5.987226525343117,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2579834216605519e-05,
+      "loss": 0.846,
+      "step": 88120
+    },
+    {
+      "epoch": 5.987566245413779,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.257771096616388e-05,
+      "loss": 0.8221,
+      "step": 88125
+    },
+    {
+      "epoch": 5.98790596548444,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.2575587715722245e-05,
+      "loss": 0.8768,
+      "step": 88130
+    },
+    {
+      "epoch": 5.9882456855551025,
+      "grad_norm": 7.875,
+      "learning_rate": 1.257346446528061e-05,
+      "loss": 0.8228,
+      "step": 88135
+    },
+    {
+      "epoch": 5.988585405625765,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2571341214838972e-05,
+      "loss": 0.8353,
+      "step": 88140
+    },
+    {
+      "epoch": 5.988925125696426,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.2569217964397337e-05,
+      "loss": 0.8903,
+      "step": 88145
+    },
+    {
+      "epoch": 5.989264845767088,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2567094713955701e-05,
+      "loss": 0.8835,
+      "step": 88150
+    },
+    {
+      "epoch": 5.98960456583775,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2564971463514067e-05,
+      "loss": 0.9114,
+      "step": 88155
+    },
+    {
+      "epoch": 5.989944285908411,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.2562848213072428e-05,
+      "loss": 0.8691,
+      "step": 88160
+    },
+    {
+      "epoch": 5.990284005979073,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.2560724962630793e-05,
+      "loss": 0.8308,
+      "step": 88165
+    },
+    {
+      "epoch": 5.990623726049735,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2558601712189157e-05,
+      "loss": 0.9055,
+      "step": 88170
+    },
+    {
+      "epoch": 5.990963446120396,
+      "grad_norm": 9.5,
+      "learning_rate": 1.255647846174752e-05,
+      "loss": 0.8617,
+      "step": 88175
+    },
+    {
+      "epoch": 5.9913031661910585,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.2554355211305885e-05,
+      "loss": 0.836,
+      "step": 88180
+    },
+    {
+      "epoch": 5.991642886261721,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.255223196086425e-05,
+      "loss": 0.8303,
+      "step": 88185
+    },
+    {
+      "epoch": 5.991982606332382,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.2550108710422612e-05,
+      "loss": 0.7836,
+      "step": 88190
+    },
+    {
+      "epoch": 5.992322326403044,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.2547985459980976e-05,
+      "loss": 0.7476,
+      "step": 88195
+    },
+    {
+      "epoch": 5.992662046473706,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2545862209539342e-05,
+      "loss": 0.9455,
+      "step": 88200
+    },
+    {
+      "epoch": 5.993001766544367,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2543738959097704e-05,
+      "loss": 0.7862,
+      "step": 88205
+    },
+    {
+      "epoch": 5.993341486615029,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2541615708656068e-05,
+      "loss": 0.8834,
+      "step": 88210
+    },
+    {
+      "epoch": 5.993681206685691,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2539492458214434e-05,
+      "loss": 0.8192,
+      "step": 88215
+    },
+    {
+      "epoch": 5.9940209267563525,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2537369207772796e-05,
+      "loss": 0.8449,
+      "step": 88220
+    },
+    {
+      "epoch": 5.9943606468270145,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.253524595733116e-05,
+      "loss": 0.9007,
+      "step": 88225
+    },
+    {
+      "epoch": 5.994700366897677,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2533122706889524e-05,
+      "loss": 0.8461,
+      "step": 88230
+    },
+    {
+      "epoch": 5.995040086968338,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2530999456447886e-05,
+      "loss": 0.8729,
+      "step": 88235
+    },
+    {
+      "epoch": 5.995379807039,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.2528876206006252e-05,
+      "loss": 0.9018,
+      "step": 88240
+    },
+    {
+      "epoch": 5.995719527109662,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2526752955564616e-05,
+      "loss": 0.8467,
+      "step": 88245
+    },
+    {
+      "epoch": 5.996059247180323,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2524629705122978e-05,
+      "loss": 0.885,
+      "step": 88250
+    },
+    {
+      "epoch": 5.996398967250985,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.2522506454681344e-05,
+      "loss": 0.8555,
+      "step": 88255
+    },
+    {
+      "epoch": 5.996738687321647,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2520383204239708e-05,
+      "loss": 0.8009,
+      "step": 88260
+    },
+    {
+      "epoch": 5.9970784073923085,
+      "grad_norm": 8.125,
+      "learning_rate": 1.251825995379807e-05,
+      "loss": 0.875,
+      "step": 88265
+    },
+    {
+      "epoch": 5.9974181274629705,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.2516136703356434e-05,
+      "loss": 0.6889,
+      "step": 88270
+    },
+    {
+      "epoch": 5.997757847533633,
+      "grad_norm": 7.875,
+      "learning_rate": 1.25140134529148e-05,
+      "loss": 0.7604,
+      "step": 88275
+    },
+    {
+      "epoch": 5.998097567604294,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.2511890202473162e-05,
+      "loss": 0.8863,
+      "step": 88280
+    },
+    {
+      "epoch": 5.998437287674956,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2509766952031526e-05,
+      "loss": 0.8398,
+      "step": 88285
+    },
+    {
+      "epoch": 5.998777007745618,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.2507643701589892e-05,
+      "loss": 0.8426,
+      "step": 88290
+    },
+    {
+      "epoch": 5.999116727816279,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.2505520451148252e-05,
+      "loss": 0.8368,
+      "step": 88295
+    },
+    {
+      "epoch": 5.999456447886941,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2503397200706618e-05,
+      "loss": 0.8004,
+      "step": 88300
+    },
+    {
+      "epoch": 5.999796167957603,
+      "grad_norm": 9.0,
+      "learning_rate": 1.2501273950264982e-05,
+      "loss": 0.8277,
+      "step": 88305
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 1.7578717470169067,
+      "eval_runtime": 146.8961,
+      "eval_samples_per_second": 70.247,
+      "eval_steps_per_second": 8.782,
+      "step": 88308
+    },
+    {
+      "epoch": 6.0001358880282645,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.2499150699823346e-05,
+      "loss": 0.8021,
+      "step": 88310
+    },
+    {
+      "epoch": 6.0004756080989265,
+      "grad_norm": 8.0,
+      "learning_rate": 1.249702744938171e-05,
+      "loss": 0.8066,
+      "step": 88315
+    },
+    {
+      "epoch": 6.000815328169589,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2494904198940074e-05,
+      "loss": 0.8475,
+      "step": 88320
+    },
+    {
+      "epoch": 6.00115504824025,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2492780948498438e-05,
+      "loss": 0.8405,
+      "step": 88325
+    },
+    {
+      "epoch": 6.001494768310912,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.24906576980568e-05,
+      "loss": 0.8341,
+      "step": 88330
+    },
+    {
+      "epoch": 6.001834488381574,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.2488534447615166e-05,
+      "loss": 0.7489,
+      "step": 88335
+    },
+    {
+      "epoch": 6.002174208452235,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.248641119717353e-05,
+      "loss": 0.7983,
+      "step": 88340
+    },
+    {
+      "epoch": 6.002513928522897,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.2484287946731893e-05,
+      "loss": 0.8263,
+      "step": 88345
+    },
+    {
+      "epoch": 6.002853648593559,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.2482164696290258e-05,
+      "loss": 0.883,
+      "step": 88350
+    },
+    {
+      "epoch": 6.0031933686642205,
+      "grad_norm": 8.0,
+      "learning_rate": 1.248004144584862e-05,
+      "loss": 0.8093,
+      "step": 88355
+    },
+    {
+      "epoch": 6.003533088734883,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2477918195406985e-05,
+      "loss": 0.8405,
+      "step": 88360
+    },
+    {
+      "epoch": 6.003872808805545,
+      "grad_norm": 7.875,
+      "learning_rate": 1.247579494496535e-05,
+      "loss": 0.7718,
+      "step": 88365
+    },
+    {
+      "epoch": 6.004212528876206,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2473671694523713e-05,
+      "loss": 0.7999,
+      "step": 88370
+    },
+    {
+      "epoch": 6.004552248946868,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2471548444082078e-05,
+      "loss": 0.8381,
+      "step": 88375
+    },
+    {
+      "epoch": 6.00489196901753,
+      "grad_norm": 8.0,
+      "learning_rate": 1.246942519364044e-05,
+      "loss": 0.8392,
+      "step": 88380
+    },
+    {
+      "epoch": 6.005231689088191,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2467301943198805e-05,
+      "loss": 0.8896,
+      "step": 88385
+    },
+    {
+      "epoch": 6.005571409158853,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.2465178692757169e-05,
+      "loss": 0.9335,
+      "step": 88390
+    },
+    {
+      "epoch": 6.005911129229515,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2463055442315533e-05,
+      "loss": 0.9417,
+      "step": 88395
+    },
+    {
+      "epoch": 6.0062508493001765,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2460932191873897e-05,
+      "loss": 0.8234,
+      "step": 88400
+    },
+    {
+      "epoch": 6.006590569370839,
+      "grad_norm": 8.375,
+      "learning_rate": 1.245880894143226e-05,
+      "loss": 0.794,
+      "step": 88405
+    },
+    {
+      "epoch": 6.006930289441501,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2456685690990625e-05,
+      "loss": 0.8414,
+      "step": 88410
+    },
+    {
+      "epoch": 6.007270009512162,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2454562440548989e-05,
+      "loss": 0.8097,
+      "step": 88415
+    },
+    {
+      "epoch": 6.007609729582824,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2452439190107353e-05,
+      "loss": 0.833,
+      "step": 88420
+    },
+    {
+      "epoch": 6.007949449653485,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2450315939665717e-05,
+      "loss": 0.8342,
+      "step": 88425
+    },
+    {
+      "epoch": 6.008289169724147,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2448192689224079e-05,
+      "loss": 0.8366,
+      "step": 88430
+    },
+    {
+      "epoch": 6.008628889794809,
+      "grad_norm": 7.625,
+      "learning_rate": 1.2446069438782445e-05,
+      "loss": 0.8291,
+      "step": 88435
+    },
+    {
+      "epoch": 6.00896860986547,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2443946188340807e-05,
+      "loss": 0.7919,
+      "step": 88440
+    },
+    {
+      "epoch": 6.0093083299361325,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2441822937899171e-05,
+      "loss": 0.9308,
+      "step": 88445
+    },
+    {
+      "epoch": 6.009648050006795,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2439699687457537e-05,
+      "loss": 0.7792,
+      "step": 88450
+    },
+    {
+      "epoch": 6.009987770077456,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2437576437015899e-05,
+      "loss": 0.7827,
+      "step": 88455
+    },
+    {
+      "epoch": 6.010327490148118,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2435453186574265e-05,
+      "loss": 0.8557,
+      "step": 88460
+    },
+    {
+      "epoch": 6.01066721021878,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2433329936132627e-05,
+      "loss": 0.8381,
+      "step": 88465
+    },
+    {
+      "epoch": 6.011006930289441,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2431206685690991e-05,
+      "loss": 0.8661,
+      "step": 88470
+    },
+    {
+      "epoch": 6.011346650360103,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2429083435249355e-05,
+      "loss": 0.8584,
+      "step": 88475
+    },
+    {
+      "epoch": 6.011686370430765,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2426960184807719e-05,
+      "loss": 0.8008,
+      "step": 88480
+    },
+    {
+      "epoch": 6.0120260905014264,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2424836934366083e-05,
+      "loss": 0.7814,
+      "step": 88485
+    },
+    {
+      "epoch": 6.0123658105720885,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2422713683924447e-05,
+      "loss": 0.8646,
+      "step": 88490
+    },
+    {
+      "epoch": 6.012705530642751,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.2420590433482811e-05,
+      "loss": 0.833,
+      "step": 88495
+    },
+    {
+      "epoch": 6.013045250713412,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.2418467183041175e-05,
+      "loss": 0.8209,
+      "step": 88500
+    },
+    {
+      "epoch": 6.013384970784074,
+      "grad_norm": 7.875,
+      "learning_rate": 1.2416343932599539e-05,
+      "loss": 0.7709,
+      "step": 88505
+    },
+    {
+      "epoch": 6.013724690854736,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2414220682157903e-05,
+      "loss": 0.7967,
+      "step": 88510
+    },
+    {
+      "epoch": 6.014064410925397,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2412097431716265e-05,
+      "loss": 0.8427,
+      "step": 88515
+    },
+    {
+      "epoch": 6.014404130996059,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2409974181274631e-05,
+      "loss": 0.8796,
+      "step": 88520
+    },
+    {
+      "epoch": 6.014743851066721,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2407850930832993e-05,
+      "loss": 0.8451,
+      "step": 88525
+    },
+    {
+      "epoch": 6.0150835711373825,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2405727680391357e-05,
+      "loss": 0.8488,
+      "step": 88530
+    },
+    {
+      "epoch": 6.0154232912080445,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2403604429949723e-05,
+      "loss": 0.8583,
+      "step": 88535
+    },
+    {
+      "epoch": 6.015763011278707,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.2401481179508085e-05,
+      "loss": 0.7233,
+      "step": 88540
+    },
+    {
+      "epoch": 6.016102731349368,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.2399357929066451e-05,
+      "loss": 0.805,
+      "step": 88545
+    },
+    {
+      "epoch": 6.01644245142003,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2397234678624813e-05,
+      "loss": 0.8383,
+      "step": 88550
+    },
+    {
+      "epoch": 6.016782171490692,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2395111428183177e-05,
+      "loss": 0.8573,
+      "step": 88555
+    },
+    {
+      "epoch": 6.017121891561353,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2392988177741541e-05,
+      "loss": 0.7581,
+      "step": 88560
+    },
+    {
+      "epoch": 6.017461611632015,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2390864927299905e-05,
+      "loss": 0.8661,
+      "step": 88565
+    },
+    {
+      "epoch": 6.017801331702677,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.238874167685827e-05,
+      "loss": 0.8198,
+      "step": 88570
+    },
+    {
+      "epoch": 6.0181410517733385,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2386618426416633e-05,
+      "loss": 0.7942,
+      "step": 88575
+    },
+    {
+      "epoch": 6.0184807718440005,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2384495175974997e-05,
+      "loss": 0.8061,
+      "step": 88580
+    },
+    {
+      "epoch": 6.018820491914663,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.2382371925533361e-05,
+      "loss": 0.8571,
+      "step": 88585
+    },
+    {
+      "epoch": 6.019160211985324,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.2380248675091725e-05,
+      "loss": 0.886,
+      "step": 88590
+    },
+    {
+      "epoch": 6.019499932055986,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.237812542465009e-05,
+      "loss": 0.7354,
+      "step": 88595
+    },
+    {
+      "epoch": 6.019839652126648,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2376002174208452e-05,
+      "loss": 0.897,
+      "step": 88600
+    },
+    {
+      "epoch": 6.020179372197309,
+      "grad_norm": 9.375,
+      "learning_rate": 1.2373878923766817e-05,
+      "loss": 0.7671,
+      "step": 88605
+    },
+    {
+      "epoch": 6.020519092267971,
+      "grad_norm": 8.125,
+      "learning_rate": 1.237175567332518e-05,
+      "loss": 0.7675,
+      "step": 88610
+    },
+    {
+      "epoch": 6.020858812338633,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2369632422883544e-05,
+      "loss": 0.8526,
+      "step": 88615
+    },
+    {
+      "epoch": 6.0211985324092945,
+      "grad_norm": 8.375,
+      "learning_rate": 1.236750917244191e-05,
+      "loss": 0.8771,
+      "step": 88620
+    },
+    {
+      "epoch": 6.0215382524799566,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2365385922000272e-05,
+      "loss": 0.8578,
+      "step": 88625
+    },
+    {
+      "epoch": 6.021877972550619,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2363262671558637e-05,
+      "loss": 0.8448,
+      "step": 88630
+    },
+    {
+      "epoch": 6.02221769262128,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2361139421117e-05,
+      "loss": 0.9137,
+      "step": 88635
+    },
+    {
+      "epoch": 6.022557412691942,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2359016170675364e-05,
+      "loss": 0.8299,
+      "step": 88640
+    },
+    {
+      "epoch": 6.022897132762604,
+      "grad_norm": 8.375,
+      "learning_rate": 1.235689292023373e-05,
+      "loss": 0.7638,
+      "step": 88645
+    },
+    {
+      "epoch": 6.023236852833265,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2354769669792092e-05,
+      "loss": 0.8923,
+      "step": 88650
+    },
+    {
+      "epoch": 6.023576572903927,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2352646419350456e-05,
+      "loss": 0.8192,
+      "step": 88655
+    },
+    {
+      "epoch": 6.023916292974589,
+      "grad_norm": 7.25,
+      "learning_rate": 1.235052316890882e-05,
+      "loss": 0.8263,
+      "step": 88660
+    },
+    {
+      "epoch": 6.0242560130452505,
+      "grad_norm": 9.125,
+      "learning_rate": 1.2348399918467184e-05,
+      "loss": 0.7795,
+      "step": 88665
+    },
+    {
+      "epoch": 6.024595733115913,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2346276668025548e-05,
+      "loss": 0.8711,
+      "step": 88670
+    },
+    {
+      "epoch": 6.024935453186575,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.2344153417583912e-05,
+      "loss": 0.8779,
+      "step": 88675
+    },
+    {
+      "epoch": 6.025275173257236,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2342030167142276e-05,
+      "loss": 0.7367,
+      "step": 88680
+    },
+    {
+      "epoch": 6.025614893327898,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2339906916700638e-05,
+      "loss": 0.8136,
+      "step": 88685
+    },
+    {
+      "epoch": 6.02595461339856,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2337783666259004e-05,
+      "loss": 0.842,
+      "step": 88690
+    },
+    {
+      "epoch": 6.026294333469221,
+      "grad_norm": 9.25,
+      "learning_rate": 1.2335660415817368e-05,
+      "loss": 0.7419,
+      "step": 88695
+    },
+    {
+      "epoch": 6.026634053539883,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.233353716537573e-05,
+      "loss": 0.9294,
+      "step": 88700
+    },
+    {
+      "epoch": 6.026973773610545,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.2331413914934096e-05,
+      "loss": 0.8197,
+      "step": 88705
+    },
+    {
+      "epoch": 6.0273134936812065,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2329290664492458e-05,
+      "loss": 0.8151,
+      "step": 88710
+    },
+    {
+      "epoch": 6.027653213751869,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2327167414050824e-05,
+      "loss": 0.7928,
+      "step": 88715
+    },
+    {
+      "epoch": 6.027992933822531,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2325044163609186e-05,
+      "loss": 0.8023,
+      "step": 88720
+    },
+    {
+      "epoch": 6.028332653893192,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.232292091316755e-05,
+      "loss": 0.8901,
+      "step": 88725
+    },
+    {
+      "epoch": 6.028672373963854,
+      "grad_norm": 7.875,
+      "learning_rate": 1.2320797662725916e-05,
+      "loss": 0.8107,
+      "step": 88730
+    },
+    {
+      "epoch": 6.029012094034516,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.2318674412284278e-05,
+      "loss": 1.1425,
+      "step": 88735
+    },
+    {
+      "epoch": 6.029351814105177,
+      "grad_norm": 7.625,
+      "learning_rate": 1.2316551161842642e-05,
+      "loss": 0.8934,
+      "step": 88740
+    },
+    {
+      "epoch": 6.029691534175839,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2314427911401006e-05,
+      "loss": 0.946,
+      "step": 88745
+    },
+    {
+      "epoch": 6.0300312542465,
+      "grad_norm": 6.40625,
+      "learning_rate": 1.231230466095937e-05,
+      "loss": 0.9218,
+      "step": 88750
+    },
+    {
+      "epoch": 6.0303709743171625,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2310181410517734e-05,
+      "loss": 0.915,
+      "step": 88755
+    },
+    {
+      "epoch": 6.030710694387825,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2308058160076098e-05,
+      "loss": 0.7919,
+      "step": 88760
+    },
+    {
+      "epoch": 6.031050414458486,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2305934909634462e-05,
+      "loss": 0.8048,
+      "step": 88765
+    },
+    {
+      "epoch": 6.031390134529148,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2303811659192824e-05,
+      "loss": 0.866,
+      "step": 88770
+    },
+    {
+      "epoch": 6.03172985459981,
+      "grad_norm": 7.5,
+      "learning_rate": 1.230168840875119e-05,
+      "loss": 0.8746,
+      "step": 88775
+    },
+    {
+      "epoch": 6.032069574670471,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.2299565158309554e-05,
+      "loss": 0.797,
+      "step": 88780
+    },
+    {
+      "epoch": 6.032409294741133,
+      "grad_norm": 6.625,
+      "learning_rate": 1.2297441907867916e-05,
+      "loss": 0.7599,
+      "step": 88785
+    },
+    {
+      "epoch": 6.032749014811795,
+      "grad_norm": 7.875,
+      "learning_rate": 1.2295318657426282e-05,
+      "loss": 0.9126,
+      "step": 88790
+    },
+    {
+      "epoch": 6.0330887348824564,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2293195406984644e-05,
+      "loss": 0.8973,
+      "step": 88795
+    },
+    {
+      "epoch": 6.0334284549531185,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.229107215654301e-05,
+      "loss": 0.8102,
+      "step": 88800
+    },
+    {
+      "epoch": 6.033768175023781,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2288948906101372e-05,
+      "loss": 0.954,
+      "step": 88805
+    },
+    {
+      "epoch": 6.034107895094442,
+      "grad_norm": 7.625,
+      "learning_rate": 1.2286825655659736e-05,
+      "loss": 0.8204,
+      "step": 88810
+    },
+    {
+      "epoch": 6.034447615165104,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.2284702405218102e-05,
+      "loss": 0.8377,
+      "step": 88815
+    },
+    {
+      "epoch": 6.034787335235766,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2282579154776464e-05,
+      "loss": 0.8546,
+      "step": 88820
+    },
+    {
+      "epoch": 6.035127055306427,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2280455904334828e-05,
+      "loss": 0.8497,
+      "step": 88825
+    },
+    {
+      "epoch": 6.035466775377089,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2278332653893192e-05,
+      "loss": 0.7596,
+      "step": 88830
+    },
+    {
+      "epoch": 6.035806495447751,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2276209403451556e-05,
+      "loss": 0.8915,
+      "step": 88835
+    },
+    {
+      "epoch": 6.0361462155184125,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.227408615300992e-05,
+      "loss": 0.8011,
+      "step": 88840
+    },
+    {
+      "epoch": 6.0364859355890745,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.2271962902568284e-05,
+      "loss": 0.8344,
+      "step": 88845
+    },
+    {
+      "epoch": 6.036825655659737,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2269839652126648e-05,
+      "loss": 0.8888,
+      "step": 88850
+    },
+    {
+      "epoch": 6.037165375730398,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.226771640168501e-05,
+      "loss": 0.9077,
+      "step": 88855
+    },
+    {
+      "epoch": 6.03750509580106,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2265593151243376e-05,
+      "loss": 0.9322,
+      "step": 88860
+    },
+    {
+      "epoch": 6.037844815871722,
+      "grad_norm": 8.5,
+      "learning_rate": 1.226346990080174e-05,
+      "loss": 0.853,
+      "step": 88865
+    },
+    {
+      "epoch": 6.038184535942383,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2261346650360103e-05,
+      "loss": 0.7986,
+      "step": 88870
+    },
+    {
+      "epoch": 6.038524256013045,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2259223399918468e-05,
+      "loss": 0.7058,
+      "step": 88875
+    },
+    {
+      "epoch": 6.038863976083707,
+      "grad_norm": 7.625,
+      "learning_rate": 1.225710014947683e-05,
+      "loss": 0.9187,
+      "step": 88880
+    },
+    {
+      "epoch": 6.0392036961543685,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2254976899035196e-05,
+      "loss": 0.8807,
+      "step": 88885
+    },
+    {
+      "epoch": 6.0395434162250305,
+      "grad_norm": 9.0,
+      "learning_rate": 1.225285364859356e-05,
+      "loss": 0.8647,
+      "step": 88890
+    },
+    {
+      "epoch": 6.039883136295693,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.2250730398151923e-05,
+      "loss": 0.7946,
+      "step": 88895
+    },
+    {
+      "epoch": 6.040222856366354,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2248607147710288e-05,
+      "loss": 0.7882,
+      "step": 88900
+    },
+    {
+      "epoch": 6.040562576437016,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.224648389726865e-05,
+      "loss": 0.8607,
+      "step": 88905
+    },
+    {
+      "epoch": 6.040902296507678,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2244360646827015e-05,
+      "loss": 0.8012,
+      "step": 88910
+    },
+    {
+      "epoch": 6.041242016578339,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2242237396385379e-05,
+      "loss": 0.7993,
+      "step": 88915
+    },
+    {
+      "epoch": 6.041581736649001,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2240114145943743e-05,
+      "loss": 0.8864,
+      "step": 88920
+    },
+    {
+      "epoch": 6.041921456719663,
+      "grad_norm": 6.0625,
+      "learning_rate": 1.2237990895502107e-05,
+      "loss": 0.7984,
+      "step": 88925
+    },
+    {
+      "epoch": 6.0422611767903245,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.223586764506047e-05,
+      "loss": 0.8254,
+      "step": 88930
+    },
+    {
+      "epoch": 6.042600896860987,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2233744394618835e-05,
+      "loss": 0.9303,
+      "step": 88935
+    },
+    {
+      "epoch": 6.042940616931649,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2231621144177199e-05,
+      "loss": 0.8701,
+      "step": 88940
+    },
+    {
+      "epoch": 6.04328033700231,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2229497893735563e-05,
+      "loss": 0.8284,
+      "step": 88945
+    },
+    {
+      "epoch": 6.043620057072972,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2227374643293927e-05,
+      "loss": 0.8657,
+      "step": 88950
+    },
+    {
+      "epoch": 6.043959777143634,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.2225251392852289e-05,
+      "loss": 0.8497,
+      "step": 88955
+    },
+    {
+      "epoch": 6.044299497214295,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2223128142410655e-05,
+      "loss": 0.9064,
+      "step": 88960
+    },
+    {
+      "epoch": 6.044639217284957,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2221004891969017e-05,
+      "loss": 0.8828,
+      "step": 88965
+    },
+    {
+      "epoch": 6.044978937355619,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2218881641527383e-05,
+      "loss": 0.8553,
+      "step": 88970
+    },
+    {
+      "epoch": 6.0453186574262805,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2216758391085747e-05,
+      "loss": 0.8672,
+      "step": 88975
+    },
+    {
+      "epoch": 6.045658377496943,
+      "grad_norm": 7.5,
+      "learning_rate": 1.221463514064411e-05,
+      "loss": 0.8219,
+      "step": 88980
+    },
+    {
+      "epoch": 6.045998097567605,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.2212511890202475e-05,
+      "loss": 0.7623,
+      "step": 88985
+    },
+    {
+      "epoch": 6.046337817638266,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2210388639760837e-05,
+      "loss": 0.8591,
+      "step": 88990
+    },
+    {
+      "epoch": 6.046677537708928,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2208265389319201e-05,
+      "loss": 0.8405,
+      "step": 88995
+    },
+    {
+      "epoch": 6.04701725777959,
+      "grad_norm": 7.875,
+      "learning_rate": 1.2206142138877565e-05,
+      "loss": 0.8794,
+      "step": 89000
+    },
+    {
+      "epoch": 6.047356977850251,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.220401888843593e-05,
+      "loss": 0.8091,
+      "step": 89005
+    },
+    {
+      "epoch": 6.047696697920913,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2201895637994293e-05,
+      "loss": 0.8217,
+      "step": 89010
+    },
+    {
+      "epoch": 6.048036417991575,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2199772387552657e-05,
+      "loss": 0.8455,
+      "step": 89015
+    },
+    {
+      "epoch": 6.0483761380622365,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2197649137111021e-05,
+      "loss": 0.8244,
+      "step": 89020
+    },
+    {
+      "epoch": 6.048715858132899,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2195525886669385e-05,
+      "loss": 0.8053,
+      "step": 89025
+    },
+    {
+      "epoch": 6.049055578203561,
+      "grad_norm": 7.875,
+      "learning_rate": 1.219340263622775e-05,
+      "loss": 0.8302,
+      "step": 89030
+    },
+    {
+      "epoch": 6.049395298274222,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.2191279385786113e-05,
+      "loss": 0.9192,
+      "step": 89035
+    },
+    {
+      "epoch": 6.049735018344884,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.2189156135344476e-05,
+      "loss": 0.919,
+      "step": 89040
+    },
+    {
+      "epoch": 6.050074738415546,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2187032884902841e-05,
+      "loss": 0.8325,
+      "step": 89045
+    },
+    {
+      "epoch": 6.050414458486207,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2184909634461204e-05,
+      "loss": 0.7822,
+      "step": 89050
+    },
+    {
+      "epoch": 6.050754178556869,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.218278638401957e-05,
+      "loss": 0.966,
+      "step": 89055
+    },
+    {
+      "epoch": 6.051093898627531,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.2180663133577933e-05,
+      "loss": 0.8314,
+      "step": 89060
+    },
+    {
+      "epoch": 6.0514336186981925,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2178539883136296e-05,
+      "loss": 0.8373,
+      "step": 89065
+    },
+    {
+      "epoch": 6.051773338768855,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2176416632694661e-05,
+      "loss": 0.8118,
+      "step": 89070
+    },
+    {
+      "epoch": 6.052113058839517,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2174293382253024e-05,
+      "loss": 0.8877,
+      "step": 89075
+    },
+    {
+      "epoch": 6.052452778910178,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.2172170131811388e-05,
+      "loss": 0.9096,
+      "step": 89080
+    },
+    {
+      "epoch": 6.05279249898084,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2170046881369753e-05,
+      "loss": 0.7052,
+      "step": 89085
+    },
+    {
+      "epoch": 6.053132219051502,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2167923630928116e-05,
+      "loss": 0.8692,
+      "step": 89090
+    },
+    {
+      "epoch": 6.053471939122163,
+      "grad_norm": 7.5,
+      "learning_rate": 1.216580038048648e-05,
+      "loss": 0.9205,
+      "step": 89095
+    },
+    {
+      "epoch": 6.053811659192825,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.2163677130044844e-05,
+      "loss": 0.7904,
+      "step": 89100
+    },
+    {
+      "epoch": 6.0541513792634865,
+      "grad_norm": 9.9375,
+      "learning_rate": 1.2161553879603208e-05,
+      "loss": 0.8931,
+      "step": 89105
+    },
+    {
+      "epoch": 6.0544910993341485,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2159430629161572e-05,
+      "loss": 0.863,
+      "step": 89110
+    },
+    {
+      "epoch": 6.054830819404811,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.2157307378719936e-05,
+      "loss": 0.9185,
+      "step": 89115
+    },
+    {
+      "epoch": 6.055170539475472,
+      "grad_norm": 8.25,
+      "learning_rate": 1.21551841282783e-05,
+      "loss": 0.9397,
+      "step": 89120
+    },
+    {
+      "epoch": 6.055510259546134,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.2153060877836662e-05,
+      "loss": 0.797,
+      "step": 89125
+    },
+    {
+      "epoch": 6.055849979616796,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2150937627395028e-05,
+      "loss": 0.7693,
+      "step": 89130
+    },
+    {
+      "epoch": 6.056189699687457,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2148814376953392e-05,
+      "loss": 0.8469,
+      "step": 89135
+    },
+    {
+      "epoch": 6.056529419758119,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2146691126511756e-05,
+      "loss": 0.8953,
+      "step": 89140
+    },
+    {
+      "epoch": 6.056869139828781,
+      "grad_norm": 7.5,
+      "learning_rate": 1.214456787607012e-05,
+      "loss": 0.8749,
+      "step": 89145
+    },
+    {
+      "epoch": 6.0572088598994425,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.2142444625628482e-05,
+      "loss": 0.8719,
+      "step": 89150
+    },
+    {
+      "epoch": 6.0575485799701045,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2140321375186848e-05,
+      "loss": 0.7915,
+      "step": 89155
+    },
+    {
+      "epoch": 6.057888300040767,
+      "grad_norm": 7.75,
+      "learning_rate": 1.213819812474521e-05,
+      "loss": 0.8709,
+      "step": 89160
+    },
+    {
+      "epoch": 6.058228020111428,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2136074874303574e-05,
+      "loss": 0.7553,
+      "step": 89165
+    },
+    {
+      "epoch": 6.05856774018209,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.213395162386194e-05,
+      "loss": 0.8378,
+      "step": 89170
+    },
+    {
+      "epoch": 6.058907460252752,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2131828373420302e-05,
+      "loss": 0.8154,
+      "step": 89175
+    },
+    {
+      "epoch": 6.059247180323413,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.2129705122978666e-05,
+      "loss": 0.8508,
+      "step": 89180
+    },
+    {
+      "epoch": 6.059586900394075,
+      "grad_norm": 8.25,
+      "learning_rate": 1.212758187253703e-05,
+      "loss": 0.9151,
+      "step": 89185
+    },
+    {
+      "epoch": 6.059926620464737,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2125458622095394e-05,
+      "loss": 0.8191,
+      "step": 89190
+    },
+    {
+      "epoch": 6.0602663405353985,
+      "grad_norm": 8.75,
+      "learning_rate": 1.2123335371653758e-05,
+      "loss": 0.8248,
+      "step": 89195
+    },
+    {
+      "epoch": 6.0606060606060606,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.2121212121212122e-05,
+      "loss": 0.886,
+      "step": 89200
+    },
+    {
+      "epoch": 6.060945780676723,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2119088870770486e-05,
+      "loss": 0.8019,
+      "step": 89205
+    },
+    {
+      "epoch": 6.061285500747384,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2116965620328848e-05,
+      "loss": 0.7667,
+      "step": 89210
+    },
+    {
+      "epoch": 6.061625220818046,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.2114842369887214e-05,
+      "loss": 0.8297,
+      "step": 89215
+    },
+    {
+      "epoch": 6.061964940888708,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2112719119445578e-05,
+      "loss": 0.8675,
+      "step": 89220
+    },
+    {
+      "epoch": 6.062304660959369,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2110595869003942e-05,
+      "loss": 0.8409,
+      "step": 89225
+    },
+    {
+      "epoch": 6.062644381030031,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2108472618562306e-05,
+      "loss": 0.8268,
+      "step": 89230
+    },
+    {
+      "epoch": 6.062984101100693,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2106349368120668e-05,
+      "loss": 0.8827,
+      "step": 89235
+    },
+    {
+      "epoch": 6.0633238211713545,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2104226117679034e-05,
+      "loss": 0.817,
+      "step": 89240
+    },
+    {
+      "epoch": 6.063663541242017,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2102102867237396e-05,
+      "loss": 0.8597,
+      "step": 89245
+    },
+    {
+      "epoch": 6.064003261312679,
+      "grad_norm": 8.375,
+      "learning_rate": 1.209997961679576e-05,
+      "loss": 0.877,
+      "step": 89250
+    },
+    {
+      "epoch": 6.06434298138334,
+      "grad_norm": 8.375,
+      "learning_rate": 1.2097856366354126e-05,
+      "loss": 0.9245,
+      "step": 89255
+    },
+    {
+      "epoch": 6.064682701454002,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2095733115912488e-05,
+      "loss": 0.8126,
+      "step": 89260
+    },
+    {
+      "epoch": 6.065022421524664,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.2093609865470852e-05,
+      "loss": 0.9355,
+      "step": 89265
+    },
+    {
+      "epoch": 6.065362141595325,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2091486615029216e-05,
+      "loss": 0.9895,
+      "step": 89270
+    },
+    {
+      "epoch": 6.065701861665987,
+      "grad_norm": 8.125,
+      "learning_rate": 1.208936336458758e-05,
+      "loss": 0.8346,
+      "step": 89275
+    },
+    {
+      "epoch": 6.066041581736649,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2087240114145944e-05,
+      "loss": 0.8101,
+      "step": 89280
+    },
+    {
+      "epoch": 6.0663813018073105,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2085116863704308e-05,
+      "loss": 0.7484,
+      "step": 89285
+    },
+    {
+      "epoch": 6.066721021877973,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2082993613262672e-05,
+      "loss": 0.8792,
+      "step": 89290
+    },
+    {
+      "epoch": 6.067060741948635,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2080870362821035e-05,
+      "loss": 0.8625,
+      "step": 89295
+    },
+    {
+      "epoch": 6.067400462019296,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.20787471123794e-05,
+      "loss": 0.8918,
+      "step": 89300
+    },
+    {
+      "epoch": 6.067740182089958,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.2076623861937764e-05,
+      "loss": 0.7808,
+      "step": 89305
+    },
+    {
+      "epoch": 6.06807990216062,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2074500611496128e-05,
+      "loss": 0.8166,
+      "step": 89310
+    },
+    {
+      "epoch": 6.068419622231281,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.2072377361054492e-05,
+      "loss": 0.7426,
+      "step": 89315
+    },
+    {
+      "epoch": 6.068759342301943,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2070254110612855e-05,
+      "loss": 0.8117,
+      "step": 89320
+    },
+    {
+      "epoch": 6.069099062372605,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.206813086017122e-05,
+      "loss": 0.9322,
+      "step": 89325
+    },
+    {
+      "epoch": 6.0694387824432665,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2066007609729583e-05,
+      "loss": 0.8677,
+      "step": 89330
+    },
+    {
+      "epoch": 6.069778502513929,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2063884359287947e-05,
+      "loss": 0.8129,
+      "step": 89335
+    },
+    {
+      "epoch": 6.070118222584591,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2061761108846312e-05,
+      "loss": 0.9188,
+      "step": 89340
+    },
+    {
+      "epoch": 6.070457942655252,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2059637858404675e-05,
+      "loss": 0.7799,
+      "step": 89345
+    },
+    {
+      "epoch": 6.070797662725914,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2057514607963039e-05,
+      "loss": 0.833,
+      "step": 89350
+    },
+    {
+      "epoch": 6.071137382796576,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.2055391357521403e-05,
+      "loss": 0.7702,
+      "step": 89355
+    },
+    {
+      "epoch": 6.071477102867237,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2053268107079767e-05,
+      "loss": 0.8199,
+      "step": 89360
+    },
+    {
+      "epoch": 6.071816822937899,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.205114485663813e-05,
+      "loss": 0.7849,
+      "step": 89365
+    },
+    {
+      "epoch": 6.072156543008561,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2049021606196495e-05,
+      "loss": 0.7909,
+      "step": 89370
+    },
+    {
+      "epoch": 6.0724962630792225,
+      "grad_norm": 7.375,
+      "learning_rate": 1.2046898355754859e-05,
+      "loss": 0.8451,
+      "step": 89375
+    },
+    {
+      "epoch": 6.072835983149885,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2044775105313221e-05,
+      "loss": 0.9383,
+      "step": 89380
+    },
+    {
+      "epoch": 6.073175703220547,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2042651854871587e-05,
+      "loss": 0.8965,
+      "step": 89385
+    },
+    {
+      "epoch": 6.073515423291208,
+      "grad_norm": 8.5,
+      "learning_rate": 1.204052860442995e-05,
+      "loss": 0.9423,
+      "step": 89390
+    },
+    {
+      "epoch": 6.07385514336187,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2038405353988315e-05,
+      "loss": 0.8558,
+      "step": 89395
+    },
+    {
+      "epoch": 6.074194863432532,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2036282103546679e-05,
+      "loss": 0.7676,
+      "step": 89400
+    },
+    {
+      "epoch": 6.074534583503193,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.2034158853105041e-05,
+      "loss": 0.9136,
+      "step": 89405
+    },
+    {
+      "epoch": 6.074874303573855,
+      "grad_norm": 7.75,
+      "learning_rate": 1.2032035602663407e-05,
+      "loss": 0.8332,
+      "step": 89410
+    },
+    {
+      "epoch": 6.075214023644517,
+      "grad_norm": 7.5,
+      "learning_rate": 1.202991235222177e-05,
+      "loss": 0.7665,
+      "step": 89415
+    },
+    {
+      "epoch": 6.0755537437151785,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.2027789101780133e-05,
+      "loss": 0.7467,
+      "step": 89420
+    },
+    {
+      "epoch": 6.075893463785841,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2025665851338499e-05,
+      "loss": 0.8429,
+      "step": 89425
+    },
+    {
+      "epoch": 6.076233183856502,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2023542600896861e-05,
+      "loss": 0.8297,
+      "step": 89430
+    },
+    {
+      "epoch": 6.076572903927164,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2021419350455225e-05,
+      "loss": 0.9118,
+      "step": 89435
+    },
+    {
+      "epoch": 6.076912623997826,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.2019296100013589e-05,
+      "loss": 0.7985,
+      "step": 89440
+    },
+    {
+      "epoch": 6.077252344068487,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2017172849571953e-05,
+      "loss": 0.8549,
+      "step": 89445
+    },
+    {
+      "epoch": 6.077592064139149,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2015049599130317e-05,
+      "loss": 0.8922,
+      "step": 89450
+    },
+    {
+      "epoch": 6.077931784209811,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.2012926348688681e-05,
+      "loss": 0.8668,
+      "step": 89455
+    },
+    {
+      "epoch": 6.0782715042804725,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2010803098247045e-05,
+      "loss": 0.9734,
+      "step": 89460
+    },
+    {
+      "epoch": 6.0786112243511345,
+      "grad_norm": 7.25,
+      "learning_rate": 1.2008679847805409e-05,
+      "loss": 0.839,
+      "step": 89465
+    },
+    {
+      "epoch": 6.078950944421797,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2006556597363773e-05,
+      "loss": 0.7976,
+      "step": 89470
+    },
+    {
+      "epoch": 6.079290664492458,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2004433346922137e-05,
+      "loss": 0.8514,
+      "step": 89475
+    },
+    {
+      "epoch": 6.07963038456312,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2002310096480501e-05,
+      "loss": 0.7494,
+      "step": 89480
+    },
+    {
+      "epoch": 6.079970104633782,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.2000186846038865e-05,
+      "loss": 0.7558,
+      "step": 89485
+    },
+    {
+      "epoch": 6.080309824704443,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.1998063595597227e-05,
+      "loss": 0.9065,
+      "step": 89490
+    },
+    {
+      "epoch": 6.080649544775105,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.1995940345155593e-05,
+      "loss": 0.8425,
+      "step": 89495
+    },
+    {
+      "epoch": 6.080989264845767,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1993817094713957e-05,
+      "loss": 0.8571,
+      "step": 89500
+    },
+    {
+      "epoch": 6.0813289849164285,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.199169384427232e-05,
+      "loss": 0.9098,
+      "step": 89505
+    },
+    {
+      "epoch": 6.081668704987091,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1989570593830685e-05,
+      "loss": 1.0229,
+      "step": 89510
+    },
+    {
+      "epoch": 6.082008425057753,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.1987447343389047e-05,
+      "loss": 0.8264,
+      "step": 89515
+    },
+    {
+      "epoch": 6.082348145128414,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1985324092947411e-05,
+      "loss": 0.9099,
+      "step": 89520
+    },
+    {
+      "epoch": 6.082687865199076,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1983200842505775e-05,
+      "loss": 0.8869,
+      "step": 89525
+    },
+    {
+      "epoch": 6.083027585269738,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.198107759206414e-05,
+      "loss": 0.8715,
+      "step": 89530
+    },
+    {
+      "epoch": 6.083367305340399,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1978954341622503e-05,
+      "loss": 0.8889,
+      "step": 89535
+    },
+    {
+      "epoch": 6.083707025411061,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1976831091180867e-05,
+      "loss": 0.7895,
+      "step": 89540
+    },
+    {
+      "epoch": 6.084046745481723,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1974707840739231e-05,
+      "loss": 0.8227,
+      "step": 89545
+    },
+    {
+      "epoch": 6.0843864655523845,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1972584590297595e-05,
+      "loss": 0.8742,
+      "step": 89550
+    },
+    {
+      "epoch": 6.084726185623047,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.197046133985596e-05,
+      "loss": 0.857,
+      "step": 89555
+    },
+    {
+      "epoch": 6.085065905693709,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1968338089414323e-05,
+      "loss": 0.8075,
+      "step": 89560
+    },
+    {
+      "epoch": 6.08540562576437,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1966214838972687e-05,
+      "loss": 0.7832,
+      "step": 89565
+    },
+    {
+      "epoch": 6.085745345835032,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1964091588531051e-05,
+      "loss": 0.8596,
+      "step": 89570
+    },
+    {
+      "epoch": 6.086085065905694,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1961968338089414e-05,
+      "loss": 0.8526,
+      "step": 89575
+    },
+    {
+      "epoch": 6.086424785976355,
+      "grad_norm": 7.875,
+      "learning_rate": 1.195984508764778e-05,
+      "loss": 0.8419,
+      "step": 89580
+    },
+    {
+      "epoch": 6.086764506047017,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1957721837206143e-05,
+      "loss": 0.7953,
+      "step": 89585
+    },
+    {
+      "epoch": 6.087104226117679,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1955598586764506e-05,
+      "loss": 0.6867,
+      "step": 89590
+    },
+    {
+      "epoch": 6.0874439461883405,
+      "grad_norm": 7.625,
+      "learning_rate": 1.1953475336322871e-05,
+      "loss": 0.8524,
+      "step": 89595
+    },
+    {
+      "epoch": 6.087783666259003,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1951352085881234e-05,
+      "loss": 0.9595,
+      "step": 89600
+    },
+    {
+      "epoch": 6.088123386329665,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1949228835439598e-05,
+      "loss": 0.8914,
+      "step": 89605
+    },
+    {
+      "epoch": 6.088463106400326,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1947105584997963e-05,
+      "loss": 0.8845,
+      "step": 89610
+    },
+    {
+      "epoch": 6.088802826470988,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1944982334556326e-05,
+      "loss": 0.8135,
+      "step": 89615
+    },
+    {
+      "epoch": 6.08914254654165,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.194285908411469e-05,
+      "loss": 0.7959,
+      "step": 89620
+    },
+    {
+      "epoch": 6.089482266612311,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.1940735833673054e-05,
+      "loss": 0.8656,
+      "step": 89625
+    },
+    {
+      "epoch": 6.089821986682973,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1938612583231418e-05,
+      "loss": 0.7189,
+      "step": 89630
+    },
+    {
+      "epoch": 6.090161706753635,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1936489332789782e-05,
+      "loss": 0.8236,
+      "step": 89635
+    },
+    {
+      "epoch": 6.0905014268242965,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1934366082348146e-05,
+      "loss": 0.7643,
+      "step": 89640
+    },
+    {
+      "epoch": 6.090841146894959,
+      "grad_norm": 8.125,
+      "learning_rate": 1.193224283190651e-05,
+      "loss": 0.8111,
+      "step": 89645
+    },
+    {
+      "epoch": 6.091180866965621,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.1930119581464874e-05,
+      "loss": 0.8415,
+      "step": 89650
+    },
+    {
+      "epoch": 6.091520587036282,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1927996331023238e-05,
+      "loss": 0.9014,
+      "step": 89655
+    },
+    {
+      "epoch": 6.091860307106944,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1925873080581602e-05,
+      "loss": 0.7901,
+      "step": 89660
+    },
+    {
+      "epoch": 6.092200027177606,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1923749830139966e-05,
+      "loss": 0.7492,
+      "step": 89665
+    },
+    {
+      "epoch": 6.092539747248267,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.192162657969833e-05,
+      "loss": 0.8821,
+      "step": 89670
+    },
+    {
+      "epoch": 6.092879467318929,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1919503329256692e-05,
+      "loss": 0.8083,
+      "step": 89675
+    },
+    {
+      "epoch": 6.093219187389591,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1917380078815058e-05,
+      "loss": 0.7448,
+      "step": 89680
+    },
+    {
+      "epoch": 6.0935589074602525,
+      "grad_norm": 8.375,
+      "learning_rate": 1.191525682837342e-05,
+      "loss": 0.7683,
+      "step": 89685
+    },
+    {
+      "epoch": 6.093898627530915,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1913133577931784e-05,
+      "loss": 0.891,
+      "step": 89690
+    },
+    {
+      "epoch": 6.094238347601577,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.191101032749015e-05,
+      "loss": 0.7607,
+      "step": 89695
+    },
+    {
+      "epoch": 6.094578067672238,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.1908887077048512e-05,
+      "loss": 0.8896,
+      "step": 89700
+    },
+    {
+      "epoch": 6.0949177877429,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1906763826606876e-05,
+      "loss": 0.9094,
+      "step": 89705
+    },
+    {
+      "epoch": 6.095257507813562,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.190464057616524e-05,
+      "loss": 0.8942,
+      "step": 89710
+    },
+    {
+      "epoch": 6.095597227884223,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1902517325723604e-05,
+      "loss": 0.8004,
+      "step": 89715
+    },
+    {
+      "epoch": 6.095936947954885,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.1900394075281968e-05,
+      "loss": 0.8907,
+      "step": 89720
+    },
+    {
+      "epoch": 6.096276668025547,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.1898270824840332e-05,
+      "loss": 0.8297,
+      "step": 89725
+    },
+    {
+      "epoch": 6.0966163880962085,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.1896147574398696e-05,
+      "loss": 0.8271,
+      "step": 89730
+    },
+    {
+      "epoch": 6.096956108166871,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.189402432395706e-05,
+      "loss": 0.8637,
+      "step": 89735
+    },
+    {
+      "epoch": 6.097295828237533,
+      "grad_norm": 7.625,
+      "learning_rate": 1.1891901073515424e-05,
+      "loss": 0.8429,
+      "step": 89740
+    },
+    {
+      "epoch": 6.097635548308194,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1889777823073788e-05,
+      "loss": 0.8332,
+      "step": 89745
+    },
+    {
+      "epoch": 6.097975268378856,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1887654572632152e-05,
+      "loss": 0.7439,
+      "step": 89750
+    },
+    {
+      "epoch": 6.098314988449518,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1885531322190516e-05,
+      "loss": 0.8345,
+      "step": 89755
+    },
+    {
+      "epoch": 6.098654708520179,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1883408071748879e-05,
+      "loss": 0.905,
+      "step": 89760
+    },
+    {
+      "epoch": 6.098994428590841,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.1881284821307244e-05,
+      "loss": 0.7867,
+      "step": 89765
+    },
+    {
+      "epoch": 6.099334148661503,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1879161570865607e-05,
+      "loss": 0.8438,
+      "step": 89770
+    },
+    {
+      "epoch": 6.0996738687321646,
+      "grad_norm": 6.25,
+      "learning_rate": 1.187703832042397e-05,
+      "loss": 0.784,
+      "step": 89775
+    },
+    {
+      "epoch": 6.100013588802827,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.1874915069982336e-05,
+      "loss": 0.7788,
+      "step": 89780
+    },
+    {
+      "epoch": 6.100353308873488,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1872791819540699e-05,
+      "loss": 0.8779,
+      "step": 89785
+    },
+    {
+      "epoch": 6.10069302894415,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1870668569099063e-05,
+      "loss": 0.8009,
+      "step": 89790
+    },
+    {
+      "epoch": 6.101032749014812,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1868545318657427e-05,
+      "loss": 0.8628,
+      "step": 89795
+    },
+    {
+      "epoch": 6.101372469085473,
+      "grad_norm": 9.0,
+      "learning_rate": 1.186642206821579e-05,
+      "loss": 0.8095,
+      "step": 89800
+    },
+    {
+      "epoch": 6.101712189156135,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.1864298817774155e-05,
+      "loss": 0.7494,
+      "step": 89805
+    },
+    {
+      "epoch": 6.102051909226797,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.1862175567332519e-05,
+      "loss": 0.774,
+      "step": 89810
+    },
+    {
+      "epoch": 6.1023916292974585,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1860052316890883e-05,
+      "loss": 0.9247,
+      "step": 89815
+    },
+    {
+      "epoch": 6.102731349368121,
+      "grad_norm": 6.34375,
+      "learning_rate": 1.1857929066449247e-05,
+      "loss": 0.756,
+      "step": 89820
+    },
+    {
+      "epoch": 6.103071069438783,
+      "grad_norm": 8.25,
+      "learning_rate": 1.185580581600761e-05,
+      "loss": 0.8685,
+      "step": 89825
+    },
+    {
+      "epoch": 6.103410789509444,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.1853682565565975e-05,
+      "loss": 0.9114,
+      "step": 89830
+    },
+    {
+      "epoch": 6.103750509580106,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.1851559315124339e-05,
+      "loss": 0.8848,
+      "step": 89835
+    },
+    {
+      "epoch": 6.104090229650768,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1849436064682703e-05,
+      "loss": 0.8282,
+      "step": 89840
+    },
+    {
+      "epoch": 6.104429949721429,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.1847312814241065e-05,
+      "loss": 0.8527,
+      "step": 89845
+    },
+    {
+      "epoch": 6.104769669792091,
+      "grad_norm": 7.75,
+      "learning_rate": 1.184518956379943e-05,
+      "loss": 0.7772,
+      "step": 89850
+    },
+    {
+      "epoch": 6.105109389862753,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1843066313357795e-05,
+      "loss": 0.8198,
+      "step": 89855
+    },
+    {
+      "epoch": 6.1054491099334145,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.1840943062916157e-05,
+      "loss": 0.8319,
+      "step": 89860
+    },
+    {
+      "epoch": 6.105788830004077,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.1838819812474523e-05,
+      "loss": 0.8492,
+      "step": 89865
+    },
+    {
+      "epoch": 6.106128550074739,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1836696562032885e-05,
+      "loss": 0.8628,
+      "step": 89870
+    },
+    {
+      "epoch": 6.1064682701454,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1834573311591249e-05,
+      "loss": 0.8883,
+      "step": 89875
+    },
+    {
+      "epoch": 6.106807990216062,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1832450061149613e-05,
+      "loss": 0.8263,
+      "step": 89880
+    },
+    {
+      "epoch": 6.107147710286724,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1830326810707977e-05,
+      "loss": 0.8527,
+      "step": 89885
+    },
+    {
+      "epoch": 6.107487430357385,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1828203560266341e-05,
+      "loss": 0.7706,
+      "step": 89890
+    },
+    {
+      "epoch": 6.107827150428047,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1826080309824705e-05,
+      "loss": 0.8447,
+      "step": 89895
+    },
+    {
+      "epoch": 6.108166870498709,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1823957059383069e-05,
+      "loss": 0.8309,
+      "step": 89900
+    },
+    {
+      "epoch": 6.1085065905693705,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1821833808941433e-05,
+      "loss": 0.8142,
+      "step": 89905
+    },
+    {
+      "epoch": 6.108846310640033,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1819710558499797e-05,
+      "loss": 0.79,
+      "step": 89910
+    },
+    {
+      "epoch": 6.109186030710695,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1817587308058161e-05,
+      "loss": 0.8795,
+      "step": 89915
+    },
+    {
+      "epoch": 6.109525750781356,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1815464057616525e-05,
+      "loss": 0.8315,
+      "step": 89920
+    },
+    {
+      "epoch": 6.109865470852018,
+      "grad_norm": 9.375,
+      "learning_rate": 1.1813340807174889e-05,
+      "loss": 0.8569,
+      "step": 89925
+    },
+    {
+      "epoch": 6.11020519092268,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1811217556733251e-05,
+      "loss": 0.8421,
+      "step": 89930
+    },
+    {
+      "epoch": 6.110544910993341,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1809094306291617e-05,
+      "loss": 0.7951,
+      "step": 89935
+    },
+    {
+      "epoch": 6.110884631064003,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1806971055849981e-05,
+      "loss": 0.8534,
+      "step": 89940
+    },
+    {
+      "epoch": 6.111224351134665,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1804847805408343e-05,
+      "loss": 0.7629,
+      "step": 89945
+    },
+    {
+      "epoch": 6.1115640712053265,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1802724554966709e-05,
+      "loss": 0.9659,
+      "step": 89950
+    },
+    {
+      "epoch": 6.111903791275989,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1800601304525071e-05,
+      "loss": 0.8025,
+      "step": 89955
+    },
+    {
+      "epoch": 6.112243511346651,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1798478054083435e-05,
+      "loss": 0.7492,
+      "step": 89960
+    },
+    {
+      "epoch": 6.112583231417312,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.17963548036418e-05,
+      "loss": 0.8789,
+      "step": 89965
+    },
+    {
+      "epoch": 6.112922951487974,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.1794231553200163e-05,
+      "loss": 0.9196,
+      "step": 89970
+    },
+    {
+      "epoch": 6.113262671558636,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1792108302758527e-05,
+      "loss": 0.8346,
+      "step": 89975
+    },
+    {
+      "epoch": 6.113602391629297,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1789985052316891e-05,
+      "loss": 0.7694,
+      "step": 89980
+    },
+    {
+      "epoch": 6.113942111699959,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1787861801875255e-05,
+      "loss": 0.8146,
+      "step": 89985
+    },
+    {
+      "epoch": 6.114281831770621,
+      "grad_norm": 8.625,
+      "learning_rate": 1.178573855143362e-05,
+      "loss": 0.7871,
+      "step": 89990
+    },
+    {
+      "epoch": 6.1146215518412825,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1783615300991983e-05,
+      "loss": 0.8752,
+      "step": 89995
+    },
+    {
+      "epoch": 6.114961271911945,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1781492050550347e-05,
+      "loss": 0.8438,
+      "step": 90000
+    },
+    {
+      "epoch": 6.115300991982607,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1779368800108711e-05,
+      "loss": 0.854,
+      "step": 90005
+    },
+    {
+      "epoch": 6.115640712053268,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1777245549667075e-05,
+      "loss": 0.8868,
+      "step": 90010
+    },
+    {
+      "epoch": 6.11598043212393,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1775122299225438e-05,
+      "loss": 0.8851,
+      "step": 90015
+    },
+    {
+      "epoch": 6.116320152194592,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1772999048783803e-05,
+      "loss": 0.788,
+      "step": 90020
+    },
+    {
+      "epoch": 6.116659872265253,
+      "grad_norm": 7.25,
+      "learning_rate": 1.1770875798342167e-05,
+      "loss": 0.9056,
+      "step": 90025
+    },
+    {
+      "epoch": 6.116999592335915,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.176875254790053e-05,
+      "loss": 0.8107,
+      "step": 90030
+    },
+    {
+      "epoch": 6.117339312406577,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1766629297458895e-05,
+      "loss": 0.8126,
+      "step": 90035
+    },
+    {
+      "epoch": 6.1176790324772385,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1764506047017258e-05,
+      "loss": 0.8211,
+      "step": 90040
+    },
+    {
+      "epoch": 6.118018752547901,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1762382796575622e-05,
+      "loss": 0.8096,
+      "step": 90045
+    },
+    {
+      "epoch": 6.118358472618563,
+      "grad_norm": 7.375,
+      "learning_rate": 1.1760259546133986e-05,
+      "loss": 0.9115,
+      "step": 90050
+    },
+    {
+      "epoch": 6.118698192689224,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.175813629569235e-05,
+      "loss": 0.9232,
+      "step": 90055
+    },
+    {
+      "epoch": 6.119037912759886,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1756013045250714e-05,
+      "loss": 0.8156,
+      "step": 90060
+    },
+    {
+      "epoch": 6.119377632830548,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1753889794809078e-05,
+      "loss": 0.9021,
+      "step": 90065
+    },
+    {
+      "epoch": 6.119717352901209,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1751766544367442e-05,
+      "loss": 0.7926,
+      "step": 90070
+    },
+    {
+      "epoch": 6.120057072971871,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1749643293925806e-05,
+      "loss": 0.8013,
+      "step": 90075
+    },
+    {
+      "epoch": 6.120396793042533,
+      "grad_norm": 8.25,
+      "learning_rate": 1.174752004348417e-05,
+      "loss": 0.7966,
+      "step": 90080
+    },
+    {
+      "epoch": 6.1207365131131946,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.1745396793042534e-05,
+      "loss": 0.7103,
+      "step": 90085
+    },
+    {
+      "epoch": 6.121076233183857,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1743273542600898e-05,
+      "loss": 0.8898,
+      "step": 90090
+    },
+    {
+      "epoch": 6.121415953254519,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1741150292159262e-05,
+      "loss": 0.8576,
+      "step": 90095
+    },
+    {
+      "epoch": 6.12175567332518,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.1739027041717624e-05,
+      "loss": 0.8138,
+      "step": 90100
+    },
+    {
+      "epoch": 6.122095393395842,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.173690379127599e-05,
+      "loss": 0.7393,
+      "step": 90105
+    },
+    {
+      "epoch": 6.122435113466503,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1734780540834354e-05,
+      "loss": 0.8186,
+      "step": 90110
+    },
+    {
+      "epoch": 6.122774833537165,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1732657290392716e-05,
+      "loss": 0.8162,
+      "step": 90115
+    },
+    {
+      "epoch": 6.123114553607827,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1730534039951082e-05,
+      "loss": 0.8294,
+      "step": 90120
+    },
+    {
+      "epoch": 6.1234542736784885,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1728410789509444e-05,
+      "loss": 0.8415,
+      "step": 90125
+    },
+    {
+      "epoch": 6.123793993749151,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1726287539067808e-05,
+      "loss": 0.8604,
+      "step": 90130
+    },
+    {
+      "epoch": 6.124133713819813,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1724164288626174e-05,
+      "loss": 0.8419,
+      "step": 90135
+    },
+    {
+      "epoch": 6.124473433890474,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1722041038184536e-05,
+      "loss": 0.9759,
+      "step": 90140
+    },
+    {
+      "epoch": 6.124813153961136,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.17199177877429e-05,
+      "loss": 0.7771,
+      "step": 90145
+    },
+    {
+      "epoch": 6.125152874031798,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.1717794537301264e-05,
+      "loss": 0.8414,
+      "step": 90150
+    },
+    {
+      "epoch": 6.125492594102459,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1715671286859628e-05,
+      "loss": 0.8157,
+      "step": 90155
+    },
+    {
+      "epoch": 6.125832314173121,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.1713548036417992e-05,
+      "loss": 0.7793,
+      "step": 90160
+    },
+    {
+      "epoch": 6.126172034243783,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1711424785976356e-05,
+      "loss": 0.9159,
+      "step": 90165
+    },
+    {
+      "epoch": 6.1265117543144445,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.170930153553472e-05,
+      "loss": 0.7636,
+      "step": 90170
+    },
+    {
+      "epoch": 6.126851474385107,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1707178285093084e-05,
+      "loss": 0.8179,
+      "step": 90175
+    },
+    {
+      "epoch": 6.127191194455769,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1705055034651448e-05,
+      "loss": 0.8361,
+      "step": 90180
+    },
+    {
+      "epoch": 6.12753091452643,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1702931784209812e-05,
+      "loss": 0.8696,
+      "step": 90185
+    },
+    {
+      "epoch": 6.127870634597092,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1700808533768176e-05,
+      "loss": 0.9553,
+      "step": 90190
+    },
+    {
+      "epoch": 6.128210354667754,
+      "grad_norm": 7.625,
+      "learning_rate": 1.169868528332654e-05,
+      "loss": 0.8295,
+      "step": 90195
+    },
+    {
+      "epoch": 6.128550074738415,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1696562032884902e-05,
+      "loss": 0.8965,
+      "step": 90200
+    },
+    {
+      "epoch": 6.128889794809077,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1694438782443268e-05,
+      "loss": 0.9731,
+      "step": 90205
+    },
+    {
+      "epoch": 6.129229514879739,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.169231553200163e-05,
+      "loss": 0.8619,
+      "step": 90210
+    },
+    {
+      "epoch": 6.1295692349504005,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1690192281559994e-05,
+      "loss": 0.8343,
+      "step": 90215
+    },
+    {
+      "epoch": 6.129908955021063,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.168806903111836e-05,
+      "loss": 0.7965,
+      "step": 90220
+    },
+    {
+      "epoch": 6.130248675091725,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.1685945780676722e-05,
+      "loss": 0.8268,
+      "step": 90225
+    },
+    {
+      "epoch": 6.130588395162386,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1683822530235086e-05,
+      "loss": 0.7922,
+      "step": 90230
+    },
+    {
+      "epoch": 6.130928115233048,
+      "grad_norm": 8.25,
+      "learning_rate": 1.168169927979345e-05,
+      "loss": 0.8591,
+      "step": 90235
+    },
+    {
+      "epoch": 6.13126783530371,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1679576029351814e-05,
+      "loss": 0.9003,
+      "step": 90240
+    },
+    {
+      "epoch": 6.131607555374371,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1677452778910178e-05,
+      "loss": 0.8955,
+      "step": 90245
+    },
+    {
+      "epoch": 6.131947275445033,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1675329528468542e-05,
+      "loss": 0.7931,
+      "step": 90250
+    },
+    {
+      "epoch": 6.132286995515695,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1673206278026906e-05,
+      "loss": 0.9153,
+      "step": 90255
+    },
+    {
+      "epoch": 6.1326267155863565,
+      "grad_norm": 8.0,
+      "learning_rate": 1.167108302758527e-05,
+      "loss": 0.866,
+      "step": 90260
+    },
+    {
+      "epoch": 6.132966435657019,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1668959777143634e-05,
+      "loss": 0.8885,
+      "step": 90265
+    },
+    {
+      "epoch": 6.133306155727681,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1666836526701998e-05,
+      "loss": 0.845,
+      "step": 90270
+    },
+    {
+      "epoch": 6.133645875798342,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1664713276260362e-05,
+      "loss": 0.8448,
+      "step": 90275
+    },
+    {
+      "epoch": 6.133985595869004,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1662590025818726e-05,
+      "loss": 0.9417,
+      "step": 90280
+    },
+    {
+      "epoch": 6.134325315939666,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1660466775377089e-05,
+      "loss": 0.8746,
+      "step": 90285
+    },
+    {
+      "epoch": 6.134665036010327,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.1658343524935454e-05,
+      "loss": 0.8097,
+      "step": 90290
+    },
+    {
+      "epoch": 6.135004756080989,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1656220274493817e-05,
+      "loss": 0.8309,
+      "step": 90295
+    },
+    {
+      "epoch": 6.135344476151651,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.165409702405218e-05,
+      "loss": 0.8246,
+      "step": 90300
+    },
+    {
+      "epoch": 6.1356841962223125,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.1651973773610546e-05,
+      "loss": 0.7922,
+      "step": 90305
+    },
+    {
+      "epoch": 6.136023916292975,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1649850523168909e-05,
+      "loss": 0.8746,
+      "step": 90310
+    },
+    {
+      "epoch": 6.136363636363637,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1647727272727273e-05,
+      "loss": 0.7678,
+      "step": 90315
+    },
+    {
+      "epoch": 6.136703356434298,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1645604022285637e-05,
+      "loss": 0.867,
+      "step": 90320
+    },
+    {
+      "epoch": 6.13704307650496,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1643480771844e-05,
+      "loss": 0.7977,
+      "step": 90325
+    },
+    {
+      "epoch": 6.137382796575622,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1641357521402366e-05,
+      "loss": 0.8413,
+      "step": 90330
+    },
+    {
+      "epoch": 6.137722516646283,
+      "grad_norm": 7.0,
+      "learning_rate": 1.1639234270960729e-05,
+      "loss": 0.8694,
+      "step": 90335
+    },
+    {
+      "epoch": 6.138062236716945,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.1637111020519093e-05,
+      "loss": 0.9105,
+      "step": 90340
+    },
+    {
+      "epoch": 6.138401956787607,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.1634987770077457e-05,
+      "loss": 0.9393,
+      "step": 90345
+    },
+    {
+      "epoch": 6.1387416768582685,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.163286451963582e-05,
+      "loss": 0.8374,
+      "step": 90350
+    },
+    {
+      "epoch": 6.139081396928931,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1630741269194185e-05,
+      "loss": 0.873,
+      "step": 90355
+    },
+    {
+      "epoch": 6.139421116999593,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1628618018752549e-05,
+      "loss": 0.9051,
+      "step": 90360
+    },
+    {
+      "epoch": 6.139760837070254,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1626494768310913e-05,
+      "loss": 0.8284,
+      "step": 90365
+    },
+    {
+      "epoch": 6.140100557140916,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1624371517869275e-05,
+      "loss": 0.8595,
+      "step": 90370
+    },
+    {
+      "epoch": 6.140440277211578,
+      "grad_norm": 7.5,
+      "learning_rate": 1.162224826742764e-05,
+      "loss": 0.8209,
+      "step": 90375
+    },
+    {
+      "epoch": 6.140779997282239,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1620125016986005e-05,
+      "loss": 0.8486,
+      "step": 90380
+    },
+    {
+      "epoch": 6.141119717352901,
+      "grad_norm": 9.0,
+      "learning_rate": 1.1618001766544367e-05,
+      "loss": 0.8272,
+      "step": 90385
+    },
+    {
+      "epoch": 6.141459437423563,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.1615878516102733e-05,
+      "loss": 0.798,
+      "step": 90390
+    },
+    {
+      "epoch": 6.141799157494225,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1613755265661095e-05,
+      "loss": 0.8755,
+      "step": 90395
+    },
+    {
+      "epoch": 6.142138877564887,
+      "grad_norm": 8.125,
+      "learning_rate": 1.161163201521946e-05,
+      "loss": 0.8398,
+      "step": 90400
+    },
+    {
+      "epoch": 6.142478597635549,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1609508764777823e-05,
+      "loss": 0.9109,
+      "step": 90405
+    },
+    {
+      "epoch": 6.14281831770621,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1607385514336187e-05,
+      "loss": 0.7589,
+      "step": 90410
+    },
+    {
+      "epoch": 6.143158037776872,
+      "grad_norm": 7.5,
+      "learning_rate": 1.1605262263894553e-05,
+      "loss": 0.9039,
+      "step": 90415
+    },
+    {
+      "epoch": 6.143497757847534,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1603139013452915e-05,
+      "loss": 0.8483,
+      "step": 90420
+    },
+    {
+      "epoch": 6.143837477918195,
+      "grad_norm": 7.625,
+      "learning_rate": 1.160101576301128e-05,
+      "loss": 0.7618,
+      "step": 90425
+    },
+    {
+      "epoch": 6.144177197988857,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1598892512569643e-05,
+      "loss": 0.8725,
+      "step": 90430
+    },
+    {
+      "epoch": 6.144516918059519,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.1596769262128007e-05,
+      "loss": 0.7469,
+      "step": 90435
+    },
+    {
+      "epoch": 6.144856638130181,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1594646011686371e-05,
+      "loss": 0.8465,
+      "step": 90440
+    },
+    {
+      "epoch": 6.145196358200843,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1592522761244735e-05,
+      "loss": 0.8103,
+      "step": 90445
+    },
+    {
+      "epoch": 6.145536078271505,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.15903995108031e-05,
+      "loss": 0.8492,
+      "step": 90450
+    },
+    {
+      "epoch": 6.145875798342166,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.1588276260361462e-05,
+      "loss": 0.8321,
+      "step": 90455
+    },
+    {
+      "epoch": 6.146215518412828,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1586153009919827e-05,
+      "loss": 0.8841,
+      "step": 90460
+    },
+    {
+      "epoch": 6.14655523848349,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1584029759478191e-05,
+      "loss": 0.8203,
+      "step": 90465
+    },
+    {
+      "epoch": 6.146894958554151,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1581906509036554e-05,
+      "loss": 0.8282,
+      "step": 90470
+    },
+    {
+      "epoch": 6.147234678624813,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.157978325859492e-05,
+      "loss": 0.8533,
+      "step": 90475
+    },
+    {
+      "epoch": 6.1475743986954745,
+      "grad_norm": 7.5,
+      "learning_rate": 1.1577660008153282e-05,
+      "loss": 0.8277,
+      "step": 90480
+    },
+    {
+      "epoch": 6.147914118766137,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1575536757711646e-05,
+      "loss": 0.8792,
+      "step": 90485
+    },
+    {
+      "epoch": 6.148253838836799,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.157341350727001e-05,
+      "loss": 0.7978,
+      "step": 90490
+    },
+    {
+      "epoch": 6.14859355890746,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1571290256828374e-05,
+      "loss": 0.8764,
+      "step": 90495
+    },
+    {
+      "epoch": 6.148933278978122,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.156916700638674e-05,
+      "loss": 0.8626,
+      "step": 90500
+    },
+    {
+      "epoch": 6.149272999048784,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1567043755945102e-05,
+      "loss": 0.8104,
+      "step": 90505
+    },
+    {
+      "epoch": 6.149612719119445,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.1564920505503466e-05,
+      "loss": 0.8545,
+      "step": 90510
+    },
+    {
+      "epoch": 6.149952439190107,
+      "grad_norm": 8.0,
+      "learning_rate": 1.156279725506183e-05,
+      "loss": 0.8484,
+      "step": 90515
+    },
+    {
+      "epoch": 6.150292159260769,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1560674004620194e-05,
+      "loss": 0.8293,
+      "step": 90520
+    },
+    {
+      "epoch": 6.1506318793314305,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1558550754178558e-05,
+      "loss": 0.8243,
+      "step": 90525
+    },
+    {
+      "epoch": 6.150971599402093,
+      "grad_norm": 7.5,
+      "learning_rate": 1.1556427503736922e-05,
+      "loss": 0.8399,
+      "step": 90530
+    },
+    {
+      "epoch": 6.151311319472755,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1554304253295286e-05,
+      "loss": 0.872,
+      "step": 90535
+    },
+    {
+      "epoch": 6.151651039543416,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.1552181002853648e-05,
+      "loss": 0.8232,
+      "step": 90540
+    },
+    {
+      "epoch": 6.151990759614078,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1550057752412014e-05,
+      "loss": 0.8207,
+      "step": 90545
+    },
+    {
+      "epoch": 6.15233047968474,
+      "grad_norm": 7.25,
+      "learning_rate": 1.1547934501970378e-05,
+      "loss": 0.876,
+      "step": 90550
+    },
+    {
+      "epoch": 6.152670199755401,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.154581125152874e-05,
+      "loss": 0.8945,
+      "step": 90555
+    },
+    {
+      "epoch": 6.153009919826063,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1543688001087106e-05,
+      "loss": 0.9307,
+      "step": 90560
+    },
+    {
+      "epoch": 6.153349639896725,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1541564750645468e-05,
+      "loss": 0.831,
+      "step": 90565
+    },
+    {
+      "epoch": 6.1536893599673865,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.1539441500203832e-05,
+      "loss": 0.8876,
+      "step": 90570
+    },
+    {
+      "epoch": 6.154029080038049,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.1537318249762198e-05,
+      "loss": 0.7879,
+      "step": 90575
+    },
+    {
+      "epoch": 6.154368800108711,
+      "grad_norm": 7.375,
+      "learning_rate": 1.153519499932056e-05,
+      "loss": 0.7937,
+      "step": 90580
+    },
+    {
+      "epoch": 6.154708520179372,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1533071748878926e-05,
+      "loss": 0.9371,
+      "step": 90585
+    },
+    {
+      "epoch": 6.155048240250034,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1530948498437288e-05,
+      "loss": 0.8913,
+      "step": 90590
+    },
+    {
+      "epoch": 6.155387960320696,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.1528825247995652e-05,
+      "loss": 0.7511,
+      "step": 90595
+    },
+    {
+      "epoch": 6.155727680391357,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1526701997554016e-05,
+      "loss": 0.9072,
+      "step": 90600
+    },
+    {
+      "epoch": 6.156067400462019,
+      "grad_norm": 7.625,
+      "learning_rate": 1.152457874711238e-05,
+      "loss": 0.7972,
+      "step": 90605
+    },
+    {
+      "epoch": 6.156407120532681,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.1522455496670744e-05,
+      "loss": 0.8832,
+      "step": 90610
+    },
+    {
+      "epoch": 6.1567468406033425,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1520332246229108e-05,
+      "loss": 0.847,
+      "step": 90615
+    },
+    {
+      "epoch": 6.157086560674005,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1518208995787472e-05,
+      "loss": 0.8345,
+      "step": 90620
+    },
+    {
+      "epoch": 6.157426280744667,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1516085745345836e-05,
+      "loss": 0.8213,
+      "step": 90625
+    },
+    {
+      "epoch": 6.157766000815328,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.15139624949042e-05,
+      "loss": 0.8252,
+      "step": 90630
+    },
+    {
+      "epoch": 6.15810572088599,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1511839244462564e-05,
+      "loss": 0.8,
+      "step": 90635
+    },
+    {
+      "epoch": 6.158445440956652,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1509715994020926e-05,
+      "loss": 0.8537,
+      "step": 90640
+    },
+    {
+      "epoch": 6.158785161027313,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1507592743579292e-05,
+      "loss": 0.7979,
+      "step": 90645
+    },
+    {
+      "epoch": 6.159124881097975,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.1505469493137654e-05,
+      "loss": 0.8193,
+      "step": 90650
+    },
+    {
+      "epoch": 6.159464601168637,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1503346242696018e-05,
+      "loss": 0.8673,
+      "step": 90655
+    },
+    {
+      "epoch": 6.1598043212392986,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1501222992254384e-05,
+      "loss": 0.9717,
+      "step": 90660
+    },
+    {
+      "epoch": 6.160144041309961,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1499099741812746e-05,
+      "loss": 0.8931,
+      "step": 90665
+    },
+    {
+      "epoch": 6.160483761380623,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1496976491371112e-05,
+      "loss": 0.8955,
+      "step": 90670
+    },
+    {
+      "epoch": 6.160823481451284,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1494853240929474e-05,
+      "loss": 0.9031,
+      "step": 90675
+    },
+    {
+      "epoch": 6.161163201521946,
+      "grad_norm": 7.375,
+      "learning_rate": 1.1492729990487838e-05,
+      "loss": 0.9161,
+      "step": 90680
+    },
+    {
+      "epoch": 6.161502921592608,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1490606740046202e-05,
+      "loss": 0.8109,
+      "step": 90685
+    },
+    {
+      "epoch": 6.161842641663269,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1488483489604566e-05,
+      "loss": 0.8488,
+      "step": 90690
+    },
+    {
+      "epoch": 6.162182361733931,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.148636023916293e-05,
+      "loss": 0.9592,
+      "step": 90695
+    },
+    {
+      "epoch": 6.162522081804593,
+      "grad_norm": 7.5,
+      "learning_rate": 1.1484236988721294e-05,
+      "loss": 0.7996,
+      "step": 90700
+    },
+    {
+      "epoch": 6.162861801875255,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1482113738279658e-05,
+      "loss": 0.955,
+      "step": 90705
+    },
+    {
+      "epoch": 6.163201521945917,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1479990487838022e-05,
+      "loss": 0.8978,
+      "step": 90710
+    },
+    {
+      "epoch": 6.163541242016579,
+      "grad_norm": 6.75,
+      "learning_rate": 1.1477867237396386e-05,
+      "loss": 0.7876,
+      "step": 90715
+    },
+    {
+      "epoch": 6.16388096208724,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.147574398695475e-05,
+      "loss": 0.8808,
+      "step": 90720
+    },
+    {
+      "epoch": 6.164220682157902,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1473620736513113e-05,
+      "loss": 0.7705,
+      "step": 90725
+    },
+    {
+      "epoch": 6.164560402228564,
+      "grad_norm": 7.375,
+      "learning_rate": 1.1471497486071478e-05,
+      "loss": 0.8307,
+      "step": 90730
+    },
+    {
+      "epoch": 6.164900122299225,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.146937423562984e-05,
+      "loss": 0.8224,
+      "step": 90735
+    },
+    {
+      "epoch": 6.165239842369887,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.1467250985188205e-05,
+      "loss": 0.7931,
+      "step": 90740
+    },
+    {
+      "epoch": 6.165579562440549,
+      "grad_norm": 8.5,
+      "learning_rate": 1.146512773474657e-05,
+      "loss": 0.8642,
+      "step": 90745
+    },
+    {
+      "epoch": 6.165919282511211,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1463004484304933e-05,
+      "loss": 0.7819,
+      "step": 90750
+    },
+    {
+      "epoch": 6.166259002581873,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1460881233863298e-05,
+      "loss": 0.8404,
+      "step": 90755
+    },
+    {
+      "epoch": 6.166598722652535,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.145875798342166e-05,
+      "loss": 0.859,
+      "step": 90760
+    },
+    {
+      "epoch": 6.166938442723196,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1456634732980025e-05,
+      "loss": 0.8973,
+      "step": 90765
+    },
+    {
+      "epoch": 6.167278162793858,
+      "grad_norm": 8.375,
+      "learning_rate": 1.145451148253839e-05,
+      "loss": 0.8151,
+      "step": 90770
+    },
+    {
+      "epoch": 6.16761788286452,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.1452388232096753e-05,
+      "loss": 0.8252,
+      "step": 90775
+    },
+    {
+      "epoch": 6.167957602935181,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1450264981655117e-05,
+      "loss": 0.8858,
+      "step": 90780
+    },
+    {
+      "epoch": 6.168297323005843,
+      "grad_norm": 8.375,
+      "learning_rate": 1.144814173121348e-05,
+      "loss": 0.8855,
+      "step": 90785
+    },
+    {
+      "epoch": 6.1686370430765045,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1446018480771845e-05,
+      "loss": 0.8441,
+      "step": 90790
+    },
+    {
+      "epoch": 6.168976763147167,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1443895230330209e-05,
+      "loss": 0.787,
+      "step": 90795
+    },
+    {
+      "epoch": 6.169316483217829,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.1441771979888573e-05,
+      "loss": 0.8073,
+      "step": 90800
+    },
+    {
+      "epoch": 6.16965620328849,
+      "grad_norm": 9.125,
+      "learning_rate": 1.1439648729446937e-05,
+      "loss": 0.8209,
+      "step": 90805
+    },
+    {
+      "epoch": 6.169995923359152,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1437525479005299e-05,
+      "loss": 0.8738,
+      "step": 90810
+    },
+    {
+      "epoch": 6.170335643429814,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.1435402228563665e-05,
+      "loss": 0.8905,
+      "step": 90815
+    },
+    {
+      "epoch": 6.170675363500475,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1433278978122029e-05,
+      "loss": 0.8639,
+      "step": 90820
+    },
+    {
+      "epoch": 6.171015083571137,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.1431155727680391e-05,
+      "loss": 0.8631,
+      "step": 90825
+    },
+    {
+      "epoch": 6.171354803641799,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1429032477238757e-05,
+      "loss": 0.7946,
+      "step": 90830
+    },
+    {
+      "epoch": 6.1716945237124605,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1426909226797119e-05,
+      "loss": 0.8389,
+      "step": 90835
+    },
+    {
+      "epoch": 6.172034243783123,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1424785976355485e-05,
+      "loss": 0.8884,
+      "step": 90840
+    },
+    {
+      "epoch": 6.172373963853785,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1422662725913847e-05,
+      "loss": 0.8013,
+      "step": 90845
+    },
+    {
+      "epoch": 6.172713683924446,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1420539475472211e-05,
+      "loss": 0.807,
+      "step": 90850
+    },
+    {
+      "epoch": 6.173053403995108,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1418416225030577e-05,
+      "loss": 0.8388,
+      "step": 90855
+    },
+    {
+      "epoch": 6.17339312406577,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1416292974588939e-05,
+      "loss": 0.7901,
+      "step": 90860
+    },
+    {
+      "epoch": 6.173732844136431,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1414169724147303e-05,
+      "loss": 0.9353,
+      "step": 90865
+    },
+    {
+      "epoch": 6.174072564207093,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1412046473705667e-05,
+      "loss": 0.8886,
+      "step": 90870
+    },
+    {
+      "epoch": 6.174412284277755,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1409923223264031e-05,
+      "loss": 0.8598,
+      "step": 90875
+    },
+    {
+      "epoch": 6.1747520043484165,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1407799972822395e-05,
+      "loss": 0.8179,
+      "step": 90880
+    },
+    {
+      "epoch": 6.175091724419079,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1405676722380759e-05,
+      "loss": 0.8423,
+      "step": 90885
+    },
+    {
+      "epoch": 6.175431444489741,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.1403553471939123e-05,
+      "loss": 0.7741,
+      "step": 90890
+    },
+    {
+      "epoch": 6.175771164560402,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.1401430221497485e-05,
+      "loss": 0.901,
+      "step": 90895
+    },
+    {
+      "epoch": 6.176110884631064,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1399306971055851e-05,
+      "loss": 0.8017,
+      "step": 90900
+    },
+    {
+      "epoch": 6.176450604701726,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.1397183720614215e-05,
+      "loss": 0.7658,
+      "step": 90905
+    },
+    {
+      "epoch": 6.176790324772387,
+      "grad_norm": 9.125,
+      "learning_rate": 1.1395060470172577e-05,
+      "loss": 0.7728,
+      "step": 90910
+    },
+    {
+      "epoch": 6.177130044843049,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1392937219730943e-05,
+      "loss": 0.8514,
+      "step": 90915
+    },
+    {
+      "epoch": 6.177469764913711,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.1390813969289305e-05,
+      "loss": 0.9165,
+      "step": 90920
+    },
+    {
+      "epoch": 6.1778094849843725,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1388690718847671e-05,
+      "loss": 0.7685,
+      "step": 90925
+    },
+    {
+      "epoch": 6.178149205055035,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1386567468406033e-05,
+      "loss": 0.9985,
+      "step": 90930
+    },
+    {
+      "epoch": 6.178488925125697,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1384444217964397e-05,
+      "loss": 0.8403,
+      "step": 90935
+    },
+    {
+      "epoch": 6.178828645196358,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1382320967522763e-05,
+      "loss": 0.9046,
+      "step": 90940
+    },
+    {
+      "epoch": 6.17916836526702,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1380197717081125e-05,
+      "loss": 0.898,
+      "step": 90945
+    },
+    {
+      "epoch": 6.179508085337682,
+      "grad_norm": 7.25,
+      "learning_rate": 1.137807446663949e-05,
+      "loss": 0.8449,
+      "step": 90950
+    },
+    {
+      "epoch": 6.179847805408343,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1375951216197853e-05,
+      "loss": 0.8931,
+      "step": 90955
+    },
+    {
+      "epoch": 6.180187525479005,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.1373827965756217e-05,
+      "loss": 0.8301,
+      "step": 90960
+    },
+    {
+      "epoch": 6.180527245549667,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1371704715314581e-05,
+      "loss": 0.8123,
+      "step": 90965
+    },
+    {
+      "epoch": 6.180866965620329,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1369581464872945e-05,
+      "loss": 0.9522,
+      "step": 90970
+    },
+    {
+      "epoch": 6.181206685690991,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.136745821443131e-05,
+      "loss": 0.904,
+      "step": 90975
+    },
+    {
+      "epoch": 6.181546405761653,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1365334963989672e-05,
+      "loss": 0.8337,
+      "step": 90980
+    },
+    {
+      "epoch": 6.181886125832314,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1363211713548037e-05,
+      "loss": 0.7822,
+      "step": 90985
+    },
+    {
+      "epoch": 6.182225845902976,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1361088463106401e-05,
+      "loss": 0.8318,
+      "step": 90990
+    },
+    {
+      "epoch": 6.182565565973638,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1358965212664764e-05,
+      "loss": 0.9082,
+      "step": 90995
+    },
+    {
+      "epoch": 6.182905286044299,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.135684196222313e-05,
+      "loss": 0.8506,
+      "step": 91000
+    },
+    {
+      "epoch": 6.183245006114961,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1354718711781492e-05,
+      "loss": 0.946,
+      "step": 91005
+    },
+    {
+      "epoch": 6.183584726185623,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1352595461339857e-05,
+      "loss": 0.8025,
+      "step": 91010
+    },
+    {
+      "epoch": 6.183924446256285,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.135047221089822e-05,
+      "loss": 0.9219,
+      "step": 91015
+    },
+    {
+      "epoch": 6.184264166326947,
+      "grad_norm": 7.625,
+      "learning_rate": 1.1348348960456584e-05,
+      "loss": 0.8775,
+      "step": 91020
+    },
+    {
+      "epoch": 6.184603886397609,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.134622571001495e-05,
+      "loss": 0.8818,
+      "step": 91025
+    },
+    {
+      "epoch": 6.18494360646827,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.1344102459573312e-05,
+      "loss": 0.7069,
+      "step": 91030
+    },
+    {
+      "epoch": 6.185283326538932,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1341979209131676e-05,
+      "loss": 0.8395,
+      "step": 91035
+    },
+    {
+      "epoch": 6.185623046609594,
+      "grad_norm": 9.5,
+      "learning_rate": 1.133985595869004e-05,
+      "loss": 0.8741,
+      "step": 91040
+    },
+    {
+      "epoch": 6.185962766680255,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1337732708248404e-05,
+      "loss": 0.87,
+      "step": 91045
+    },
+    {
+      "epoch": 6.186302486750917,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1335609457806768e-05,
+      "loss": 0.8449,
+      "step": 91050
+    },
+    {
+      "epoch": 6.186642206821579,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1333486207365132e-05,
+      "loss": 0.8288,
+      "step": 91055
+    },
+    {
+      "epoch": 6.186981926892241,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1331362956923496e-05,
+      "loss": 0.7961,
+      "step": 91060
+    },
+    {
+      "epoch": 6.187321646962903,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1329239706481858e-05,
+      "loss": 0.8456,
+      "step": 91065
+    },
+    {
+      "epoch": 6.187661367033565,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1327116456040224e-05,
+      "loss": 0.772,
+      "step": 91070
+    },
+    {
+      "epoch": 6.188001087104226,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1324993205598588e-05,
+      "loss": 0.8083,
+      "step": 91075
+    },
+    {
+      "epoch": 6.188340807174888,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.132286995515695e-05,
+      "loss": 0.8724,
+      "step": 91080
+    },
+    {
+      "epoch": 6.18868052724555,
+      "grad_norm": 8.875,
+      "learning_rate": 1.1320746704715316e-05,
+      "loss": 0.8172,
+      "step": 91085
+    },
+    {
+      "epoch": 6.189020247316211,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1318623454273678e-05,
+      "loss": 0.8993,
+      "step": 91090
+    },
+    {
+      "epoch": 6.189359967386873,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1316500203832044e-05,
+      "loss": 0.9007,
+      "step": 91095
+    },
+    {
+      "epoch": 6.189699687457535,
+      "grad_norm": 7.5,
+      "learning_rate": 1.1314376953390408e-05,
+      "loss": 0.8636,
+      "step": 91100
+    },
+    {
+      "epoch": 6.190039407528197,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.131225370294877e-05,
+      "loss": 0.8052,
+      "step": 91105
+    },
+    {
+      "epoch": 6.190379127598859,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1310130452507136e-05,
+      "loss": 0.8283,
+      "step": 91110
+    },
+    {
+      "epoch": 6.190718847669521,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1308007202065498e-05,
+      "loss": 0.9095,
+      "step": 91115
+    },
+    {
+      "epoch": 6.191058567740182,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1305883951623862e-05,
+      "loss": 0.8543,
+      "step": 91120
+    },
+    {
+      "epoch": 6.191398287810844,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1303760701182226e-05,
+      "loss": 0.838,
+      "step": 91125
+    },
+    {
+      "epoch": 6.191738007881506,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.130163745074059e-05,
+      "loss": 0.8544,
+      "step": 91130
+    },
+    {
+      "epoch": 6.192077727952167,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.1299514200298954e-05,
+      "loss": 0.8598,
+      "step": 91135
+    },
+    {
+      "epoch": 6.192417448022829,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1297390949857318e-05,
+      "loss": 0.9122,
+      "step": 91140
+    },
+    {
+      "epoch": 6.192757168093491,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1295267699415682e-05,
+      "loss": 0.8105,
+      "step": 91145
+    },
+    {
+      "epoch": 6.193096888164153,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1293144448974046e-05,
+      "loss": 0.8743,
+      "step": 91150
+    },
+    {
+      "epoch": 6.193436608234815,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.129102119853241e-05,
+      "loss": 0.8726,
+      "step": 91155
+    },
+    {
+      "epoch": 6.193776328305476,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1288897948090774e-05,
+      "loss": 0.8254,
+      "step": 91160
+    },
+    {
+      "epoch": 6.194116048376138,
+      "grad_norm": 7.375,
+      "learning_rate": 1.1286774697649137e-05,
+      "loss": 0.8704,
+      "step": 91165
+    },
+    {
+      "epoch": 6.1944557684468,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1284651447207502e-05,
+      "loss": 0.7804,
+      "step": 91170
+    },
+    {
+      "epoch": 6.194795488517461,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1282528196765865e-05,
+      "loss": 0.7802,
+      "step": 91175
+    },
+    {
+      "epoch": 6.195135208588123,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.128040494632423e-05,
+      "loss": 0.8907,
+      "step": 91180
+    },
+    {
+      "epoch": 6.195474928658785,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1278281695882594e-05,
+      "loss": 0.793,
+      "step": 91185
+    },
+    {
+      "epoch": 6.1958146487294465,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1276158445440957e-05,
+      "loss": 0.8377,
+      "step": 91190
+    },
+    {
+      "epoch": 6.196154368800109,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.1274035194999322e-05,
+      "loss": 0.9987,
+      "step": 91195
+    },
+    {
+      "epoch": 6.196494088870771,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.1271911944557685e-05,
+      "loss": 0.8686,
+      "step": 91200
+    },
+    {
+      "epoch": 6.196833808941432,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1269788694116049e-05,
+      "loss": 0.8254,
+      "step": 91205
+    },
+    {
+      "epoch": 6.197173529012094,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1267665443674413e-05,
+      "loss": 0.7892,
+      "step": 91210
+    },
+    {
+      "epoch": 6.197513249082756,
+      "grad_norm": 9.75,
+      "learning_rate": 1.1265542193232777e-05,
+      "loss": 0.855,
+      "step": 91215
+    },
+    {
+      "epoch": 6.197852969153417,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.126341894279114e-05,
+      "loss": 0.8057,
+      "step": 91220
+    },
+    {
+      "epoch": 6.198192689224079,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1261295692349505e-05,
+      "loss": 0.8491,
+      "step": 91225
+    },
+    {
+      "epoch": 6.198532409294741,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1259172441907869e-05,
+      "loss": 0.8666,
+      "step": 91230
+    },
+    {
+      "epoch": 6.1988721293654026,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1257049191466233e-05,
+      "loss": 0.834,
+      "step": 91235
+    },
+    {
+      "epoch": 6.199211849436065,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1254925941024597e-05,
+      "loss": 0.8624,
+      "step": 91240
+    },
+    {
+      "epoch": 6.199551569506727,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.125280269058296e-05,
+      "loss": 0.886,
+      "step": 91245
+    },
+    {
+      "epoch": 6.199891289577388,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1250679440141323e-05,
+      "loss": 0.7926,
+      "step": 91250
+    },
+    {
+      "epoch": 6.20023100964805,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1248556189699689e-05,
+      "loss": 0.7932,
+      "step": 91255
+    },
+    {
+      "epoch": 6.200570729718712,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1246432939258051e-05,
+      "loss": 0.8247,
+      "step": 91260
+    },
+    {
+      "epoch": 6.200910449789373,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1244309688816417e-05,
+      "loss": 0.8375,
+      "step": 91265
+    },
+    {
+      "epoch": 6.201250169860035,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.124218643837478e-05,
+      "loss": 0.8931,
+      "step": 91270
+    },
+    {
+      "epoch": 6.201589889930697,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1240063187933143e-05,
+      "loss": 0.8623,
+      "step": 91275
+    },
+    {
+      "epoch": 6.201929610001359,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1237939937491509e-05,
+      "loss": 0.9384,
+      "step": 91280
+    },
+    {
+      "epoch": 6.202269330072021,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.1235816687049871e-05,
+      "loss": 0.8775,
+      "step": 91285
+    },
+    {
+      "epoch": 6.202609050142683,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.1233693436608235e-05,
+      "loss": 0.8491,
+      "step": 91290
+    },
+    {
+      "epoch": 6.202948770213344,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.12315701861666e-05,
+      "loss": 0.8184,
+      "step": 91295
+    },
+    {
+      "epoch": 6.203288490284006,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1229446935724963e-05,
+      "loss": 0.7575,
+      "step": 91300
+    },
+    {
+      "epoch": 6.203628210354668,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1227323685283327e-05,
+      "loss": 0.9522,
+      "step": 91305
+    },
+    {
+      "epoch": 6.203967930425329,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1225200434841691e-05,
+      "loss": 0.8495,
+      "step": 91310
+    },
+    {
+      "epoch": 6.204307650495991,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1223077184400055e-05,
+      "loss": 0.7979,
+      "step": 91315
+    },
+    {
+      "epoch": 6.204647370566653,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.1220953933958419e-05,
+      "loss": 0.8755,
+      "step": 91320
+    },
+    {
+      "epoch": 6.204987090637315,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1218830683516783e-05,
+      "loss": 0.7329,
+      "step": 91325
+    },
+    {
+      "epoch": 6.205326810707977,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.1216707433075147e-05,
+      "loss": 0.8519,
+      "step": 91330
+    },
+    {
+      "epoch": 6.205666530778639,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.121458418263351e-05,
+      "loss": 0.8565,
+      "step": 91335
+    },
+    {
+      "epoch": 6.2060062508493,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.1212460932191875e-05,
+      "loss": 0.8342,
+      "step": 91340
+    },
+    {
+      "epoch": 6.206345970919962,
+      "grad_norm": 7.125,
+      "learning_rate": 1.1210337681750239e-05,
+      "loss": 0.8432,
+      "step": 91345
+    },
+    {
+      "epoch": 6.206685690990624,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.1208214431308603e-05,
+      "loss": 0.8102,
+      "step": 91350
+    },
+    {
+      "epoch": 6.207025411061285,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1206091180866967e-05,
+      "loss": 0.796,
+      "step": 91355
+    },
+    {
+      "epoch": 6.207365131131947,
+      "grad_norm": 7.625,
+      "learning_rate": 1.120396793042533e-05,
+      "loss": 0.7338,
+      "step": 91360
+    },
+    {
+      "epoch": 6.207704851202609,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1201844679983695e-05,
+      "loss": 0.8212,
+      "step": 91365
+    },
+    {
+      "epoch": 6.208044571273271,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1199721429542057e-05,
+      "loss": 1.0071,
+      "step": 91370
+    },
+    {
+      "epoch": 6.208384291343933,
+      "grad_norm": 7.5,
+      "learning_rate": 1.1197598179100421e-05,
+      "loss": 0.8772,
+      "step": 91375
+    },
+    {
+      "epoch": 6.208724011414595,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1195474928658787e-05,
+      "loss": 0.7953,
+      "step": 91380
+    },
+    {
+      "epoch": 6.209063731485256,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.119335167821715e-05,
+      "loss": 0.9646,
+      "step": 91385
+    },
+    {
+      "epoch": 6.209403451555918,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1191228427775513e-05,
+      "loss": 0.8905,
+      "step": 91390
+    },
+    {
+      "epoch": 6.20974317162658,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1189105177333877e-05,
+      "loss": 0.8223,
+      "step": 91395
+    },
+    {
+      "epoch": 6.210082891697241,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1186981926892241e-05,
+      "loss": 0.8373,
+      "step": 91400
+    },
+    {
+      "epoch": 6.210422611767903,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.1184858676450605e-05,
+      "loss": 0.8282,
+      "step": 91405
+    },
+    {
+      "epoch": 6.210762331838565,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.118273542600897e-05,
+      "loss": 0.7857,
+      "step": 91410
+    },
+    {
+      "epoch": 6.211102051909227,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1180612175567333e-05,
+      "loss": 0.7883,
+      "step": 91415
+    },
+    {
+      "epoch": 6.211441771979889,
+      "grad_norm": 7.625,
+      "learning_rate": 1.1178488925125696e-05,
+      "loss": 0.8412,
+      "step": 91420
+    },
+    {
+      "epoch": 6.211781492050551,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1176365674684061e-05,
+      "loss": 0.7604,
+      "step": 91425
+    },
+    {
+      "epoch": 6.212121212121212,
+      "grad_norm": 9.625,
+      "learning_rate": 1.1174242424242425e-05,
+      "loss": 0.8292,
+      "step": 91430
+    },
+    {
+      "epoch": 6.212460932191874,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.117211917380079e-05,
+      "loss": 0.9136,
+      "step": 91435
+    },
+    {
+      "epoch": 6.212800652262536,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.1169995923359153e-05,
+      "loss": 0.8956,
+      "step": 91440
+    },
+    {
+      "epoch": 6.213140372333197,
+      "grad_norm": 6.75,
+      "learning_rate": 1.1167872672917516e-05,
+      "loss": 0.8303,
+      "step": 91445
+    },
+    {
+      "epoch": 6.213480092403859,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.1165749422475881e-05,
+      "loss": 0.8993,
+      "step": 91450
+    },
+    {
+      "epoch": 6.213819812474521,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1163626172034244e-05,
+      "loss": 0.7988,
+      "step": 91455
+    },
+    {
+      "epoch": 6.214159532545183,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1161502921592608e-05,
+      "loss": 1.0177,
+      "step": 91460
+    },
+    {
+      "epoch": 6.214499252615845,
+      "grad_norm": 9.125,
+      "learning_rate": 1.1159379671150973e-05,
+      "loss": 0.8162,
+      "step": 91465
+    },
+    {
+      "epoch": 6.214838972686506,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1157256420709336e-05,
+      "loss": 0.728,
+      "step": 91470
+    },
+    {
+      "epoch": 6.215178692757168,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.11551331702677e-05,
+      "loss": 0.7908,
+      "step": 91475
+    },
+    {
+      "epoch": 6.21551841282783,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1153009919826064e-05,
+      "loss": 0.8488,
+      "step": 91480
+    },
+    {
+      "epoch": 6.215858132898491,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1150886669384428e-05,
+      "loss": 0.8972,
+      "step": 91485
+    },
+    {
+      "epoch": 6.216197852969153,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1148763418942792e-05,
+      "loss": 0.8389,
+      "step": 91490
+    },
+    {
+      "epoch": 6.216537573039815,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.1146640168501156e-05,
+      "loss": 0.8514,
+      "step": 91495
+    },
+    {
+      "epoch": 6.2168772931104765,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.114451691805952e-05,
+      "loss": 0.9426,
+      "step": 91500
+    },
+    {
+      "epoch": 6.217217013181139,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1142393667617882e-05,
+      "loss": 0.9039,
+      "step": 91505
+    },
+    {
+      "epoch": 6.217556733251801,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1140270417176248e-05,
+      "loss": 0.7135,
+      "step": 91510
+    },
+    {
+      "epoch": 6.217896453322462,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1138147166734612e-05,
+      "loss": 0.7963,
+      "step": 91515
+    },
+    {
+      "epoch": 6.218236173393124,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1136023916292976e-05,
+      "loss": 0.7728,
+      "step": 91520
+    },
+    {
+      "epoch": 6.218575893463786,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.113390066585134e-05,
+      "loss": 0.8464,
+      "step": 91525
+    },
+    {
+      "epoch": 6.218915613534447,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1131777415409702e-05,
+      "loss": 0.8403,
+      "step": 91530
+    },
+    {
+      "epoch": 6.219255333605109,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1129654164968068e-05,
+      "loss": 0.8173,
+      "step": 91535
+    },
+    {
+      "epoch": 6.219595053675771,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1127530914526432e-05,
+      "loss": 0.842,
+      "step": 91540
+    },
+    {
+      "epoch": 6.2199347737464326,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1125407664084794e-05,
+      "loss": 0.7902,
+      "step": 91545
+    },
+    {
+      "epoch": 6.220274493817095,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.112328441364316e-05,
+      "loss": 0.8172,
+      "step": 91550
+    },
+    {
+      "epoch": 6.220614213887757,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1121161163201522e-05,
+      "loss": 0.7977,
+      "step": 91555
+    },
+    {
+      "epoch": 6.220953933958418,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1119037912759886e-05,
+      "loss": 0.8863,
+      "step": 91560
+    },
+    {
+      "epoch": 6.22129365402908,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.111691466231825e-05,
+      "loss": 0.8429,
+      "step": 91565
+    },
+    {
+      "epoch": 6.221633374099742,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.1114791411876614e-05,
+      "loss": 0.7725,
+      "step": 91570
+    },
+    {
+      "epoch": 6.221973094170403,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.1112668161434978e-05,
+      "loss": 0.7379,
+      "step": 91575
+    },
+    {
+      "epoch": 6.222312814241065,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1110544910993342e-05,
+      "loss": 0.8147,
+      "step": 91580
+    },
+    {
+      "epoch": 6.222652534311727,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1108421660551706e-05,
+      "loss": 0.7897,
+      "step": 91585
+    },
+    {
+      "epoch": 6.222992254382389,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.110629841011007e-05,
+      "loss": 0.8291,
+      "step": 91590
+    },
+    {
+      "epoch": 6.223331974453051,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1104175159668434e-05,
+      "loss": 0.8832,
+      "step": 91595
+    },
+    {
+      "epoch": 6.223671694523713,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1102051909226798e-05,
+      "loss": 0.8562,
+      "step": 91600
+    },
+    {
+      "epoch": 6.224011414594374,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1099928658785162e-05,
+      "loss": 0.8522,
+      "step": 91605
+    },
+    {
+      "epoch": 6.224351134665036,
+      "grad_norm": 7.75,
+      "learning_rate": 1.1097805408343526e-05,
+      "loss": 0.7822,
+      "step": 91610
+    },
+    {
+      "epoch": 6.224690854735698,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1095682157901888e-05,
+      "loss": 0.8888,
+      "step": 91615
+    },
+    {
+      "epoch": 6.225030574806359,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1093558907460254e-05,
+      "loss": 0.7106,
+      "step": 91620
+    },
+    {
+      "epoch": 6.225370294877021,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1091435657018618e-05,
+      "loss": 0.9234,
+      "step": 91625
+    },
+    {
+      "epoch": 6.225710014947683,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.108931240657698e-05,
+      "loss": 0.8699,
+      "step": 91630
+    },
+    {
+      "epoch": 6.226049735018345,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.1087189156135346e-05,
+      "loss": 0.8533,
+      "step": 91635
+    },
+    {
+      "epoch": 6.226389455089007,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1085065905693708e-05,
+      "loss": 0.8123,
+      "step": 91640
+    },
+    {
+      "epoch": 6.226729175159669,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1082942655252072e-05,
+      "loss": 0.8111,
+      "step": 91645
+    },
+    {
+      "epoch": 6.22706889523033,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1080819404810436e-05,
+      "loss": 0.8232,
+      "step": 91650
+    },
+    {
+      "epoch": 6.227408615300992,
+      "grad_norm": 8.25,
+      "learning_rate": 1.10786961543688e-05,
+      "loss": 0.8724,
+      "step": 91655
+    },
+    {
+      "epoch": 6.227748335371654,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1076572903927164e-05,
+      "loss": 0.8712,
+      "step": 91660
+    },
+    {
+      "epoch": 6.228088055442315,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1074449653485528e-05,
+      "loss": 0.8719,
+      "step": 91665
+    },
+    {
+      "epoch": 6.228427775512977,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1072326403043892e-05,
+      "loss": 0.7246,
+      "step": 91670
+    },
+    {
+      "epoch": 6.228767495583639,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1070203152602256e-05,
+      "loss": 0.8684,
+      "step": 91675
+    },
+    {
+      "epoch": 6.229107215654301,
+      "grad_norm": 8.375,
+      "learning_rate": 1.106807990216062e-05,
+      "loss": 0.9501,
+      "step": 91680
+    },
+    {
+      "epoch": 6.229446935724963,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1065956651718984e-05,
+      "loss": 0.9006,
+      "step": 91685
+    },
+    {
+      "epoch": 6.229786655795625,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1063833401277348e-05,
+      "loss": 0.8709,
+      "step": 91690
+    },
+    {
+      "epoch": 6.230126375866286,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1061710150835712e-05,
+      "loss": 0.8733,
+      "step": 91695
+    },
+    {
+      "epoch": 6.230466095936948,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1059586900394075e-05,
+      "loss": 0.8383,
+      "step": 91700
+    },
+    {
+      "epoch": 6.23080581600761,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.105746364995244e-05,
+      "loss": 0.8926,
+      "step": 91705
+    },
+    {
+      "epoch": 6.231145536078271,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1055340399510804e-05,
+      "loss": 0.8803,
+      "step": 91710
+    },
+    {
+      "epoch": 6.231485256148933,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1053217149069167e-05,
+      "loss": 0.8039,
+      "step": 91715
+    },
+    {
+      "epoch": 6.231824976219595,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.1051093898627532e-05,
+      "loss": 0.7526,
+      "step": 91720
+    },
+    {
+      "epoch": 6.232164696290257,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1048970648185895e-05,
+      "loss": 0.8518,
+      "step": 91725
+    },
+    {
+      "epoch": 6.232504416360919,
+      "grad_norm": 7.125,
+      "learning_rate": 1.1046847397744259e-05,
+      "loss": 0.8314,
+      "step": 91730
+    },
+    {
+      "epoch": 6.232844136431581,
+      "grad_norm": 9.0,
+      "learning_rate": 1.1044724147302623e-05,
+      "loss": 0.9316,
+      "step": 91735
+    },
+    {
+      "epoch": 6.233183856502242,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.1042600896860987e-05,
+      "loss": 0.8437,
+      "step": 91740
+    },
+    {
+      "epoch": 6.233523576572904,
+      "grad_norm": 8.375,
+      "learning_rate": 1.104047764641935e-05,
+      "loss": 0.7938,
+      "step": 91745
+    },
+    {
+      "epoch": 6.233863296643566,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.1038354395977715e-05,
+      "loss": 0.8582,
+      "step": 91750
+    },
+    {
+      "epoch": 6.234203016714227,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.1036231145536079e-05,
+      "loss": 0.9113,
+      "step": 91755
+    },
+    {
+      "epoch": 6.234542736784889,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1034107895094443e-05,
+      "loss": 0.8777,
+      "step": 91760
+    },
+    {
+      "epoch": 6.234882456855551,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1031984644652807e-05,
+      "loss": 0.8847,
+      "step": 91765
+    },
+    {
+      "epoch": 6.235222176926213,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.102986139421117e-05,
+      "loss": 0.9077,
+      "step": 91770
+    },
+    {
+      "epoch": 6.235561896996875,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.1027738143769535e-05,
+      "loss": 0.8118,
+      "step": 91775
+    },
+    {
+      "epoch": 6.235901617067537,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.1025614893327899e-05,
+      "loss": 0.7362,
+      "step": 91780
+    },
+    {
+      "epoch": 6.236241337138198,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1023491642886261e-05,
+      "loss": 0.7385,
+      "step": 91785
+    },
+    {
+      "epoch": 6.23658105720886,
+      "grad_norm": 9.5,
+      "learning_rate": 1.1021368392444627e-05,
+      "loss": 0.854,
+      "step": 91790
+    },
+    {
+      "epoch": 6.236920777279522,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.101924514200299e-05,
+      "loss": 0.8306,
+      "step": 91795
+    },
+    {
+      "epoch": 6.237260497350183,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1017121891561353e-05,
+      "loss": 0.834,
+      "step": 91800
+    },
+    {
+      "epoch": 6.237600217420845,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1014998641119719e-05,
+      "loss": 0.9029,
+      "step": 91805
+    },
+    {
+      "epoch": 6.237939937491507,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1012875390678081e-05,
+      "loss": 0.7762,
+      "step": 91810
+    },
+    {
+      "epoch": 6.238279657562169,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1010752140236445e-05,
+      "loss": 0.9171,
+      "step": 91815
+    },
+    {
+      "epoch": 6.238619377632831,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.100862888979481e-05,
+      "loss": 0.8353,
+      "step": 91820
+    },
+    {
+      "epoch": 6.238959097703493,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1006505639353173e-05,
+      "loss": 0.823,
+      "step": 91825
+    },
+    {
+      "epoch": 6.239298817774154,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1004382388911537e-05,
+      "loss": 0.7015,
+      "step": 91830
+    },
+    {
+      "epoch": 6.239638537844816,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1002259138469901e-05,
+      "loss": 0.8213,
+      "step": 91835
+    },
+    {
+      "epoch": 6.239978257915477,
+      "grad_norm": 7.875,
+      "learning_rate": 1.1000135888028265e-05,
+      "loss": 0.823,
+      "step": 91840
+    },
+    {
+      "epoch": 6.240317977986139,
+      "grad_norm": 8.25,
+      "learning_rate": 1.099801263758663e-05,
+      "loss": 0.875,
+      "step": 91845
+    },
+    {
+      "epoch": 6.240657698056801,
+      "grad_norm": 8.875,
+      "learning_rate": 1.0995889387144993e-05,
+      "loss": 0.8706,
+      "step": 91850
+    },
+    {
+      "epoch": 6.240997418127463,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0993766136703357e-05,
+      "loss": 0.9325,
+      "step": 91855
+    },
+    {
+      "epoch": 6.241337138198125,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0991642886261721e-05,
+      "loss": 0.8029,
+      "step": 91860
+    },
+    {
+      "epoch": 6.241676858268787,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0989519635820085e-05,
+      "loss": 0.8667,
+      "step": 91865
+    },
+    {
+      "epoch": 6.242016578339448,
+      "grad_norm": 7.5,
+      "learning_rate": 1.098739638537845e-05,
+      "loss": 0.7899,
+      "step": 91870
+    },
+    {
+      "epoch": 6.24235629841011,
+      "grad_norm": 9.0,
+      "learning_rate": 1.0985273134936813e-05,
+      "loss": 0.8369,
+      "step": 91875
+    },
+    {
+      "epoch": 6.242696018480772,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.0983149884495177e-05,
+      "loss": 0.7892,
+      "step": 91880
+    },
+    {
+      "epoch": 6.243035738551433,
+      "grad_norm": 8.125,
+      "learning_rate": 1.098102663405354e-05,
+      "loss": 0.8628,
+      "step": 91885
+    },
+    {
+      "epoch": 6.243375458622095,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0978903383611905e-05,
+      "loss": 0.8784,
+      "step": 91890
+    },
+    {
+      "epoch": 6.243715178692757,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0976780133170268e-05,
+      "loss": 0.8482,
+      "step": 91895
+    },
+    {
+      "epoch": 6.244054898763419,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0974656882728632e-05,
+      "loss": 0.8461,
+      "step": 91900
+    },
+    {
+      "epoch": 6.244394618834081,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0972533632286997e-05,
+      "loss": 0.9048,
+      "step": 91905
+    },
+    {
+      "epoch": 6.244734338904743,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.097041038184536e-05,
+      "loss": 0.8585,
+      "step": 91910
+    },
+    {
+      "epoch": 6.245074058975404,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.0968287131403724e-05,
+      "loss": 0.8234,
+      "step": 91915
+    },
+    {
+      "epoch": 6.245413779046066,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0966163880962088e-05,
+      "loss": 0.7887,
+      "step": 91920
+    },
+    {
+      "epoch": 6.245753499116728,
+      "grad_norm": 7.375,
+      "learning_rate": 1.0964040630520452e-05,
+      "loss": 0.8265,
+      "step": 91925
+    },
+    {
+      "epoch": 6.246093219187389,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0961917380078816e-05,
+      "loss": 0.7569,
+      "step": 91930
+    },
+    {
+      "epoch": 6.246432939258051,
+      "grad_norm": 8.25,
+      "learning_rate": 1.095979412963718e-05,
+      "loss": 0.8362,
+      "step": 91935
+    },
+    {
+      "epoch": 6.246772659328713,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0957670879195544e-05,
+      "loss": 0.8245,
+      "step": 91940
+    },
+    {
+      "epoch": 6.247112379399375,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0955547628753908e-05,
+      "loss": 0.8446,
+      "step": 91945
+    },
+    {
+      "epoch": 6.247452099470037,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0953424378312272e-05,
+      "loss": 0.7788,
+      "step": 91950
+    },
+    {
+      "epoch": 6.247791819540699,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0951301127870636e-05,
+      "loss": 0.8401,
+      "step": 91955
+    },
+    {
+      "epoch": 6.24813153961136,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.0949177877429e-05,
+      "loss": 0.826,
+      "step": 91960
+    },
+    {
+      "epoch": 6.248471259682022,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0947054626987364e-05,
+      "loss": 0.7668,
+      "step": 91965
+    },
+    {
+      "epoch": 6.248810979752684,
+      "grad_norm": 7.5,
+      "learning_rate": 1.0944931376545726e-05,
+      "loss": 0.8109,
+      "step": 91970
+    },
+    {
+      "epoch": 6.249150699823345,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0942808126104092e-05,
+      "loss": 0.7568,
+      "step": 91975
+    },
+    {
+      "epoch": 6.249490419894007,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0940684875662454e-05,
+      "loss": 0.8462,
+      "step": 91980
+    },
+    {
+      "epoch": 6.249830139964669,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0938561625220818e-05,
+      "loss": 0.8799,
+      "step": 91985
+    },
+    {
+      "epoch": 6.250169860035331,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0936438374779184e-05,
+      "loss": 0.8621,
+      "step": 91990
+    },
+    {
+      "epoch": 6.250509580105993,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0934315124337546e-05,
+      "loss": 0.8651,
+      "step": 91995
+    },
+    {
+      "epoch": 6.250849300176655,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.093219187389591e-05,
+      "loss": 0.9633,
+      "step": 92000
+    },
+    {
+      "epoch": 6.251189020247316,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0930068623454274e-05,
+      "loss": 0.9729,
+      "step": 92005
+    },
+    {
+      "epoch": 6.251528740317978,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0927945373012638e-05,
+      "loss": 0.8483,
+      "step": 92010
+    },
+    {
+      "epoch": 6.25186846038864,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.0925822122571002e-05,
+      "loss": 0.9103,
+      "step": 92015
+    },
+    {
+      "epoch": 6.252208180459301,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0923698872129366e-05,
+      "loss": 0.7952,
+      "step": 92020
+    },
+    {
+      "epoch": 6.252547900529963,
+      "grad_norm": 10.1875,
+      "learning_rate": 1.092157562168773e-05,
+      "loss": 0.9,
+      "step": 92025
+    },
+    {
+      "epoch": 6.252887620600625,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0919452371246092e-05,
+      "loss": 0.7749,
+      "step": 92030
+    },
+    {
+      "epoch": 6.253227340671287,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0917329120804458e-05,
+      "loss": 0.8128,
+      "step": 92035
+    },
+    {
+      "epoch": 6.253567060741949,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0915205870362822e-05,
+      "loss": 0.8763,
+      "step": 92040
+    },
+    {
+      "epoch": 6.253906780812611,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0913082619921186e-05,
+      "loss": 0.8327,
+      "step": 92045
+    },
+    {
+      "epoch": 6.254246500883272,
+      "grad_norm": 8.125,
+      "learning_rate": 1.091095936947955e-05,
+      "loss": 0.9112,
+      "step": 92050
+    },
+    {
+      "epoch": 6.254586220953934,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0908836119037912e-05,
+      "loss": 0.8691,
+      "step": 92055
+    },
+    {
+      "epoch": 6.254925941024596,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0906712868596278e-05,
+      "loss": 0.8135,
+      "step": 92060
+    },
+    {
+      "epoch": 6.255265661095257,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0904589618154642e-05,
+      "loss": 0.781,
+      "step": 92065
+    },
+    {
+      "epoch": 6.255605381165919,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0902466367713004e-05,
+      "loss": 0.7646,
+      "step": 92070
+    },
+    {
+      "epoch": 6.255945101236581,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.090034311727137e-05,
+      "loss": 0.8525,
+      "step": 92075
+    },
+    {
+      "epoch": 6.256284821307243,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0898219866829732e-05,
+      "loss": 0.7766,
+      "step": 92080
+    },
+    {
+      "epoch": 6.256624541377905,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0896096616388096e-05,
+      "loss": 0.7907,
+      "step": 92085
+    },
+    {
+      "epoch": 6.256964261448567,
+      "grad_norm": 8.0,
+      "learning_rate": 1.089397336594646e-05,
+      "loss": 0.8504,
+      "step": 92090
+    },
+    {
+      "epoch": 6.257303981519228,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0891850115504824e-05,
+      "loss": 0.8039,
+      "step": 92095
+    },
+    {
+      "epoch": 6.25764370158989,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0889726865063188e-05,
+      "loss": 0.9074,
+      "step": 92100
+    },
+    {
+      "epoch": 6.257983421660552,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.0887603614621552e-05,
+      "loss": 0.9104,
+      "step": 92105
+    },
+    {
+      "epoch": 6.258323141731213,
+      "grad_norm": 6.875,
+      "learning_rate": 1.0885480364179916e-05,
+      "loss": 0.8032,
+      "step": 92110
+    },
+    {
+      "epoch": 6.258662861801875,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.088335711373828e-05,
+      "loss": 0.9212,
+      "step": 92115
+    },
+    {
+      "epoch": 6.259002581872537,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0881233863296644e-05,
+      "loss": 0.9739,
+      "step": 92120
+    },
+    {
+      "epoch": 6.259342301943199,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0879110612855008e-05,
+      "loss": 0.808,
+      "step": 92125
+    },
+    {
+      "epoch": 6.259682022013861,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0876987362413372e-05,
+      "loss": 0.8229,
+      "step": 92130
+    },
+    {
+      "epoch": 6.260021742084522,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0874864111971736e-05,
+      "loss": 0.8247,
+      "step": 92135
+    },
+    {
+      "epoch": 6.260361462155184,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0872740861530099e-05,
+      "loss": 0.8364,
+      "step": 92140
+    },
+    {
+      "epoch": 6.260701182225846,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.0870617611088464e-05,
+      "loss": 0.8213,
+      "step": 92145
+    },
+    {
+      "epoch": 6.261040902296507,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0868494360646828e-05,
+      "loss": 0.8817,
+      "step": 92150
+    },
+    {
+      "epoch": 6.261380622367169,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.086637111020519e-05,
+      "loss": 0.8599,
+      "step": 92155
+    },
+    {
+      "epoch": 6.261720342437831,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.0864247859763556e-05,
+      "loss": 0.8934,
+      "step": 92160
+    },
+    {
+      "epoch": 6.262060062508493,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0862124609321919e-05,
+      "loss": 0.8894,
+      "step": 92165
+    },
+    {
+      "epoch": 6.262399782579155,
+      "grad_norm": 8.875,
+      "learning_rate": 1.0860001358880283e-05,
+      "loss": 0.7783,
+      "step": 92170
+    },
+    {
+      "epoch": 6.262739502649817,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0857878108438647e-05,
+      "loss": 0.8829,
+      "step": 92175
+    },
+    {
+      "epoch": 6.263079222720478,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.085575485799701e-05,
+      "loss": 0.8258,
+      "step": 92180
+    },
+    {
+      "epoch": 6.26341894279114,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.0853631607555375e-05,
+      "loss": 0.8369,
+      "step": 92185
+    },
+    {
+      "epoch": 6.263758662861802,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0851508357113739e-05,
+      "loss": 0.8291,
+      "step": 92190
+    },
+    {
+      "epoch": 6.264098382932463,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0849385106672103e-05,
+      "loss": 0.8556,
+      "step": 92195
+    },
+    {
+      "epoch": 6.264438103003125,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0847261856230467e-05,
+      "loss": 0.8711,
+      "step": 92200
+    },
+    {
+      "epoch": 6.264777823073787,
+      "grad_norm": 8.375,
+      "learning_rate": 1.084513860578883e-05,
+      "loss": 0.8187,
+      "step": 92205
+    },
+    {
+      "epoch": 6.265117543144449,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0843015355347195e-05,
+      "loss": 0.9282,
+      "step": 92210
+    },
+    {
+      "epoch": 6.265457263215111,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0840892104905559e-05,
+      "loss": 0.7797,
+      "step": 92215
+    },
+    {
+      "epoch": 6.265796983285773,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0838768854463923e-05,
+      "loss": 0.9173,
+      "step": 92220
+    },
+    {
+      "epoch": 6.266136703356434,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.0836645604022285e-05,
+      "loss": 0.8121,
+      "step": 92225
+    },
+    {
+      "epoch": 6.266476423427096,
+      "grad_norm": 7.75,
+      "learning_rate": 1.083452235358065e-05,
+      "loss": 0.78,
+      "step": 92230
+    },
+    {
+      "epoch": 6.266816143497758,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0832399103139015e-05,
+      "loss": 0.8301,
+      "step": 92235
+    },
+    {
+      "epoch": 6.267155863568419,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0830275852697377e-05,
+      "loss": 0.8181,
+      "step": 92240
+    },
+    {
+      "epoch": 6.267495583639081,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0828152602255743e-05,
+      "loss": 0.9049,
+      "step": 92245
+    },
+    {
+      "epoch": 6.267835303709743,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0826029351814105e-05,
+      "loss": 0.8089,
+      "step": 92250
+    },
+    {
+      "epoch": 6.268175023780405,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0823906101372469e-05,
+      "loss": 0.8921,
+      "step": 92255
+    },
+    {
+      "epoch": 6.268514743851067,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0821782850930835e-05,
+      "loss": 0.8982,
+      "step": 92260
+    },
+    {
+      "epoch": 6.268854463921729,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0819659600489197e-05,
+      "loss": 0.7498,
+      "step": 92265
+    },
+    {
+      "epoch": 6.26919418399239,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0817536350047561e-05,
+      "loss": 0.8157,
+      "step": 92270
+    },
+    {
+      "epoch": 6.269533904063052,
+      "grad_norm": 7.5,
+      "learning_rate": 1.0815413099605925e-05,
+      "loss": 0.8914,
+      "step": 92275
+    },
+    {
+      "epoch": 6.269873624133714,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0813289849164289e-05,
+      "loss": 0.9954,
+      "step": 92280
+    },
+    {
+      "epoch": 6.270213344204375,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0811166598722653e-05,
+      "loss": 0.9361,
+      "step": 92285
+    },
+    {
+      "epoch": 6.270553064275037,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0809043348281017e-05,
+      "loss": 0.9131,
+      "step": 92290
+    },
+    {
+      "epoch": 6.270892784345699,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0806920097839381e-05,
+      "loss": 0.831,
+      "step": 92295
+    },
+    {
+      "epoch": 6.271232504416361,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0804796847397745e-05,
+      "loss": 0.7566,
+      "step": 92300
+    },
+    {
+      "epoch": 6.271572224487023,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0802673596956109e-05,
+      "loss": 0.7878,
+      "step": 92305
+    },
+    {
+      "epoch": 6.271911944557685,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0800550346514473e-05,
+      "loss": 0.9947,
+      "step": 92310
+    },
+    {
+      "epoch": 6.272251664628346,
+      "grad_norm": 9.625,
+      "learning_rate": 1.0798427096072837e-05,
+      "loss": 0.8502,
+      "step": 92315
+    },
+    {
+      "epoch": 6.272591384699008,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0796303845631201e-05,
+      "loss": 0.8526,
+      "step": 92320
+    },
+    {
+      "epoch": 6.27293110476967,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0794180595189563e-05,
+      "loss": 0.8613,
+      "step": 92325
+    },
+    {
+      "epoch": 6.273270824840331,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.0792057344747929e-05,
+      "loss": 0.8356,
+      "step": 92330
+    },
+    {
+      "epoch": 6.273610544910993,
+      "grad_norm": 9.25,
+      "learning_rate": 1.0789934094306291e-05,
+      "loss": 0.8512,
+      "step": 92335
+    },
+    {
+      "epoch": 6.273950264981655,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0787810843864655e-05,
+      "loss": 0.6769,
+      "step": 92340
+    },
+    {
+      "epoch": 6.274289985052317,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0785687593423021e-05,
+      "loss": 0.8407,
+      "step": 92345
+    },
+    {
+      "epoch": 6.274629705122979,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.0783564342981383e-05,
+      "loss": 0.8875,
+      "step": 92350
+    },
+    {
+      "epoch": 6.274969425193641,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0781441092539747e-05,
+      "loss": 0.8003,
+      "step": 92355
+    },
+    {
+      "epoch": 6.275309145264302,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0779317842098111e-05,
+      "loss": 0.8423,
+      "step": 92360
+    },
+    {
+      "epoch": 6.275648865334964,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0777194591656475e-05,
+      "loss": 0.7928,
+      "step": 92365
+    },
+    {
+      "epoch": 6.275988585405626,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.077507134121484e-05,
+      "loss": 0.9217,
+      "step": 92370
+    },
+    {
+      "epoch": 6.276328305476287,
+      "grad_norm": 7.25,
+      "learning_rate": 1.0772948090773203e-05,
+      "loss": 0.8184,
+      "step": 92375
+    },
+    {
+      "epoch": 6.276668025546949,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0770824840331567e-05,
+      "loss": 0.7714,
+      "step": 92380
+    },
+    {
+      "epoch": 6.277007745617611,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0768701589889931e-05,
+      "loss": 0.9161,
+      "step": 92385
+    },
+    {
+      "epoch": 6.277347465688273,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0766578339448295e-05,
+      "loss": 0.7974,
+      "step": 92390
+    },
+    {
+      "epoch": 6.277687185758935,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.076445508900666e-05,
+      "loss": 0.8709,
+      "step": 92395
+    },
+    {
+      "epoch": 6.278026905829597,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0762331838565023e-05,
+      "loss": 0.8161,
+      "step": 92400
+    },
+    {
+      "epoch": 6.278366625900258,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0760208588123387e-05,
+      "loss": 0.7909,
+      "step": 92405
+    },
+    {
+      "epoch": 6.27870634597092,
+      "grad_norm": 9.125,
+      "learning_rate": 1.075808533768175e-05,
+      "loss": 0.9408,
+      "step": 92410
+    },
+    {
+      "epoch": 6.279046066041582,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0755962087240115e-05,
+      "loss": 0.821,
+      "step": 92415
+    },
+    {
+      "epoch": 6.279385786112243,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0753838836798478e-05,
+      "loss": 0.783,
+      "step": 92420
+    },
+    {
+      "epoch": 6.279725506182905,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0751715586356842e-05,
+      "loss": 0.9112,
+      "step": 92425
+    },
+    {
+      "epoch": 6.2800652262535674,
+      "grad_norm": 8.875,
+      "learning_rate": 1.0749592335915207e-05,
+      "loss": 0.8211,
+      "step": 92430
+    },
+    {
+      "epoch": 6.280404946324229,
+      "grad_norm": 8.375,
+      "learning_rate": 1.074746908547357e-05,
+      "loss": 0.9108,
+      "step": 92435
+    },
+    {
+      "epoch": 6.280744666394891,
+      "grad_norm": 10.4375,
+      "learning_rate": 1.0745345835031934e-05,
+      "loss": 0.8323,
+      "step": 92440
+    },
+    {
+      "epoch": 6.281084386465553,
+      "grad_norm": 9.625,
+      "learning_rate": 1.0743222584590298e-05,
+      "loss": 0.8695,
+      "step": 92445
+    },
+    {
+      "epoch": 6.281424106536214,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0741099334148662e-05,
+      "loss": 0.8808,
+      "step": 92450
+    },
+    {
+      "epoch": 6.281763826606876,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0738976083707026e-05,
+      "loss": 0.8935,
+      "step": 92455
+    },
+    {
+      "epoch": 6.282103546677538,
+      "grad_norm": 8.875,
+      "learning_rate": 1.073685283326539e-05,
+      "loss": 0.8311,
+      "step": 92460
+    },
+    {
+      "epoch": 6.282443266748199,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0734729582823754e-05,
+      "loss": 0.7853,
+      "step": 92465
+    },
+    {
+      "epoch": 6.282782986818861,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0732606332382118e-05,
+      "loss": 0.8821,
+      "step": 92470
+    },
+    {
+      "epoch": 6.2831227068895235,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0730483081940482e-05,
+      "loss": 0.902,
+      "step": 92475
+    },
+    {
+      "epoch": 6.283462426960185,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0728359831498846e-05,
+      "loss": 0.8642,
+      "step": 92480
+    },
+    {
+      "epoch": 6.283802147030847,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.072623658105721e-05,
+      "loss": 0.8327,
+      "step": 92485
+    },
+    {
+      "epoch": 6.284141867101509,
+      "grad_norm": 7.375,
+      "learning_rate": 1.0724113330615574e-05,
+      "loss": 0.8387,
+      "step": 92490
+    },
+    {
+      "epoch": 6.28448158717217,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.0721990080173936e-05,
+      "loss": 0.8266,
+      "step": 92495
+    },
+    {
+      "epoch": 6.284821307242832,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.0719866829732302e-05,
+      "loss": 0.7227,
+      "step": 92500
+    },
+    {
+      "epoch": 6.285161027313494,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0717743579290664e-05,
+      "loss": 0.8391,
+      "step": 92505
+    },
+    {
+      "epoch": 6.285500747384155,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0715620328849028e-05,
+      "loss": 0.8466,
+      "step": 92510
+    },
+    {
+      "epoch": 6.285840467454817,
+      "grad_norm": 6.875,
+      "learning_rate": 1.0713497078407394e-05,
+      "loss": 0.8535,
+      "step": 92515
+    },
+    {
+      "epoch": 6.2861801875254795,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0711373827965756e-05,
+      "loss": 0.8626,
+      "step": 92520
+    },
+    {
+      "epoch": 6.286519907596141,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.070925057752412e-05,
+      "loss": 0.8166,
+      "step": 92525
+    },
+    {
+      "epoch": 6.286859627666803,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0707127327082484e-05,
+      "loss": 0.9481,
+      "step": 92530
+    },
+    {
+      "epoch": 6.287199347737464,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.0705004076640848e-05,
+      "loss": 0.8343,
+      "step": 92535
+    },
+    {
+      "epoch": 6.287539067808126,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0702880826199212e-05,
+      "loss": 0.8521,
+      "step": 92540
+    },
+    {
+      "epoch": 6.287878787878788,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0700757575757576e-05,
+      "loss": 0.8677,
+      "step": 92545
+    },
+    {
+      "epoch": 6.288218507949449,
+      "grad_norm": 8.5,
+      "learning_rate": 1.069863432531594e-05,
+      "loss": 0.9204,
+      "step": 92550
+    },
+    {
+      "epoch": 6.288558228020111,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.0696511074874304e-05,
+      "loss": 0.8518,
+      "step": 92555
+    },
+    {
+      "epoch": 6.288897948090773,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0694387824432668e-05,
+      "loss": 0.9644,
+      "step": 92560
+    },
+    {
+      "epoch": 6.289237668161435,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.0692264573991032e-05,
+      "loss": 0.7817,
+      "step": 92565
+    },
+    {
+      "epoch": 6.289577388232097,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0690141323549396e-05,
+      "loss": 0.8958,
+      "step": 92570
+    },
+    {
+      "epoch": 6.289917108302759,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.068801807310776e-05,
+      "loss": 0.827,
+      "step": 92575
+    },
+    {
+      "epoch": 6.29025682837342,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0685894822666122e-05,
+      "loss": 0.8539,
+      "step": 92580
+    },
+    {
+      "epoch": 6.290596548444082,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.0683771572224488e-05,
+      "loss": 0.8391,
+      "step": 92585
+    },
+    {
+      "epoch": 6.290936268514744,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.0681648321782852e-05,
+      "loss": 0.8776,
+      "step": 92590
+    },
+    {
+      "epoch": 6.291275988585405,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0679525071341214e-05,
+      "loss": 0.7292,
+      "step": 92595
+    },
+    {
+      "epoch": 6.291615708656067,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.067740182089958e-05,
+      "loss": 0.8399,
+      "step": 92600
+    },
+    {
+      "epoch": 6.291955428726729,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.0675278570457943e-05,
+      "loss": 0.8408,
+      "step": 92605
+    },
+    {
+      "epoch": 6.292295148797391,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0673155320016307e-05,
+      "loss": 0.8946,
+      "step": 92610
+    },
+    {
+      "epoch": 6.292634868868053,
+      "grad_norm": 8.625,
+      "learning_rate": 1.067103206957467e-05,
+      "loss": 0.8722,
+      "step": 92615
+    },
+    {
+      "epoch": 6.292974588938715,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0668908819133035e-05,
+      "loss": 0.9,
+      "step": 92620
+    },
+    {
+      "epoch": 6.293314309009376,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0666785568691399e-05,
+      "loss": 0.7412,
+      "step": 92625
+    },
+    {
+      "epoch": 6.293654029080038,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0664662318249763e-05,
+      "loss": 0.8885,
+      "step": 92630
+    },
+    {
+      "epoch": 6.2939937491507,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.0662539067808127e-05,
+      "loss": 0.9064,
+      "step": 92635
+    },
+    {
+      "epoch": 6.294333469221361,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.066041581736649e-05,
+      "loss": 0.8981,
+      "step": 92640
+    },
+    {
+      "epoch": 6.294673189292023,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0658292566924855e-05,
+      "loss": 0.8351,
+      "step": 92645
+    },
+    {
+      "epoch": 6.295012909362685,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.0656169316483219e-05,
+      "loss": 0.7771,
+      "step": 92650
+    },
+    {
+      "epoch": 6.295352629433347,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0654046066041583e-05,
+      "loss": 0.9571,
+      "step": 92655
+    },
+    {
+      "epoch": 6.295692349504009,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0651922815599947e-05,
+      "loss": 0.9309,
+      "step": 92660
+    },
+    {
+      "epoch": 6.296032069574671,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0649799565158309e-05,
+      "loss": 0.8746,
+      "step": 92665
+    },
+    {
+      "epoch": 6.296371789645332,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0647676314716675e-05,
+      "loss": 0.8826,
+      "step": 92670
+    },
+    {
+      "epoch": 6.296711509715994,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0645553064275039e-05,
+      "loss": 0.7978,
+      "step": 92675
+    },
+    {
+      "epoch": 6.297051229786656,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0643429813833401e-05,
+      "loss": 0.856,
+      "step": 92680
+    },
+    {
+      "epoch": 6.297390949857317,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0641306563391767e-05,
+      "loss": 0.8475,
+      "step": 92685
+    },
+    {
+      "epoch": 6.297730669927979,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0639183312950129e-05,
+      "loss": 0.8879,
+      "step": 92690
+    },
+    {
+      "epoch": 6.298070389998641,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0637060062508493e-05,
+      "loss": 0.8099,
+      "step": 92695
+    },
+    {
+      "epoch": 6.298410110069303,
+      "grad_norm": 6.625,
+      "learning_rate": 1.0634936812066857e-05,
+      "loss": 0.8408,
+      "step": 92700
+    },
+    {
+      "epoch": 6.298749830139965,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.0632813561625221e-05,
+      "loss": 0.7829,
+      "step": 92705
+    },
+    {
+      "epoch": 6.299089550210627,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0630690311183585e-05,
+      "loss": 0.8417,
+      "step": 92710
+    },
+    {
+      "epoch": 6.299429270281288,
+      "grad_norm": 9.5625,
+      "learning_rate": 1.0628567060741949e-05,
+      "loss": 0.947,
+      "step": 92715
+    },
+    {
+      "epoch": 6.29976899035195,
+      "grad_norm": 7.25,
+      "learning_rate": 1.0626443810300313e-05,
+      "loss": 0.8983,
+      "step": 92720
+    },
+    {
+      "epoch": 6.300108710422612,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0624320559858677e-05,
+      "loss": 0.8696,
+      "step": 92725
+    },
+    {
+      "epoch": 6.300448430493273,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0622197309417041e-05,
+      "loss": 0.8386,
+      "step": 92730
+    },
+    {
+      "epoch": 6.300788150563935,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0620074058975405e-05,
+      "loss": 0.8248,
+      "step": 92735
+    },
+    {
+      "epoch": 6.3011278706345974,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0617950808533769e-05,
+      "loss": 0.8544,
+      "step": 92740
+    },
+    {
+      "epoch": 6.301467590705259,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0615827558092133e-05,
+      "loss": 0.9254,
+      "step": 92745
+    },
+    {
+      "epoch": 6.301807310775921,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.0613704307650495e-05,
+      "loss": 0.808,
+      "step": 92750
+    },
+    {
+      "epoch": 6.302147030846583,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0611581057208861e-05,
+      "loss": 0.8012,
+      "step": 92755
+    },
+    {
+      "epoch": 6.302486750917244,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0609457806767225e-05,
+      "loss": 0.87,
+      "step": 92760
+    },
+    {
+      "epoch": 6.302826470987906,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0607334556325587e-05,
+      "loss": 0.8361,
+      "step": 92765
+    },
+    {
+      "epoch": 6.303166191058568,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0605211305883953e-05,
+      "loss": 0.8221,
+      "step": 92770
+    },
+    {
+      "epoch": 6.303505911129229,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0603088055442315e-05,
+      "loss": 0.8559,
+      "step": 92775
+    },
+    {
+      "epoch": 6.303845631199891,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.060096480500068e-05,
+      "loss": 0.8676,
+      "step": 92780
+    },
+    {
+      "epoch": 6.3041853512705535,
+      "grad_norm": 6.3125,
+      "learning_rate": 1.0598841554559045e-05,
+      "loss": 0.8557,
+      "step": 92785
+    },
+    {
+      "epoch": 6.304525071341215,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0596718304117407e-05,
+      "loss": 0.9274,
+      "step": 92790
+    },
+    {
+      "epoch": 6.304864791411877,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0594595053675771e-05,
+      "loss": 0.9526,
+      "step": 92795
+    },
+    {
+      "epoch": 6.305204511482539,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0592471803234135e-05,
+      "loss": 0.7947,
+      "step": 92800
+    },
+    {
+      "epoch": 6.3055442315532,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.05903485527925e-05,
+      "loss": 0.903,
+      "step": 92805
+    },
+    {
+      "epoch": 6.305883951623862,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0588225302350863e-05,
+      "loss": 0.7157,
+      "step": 92810
+    },
+    {
+      "epoch": 6.306223671694523,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0586102051909227e-05,
+      "loss": 0.8904,
+      "step": 92815
+    },
+    {
+      "epoch": 6.306563391765185,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0583978801467591e-05,
+      "loss": 0.7579,
+      "step": 92820
+    },
+    {
+      "epoch": 6.306903111835847,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0581855551025955e-05,
+      "loss": 0.8576,
+      "step": 92825
+    },
+    {
+      "epoch": 6.307242831906509,
+      "grad_norm": 8.25,
+      "learning_rate": 1.057973230058432e-05,
+      "loss": 0.8716,
+      "step": 92830
+    },
+    {
+      "epoch": 6.307582551977171,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0577609050142683e-05,
+      "loss": 0.8284,
+      "step": 92835
+    },
+    {
+      "epoch": 6.307922272047833,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0575485799701047e-05,
+      "loss": 0.9126,
+      "step": 92840
+    },
+    {
+      "epoch": 6.308261992118494,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.0573362549259411e-05,
+      "loss": 0.8935,
+      "step": 92845
+    },
+    {
+      "epoch": 6.308601712189156,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.0571239298817774e-05,
+      "loss": 0.8655,
+      "step": 92850
+    },
+    {
+      "epoch": 6.308941432259818,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.056911604837614e-05,
+      "loss": 0.8349,
+      "step": 92855
+    },
+    {
+      "epoch": 6.309281152330479,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0566992797934502e-05,
+      "loss": 0.8744,
+      "step": 92860
+    },
+    {
+      "epoch": 6.309620872401141,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0564869547492866e-05,
+      "loss": 0.8926,
+      "step": 92865
+    },
+    {
+      "epoch": 6.309960592471803,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0562746297051231e-05,
+      "loss": 0.8169,
+      "step": 92870
+    },
+    {
+      "epoch": 6.310300312542465,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0560623046609594e-05,
+      "loss": 0.8619,
+      "step": 92875
+    },
+    {
+      "epoch": 6.310640032613127,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0558499796167958e-05,
+      "loss": 0.8366,
+      "step": 92880
+    },
+    {
+      "epoch": 6.310979752683789,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0556376545726322e-05,
+      "loss": 0.858,
+      "step": 92885
+    },
+    {
+      "epoch": 6.31131947275445,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0554253295284686e-05,
+      "loss": 0.8436,
+      "step": 92890
+    },
+    {
+      "epoch": 6.311659192825112,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.055213004484305e-05,
+      "loss": 0.8304,
+      "step": 92895
+    },
+    {
+      "epoch": 6.311998912895774,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0550006794401414e-05,
+      "loss": 0.8303,
+      "step": 92900
+    },
+    {
+      "epoch": 6.312338632966435,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0547883543959778e-05,
+      "loss": 0.7898,
+      "step": 92905
+    },
+    {
+      "epoch": 6.312678353037097,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0545760293518142e-05,
+      "loss": 0.7752,
+      "step": 92910
+    },
+    {
+      "epoch": 6.313018073107759,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0543637043076506e-05,
+      "loss": 0.823,
+      "step": 92915
+    },
+    {
+      "epoch": 6.313357793178421,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.054151379263487e-05,
+      "loss": 0.7204,
+      "step": 92920
+    },
+    {
+      "epoch": 6.313697513249083,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0539390542193234e-05,
+      "loss": 0.8083,
+      "step": 92925
+    },
+    {
+      "epoch": 6.314037233319745,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0537267291751598e-05,
+      "loss": 0.832,
+      "step": 92930
+    },
+    {
+      "epoch": 6.314376953390406,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.053514404130996e-05,
+      "loss": 0.7587,
+      "step": 92935
+    },
+    {
+      "epoch": 6.314716673461068,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0533020790868326e-05,
+      "loss": 0.8723,
+      "step": 92940
+    },
+    {
+      "epoch": 6.31505639353173,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0530897540426688e-05,
+      "loss": 0.9012,
+      "step": 92945
+    },
+    {
+      "epoch": 6.315396113602391,
+      "grad_norm": 7.5,
+      "learning_rate": 1.0528774289985052e-05,
+      "loss": 0.7566,
+      "step": 92950
+    },
+    {
+      "epoch": 6.315735833673053,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0526651039543418e-05,
+      "loss": 0.775,
+      "step": 92955
+    },
+    {
+      "epoch": 6.316075553743715,
+      "grad_norm": 8.25,
+      "learning_rate": 1.052452778910178e-05,
+      "loss": 0.8373,
+      "step": 92960
+    },
+    {
+      "epoch": 6.316415273814377,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0522404538660144e-05,
+      "loss": 0.8375,
+      "step": 92965
+    },
+    {
+      "epoch": 6.316754993885039,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0520281288218508e-05,
+      "loss": 0.7855,
+      "step": 92970
+    },
+    {
+      "epoch": 6.317094713955701,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0518158037776872e-05,
+      "loss": 0.8708,
+      "step": 92975
+    },
+    {
+      "epoch": 6.317434434026362,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0516034787335238e-05,
+      "loss": 0.8379,
+      "step": 92980
+    },
+    {
+      "epoch": 6.317774154097024,
+      "grad_norm": 7.875,
+      "learning_rate": 1.05139115368936e-05,
+      "loss": 0.7743,
+      "step": 92985
+    },
+    {
+      "epoch": 6.318113874167686,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0511788286451964e-05,
+      "loss": 0.8578,
+      "step": 92990
+    },
+    {
+      "epoch": 6.318453594238347,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.0509665036010328e-05,
+      "loss": 0.8529,
+      "step": 92995
+    },
+    {
+      "epoch": 6.318793314309009,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0507541785568692e-05,
+      "loss": 0.8409,
+      "step": 93000
+    },
+    {
+      "epoch": 6.319133034379671,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0505418535127056e-05,
+      "loss": 0.8892,
+      "step": 93005
+    },
+    {
+      "epoch": 6.319472754450333,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.050329528468542e-05,
+      "loss": 0.8514,
+      "step": 93010
+    },
+    {
+      "epoch": 6.319812474520995,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0501172034243784e-05,
+      "loss": 0.6985,
+      "step": 93015
+    },
+    {
+      "epoch": 6.320152194591657,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0499048783802146e-05,
+      "loss": 0.8412,
+      "step": 93020
+    },
+    {
+      "epoch": 6.320491914662318,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0496925533360512e-05,
+      "loss": 0.7412,
+      "step": 93025
+    },
+    {
+      "epoch": 6.32083163473298,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0494802282918876e-05,
+      "loss": 0.8148,
+      "step": 93030
+    },
+    {
+      "epoch": 6.321171354803642,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0492679032477238e-05,
+      "loss": 0.8613,
+      "step": 93035
+    },
+    {
+      "epoch": 6.321511074874303,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0490555782035604e-05,
+      "loss": 0.8166,
+      "step": 93040
+    },
+    {
+      "epoch": 6.321850794944965,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0488432531593966e-05,
+      "loss": 0.9718,
+      "step": 93045
+    },
+    {
+      "epoch": 6.3221905150156275,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.048630928115233e-05,
+      "loss": 0.7457,
+      "step": 93050
+    },
+    {
+      "epoch": 6.322530235086289,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0484186030710694e-05,
+      "loss": 0.7728,
+      "step": 93055
+    },
+    {
+      "epoch": 6.322869955156951,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0482062780269058e-05,
+      "loss": 0.8662,
+      "step": 93060
+    },
+    {
+      "epoch": 6.323209675227613,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.0479939529827424e-05,
+      "loss": 0.9056,
+      "step": 93065
+    },
+    {
+      "epoch": 6.323549395298274,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0477816279385786e-05,
+      "loss": 0.819,
+      "step": 93070
+    },
+    {
+      "epoch": 6.323889115368936,
+      "grad_norm": 8.0,
+      "learning_rate": 1.047569302894415e-05,
+      "loss": 0.8708,
+      "step": 93075
+    },
+    {
+      "epoch": 6.324228835439598,
+      "grad_norm": 6.75,
+      "learning_rate": 1.0473569778502514e-05,
+      "loss": 0.8284,
+      "step": 93080
+    },
+    {
+      "epoch": 6.324568555510259,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0471446528060878e-05,
+      "loss": 1.1023,
+      "step": 93085
+    },
+    {
+      "epoch": 6.324908275580921,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0469323277619242e-05,
+      "loss": 0.8501,
+      "step": 93090
+    },
+    {
+      "epoch": 6.3252479956515835,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0467200027177606e-05,
+      "loss": 0.8438,
+      "step": 93095
+    },
+    {
+      "epoch": 6.325587715722245,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.046507677673597e-05,
+      "loss": 0.8444,
+      "step": 93100
+    },
+    {
+      "epoch": 6.325927435792907,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0462953526294333e-05,
+      "loss": 0.903,
+      "step": 93105
+    },
+    {
+      "epoch": 6.326267155863569,
+      "grad_norm": 9.625,
+      "learning_rate": 1.0460830275852698e-05,
+      "loss": 0.7958,
+      "step": 93110
+    },
+    {
+      "epoch": 6.32660687593423,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0458707025411062e-05,
+      "loss": 0.807,
+      "step": 93115
+    },
+    {
+      "epoch": 6.326946596004892,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.0456583774969425e-05,
+      "loss": 0.8869,
+      "step": 93120
+    },
+    {
+      "epoch": 6.327286316075554,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.045446052452779e-05,
+      "loss": 0.8582,
+      "step": 93125
+    },
+    {
+      "epoch": 6.327626036146215,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0452337274086153e-05,
+      "loss": 0.8872,
+      "step": 93130
+    },
+    {
+      "epoch": 6.327965756216877,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0450214023644517e-05,
+      "loss": 0.8353,
+      "step": 93135
+    },
+    {
+      "epoch": 6.3283054762875395,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.044809077320288e-05,
+      "loss": 0.8338,
+      "step": 93140
+    },
+    {
+      "epoch": 6.328645196358201,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0445967522761245e-05,
+      "loss": 0.8148,
+      "step": 93145
+    },
+    {
+      "epoch": 6.328984916428863,
+      "grad_norm": 8.0,
+      "learning_rate": 1.044384427231961e-05,
+      "loss": 0.8933,
+      "step": 93150
+    },
+    {
+      "epoch": 6.329324636499525,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0441721021877973e-05,
+      "loss": 0.9363,
+      "step": 93155
+    },
+    {
+      "epoch": 6.329664356570186,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0439597771436337e-05,
+      "loss": 0.8875,
+      "step": 93160
+    },
+    {
+      "epoch": 6.330004076640848,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.04374745209947e-05,
+      "loss": 0.7757,
+      "step": 93165
+    },
+    {
+      "epoch": 6.33034379671151,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0435351270553065e-05,
+      "loss": 0.8129,
+      "step": 93170
+    },
+    {
+      "epoch": 6.330683516782171,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0433228020111429e-05,
+      "loss": 0.8296,
+      "step": 93175
+    },
+    {
+      "epoch": 6.331023236852833,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0431104769669793e-05,
+      "loss": 0.9194,
+      "step": 93180
+    },
+    {
+      "epoch": 6.3313629569234955,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0428981519228157e-05,
+      "loss": 0.7559,
+      "step": 93185
+    },
+    {
+      "epoch": 6.331702676994157,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.0426858268786519e-05,
+      "loss": 0.8167,
+      "step": 93190
+    },
+    {
+      "epoch": 6.332042397064819,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0424735018344885e-05,
+      "loss": 0.809,
+      "step": 93195
+    },
+    {
+      "epoch": 6.332382117135481,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0422611767903249e-05,
+      "loss": 0.9153,
+      "step": 93200
+    },
+    {
+      "epoch": 6.332721837206142,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0420488517461611e-05,
+      "loss": 0.8149,
+      "step": 93205
+    },
+    {
+      "epoch": 6.333061557276804,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0418365267019977e-05,
+      "loss": 0.8456,
+      "step": 93210
+    },
+    {
+      "epoch": 6.333401277347465,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0416242016578339e-05,
+      "loss": 0.8668,
+      "step": 93215
+    },
+    {
+      "epoch": 6.333740997418127,
+      "grad_norm": 9.25,
+      "learning_rate": 1.0414118766136703e-05,
+      "loss": 0.7735,
+      "step": 93220
+    },
+    {
+      "epoch": 6.334080717488789,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0411995515695069e-05,
+      "loss": 0.8087,
+      "step": 93225
+    },
+    {
+      "epoch": 6.334420437559451,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0409872265253431e-05,
+      "loss": 0.8605,
+      "step": 93230
+    },
+    {
+      "epoch": 6.334760157630113,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0407749014811797e-05,
+      "loss": 0.8701,
+      "step": 93235
+    },
+    {
+      "epoch": 6.335099877700775,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0405625764370159e-05,
+      "loss": 0.7897,
+      "step": 93240
+    },
+    {
+      "epoch": 6.335439597771436,
+      "grad_norm": 9.25,
+      "learning_rate": 1.0403502513928523e-05,
+      "loss": 0.8629,
+      "step": 93245
+    },
+    {
+      "epoch": 6.335779317842098,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0401379263486887e-05,
+      "loss": 0.8056,
+      "step": 93250
+    },
+    {
+      "epoch": 6.33611903791276,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0399256013045251e-05,
+      "loss": 0.9184,
+      "step": 93255
+    },
+    {
+      "epoch": 6.336458757983421,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0397132762603615e-05,
+      "loss": 0.8637,
+      "step": 93260
+    },
+    {
+      "epoch": 6.336798478054083,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.039500951216198e-05,
+      "loss": 0.8324,
+      "step": 93265
+    },
+    {
+      "epoch": 6.337138198124745,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0392886261720343e-05,
+      "loss": 0.8926,
+      "step": 93270
+    },
+    {
+      "epoch": 6.337477918195407,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0390763011278707e-05,
+      "loss": 0.8651,
+      "step": 93275
+    },
+    {
+      "epoch": 6.337817638266069,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0388639760837071e-05,
+      "loss": 0.8411,
+      "step": 93280
+    },
+    {
+      "epoch": 6.338157358336731,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0386516510395435e-05,
+      "loss": 0.7808,
+      "step": 93285
+    },
+    {
+      "epoch": 6.338497078407392,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.0384393259953797e-05,
+      "loss": 0.8525,
+      "step": 93290
+    },
+    {
+      "epoch": 6.338836798478054,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0382270009512163e-05,
+      "loss": 0.8706,
+      "step": 93295
+    },
+    {
+      "epoch": 6.339176518548716,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.0380146759070525e-05,
+      "loss": 0.8403,
+      "step": 93300
+    },
+    {
+      "epoch": 6.339516238619377,
+      "grad_norm": 8.375,
+      "learning_rate": 1.037802350862889e-05,
+      "loss": 0.8847,
+      "step": 93305
+    },
+    {
+      "epoch": 6.339855958690039,
+      "grad_norm": 8.75,
+      "learning_rate": 1.0375900258187255e-05,
+      "loss": 0.873,
+      "step": 93310
+    },
+    {
+      "epoch": 6.3401956787607014,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.0373777007745618e-05,
+      "loss": 0.8242,
+      "step": 93315
+    },
+    {
+      "epoch": 6.340535398831363,
+      "grad_norm": 7.5,
+      "learning_rate": 1.0371653757303983e-05,
+      "loss": 0.8188,
+      "step": 93320
+    },
+    {
+      "epoch": 6.340875118902025,
+      "grad_norm": 7.5,
+      "learning_rate": 1.0369530506862346e-05,
+      "loss": 0.8086,
+      "step": 93325
+    },
+    {
+      "epoch": 6.341214838972687,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.036740725642071e-05,
+      "loss": 0.8479,
+      "step": 93330
+    },
+    {
+      "epoch": 6.341554559043348,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0365284005979074e-05,
+      "loss": 0.8822,
+      "step": 93335
+    },
+    {
+      "epoch": 6.34189427911401,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0363160755537438e-05,
+      "loss": 0.8088,
+      "step": 93340
+    },
+    {
+      "epoch": 6.342233999184672,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0361037505095802e-05,
+      "loss": 0.8804,
+      "step": 93345
+    },
+    {
+      "epoch": 6.342573719255333,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0358914254654166e-05,
+      "loss": 0.8598,
+      "step": 93350
+    },
+    {
+      "epoch": 6.342913439325995,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.035679100421253e-05,
+      "loss": 0.8146,
+      "step": 93355
+    },
+    {
+      "epoch": 6.3432531593966575,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.0354667753770894e-05,
+      "loss": 0.8531,
+      "step": 93360
+    },
+    {
+      "epoch": 6.343592879467319,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.0352544503329258e-05,
+      "loss": 0.7406,
+      "step": 93365
+    },
+    {
+      "epoch": 6.343932599537981,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0350421252887622e-05,
+      "loss": 0.84,
+      "step": 93370
+    },
+    {
+      "epoch": 6.344272319608643,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0348298002445984e-05,
+      "loss": 0.9213,
+      "step": 93375
+    },
+    {
+      "epoch": 6.344612039679304,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.034617475200435e-05,
+      "loss": 0.92,
+      "step": 93380
+    },
+    {
+      "epoch": 6.344951759749966,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0344051501562712e-05,
+      "loss": 0.8122,
+      "step": 93385
+    },
+    {
+      "epoch": 6.345291479820628,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0341928251121076e-05,
+      "loss": 0.8251,
+      "step": 93390
+    },
+    {
+      "epoch": 6.345631199891289,
+      "grad_norm": 8.75,
+      "learning_rate": 1.0339805000679442e-05,
+      "loss": 0.9008,
+      "step": 93395
+    },
+    {
+      "epoch": 6.345970919961951,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0337681750237804e-05,
+      "loss": 0.8611,
+      "step": 93400
+    },
+    {
+      "epoch": 6.3463106400326135,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.033555849979617e-05,
+      "loss": 0.8831,
+      "step": 93405
+    },
+    {
+      "epoch": 6.346650360103275,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0333435249354532e-05,
+      "loss": 0.8818,
+      "step": 93410
+    },
+    {
+      "epoch": 6.346990080173937,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0331311998912896e-05,
+      "loss": 0.7854,
+      "step": 93415
+    },
+    {
+      "epoch": 6.347329800244599,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.032918874847126e-05,
+      "loss": 0.9232,
+      "step": 93420
+    },
+    {
+      "epoch": 6.34766952031526,
+      "grad_norm": 7.125,
+      "learning_rate": 1.0327065498029624e-05,
+      "loss": 0.8882,
+      "step": 93425
+    },
+    {
+      "epoch": 6.348009240385922,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0324942247587988e-05,
+      "loss": 0.8322,
+      "step": 93430
+    },
+    {
+      "epoch": 6.348348960456584,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0322818997146352e-05,
+      "loss": 0.9112,
+      "step": 93435
+    },
+    {
+      "epoch": 6.348688680527245,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0320695746704716e-05,
+      "loss": 0.9233,
+      "step": 93440
+    },
+    {
+      "epoch": 6.349028400597907,
+      "grad_norm": 8.75,
+      "learning_rate": 1.031857249626308e-05,
+      "loss": 0.8027,
+      "step": 93445
+    },
+    {
+      "epoch": 6.3493681206685695,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.0316449245821444e-05,
+      "loss": 0.8251,
+      "step": 93450
+    },
+    {
+      "epoch": 6.349707840739231,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0314325995379808e-05,
+      "loss": 0.8671,
+      "step": 93455
+    },
+    {
+      "epoch": 6.350047560809893,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.031220274493817e-05,
+      "loss": 0.8798,
+      "step": 93460
+    },
+    {
+      "epoch": 6.350387280880555,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0310079494496536e-05,
+      "loss": 0.8141,
+      "step": 93465
+    },
+    {
+      "epoch": 6.350727000951216,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0307956244054898e-05,
+      "loss": 0.796,
+      "step": 93470
+    },
+    {
+      "epoch": 6.351066721021878,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0305832993613262e-05,
+      "loss": 0.9873,
+      "step": 93475
+    },
+    {
+      "epoch": 6.35140644109254,
+      "grad_norm": 8.875,
+      "learning_rate": 1.0303709743171628e-05,
+      "loss": 0.8661,
+      "step": 93480
+    },
+    {
+      "epoch": 6.351746161163201,
+      "grad_norm": 6.75,
+      "learning_rate": 1.030158649272999e-05,
+      "loss": 0.7703,
+      "step": 93485
+    },
+    {
+      "epoch": 6.352085881233863,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0299463242288356e-05,
+      "loss": 0.8172,
+      "step": 93490
+    },
+    {
+      "epoch": 6.352425601304525,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0297339991846718e-05,
+      "loss": 0.8471,
+      "step": 93495
+    },
+    {
+      "epoch": 6.352765321375187,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0295216741405082e-05,
+      "loss": 0.8249,
+      "step": 93500
+    },
+    {
+      "epoch": 6.353105041445849,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0293093490963448e-05,
+      "loss": 0.8305,
+      "step": 93505
+    },
+    {
+      "epoch": 6.35344476151651,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.029097024052181e-05,
+      "loss": 0.747,
+      "step": 93510
+    },
+    {
+      "epoch": 6.353784481587172,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0288846990080174e-05,
+      "loss": 0.7497,
+      "step": 93515
+    },
+    {
+      "epoch": 6.354124201657834,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.0286723739638538e-05,
+      "loss": 0.8197,
+      "step": 93520
+    },
+    {
+      "epoch": 6.354463921728495,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0284600489196902e-05,
+      "loss": 0.9201,
+      "step": 93525
+    },
+    {
+      "epoch": 6.354803641799157,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0282477238755266e-05,
+      "loss": 0.7885,
+      "step": 93530
+    },
+    {
+      "epoch": 6.355143361869819,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.028035398831363e-05,
+      "loss": 0.8094,
+      "step": 93535
+    },
+    {
+      "epoch": 6.355483081940481,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.0278230737871994e-05,
+      "loss": 0.7699,
+      "step": 93540
+    },
+    {
+      "epoch": 6.355822802011143,
+      "grad_norm": 9.125,
+      "learning_rate": 1.0276107487430357e-05,
+      "loss": 0.8481,
+      "step": 93545
+    },
+    {
+      "epoch": 6.356162522081805,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.0273984236988722e-05,
+      "loss": 0.8772,
+      "step": 93550
+    },
+    {
+      "epoch": 6.356502242152466,
+      "grad_norm": 7.375,
+      "learning_rate": 1.0271860986547086e-05,
+      "loss": 0.8049,
+      "step": 93555
+    },
+    {
+      "epoch": 6.356841962223128,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0269737736105449e-05,
+      "loss": 0.7688,
+      "step": 93560
+    },
+    {
+      "epoch": 6.35718168229379,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.0267614485663814e-05,
+      "loss": 0.8952,
+      "step": 93565
+    },
+    {
+      "epoch": 6.357521402364451,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.0265491235222177e-05,
+      "loss": 0.8716,
+      "step": 93570
+    },
+    {
+      "epoch": 6.357861122435113,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0263367984780542e-05,
+      "loss": 0.8089,
+      "step": 93575
+    },
+    {
+      "epoch": 6.358200842505775,
+      "grad_norm": 6.75,
+      "learning_rate": 1.0261244734338905e-05,
+      "loss": 0.8503,
+      "step": 93580
+    },
+    {
+      "epoch": 6.358540562576437,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0259121483897269e-05,
+      "loss": 0.8338,
+      "step": 93585
+    },
+    {
+      "epoch": 6.358880282647099,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0256998233455634e-05,
+      "loss": 0.7976,
+      "step": 93590
+    },
+    {
+      "epoch": 6.359220002717761,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0254874983013997e-05,
+      "loss": 0.8554,
+      "step": 93595
+    },
+    {
+      "epoch": 6.359559722788422,
+      "grad_norm": 7.75,
+      "learning_rate": 1.025275173257236e-05,
+      "loss": 0.8362,
+      "step": 93600
+    },
+    {
+      "epoch": 6.359899442859084,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0250628482130725e-05,
+      "loss": 0.7554,
+      "step": 93605
+    },
+    {
+      "epoch": 6.360239162929746,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.0248505231689089e-05,
+      "loss": 0.827,
+      "step": 93610
+    },
+    {
+      "epoch": 6.360578883000407,
+      "grad_norm": 5.90625,
+      "learning_rate": 1.0246381981247453e-05,
+      "loss": 0.8761,
+      "step": 93615
+    },
+    {
+      "epoch": 6.360918603071069,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0244258730805817e-05,
+      "loss": 0.7744,
+      "step": 93620
+    },
+    {
+      "epoch": 6.3612583231417315,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.024213548036418e-05,
+      "loss": 0.8167,
+      "step": 93625
+    },
+    {
+      "epoch": 6.361598043212393,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0240012229922543e-05,
+      "loss": 0.8433,
+      "step": 93630
+    },
+    {
+      "epoch": 6.361937763283055,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0237888979480909e-05,
+      "loss": 0.8429,
+      "step": 93635
+    },
+    {
+      "epoch": 6.362277483353717,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.0235765729039273e-05,
+      "loss": 0.8237,
+      "step": 93640
+    },
+    {
+      "epoch": 6.362617203424378,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0233642478597635e-05,
+      "loss": 0.8583,
+      "step": 93645
+    },
+    {
+      "epoch": 6.36295692349504,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0231519228156e-05,
+      "loss": 0.771,
+      "step": 93650
+    },
+    {
+      "epoch": 6.363296643565702,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0229395977714363e-05,
+      "loss": 0.8989,
+      "step": 93655
+    },
+    {
+      "epoch": 6.363636363636363,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0227272727272729e-05,
+      "loss": 0.7519,
+      "step": 93660
+    },
+    {
+      "epoch": 6.363976083707025,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0225149476831091e-05,
+      "loss": 0.8319,
+      "step": 93665
+    },
+    {
+      "epoch": 6.3643158037776875,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.0223026226389455e-05,
+      "loss": 0.8378,
+      "step": 93670
+    },
+    {
+      "epoch": 6.364655523848349,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.022090297594782e-05,
+      "loss": 0.812,
+      "step": 93675
+    },
+    {
+      "epoch": 6.364995243919011,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0218779725506183e-05,
+      "loss": 0.7991,
+      "step": 93680
+    },
+    {
+      "epoch": 6.365334963989673,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0216656475064547e-05,
+      "loss": 0.8726,
+      "step": 93685
+    },
+    {
+      "epoch": 6.365674684060334,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.0214533224622911e-05,
+      "loss": 0.9425,
+      "step": 93690
+    },
+    {
+      "epoch": 6.366014404130996,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0212409974181275e-05,
+      "loss": 0.828,
+      "step": 93695
+    },
+    {
+      "epoch": 6.366354124201658,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0210286723739639e-05,
+      "loss": 0.8504,
+      "step": 93700
+    },
+    {
+      "epoch": 6.366693844272319,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.0208163473298003e-05,
+      "loss": 0.8439,
+      "step": 93705
+    },
+    {
+      "epoch": 6.367033564342981,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.0206040222856367e-05,
+      "loss": 0.8371,
+      "step": 93710
+    },
+    {
+      "epoch": 6.3673732844136435,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.020391697241473e-05,
+      "loss": 0.786,
+      "step": 93715
+    },
+    {
+      "epoch": 6.367713004484305,
+      "grad_norm": 7.375,
+      "learning_rate": 1.0201793721973095e-05,
+      "loss": 0.7641,
+      "step": 93720
+    },
+    {
+      "epoch": 6.368052724554967,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0199670471531459e-05,
+      "loss": 0.7828,
+      "step": 93725
+    },
+    {
+      "epoch": 6.368392444625629,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0197547221089821e-05,
+      "loss": 0.929,
+      "step": 93730
+    },
+    {
+      "epoch": 6.36873216469629,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0195423970648187e-05,
+      "loss": 0.8223,
+      "step": 93735
+    },
+    {
+      "epoch": 6.369071884766952,
+      "grad_norm": 8.125,
+      "learning_rate": 1.019330072020655e-05,
+      "loss": 0.8746,
+      "step": 93740
+    },
+    {
+      "epoch": 6.369411604837614,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0191177469764915e-05,
+      "loss": 0.8711,
+      "step": 93745
+    },
+    {
+      "epoch": 6.369751324908275,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0189054219323279e-05,
+      "loss": 0.8298,
+      "step": 93750
+    },
+    {
+      "epoch": 6.370091044978937,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0186930968881641e-05,
+      "loss": 0.8901,
+      "step": 93755
+    },
+    {
+      "epoch": 6.3704307650495995,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.0184807718440007e-05,
+      "loss": 0.8298,
+      "step": 93760
+    },
+    {
+      "epoch": 6.370770485120261,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.018268446799837e-05,
+      "loss": 0.8252,
+      "step": 93765
+    },
+    {
+      "epoch": 6.371110205190923,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0180561217556733e-05,
+      "loss": 0.876,
+      "step": 93770
+    },
+    {
+      "epoch": 6.371449925261585,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0178437967115097e-05,
+      "loss": 0.8716,
+      "step": 93775
+    },
+    {
+      "epoch": 6.371789645332246,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.0176314716673461e-05,
+      "loss": 0.8714,
+      "step": 93780
+    },
+    {
+      "epoch": 6.372129365402908,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.0174191466231825e-05,
+      "loss": 0.7761,
+      "step": 93785
+    },
+    {
+      "epoch": 6.37246908547357,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.017206821579019e-05,
+      "loss": 0.7977,
+      "step": 93790
+    },
+    {
+      "epoch": 6.372808805544231,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0169944965348553e-05,
+      "loss": 0.8301,
+      "step": 93795
+    },
+    {
+      "epoch": 6.373148525614893,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0167821714906917e-05,
+      "loss": 0.7916,
+      "step": 93800
+    },
+    {
+      "epoch": 6.3734882456855555,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0165698464465281e-05,
+      "loss": 0.9151,
+      "step": 93805
+    },
+    {
+      "epoch": 6.373827965756217,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0163575214023645e-05,
+      "loss": 0.8831,
+      "step": 93810
+    },
+    {
+      "epoch": 6.374167685826879,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0161451963582008e-05,
+      "loss": 0.8547,
+      "step": 93815
+    },
+    {
+      "epoch": 6.374507405897541,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0159328713140373e-05,
+      "loss": 0.8956,
+      "step": 93820
+    },
+    {
+      "epoch": 6.374847125968202,
+      "grad_norm": 9.0,
+      "learning_rate": 1.0157205462698736e-05,
+      "loss": 0.9072,
+      "step": 93825
+    },
+    {
+      "epoch": 6.375186846038864,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0155082212257101e-05,
+      "loss": 0.8593,
+      "step": 93830
+    },
+    {
+      "epoch": 6.375526566109526,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.0152958961815465e-05,
+      "loss": 0.784,
+      "step": 93835
+    },
+    {
+      "epoch": 6.375866286180187,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0150835711373828e-05,
+      "loss": 0.8397,
+      "step": 93840
+    },
+    {
+      "epoch": 6.376206006250849,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0148712460932193e-05,
+      "loss": 0.9209,
+      "step": 93845
+    },
+    {
+      "epoch": 6.3765457263215115,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0146589210490556e-05,
+      "loss": 0.9431,
+      "step": 93850
+    },
+    {
+      "epoch": 6.376885446392173,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.014446596004892e-05,
+      "loss": 0.9454,
+      "step": 93855
+    },
+    {
+      "epoch": 6.377225166462835,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0142342709607284e-05,
+      "loss": 0.9258,
+      "step": 93860
+    },
+    {
+      "epoch": 6.377564886533497,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0140219459165648e-05,
+      "loss": 0.7595,
+      "step": 93865
+    },
+    {
+      "epoch": 6.377904606604158,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0138096208724012e-05,
+      "loss": 0.8394,
+      "step": 93870
+    },
+    {
+      "epoch": 6.37824432667482,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.0135972958282376e-05,
+      "loss": 0.7993,
+      "step": 93875
+    },
+    {
+      "epoch": 6.378584046745482,
+      "grad_norm": 7.0,
+      "learning_rate": 1.013384970784074e-05,
+      "loss": 0.7856,
+      "step": 93880
+    },
+    {
+      "epoch": 6.378923766816143,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0131726457399104e-05,
+      "loss": 0.9828,
+      "step": 93885
+    },
+    {
+      "epoch": 6.3792634868868054,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0129603206957468e-05,
+      "loss": 0.7515,
+      "step": 93890
+    },
+    {
+      "epoch": 6.379603206957467,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0127479956515832e-05,
+      "loss": 0.8516,
+      "step": 93895
+    },
+    {
+      "epoch": 6.379942927028129,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0125356706074194e-05,
+      "loss": 0.7063,
+      "step": 93900
+    },
+    {
+      "epoch": 6.380282647098791,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.012323345563256e-05,
+      "loss": 0.7685,
+      "step": 93905
+    },
+    {
+      "epoch": 6.380622367169452,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.0121110205190922e-05,
+      "loss": 0.9089,
+      "step": 93910
+    },
+    {
+      "epoch": 6.380962087240114,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.0118986954749288e-05,
+      "loss": 0.8384,
+      "step": 93915
+    },
+    {
+      "epoch": 6.381301807310776,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0116863704307652e-05,
+      "loss": 0.7761,
+      "step": 93920
+    },
+    {
+      "epoch": 6.381641527381437,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0114740453866014e-05,
+      "loss": 0.7984,
+      "step": 93925
+    },
+    {
+      "epoch": 6.381981247452099,
+      "grad_norm": 8.5,
+      "learning_rate": 1.011261720342438e-05,
+      "loss": 0.8967,
+      "step": 93930
+    },
+    {
+      "epoch": 6.3823209675227615,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.0110493952982742e-05,
+      "loss": 0.842,
+      "step": 93935
+    },
+    {
+      "epoch": 6.382660687593423,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0108370702541106e-05,
+      "loss": 0.9597,
+      "step": 93940
+    },
+    {
+      "epoch": 6.383000407664085,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0106247452099472e-05,
+      "loss": 0.8499,
+      "step": 93945
+    },
+    {
+      "epoch": 6.383340127734747,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.0104124201657834e-05,
+      "loss": 0.8087,
+      "step": 93950
+    },
+    {
+      "epoch": 6.383679847805408,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0102000951216198e-05,
+      "loss": 0.813,
+      "step": 93955
+    },
+    {
+      "epoch": 6.38401956787607,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0099877700774562e-05,
+      "loss": 0.7333,
+      "step": 93960
+    },
+    {
+      "epoch": 6.384359287946732,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0097754450332926e-05,
+      "loss": 0.8776,
+      "step": 93965
+    },
+    {
+      "epoch": 6.384699008017393,
+      "grad_norm": 8.25,
+      "learning_rate": 1.009563119989129e-05,
+      "loss": 0.8133,
+      "step": 93970
+    },
+    {
+      "epoch": 6.385038728088055,
+      "grad_norm": 6.625,
+      "learning_rate": 1.0093507949449654e-05,
+      "loss": 0.8198,
+      "step": 93975
+    },
+    {
+      "epoch": 6.3853784481587175,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0091384699008018e-05,
+      "loss": 0.8583,
+      "step": 93980
+    },
+    {
+      "epoch": 6.385718168229379,
+      "grad_norm": 8.375,
+      "learning_rate": 1.008926144856638e-05,
+      "loss": 0.8052,
+      "step": 93985
+    },
+    {
+      "epoch": 6.386057888300041,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0087138198124746e-05,
+      "loss": 0.8197,
+      "step": 93990
+    },
+    {
+      "epoch": 6.386397608370703,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.008501494768311e-05,
+      "loss": 0.8221,
+      "step": 93995
+    },
+    {
+      "epoch": 6.386737328441364,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0082891697241474e-05,
+      "loss": 0.8209,
+      "step": 94000
+    },
+    {
+      "epoch": 6.387077048512026,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0080768446799838e-05,
+      "loss": 0.8136,
+      "step": 94005
+    },
+    {
+      "epoch": 6.387416768582688,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.00786451963582e-05,
+      "loss": 0.8661,
+      "step": 94010
+    },
+    {
+      "epoch": 6.387756488653349,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0076521945916566e-05,
+      "loss": 0.87,
+      "step": 94015
+    },
+    {
+      "epoch": 6.388096208724011,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0074398695474928e-05,
+      "loss": 0.883,
+      "step": 94020
+    },
+    {
+      "epoch": 6.3884359287946735,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0072275445033292e-05,
+      "loss": 0.9037,
+      "step": 94025
+    },
+    {
+      "epoch": 6.388775648865335,
+      "grad_norm": 5.96875,
+      "learning_rate": 1.0070152194591658e-05,
+      "loss": 0.843,
+      "step": 94030
+    },
+    {
+      "epoch": 6.389115368935997,
+      "grad_norm": 6.9375,
+      "learning_rate": 1.006802894415002e-05,
+      "loss": 0.9787,
+      "step": 94035
+    },
+    {
+      "epoch": 6.389455089006659,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.0065905693708385e-05,
+      "loss": 0.8614,
+      "step": 94040
+    },
+    {
+      "epoch": 6.38979480907732,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0063782443266749e-05,
+      "loss": 0.7939,
+      "step": 94045
+    },
+    {
+      "epoch": 6.390134529147982,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.0061659192825113e-05,
+      "loss": 0.8017,
+      "step": 94050
+    },
+    {
+      "epoch": 6.390474249218644,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0059535942383477e-05,
+      "loss": 0.8339,
+      "step": 94055
+    },
+    {
+      "epoch": 6.390813969289305,
+      "grad_norm": 8.5,
+      "learning_rate": 1.005741269194184e-05,
+      "loss": 0.7691,
+      "step": 94060
+    },
+    {
+      "epoch": 6.391153689359967,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0055289441500205e-05,
+      "loss": 0.8094,
+      "step": 94065
+    },
+    {
+      "epoch": 6.3914934094306295,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0053166191058567e-05,
+      "loss": 0.9426,
+      "step": 94070
+    },
+    {
+      "epoch": 6.391833129501291,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0051042940616933e-05,
+      "loss": 0.8549,
+      "step": 94075
+    },
+    {
+      "epoch": 6.392172849571953,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0048919690175297e-05,
+      "loss": 0.8639,
+      "step": 94080
+    },
+    {
+      "epoch": 6.392512569642615,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.004679643973366e-05,
+      "loss": 0.9249,
+      "step": 94085
+    },
+    {
+      "epoch": 6.392852289713276,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.0044673189292025e-05,
+      "loss": 0.8382,
+      "step": 94090
+    },
+    {
+      "epoch": 6.393192009783938,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0042549938850387e-05,
+      "loss": 0.8168,
+      "step": 94095
+    },
+    {
+      "epoch": 6.3935317298546,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.0040426688408753e-05,
+      "loss": 0.8914,
+      "step": 94100
+    },
+    {
+      "epoch": 6.393871449925261,
+      "grad_norm": 7.5,
+      "learning_rate": 1.0038303437967115e-05,
+      "loss": 0.7804,
+      "step": 94105
+    },
+    {
+      "epoch": 6.394211169995923,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0036180187525479e-05,
+      "loss": 0.8566,
+      "step": 94110
+    },
+    {
+      "epoch": 6.3945508900665855,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0034056937083845e-05,
+      "loss": 0.8148,
+      "step": 94115
+    },
+    {
+      "epoch": 6.394890610137247,
+      "grad_norm": 9.3125,
+      "learning_rate": 1.0031933686642207e-05,
+      "loss": 0.7974,
+      "step": 94120
+    },
+    {
+      "epoch": 6.395230330207909,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.0029810436200571e-05,
+      "loss": 0.8816,
+      "step": 94125
+    },
+    {
+      "epoch": 6.395570050278571,
+      "grad_norm": 8.625,
+      "learning_rate": 1.0027687185758935e-05,
+      "loss": 0.8776,
+      "step": 94130
+    },
+    {
+      "epoch": 6.395909770349232,
+      "grad_norm": 8.75,
+      "learning_rate": 1.0025563935317299e-05,
+      "loss": 0.8345,
+      "step": 94135
+    },
+    {
+      "epoch": 6.396249490419894,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.0023440684875663e-05,
+      "loss": 0.7369,
+      "step": 94140
+    },
+    {
+      "epoch": 6.396589210490556,
+      "grad_norm": 7.125,
+      "learning_rate": 1.0021317434434027e-05,
+      "loss": 0.7341,
+      "step": 94145
+    },
+    {
+      "epoch": 6.396928930561217,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0019194183992391e-05,
+      "loss": 0.8248,
+      "step": 94150
+    },
+    {
+      "epoch": 6.397268650631879,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0017070933550753e-05,
+      "loss": 0.8098,
+      "step": 94155
+    },
+    {
+      "epoch": 6.3976083707025415,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0014947683109119e-05,
+      "loss": 0.8871,
+      "step": 94160
+    },
+    {
+      "epoch": 6.397948090773203,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0012824432667483e-05,
+      "loss": 0.8752,
+      "step": 94165
+    },
+    {
+      "epoch": 6.398287810843865,
+      "grad_norm": 9.125,
+      "learning_rate": 1.0010701182225847e-05,
+      "loss": 0.8583,
+      "step": 94170
+    },
+    {
+      "epoch": 6.398627530914526,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0008577931784211e-05,
+      "loss": 0.8315,
+      "step": 94175
+    },
+    {
+      "epoch": 6.398967250985188,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.0006454681342573e-05,
+      "loss": 0.8023,
+      "step": 94180
+    },
+    {
+      "epoch": 6.39930697105585,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.0004331430900939e-05,
+      "loss": 0.8781,
+      "step": 94185
+    },
+    {
+      "epoch": 6.399646691126511,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0002208180459301e-05,
+      "loss": 0.9017,
+      "step": 94190
+    },
+    {
+      "epoch": 6.399986411197173,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0000084930017665e-05,
+      "loss": 0.9786,
+      "step": 94195
+    },
+    {
+      "epoch": 6.4003261312678354,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.997961679576031e-06,
+      "loss": 0.8154,
+      "step": 94200
+    },
+    {
+      "epoch": 6.400665851338497,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.995838429134393e-06,
+      "loss": 0.7736,
+      "step": 94205
+    },
+    {
+      "epoch": 6.401005571409159,
+      "grad_norm": 7.75,
+      "learning_rate": 9.993715178692757e-06,
+      "loss": 0.8478,
+      "step": 94210
+    },
+    {
+      "epoch": 6.401345291479821,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.991591928251121e-06,
+      "loss": 0.8048,
+      "step": 94215
+    },
+    {
+      "epoch": 6.401685011550482,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.989468677809485e-06,
+      "loss": 0.9043,
+      "step": 94220
+    },
+    {
+      "epoch": 6.402024731621144,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.98734542736785e-06,
+      "loss": 0.8902,
+      "step": 94225
+    },
+    {
+      "epoch": 6.402364451691806,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.985222176926213e-06,
+      "loss": 0.8065,
+      "step": 94230
+    },
+    {
+      "epoch": 6.402704171762467,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.983098926484577e-06,
+      "loss": 0.8187,
+      "step": 94235
+    },
+    {
+      "epoch": 6.403043891833129,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.98097567604294e-06,
+      "loss": 0.7905,
+      "step": 94240
+    },
+    {
+      "epoch": 6.4033836119037915,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.978852425601305e-06,
+      "loss": 0.8862,
+      "step": 94245
+    },
+    {
+      "epoch": 6.403723331974453,
+      "grad_norm": 7.875,
+      "learning_rate": 9.97672917515967e-06,
+      "loss": 0.9123,
+      "step": 94250
+    },
+    {
+      "epoch": 6.404063052045115,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.974605924718033e-06,
+      "loss": 0.952,
+      "step": 94255
+    },
+    {
+      "epoch": 6.404402772115777,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.972482674276397e-06,
+      "loss": 0.8463,
+      "step": 94260
+    },
+    {
+      "epoch": 6.404742492186438,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.97035942383476e-06,
+      "loss": 0.8694,
+      "step": 94265
+    },
+    {
+      "epoch": 6.4050822122571,
+      "grad_norm": 8.625,
+      "learning_rate": 9.968236173393125e-06,
+      "loss": 0.8641,
+      "step": 94270
+    },
+    {
+      "epoch": 6.405421932327762,
+      "grad_norm": 8.375,
+      "learning_rate": 9.96611292295149e-06,
+      "loss": 0.8193,
+      "step": 94275
+    },
+    {
+      "epoch": 6.405761652398423,
+      "grad_norm": 8.0,
+      "learning_rate": 9.963989672509852e-06,
+      "loss": 0.7734,
+      "step": 94280
+    },
+    {
+      "epoch": 6.406101372469085,
+      "grad_norm": 7.875,
+      "learning_rate": 9.961866422068217e-06,
+      "loss": 0.9018,
+      "step": 94285
+    },
+    {
+      "epoch": 6.4064410925397475,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.95974317162658e-06,
+      "loss": 0.7236,
+      "step": 94290
+    },
+    {
+      "epoch": 6.406780812610409,
+      "grad_norm": 7.5,
+      "learning_rate": 9.957619921184944e-06,
+      "loss": 0.8093,
+      "step": 94295
+    },
+    {
+      "epoch": 6.407120532681071,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.955496670743308e-06,
+      "loss": 0.8865,
+      "step": 94300
+    },
+    {
+      "epoch": 6.407460252751733,
+      "grad_norm": 9.0,
+      "learning_rate": 9.953373420301672e-06,
+      "loss": 0.9059,
+      "step": 94305
+    },
+    {
+      "epoch": 6.407799972822394,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.951250169860036e-06,
+      "loss": 0.7632,
+      "step": 94310
+    },
+    {
+      "epoch": 6.408139692893056,
+      "grad_norm": 8.25,
+      "learning_rate": 9.9491269194184e-06,
+      "loss": 0.8238,
+      "step": 94315
+    },
+    {
+      "epoch": 6.408479412963718,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.947003668976764e-06,
+      "loss": 0.8227,
+      "step": 94320
+    },
+    {
+      "epoch": 6.408819133034379,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.944880418535128e-06,
+      "loss": 0.8961,
+      "step": 94325
+    },
+    {
+      "epoch": 6.409158853105041,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.942757168093492e-06,
+      "loss": 0.8766,
+      "step": 94330
+    },
+    {
+      "epoch": 6.4094985731757035,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.940633917651856e-06,
+      "loss": 0.7992,
+      "step": 94335
+    },
+    {
+      "epoch": 6.409838293246365,
+      "grad_norm": 8.375,
+      "learning_rate": 9.93851066721022e-06,
+      "loss": 0.8503,
+      "step": 94340
+    },
+    {
+      "epoch": 6.410178013317027,
+      "grad_norm": 7.125,
+      "learning_rate": 9.936387416768584e-06,
+      "loss": 0.9628,
+      "step": 94345
+    },
+    {
+      "epoch": 6.410517733387689,
+      "grad_norm": 8.625,
+      "learning_rate": 9.934264166326946e-06,
+      "loss": 0.8454,
+      "step": 94350
+    },
+    {
+      "epoch": 6.41085745345835,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.932140915885312e-06,
+      "loss": 0.8978,
+      "step": 94355
+    },
+    {
+      "epoch": 6.411197173529012,
+      "grad_norm": 9.0,
+      "learning_rate": 9.930017665443676e-06,
+      "loss": 0.8974,
+      "step": 94360
+    },
+    {
+      "epoch": 6.411536893599674,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.927894415002038e-06,
+      "loss": 0.8023,
+      "step": 94365
+    },
+    {
+      "epoch": 6.411876613670335,
+      "grad_norm": 8.125,
+      "learning_rate": 9.925771164560404e-06,
+      "loss": 0.8254,
+      "step": 94370
+    },
+    {
+      "epoch": 6.412216333740997,
+      "grad_norm": 8.375,
+      "learning_rate": 9.923647914118766e-06,
+      "loss": 0.8216,
+      "step": 94375
+    },
+    {
+      "epoch": 6.4125560538116595,
+      "grad_norm": 7.75,
+      "learning_rate": 9.92152466367713e-06,
+      "loss": 0.9402,
+      "step": 94380
+    },
+    {
+      "epoch": 6.412895773882321,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.919401413235494e-06,
+      "loss": 0.8383,
+      "step": 94385
+    },
+    {
+      "epoch": 6.413235493952983,
+      "grad_norm": 7.5,
+      "learning_rate": 9.917278162793858e-06,
+      "loss": 0.7229,
+      "step": 94390
+    },
+    {
+      "epoch": 6.413575214023645,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.915154912352222e-06,
+      "loss": 0.8574,
+      "step": 94395
+    },
+    {
+      "epoch": 6.413914934094306,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.913031661910586e-06,
+      "loss": 0.7849,
+      "step": 94400
+    },
+    {
+      "epoch": 6.414254654164968,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.91090841146895e-06,
+      "loss": 0.8242,
+      "step": 94405
+    },
+    {
+      "epoch": 6.41459437423563,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.908785161027314e-06,
+      "loss": 0.9494,
+      "step": 94410
+    },
+    {
+      "epoch": 6.414934094306291,
+      "grad_norm": 8.5,
+      "learning_rate": 9.906661910585678e-06,
+      "loss": 0.9474,
+      "step": 94415
+    },
+    {
+      "epoch": 6.415273814376953,
+      "grad_norm": 7.25,
+      "learning_rate": 9.904538660144042e-06,
+      "loss": 0.7628,
+      "step": 94420
+    },
+    {
+      "epoch": 6.4156135344476155,
+      "grad_norm": 7.75,
+      "learning_rate": 9.902415409702406e-06,
+      "loss": 0.8183,
+      "step": 94425
+    },
+    {
+      "epoch": 6.415953254518277,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.90029215926077e-06,
+      "loss": 0.9199,
+      "step": 94430
+    },
+    {
+      "epoch": 6.416292974588939,
+      "grad_norm": 8.25,
+      "learning_rate": 9.898168908819132e-06,
+      "loss": 0.8878,
+      "step": 94435
+    },
+    {
+      "epoch": 6.416632694659601,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.896045658377498e-06,
+      "loss": 0.887,
+      "step": 94440
+    },
+    {
+      "epoch": 6.416972414730262,
+      "grad_norm": 8.375,
+      "learning_rate": 9.893922407935862e-06,
+      "loss": 0.8619,
+      "step": 94445
+    },
+    {
+      "epoch": 6.417312134800924,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.891799157494224e-06,
+      "loss": 0.8052,
+      "step": 94450
+    },
+    {
+      "epoch": 6.417651854871586,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.88967590705259e-06,
+      "loss": 0.8991,
+      "step": 94455
+    },
+    {
+      "epoch": 6.417991574942247,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.887552656610952e-06,
+      "loss": 0.868,
+      "step": 94460
+    },
+    {
+      "epoch": 6.418331295012909,
+      "grad_norm": 8.75,
+      "learning_rate": 9.885429406169316e-06,
+      "loss": 0.8581,
+      "step": 94465
+    },
+    {
+      "epoch": 6.4186710150835715,
+      "grad_norm": 6.875,
+      "learning_rate": 9.883306155727682e-06,
+      "loss": 0.8951,
+      "step": 94470
+    },
+    {
+      "epoch": 6.419010735154233,
+      "grad_norm": 7.875,
+      "learning_rate": 9.881182905286044e-06,
+      "loss": 0.8609,
+      "step": 94475
+    },
+    {
+      "epoch": 6.419350455224895,
+      "grad_norm": 6.46875,
+      "learning_rate": 9.879059654844408e-06,
+      "loss": 0.9733,
+      "step": 94480
+    },
+    {
+      "epoch": 6.419690175295557,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.876936404402772e-06,
+      "loss": 0.8124,
+      "step": 94485
+    },
+    {
+      "epoch": 6.420029895366218,
+      "grad_norm": 8.375,
+      "learning_rate": 9.874813153961136e-06,
+      "loss": 0.8721,
+      "step": 94490
+    },
+    {
+      "epoch": 6.42036961543688,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.8726899035195e-06,
+      "loss": 0.6806,
+      "step": 94495
+    },
+    {
+      "epoch": 6.420709335507542,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.870566653077864e-06,
+      "loss": 0.9196,
+      "step": 94500
+    },
+    {
+      "epoch": 6.421049055578203,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.868443402636228e-06,
+      "loss": 0.8156,
+      "step": 94505
+    },
+    {
+      "epoch": 6.4213887756488655,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.866320152194592e-06,
+      "loss": 0.7511,
+      "step": 94510
+    },
+    {
+      "epoch": 6.4217284957195275,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.864196901752956e-06,
+      "loss": 0.9319,
+      "step": 94515
+    },
+    {
+      "epoch": 6.422068215790189,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.86207365131132e-06,
+      "loss": 0.8754,
+      "step": 94520
+    },
+    {
+      "epoch": 6.422407935860851,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.859950400869684e-06,
+      "loss": 0.8,
+      "step": 94525
+    },
+    {
+      "epoch": 6.422747655931513,
+      "grad_norm": 9.25,
+      "learning_rate": 9.857827150428048e-06,
+      "loss": 0.9,
+      "step": 94530
+    },
+    {
+      "epoch": 6.423087376002174,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.85570389998641e-06,
+      "loss": 0.8522,
+      "step": 94535
+    },
+    {
+      "epoch": 6.423427096072836,
+      "grad_norm": 7.75,
+      "learning_rate": 9.853580649544776e-06,
+      "loss": 0.9464,
+      "step": 94540
+    },
+    {
+      "epoch": 6.423766816143498,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.851457399103139e-06,
+      "loss": 0.788,
+      "step": 94545
+    },
+    {
+      "epoch": 6.424106536214159,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.849334148661503e-06,
+      "loss": 0.7964,
+      "step": 94550
+    },
+    {
+      "epoch": 6.4244462562848215,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.847210898219868e-06,
+      "loss": 0.8757,
+      "step": 94555
+    },
+    {
+      "epoch": 6.4247859763554835,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.84508764777823e-06,
+      "loss": 0.9133,
+      "step": 94560
+    },
+    {
+      "epoch": 6.425125696426145,
+      "grad_norm": 8.875,
+      "learning_rate": 9.842964397336595e-06,
+      "loss": 0.8611,
+      "step": 94565
+    },
+    {
+      "epoch": 6.425465416496807,
+      "grad_norm": 9.25,
+      "learning_rate": 9.840841146894959e-06,
+      "loss": 0.7861,
+      "step": 94570
+    },
+    {
+      "epoch": 6.425805136567468,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.838717896453323e-06,
+      "loss": 0.8445,
+      "step": 94575
+    },
+    {
+      "epoch": 6.42614485663813,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.836594646011687e-06,
+      "loss": 0.8037,
+      "step": 94580
+    },
+    {
+      "epoch": 6.426484576708792,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.83447139557005e-06,
+      "loss": 1.0165,
+      "step": 94585
+    },
+    {
+      "epoch": 6.426824296779453,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.832348145128415e-06,
+      "loss": 0.9219,
+      "step": 94590
+    },
+    {
+      "epoch": 6.427164016850115,
+      "grad_norm": 8.875,
+      "learning_rate": 9.830224894686779e-06,
+      "loss": 0.9414,
+      "step": 94595
+    },
+    {
+      "epoch": 6.4275037369207775,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.828101644245143e-06,
+      "loss": 0.8926,
+      "step": 94600
+    },
+    {
+      "epoch": 6.427843456991439,
+      "grad_norm": 8.5,
+      "learning_rate": 9.825978393803507e-06,
+      "loss": 0.9322,
+      "step": 94605
+    },
+    {
+      "epoch": 6.428183177062101,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.82385514336187e-06,
+      "loss": 0.9017,
+      "step": 94610
+    },
+    {
+      "epoch": 6.428522897132763,
+      "grad_norm": 8.0,
+      "learning_rate": 9.821731892920235e-06,
+      "loss": 0.7856,
+      "step": 94615
+    },
+    {
+      "epoch": 6.428862617203424,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.819608642478597e-06,
+      "loss": 0.8432,
+      "step": 94620
+    },
+    {
+      "epoch": 6.429202337274086,
+      "grad_norm": 7.75,
+      "learning_rate": 9.817485392036963e-06,
+      "loss": 0.7859,
+      "step": 94625
+    },
+    {
+      "epoch": 6.429542057344748,
+      "grad_norm": 6.875,
+      "learning_rate": 9.815362141595325e-06,
+      "loss": 0.9272,
+      "step": 94630
+    },
+    {
+      "epoch": 6.429881777415409,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.813238891153689e-06,
+      "loss": 0.7902,
+      "step": 94635
+    },
+    {
+      "epoch": 6.430221497486071,
+      "grad_norm": 8.375,
+      "learning_rate": 9.811115640712055e-06,
+      "loss": 0.8377,
+      "step": 94640
+    },
+    {
+      "epoch": 6.4305612175567335,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.808992390270417e-06,
+      "loss": 0.8905,
+      "step": 94645
+    },
+    {
+      "epoch": 6.430900937627395,
+      "grad_norm": 7.375,
+      "learning_rate": 9.806869139828781e-06,
+      "loss": 0.8822,
+      "step": 94650
+    },
+    {
+      "epoch": 6.431240657698057,
+      "grad_norm": 6.6875,
+      "learning_rate": 9.804745889387145e-06,
+      "loss": 0.7402,
+      "step": 94655
+    },
+    {
+      "epoch": 6.431580377768719,
+      "grad_norm": 8.625,
+      "learning_rate": 9.802622638945509e-06,
+      "loss": 0.7708,
+      "step": 94660
+    },
+    {
+      "epoch": 6.43192009783938,
+      "grad_norm": 7.75,
+      "learning_rate": 9.800499388503873e-06,
+      "loss": 0.8306,
+      "step": 94665
+    },
+    {
+      "epoch": 6.432259817910042,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.798376138062237e-06,
+      "loss": 0.8144,
+      "step": 94670
+    },
+    {
+      "epoch": 6.432599537980704,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.796252887620601e-06,
+      "loss": 0.93,
+      "step": 94675
+    },
+    {
+      "epoch": 6.432939258051365,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.794129637178965e-06,
+      "loss": 0.8592,
+      "step": 94680
+    },
+    {
+      "epoch": 6.433278978122027,
+      "grad_norm": 6.75,
+      "learning_rate": 9.792006386737329e-06,
+      "loss": 0.7625,
+      "step": 94685
+    },
+    {
+      "epoch": 6.4336186981926895,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.789883136295693e-06,
+      "loss": 0.7987,
+      "step": 94690
+    },
+    {
+      "epoch": 6.433958418263351,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.787759885854057e-06,
+      "loss": 0.7799,
+      "step": 94695
+    },
+    {
+      "epoch": 6.434298138334013,
+      "grad_norm": 8.25,
+      "learning_rate": 9.785636635412421e-06,
+      "loss": 0.8501,
+      "step": 94700
+    },
+    {
+      "epoch": 6.434637858404675,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.783513384970783e-06,
+      "loss": 0.8684,
+      "step": 94705
+    },
+    {
+      "epoch": 6.434977578475336,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.78139013452915e-06,
+      "loss": 0.9834,
+      "step": 94710
+    },
+    {
+      "epoch": 6.435317298545998,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.779266884087513e-06,
+      "loss": 0.8673,
+      "step": 94715
+    },
+    {
+      "epoch": 6.43565701861666,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.777143633645875e-06,
+      "loss": 0.8764,
+      "step": 94720
+    },
+    {
+      "epoch": 6.435996738687321,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.775020383204241e-06,
+      "loss": 0.8612,
+      "step": 94725
+    },
+    {
+      "epoch": 6.436336458757983,
+      "grad_norm": 9.0,
+      "learning_rate": 9.772897132762603e-06,
+      "loss": 0.8999,
+      "step": 94730
+    },
+    {
+      "epoch": 6.4366761788286455,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.770773882320967e-06,
+      "loss": 0.8968,
+      "step": 94735
+    },
+    {
+      "epoch": 6.437015898899307,
+      "grad_norm": 9.0,
+      "learning_rate": 9.768650631879331e-06,
+      "loss": 0.8897,
+      "step": 94740
+    },
+    {
+      "epoch": 6.437355618969969,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.766527381437696e-06,
+      "loss": 0.7522,
+      "step": 94745
+    },
+    {
+      "epoch": 6.437695339040631,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.76440413099606e-06,
+      "loss": 0.8754,
+      "step": 94750
+    },
+    {
+      "epoch": 6.438035059111292,
+      "grad_norm": 8.125,
+      "learning_rate": 9.762280880554424e-06,
+      "loss": 0.8296,
+      "step": 94755
+    },
+    {
+      "epoch": 6.438374779181954,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.760157630112788e-06,
+      "loss": 0.7776,
+      "step": 94760
+    },
+    {
+      "epoch": 6.438714499252616,
+      "grad_norm": 8.5,
+      "learning_rate": 9.758034379671152e-06,
+      "loss": 0.8939,
+      "step": 94765
+    },
+    {
+      "epoch": 6.439054219323277,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.755911129229516e-06,
+      "loss": 0.8277,
+      "step": 94770
+    },
+    {
+      "epoch": 6.4393939393939394,
+      "grad_norm": 8.875,
+      "learning_rate": 9.75378787878788e-06,
+      "loss": 0.9363,
+      "step": 94775
+    },
+    {
+      "epoch": 6.4397336594646015,
+      "grad_norm": 7.5,
+      "learning_rate": 9.751664628346244e-06,
+      "loss": 0.8895,
+      "step": 94780
+    },
+    {
+      "epoch": 6.440073379535263,
+      "grad_norm": 8.25,
+      "learning_rate": 9.749541377904608e-06,
+      "loss": 0.8764,
+      "step": 94785
+    },
+    {
+      "epoch": 6.440413099605925,
+      "grad_norm": 6.71875,
+      "learning_rate": 9.74741812746297e-06,
+      "loss": 0.7954,
+      "step": 94790
+    },
+    {
+      "epoch": 6.440752819676587,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.745294877021336e-06,
+      "loss": 0.8214,
+      "step": 94795
+    },
+    {
+      "epoch": 6.441092539747248,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.7431716265797e-06,
+      "loss": 0.7262,
+      "step": 94800
+    },
+    {
+      "epoch": 6.44143225981791,
+      "grad_norm": 8.625,
+      "learning_rate": 9.741048376138062e-06,
+      "loss": 0.7961,
+      "step": 94805
+    },
+    {
+      "epoch": 6.441771979888572,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.738925125696428e-06,
+      "loss": 0.8235,
+      "step": 94810
+    },
+    {
+      "epoch": 6.442111699959233,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.73680187525479e-06,
+      "loss": 0.827,
+      "step": 94815
+    },
+    {
+      "epoch": 6.4424514200298955,
+      "grad_norm": 7.875,
+      "learning_rate": 9.734678624813154e-06,
+      "loss": 0.8542,
+      "step": 94820
+    },
+    {
+      "epoch": 6.4427911401005575,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.732555374371518e-06,
+      "loss": 0.9031,
+      "step": 94825
+    },
+    {
+      "epoch": 6.443130860171219,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.730432123929882e-06,
+      "loss": 0.9155,
+      "step": 94830
+    },
+    {
+      "epoch": 6.443470580241881,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.728308873488246e-06,
+      "loss": 0.8064,
+      "step": 94835
+    },
+    {
+      "epoch": 6.443810300312543,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.72618562304661e-06,
+      "loss": 0.8314,
+      "step": 94840
+    },
+    {
+      "epoch": 6.444150020383204,
+      "grad_norm": 7.625,
+      "learning_rate": 9.724062372604974e-06,
+      "loss": 0.8707,
+      "step": 94845
+    },
+    {
+      "epoch": 6.444489740453866,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.721939122163338e-06,
+      "loss": 0.864,
+      "step": 94850
+    },
+    {
+      "epoch": 6.444829460524527,
+      "grad_norm": 8.125,
+      "learning_rate": 9.719815871721702e-06,
+      "loss": 0.8757,
+      "step": 94855
+    },
+    {
+      "epoch": 6.445169180595189,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.717692621280066e-06,
+      "loss": 0.8668,
+      "step": 94860
+    },
+    {
+      "epoch": 6.4455089006658515,
+      "grad_norm": 6.25,
+      "learning_rate": 9.71556937083843e-06,
+      "loss": 0.82,
+      "step": 94865
+    },
+    {
+      "epoch": 6.445848620736513,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.713446120396794e-06,
+      "loss": 0.8602,
+      "step": 94870
+    },
+    {
+      "epoch": 6.446188340807175,
+      "grad_norm": 8.375,
+      "learning_rate": 9.711322869955156e-06,
+      "loss": 0.8508,
+      "step": 94875
+    },
+    {
+      "epoch": 6.446528060877837,
+      "grad_norm": 8.5,
+      "learning_rate": 9.709199619513522e-06,
+      "loss": 0.8364,
+      "step": 94880
+    },
+    {
+      "epoch": 6.446867780948498,
+      "grad_norm": 8.5,
+      "learning_rate": 9.707076369071886e-06,
+      "loss": 0.8505,
+      "step": 94885
+    },
+    {
+      "epoch": 6.44720750101916,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.704953118630248e-06,
+      "loss": 0.8154,
+      "step": 94890
+    },
+    {
+      "epoch": 6.447547221089822,
+      "grad_norm": 9.3125,
+      "learning_rate": 9.702829868188614e-06,
+      "loss": 0.9365,
+      "step": 94895
+    },
+    {
+      "epoch": 6.447886941160483,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.700706617746976e-06,
+      "loss": 0.8089,
+      "step": 94900
+    },
+    {
+      "epoch": 6.448226661231145,
+      "grad_norm": 7.625,
+      "learning_rate": 9.69858336730534e-06,
+      "loss": 0.8696,
+      "step": 94905
+    },
+    {
+      "epoch": 6.4485663813018075,
+      "grad_norm": 8.125,
+      "learning_rate": 9.696460116863706e-06,
+      "loss": 0.7938,
+      "step": 94910
+    },
+    {
+      "epoch": 6.448906101372469,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.694336866422068e-06,
+      "loss": 0.7996,
+      "step": 94915
+    },
+    {
+      "epoch": 6.449245821443131,
+      "grad_norm": 8.25,
+      "learning_rate": 9.692213615980432e-06,
+      "loss": 0.8938,
+      "step": 94920
+    },
+    {
+      "epoch": 6.449585541513793,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.690090365538796e-06,
+      "loss": 0.9236,
+      "step": 94925
+    },
+    {
+      "epoch": 6.449925261584454,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.68796711509716e-06,
+      "loss": 0.8967,
+      "step": 94930
+    },
+    {
+      "epoch": 6.450264981655116,
+      "grad_norm": 6.65625,
+      "learning_rate": 9.685843864655524e-06,
+      "loss": 0.9069,
+      "step": 94935
+    },
+    {
+      "epoch": 6.450604701725778,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.683720614213888e-06,
+      "loss": 0.842,
+      "step": 94940
+    },
+    {
+      "epoch": 6.450944421796439,
+      "grad_norm": 7.625,
+      "learning_rate": 9.681597363772252e-06,
+      "loss": 0.8173,
+      "step": 94945
+    },
+    {
+      "epoch": 6.451284141867101,
+      "grad_norm": 8.25,
+      "learning_rate": 9.679474113330616e-06,
+      "loss": 0.8833,
+      "step": 94950
+    },
+    {
+      "epoch": 6.4516238619377635,
+      "grad_norm": 9.9375,
+      "learning_rate": 9.67735086288898e-06,
+      "loss": 0.8313,
+      "step": 94955
+    },
+    {
+      "epoch": 6.451963582008425,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.675227612447344e-06,
+      "loss": 0.9301,
+      "step": 94960
+    },
+    {
+      "epoch": 6.452303302079087,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.673104362005708e-06,
+      "loss": 0.9153,
+      "step": 94965
+    },
+    {
+      "epoch": 6.452643022149749,
+      "grad_norm": 8.25,
+      "learning_rate": 9.670981111564072e-06,
+      "loss": 1.0021,
+      "step": 94970
+    },
+    {
+      "epoch": 6.45298274222041,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.668857861122435e-06,
+      "loss": 0.8137,
+      "step": 94975
+    },
+    {
+      "epoch": 6.453322462291072,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.6667346106808e-06,
+      "loss": 0.7721,
+      "step": 94980
+    },
+    {
+      "epoch": 6.453662182361734,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.664611360239163e-06,
+      "loss": 0.8119,
+      "step": 94985
+    },
+    {
+      "epoch": 6.454001902432395,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.662488109797527e-06,
+      "loss": 0.8686,
+      "step": 94990
+    },
+    {
+      "epoch": 6.454341622503057,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.660364859355892e-06,
+      "loss": 0.8902,
+      "step": 94995
+    },
+    {
+      "epoch": 6.4546813425737195,
+      "grad_norm": 8.0,
+      "learning_rate": 9.658241608914255e-06,
+      "loss": 0.8987,
+      "step": 95000
+    },
+    {
+      "epoch": 6.455021062644381,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.656118358472619e-06,
+      "loss": 0.8008,
+      "step": 95005
+    },
+    {
+      "epoch": 6.455360782715043,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.653995108030983e-06,
+      "loss": 0.889,
+      "step": 95010
+    },
+    {
+      "epoch": 6.455700502785705,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.651871857589347e-06,
+      "loss": 0.8272,
+      "step": 95015
+    },
+    {
+      "epoch": 6.456040222856366,
+      "grad_norm": 8.5,
+      "learning_rate": 9.64974860714771e-06,
+      "loss": 0.8392,
+      "step": 95020
+    },
+    {
+      "epoch": 6.456379942927028,
+      "grad_norm": 8.375,
+      "learning_rate": 9.647625356706075e-06,
+      "loss": 0.8783,
+      "step": 95025
+    },
+    {
+      "epoch": 6.45671966299769,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.645502106264439e-06,
+      "loss": 0.7613,
+      "step": 95030
+    },
+    {
+      "epoch": 6.457059383068351,
+      "grad_norm": 8.625,
+      "learning_rate": 9.643378855822803e-06,
+      "loss": 0.7859,
+      "step": 95035
+    },
+    {
+      "epoch": 6.457399103139013,
+      "grad_norm": 8.75,
+      "learning_rate": 9.641255605381167e-06,
+      "loss": 0.8477,
+      "step": 95040
+    },
+    {
+      "epoch": 6.4577388232096755,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.63913235493953e-06,
+      "loss": 0.8225,
+      "step": 95045
+    },
+    {
+      "epoch": 6.458078543280337,
+      "grad_norm": 8.625,
+      "learning_rate": 9.637009104497895e-06,
+      "loss": 1.0415,
+      "step": 95050
+    },
+    {
+      "epoch": 6.458418263350999,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.634885854056259e-06,
+      "loss": 0.763,
+      "step": 95055
+    },
+    {
+      "epoch": 6.458757983421661,
+      "grad_norm": 7.5,
+      "learning_rate": 9.632762603614621e-06,
+      "loss": 0.8043,
+      "step": 95060
+    },
+    {
+      "epoch": 6.459097703492322,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.630639353172987e-06,
+      "loss": 0.8401,
+      "step": 95065
+    },
+    {
+      "epoch": 6.459437423562984,
+      "grad_norm": 8.25,
+      "learning_rate": 9.628516102731349e-06,
+      "loss": 0.8236,
+      "step": 95070
+    },
+    {
+      "epoch": 6.459777143633646,
+      "grad_norm": 8.625,
+      "learning_rate": 9.626392852289713e-06,
+      "loss": 0.9023,
+      "step": 95075
+    },
+    {
+      "epoch": 6.460116863704307,
+      "grad_norm": 7.75,
+      "learning_rate": 9.624269601848079e-06,
+      "loss": 0.8788,
+      "step": 95080
+    },
+    {
+      "epoch": 6.4604565837749695,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.622146351406441e-06,
+      "loss": 0.7988,
+      "step": 95085
+    },
+    {
+      "epoch": 6.4607963038456315,
+      "grad_norm": 8.5,
+      "learning_rate": 9.620023100964805e-06,
+      "loss": 0.9106,
+      "step": 95090
+    },
+    {
+      "epoch": 6.461136023916293,
+      "grad_norm": 7.21875,
+      "learning_rate": 9.617899850523169e-06,
+      "loss": 0.8986,
+      "step": 95095
+    },
+    {
+      "epoch": 6.461475743986955,
+      "grad_norm": 8.625,
+      "learning_rate": 9.615776600081533e-06,
+      "loss": 0.899,
+      "step": 95100
+    },
+    {
+      "epoch": 6.461815464057617,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.613653349639897e-06,
+      "loss": 0.7755,
+      "step": 95105
+    },
+    {
+      "epoch": 6.462155184128278,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.611530099198261e-06,
+      "loss": 0.7477,
+      "step": 95110
+    },
+    {
+      "epoch": 6.46249490419894,
+      "grad_norm": 7.1875,
+      "learning_rate": 9.609406848756625e-06,
+      "loss": 0.872,
+      "step": 95115
+    },
+    {
+      "epoch": 6.462834624269602,
+      "grad_norm": 8.125,
+      "learning_rate": 9.607283598314989e-06,
+      "loss": 0.8431,
+      "step": 95120
+    },
+    {
+      "epoch": 6.463174344340263,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.605160347873353e-06,
+      "loss": 0.9384,
+      "step": 95125
+    },
+    {
+      "epoch": 6.4635140644109255,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.603037097431717e-06,
+      "loss": 0.8301,
+      "step": 95130
+    },
+    {
+      "epoch": 6.4638537844815875,
+      "grad_norm": 8.875,
+      "learning_rate": 9.600913846990081e-06,
+      "loss": 0.857,
+      "step": 95135
+    },
+    {
+      "epoch": 6.464193504552249,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.598790596548445e-06,
+      "loss": 0.8094,
+      "step": 95140
+    },
+    {
+      "epoch": 6.464533224622911,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.596667346106807e-06,
+      "loss": 0.7632,
+      "step": 95145
+    },
+    {
+      "epoch": 6.464872944693573,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.594544095665173e-06,
+      "loss": 0.9071,
+      "step": 95150
+    },
+    {
+      "epoch": 6.465212664764234,
+      "grad_norm": 7.375,
+      "learning_rate": 9.592420845223535e-06,
+      "loss": 0.7438,
+      "step": 95155
+    },
+    {
+      "epoch": 6.465552384834896,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.5902975947819e-06,
+      "loss": 0.8727,
+      "step": 95160
+    },
+    {
+      "epoch": 6.465892104905558,
+      "grad_norm": 8.75,
+      "learning_rate": 9.588174344340265e-06,
+      "loss": 0.7753,
+      "step": 95165
+    },
+    {
+      "epoch": 6.466231824976219,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.586051093898627e-06,
+      "loss": 0.7402,
+      "step": 95170
+    },
+    {
+      "epoch": 6.4665715450468815,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.583927843456991e-06,
+      "loss": 0.922,
+      "step": 95175
+    },
+    {
+      "epoch": 6.4669112651175436,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.581804593015355e-06,
+      "loss": 0.7459,
+      "step": 95180
+    },
+    {
+      "epoch": 6.467250985188205,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.57968134257372e-06,
+      "loss": 0.8368,
+      "step": 95185
+    },
+    {
+      "epoch": 6.467590705258867,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.577558092132085e-06,
+      "loss": 0.7947,
+      "step": 95190
+    },
+    {
+      "epoch": 6.467930425329529,
+      "grad_norm": 8.25,
+      "learning_rate": 9.575434841690447e-06,
+      "loss": 0.8166,
+      "step": 95195
+    },
+    {
+      "epoch": 6.46827014540019,
+      "grad_norm": 8.375,
+      "learning_rate": 9.573311591248811e-06,
+      "loss": 0.8475,
+      "step": 95200
+    },
+    {
+      "epoch": 6.468609865470852,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.571188340807175e-06,
+      "loss": 0.8272,
+      "step": 95205
+    },
+    {
+      "epoch": 6.468949585541514,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.56906509036554e-06,
+      "loss": 0.81,
+      "step": 95210
+    },
+    {
+      "epoch": 6.469289305612175,
+      "grad_norm": 8.75,
+      "learning_rate": 9.566941839923903e-06,
+      "loss": 0.8203,
+      "step": 95215
+    },
+    {
+      "epoch": 6.4696290256828375,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.564818589482267e-06,
+      "loss": 0.7977,
+      "step": 95220
+    },
+    {
+      "epoch": 6.4699687457535,
+      "grad_norm": 8.125,
+      "learning_rate": 9.562695339040631e-06,
+      "loss": 0.8559,
+      "step": 95225
+    },
+    {
+      "epoch": 6.470308465824161,
+      "grad_norm": 8.0,
+      "learning_rate": 9.560572088598994e-06,
+      "loss": 0.8126,
+      "step": 95230
+    },
+    {
+      "epoch": 6.470648185894823,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.55844883815736e-06,
+      "loss": 1.0108,
+      "step": 95235
+    },
+    {
+      "epoch": 6.470987905965485,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.556325587715723e-06,
+      "loss": 0.923,
+      "step": 95240
+    },
+    {
+      "epoch": 6.471327626036146,
+      "grad_norm": 9.125,
+      "learning_rate": 9.554202337274086e-06,
+      "loss": 0.8779,
+      "step": 95245
+    },
+    {
+      "epoch": 6.471667346106808,
+      "grad_norm": 8.5,
+      "learning_rate": 9.552079086832451e-06,
+      "loss": 0.8585,
+      "step": 95250
+    },
+    {
+      "epoch": 6.47200706617747,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.549955836390814e-06,
+      "loss": 0.7821,
+      "step": 95255
+    },
+    {
+      "epoch": 6.472346786248131,
+      "grad_norm": 8.625,
+      "learning_rate": 9.547832585949178e-06,
+      "loss": 0.814,
+      "step": 95260
+    },
+    {
+      "epoch": 6.4726865063187935,
+      "grad_norm": 8.375,
+      "learning_rate": 9.545709335507542e-06,
+      "loss": 0.8964,
+      "step": 95265
+    },
+    {
+      "epoch": 6.473026226389455,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.543586085065906e-06,
+      "loss": 0.844,
+      "step": 95270
+    },
+    {
+      "epoch": 6.473365946460117,
+      "grad_norm": 8.25,
+      "learning_rate": 9.541462834624271e-06,
+      "loss": 0.8437,
+      "step": 95275
+    },
+    {
+      "epoch": 6.473705666530779,
+      "grad_norm": 8.0,
+      "learning_rate": 9.539339584182634e-06,
+      "loss": 0.8572,
+      "step": 95280
+    },
+    {
+      "epoch": 6.47404538660144,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.537216333740998e-06,
+      "loss": 0.8536,
+      "step": 95285
+    },
+    {
+      "epoch": 6.474385106672102,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.535093083299362e-06,
+      "loss": 0.8378,
+      "step": 95290
+    },
+    {
+      "epoch": 6.474724826742764,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.532969832857726e-06,
+      "loss": 0.7523,
+      "step": 95295
+    },
+    {
+      "epoch": 6.475064546813425,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.53084658241609e-06,
+      "loss": 0.8956,
+      "step": 95300
+    },
+    {
+      "epoch": 6.475404266884087,
+      "grad_norm": 7.3125,
+      "learning_rate": 9.528723331974454e-06,
+      "loss": 0.8406,
+      "step": 95305
+    },
+    {
+      "epoch": 6.4757439869547495,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.526600081532818e-06,
+      "loss": 0.9309,
+      "step": 95310
+    },
+    {
+      "epoch": 6.476083707025411,
+      "grad_norm": 8.625,
+      "learning_rate": 9.52447683109118e-06,
+      "loss": 0.8707,
+      "step": 95315
+    },
+    {
+      "epoch": 6.476423427096073,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.522353580649546e-06,
+      "loss": 0.7956,
+      "step": 95320
+    },
+    {
+      "epoch": 6.476763147166735,
+      "grad_norm": 8.25,
+      "learning_rate": 9.52023033020791e-06,
+      "loss": 0.8736,
+      "step": 95325
+    },
+    {
+      "epoch": 6.477102867237396,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.518107079766272e-06,
+      "loss": 0.9375,
+      "step": 95330
+    },
+    {
+      "epoch": 6.477442587308058,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.515983829324638e-06,
+      "loss": 0.9344,
+      "step": 95335
+    },
+    {
+      "epoch": 6.47778230737872,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.513860578883e-06,
+      "loss": 0.828,
+      "step": 95340
+    },
+    {
+      "epoch": 6.478122027449381,
+      "grad_norm": 7.625,
+      "learning_rate": 9.511737328441364e-06,
+      "loss": 0.8754,
+      "step": 95345
+    },
+    {
+      "epoch": 6.4784617475200434,
+      "grad_norm": 8.5,
+      "learning_rate": 9.509614077999728e-06,
+      "loss": 0.8322,
+      "step": 95350
+    },
+    {
+      "epoch": 6.4788014675907055,
+      "grad_norm": 7.875,
+      "learning_rate": 9.507490827558092e-06,
+      "loss": 0.876,
+      "step": 95355
+    },
+    {
+      "epoch": 6.479141187661367,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.505367577116458e-06,
+      "loss": 0.9997,
+      "step": 95360
+    },
+    {
+      "epoch": 6.479480907732029,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.50324432667482e-06,
+      "loss": 0.8238,
+      "step": 95365
+    },
+    {
+      "epoch": 6.479820627802691,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.501121076233184e-06,
+      "loss": 0.8752,
+      "step": 95370
+    },
+    {
+      "epoch": 6.480160347873352,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.498997825791548e-06,
+      "loss": 0.7848,
+      "step": 95375
+    },
+    {
+      "epoch": 6.480500067944014,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.496874575349912e-06,
+      "loss": 0.8812,
+      "step": 95380
+    },
+    {
+      "epoch": 6.480839788014676,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.494751324908276e-06,
+      "loss": 0.8423,
+      "step": 95385
+    },
+    {
+      "epoch": 6.481179508085337,
+      "grad_norm": 8.875,
+      "learning_rate": 9.49262807446664e-06,
+      "loss": 0.8887,
+      "step": 95390
+    },
+    {
+      "epoch": 6.4815192281559995,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.490504824025004e-06,
+      "loss": 0.9175,
+      "step": 95395
+    },
+    {
+      "epoch": 6.4818589482266615,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.488381573583366e-06,
+      "loss": 0.7738,
+      "step": 95400
+    },
+    {
+      "epoch": 6.482198668297323,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.486258323141732e-06,
+      "loss": 0.9098,
+      "step": 95405
+    },
+    {
+      "epoch": 6.482538388367985,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.484135072700096e-06,
+      "loss": 0.7897,
+      "step": 95410
+    },
+    {
+      "epoch": 6.482878108438647,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.482011822258458e-06,
+      "loss": 0.8301,
+      "step": 95415
+    },
+    {
+      "epoch": 6.483217828509308,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.479888571816824e-06,
+      "loss": 0.8379,
+      "step": 95420
+    },
+    {
+      "epoch": 6.48355754857997,
+      "grad_norm": 8.625,
+      "learning_rate": 9.477765321375186e-06,
+      "loss": 0.9151,
+      "step": 95425
+    },
+    {
+      "epoch": 6.483897268650632,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.47564207093355e-06,
+      "loss": 0.8232,
+      "step": 95430
+    },
+    {
+      "epoch": 6.484236988721293,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.473518820491916e-06,
+      "loss": 0.7771,
+      "step": 95435
+    },
+    {
+      "epoch": 6.4845767087919555,
+      "grad_norm": 8.75,
+      "learning_rate": 9.471395570050278e-06,
+      "loss": 0.857,
+      "step": 95440
+    },
+    {
+      "epoch": 6.4849164288626175,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.469272319608644e-06,
+      "loss": 0.7771,
+      "step": 95445
+    },
+    {
+      "epoch": 6.485256148933279,
+      "grad_norm": 8.25,
+      "learning_rate": 9.467149069167006e-06,
+      "loss": 0.7812,
+      "step": 95450
+    },
+    {
+      "epoch": 6.485595869003941,
+      "grad_norm": 8.25,
+      "learning_rate": 9.46502581872537e-06,
+      "loss": 0.8609,
+      "step": 95455
+    },
+    {
+      "epoch": 6.485935589074603,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.462902568283735e-06,
+      "loss": 0.7724,
+      "step": 95460
+    },
+    {
+      "epoch": 6.486275309145264,
+      "grad_norm": 8.625,
+      "learning_rate": 9.460779317842099e-06,
+      "loss": 0.906,
+      "step": 95465
+    },
+    {
+      "epoch": 6.486615029215926,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.458656067400463e-06,
+      "loss": 0.7881,
+      "step": 95470
+    },
+    {
+      "epoch": 6.486954749286588,
+      "grad_norm": 7.375,
+      "learning_rate": 9.456532816958827e-06,
+      "loss": 0.8351,
+      "step": 95475
+    },
+    {
+      "epoch": 6.487294469357249,
+      "grad_norm": 7.125,
+      "learning_rate": 9.45440956651719e-06,
+      "loss": 0.8257,
+      "step": 95480
+    },
+    {
+      "epoch": 6.4876341894279115,
+      "grad_norm": 7.5,
+      "learning_rate": 9.452286316075555e-06,
+      "loss": 0.8523,
+      "step": 95485
+    },
+    {
+      "epoch": 6.4879739094985736,
+      "grad_norm": 9.5,
+      "learning_rate": 9.450163065633919e-06,
+      "loss": 0.7324,
+      "step": 95490
+    },
+    {
+      "epoch": 6.488313629569235,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.448039815192283e-06,
+      "loss": 0.8925,
+      "step": 95495
+    },
+    {
+      "epoch": 6.488653349639897,
+      "grad_norm": 8.625,
+      "learning_rate": 9.445916564750645e-06,
+      "loss": 0.7617,
+      "step": 95500
+    },
+    {
+      "epoch": 6.488993069710559,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.44379331430901e-06,
+      "loss": 0.716,
+      "step": 95505
+    },
+    {
+      "epoch": 6.48933278978122,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.441670063867373e-06,
+      "loss": 0.9319,
+      "step": 95510
+    },
+    {
+      "epoch": 6.489672509851882,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.439546813425737e-06,
+      "loss": 0.8555,
+      "step": 95515
+    },
+    {
+      "epoch": 6.490012229922544,
+      "grad_norm": 7.75,
+      "learning_rate": 9.437423562984103e-06,
+      "loss": 0.9692,
+      "step": 95520
+    },
+    {
+      "epoch": 6.490351949993205,
+      "grad_norm": 8.0,
+      "learning_rate": 9.435300312542465e-06,
+      "loss": 0.8137,
+      "step": 95525
+    },
+    {
+      "epoch": 6.4906916700638675,
+      "grad_norm": 8.5,
+      "learning_rate": 9.43317706210083e-06,
+      "loss": 0.8344,
+      "step": 95530
+    },
+    {
+      "epoch": 6.491031390134529,
+      "grad_norm": 8.375,
+      "learning_rate": 9.431053811659193e-06,
+      "loss": 0.8408,
+      "step": 95535
+    },
+    {
+      "epoch": 6.491371110205191,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.428930561217557e-06,
+      "loss": 0.832,
+      "step": 95540
+    },
+    {
+      "epoch": 6.491710830275853,
+      "grad_norm": 7.3125,
+      "learning_rate": 9.426807310775921e-06,
+      "loss": 0.8537,
+      "step": 95545
+    },
+    {
+      "epoch": 6.492050550346514,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.424684060334285e-06,
+      "loss": 0.85,
+      "step": 95550
+    },
+    {
+      "epoch": 6.492390270417176,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.422560809892649e-06,
+      "loss": 0.793,
+      "step": 95555
+    },
+    {
+      "epoch": 6.492729990487838,
+      "grad_norm": 6.96875,
+      "learning_rate": 9.420437559451013e-06,
+      "loss": 0.9238,
+      "step": 95560
+    },
+    {
+      "epoch": 6.493069710558499,
+      "grad_norm": 7.0,
+      "learning_rate": 9.418314309009377e-06,
+      "loss": 0.807,
+      "step": 95565
+    },
+    {
+      "epoch": 6.493409430629161,
+      "grad_norm": 8.0,
+      "learning_rate": 9.416191058567741e-06,
+      "loss": 0.8985,
+      "step": 95570
+    },
+    {
+      "epoch": 6.4937491506998235,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.414067808126105e-06,
+      "loss": 0.9574,
+      "step": 95575
+    },
+    {
+      "epoch": 6.494088870770485,
+      "grad_norm": 7.15625,
+      "learning_rate": 9.411944557684469e-06,
+      "loss": 0.8657,
+      "step": 95580
+    },
+    {
+      "epoch": 6.494428590841147,
+      "grad_norm": 7.375,
+      "learning_rate": 9.409821307242831e-06,
+      "loss": 0.7908,
+      "step": 95585
+    },
+    {
+      "epoch": 6.494768310911809,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.407698056801197e-06,
+      "loss": 0.7978,
+      "step": 95590
+    },
+    {
+      "epoch": 6.49510803098247,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.40557480635956e-06,
+      "loss": 0.8457,
+      "step": 95595
+    },
+    {
+      "epoch": 6.495447751053132,
+      "grad_norm": 7.75,
+      "learning_rate": 9.403451555917923e-06,
+      "loss": 0.8445,
+      "step": 95600
+    },
+    {
+      "epoch": 6.495787471123794,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.401328305476289e-06,
+      "loss": 0.8251,
+      "step": 95605
+    },
+    {
+      "epoch": 6.496127191194455,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.399205055034651e-06,
+      "loss": 0.9332,
+      "step": 95610
+    },
+    {
+      "epoch": 6.496466911265117,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.397081804593017e-06,
+      "loss": 0.855,
+      "step": 95615
+    },
+    {
+      "epoch": 6.4968066313357795,
+      "grad_norm": 7.21875,
+      "learning_rate": 9.39495855415138e-06,
+      "loss": 0.8641,
+      "step": 95620
+    },
+    {
+      "epoch": 6.497146351406441,
+      "grad_norm": 8.0,
+      "learning_rate": 9.392835303709743e-06,
+      "loss": 0.808,
+      "step": 95625
+    },
+    {
+      "epoch": 6.497486071477103,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.390712053268109e-06,
+      "loss": 0.8359,
+      "step": 95630
+    },
+    {
+      "epoch": 6.497825791547765,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.388588802826471e-06,
+      "loss": 0.8996,
+      "step": 95635
+    },
+    {
+      "epoch": 6.498165511618426,
+      "grad_norm": 6.15625,
+      "learning_rate": 9.386465552384835e-06,
+      "loss": 0.8693,
+      "step": 95640
+    },
+    {
+      "epoch": 6.498505231689088,
+      "grad_norm": 8.375,
+      "learning_rate": 9.3843423019432e-06,
+      "loss": 0.9402,
+      "step": 95645
+    },
+    {
+      "epoch": 6.49884495175975,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.382219051501563e-06,
+      "loss": 0.8843,
+      "step": 95650
+    },
+    {
+      "epoch": 6.499184671830411,
+      "grad_norm": 7.75,
+      "learning_rate": 9.380095801059927e-06,
+      "loss": 0.7854,
+      "step": 95655
+    },
+    {
+      "epoch": 6.4995243919010735,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.377972550618291e-06,
+      "loss": 0.7859,
+      "step": 95660
+    },
+    {
+      "epoch": 6.4998641119717355,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.375849300176655e-06,
+      "loss": 0.8832,
+      "step": 95665
+    },
+    {
+      "epoch": 6.500203832042397,
+      "grad_norm": 8.375,
+      "learning_rate": 9.373726049735018e-06,
+      "loss": 0.7984,
+      "step": 95670
+    },
+    {
+      "epoch": 6.500543552113059,
+      "grad_norm": 8.125,
+      "learning_rate": 9.371602799293383e-06,
+      "loss": 0.8606,
+      "step": 95675
+    },
+    {
+      "epoch": 6.500883272183721,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.369479548851747e-06,
+      "loss": 0.8071,
+      "step": 95680
+    },
+    {
+      "epoch": 6.501222992254382,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.36735629841011e-06,
+      "loss": 0.7617,
+      "step": 95685
+    },
+    {
+      "epoch": 6.501562712325044,
+      "grad_norm": 8.9375,
+      "learning_rate": 9.365233047968475e-06,
+      "loss": 0.9633,
+      "step": 95690
+    },
+    {
+      "epoch": 6.501902432395706,
+      "grad_norm": 7.40625,
+      "learning_rate": 9.363109797526838e-06,
+      "loss": 0.8552,
+      "step": 95695
+    },
+    {
+      "epoch": 6.502242152466367,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.360986547085203e-06,
+      "loss": 0.803,
+      "step": 95700
+    },
+    {
+      "epoch": 6.5025818725370295,
+      "grad_norm": 8.375,
+      "learning_rate": 9.358863296643566e-06,
+      "loss": 0.8232,
+      "step": 95705
+    },
+    {
+      "epoch": 6.5029215926076915,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.35674004620193e-06,
+      "loss": 0.961,
+      "step": 95710
+    },
+    {
+      "epoch": 6.503261312678353,
+      "grad_norm": 7.125,
+      "learning_rate": 9.354616795760295e-06,
+      "loss": 0.7504,
+      "step": 95715
+    },
+    {
+      "epoch": 6.503601032749015,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.352493545318658e-06,
+      "loss": 0.8532,
+      "step": 95720
+    },
+    {
+      "epoch": 6.503940752819677,
+      "grad_norm": 8.5,
+      "learning_rate": 9.350370294877022e-06,
+      "loss": 0.8795,
+      "step": 95725
+    },
+    {
+      "epoch": 6.504280472890338,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.348247044435386e-06,
+      "loss": 0.8489,
+      "step": 95730
+    },
+    {
+      "epoch": 6.504620192961,
+      "grad_norm": 8.875,
+      "learning_rate": 9.34612379399375e-06,
+      "loss": 0.9646,
+      "step": 95735
+    },
+    {
+      "epoch": 6.504959913031662,
+      "grad_norm": 9.0,
+      "learning_rate": 9.344000543552114e-06,
+      "loss": 0.8344,
+      "step": 95740
+    },
+    {
+      "epoch": 6.505299633102323,
+      "grad_norm": 8.125,
+      "learning_rate": 9.341877293110478e-06,
+      "loss": 0.8918,
+      "step": 95745
+    },
+    {
+      "epoch": 6.5056393531729855,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.339754042668842e-06,
+      "loss": 0.8544,
+      "step": 95750
+    },
+    {
+      "epoch": 6.5059790732436475,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.337630792227204e-06,
+      "loss": 0.7904,
+      "step": 95755
+    },
+    {
+      "epoch": 6.506318793314309,
+      "grad_norm": 8.125,
+      "learning_rate": 9.33550754178557e-06,
+      "loss": 0.8164,
+      "step": 95760
+    },
+    {
+      "epoch": 6.506658513384971,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.333384291343934e-06,
+      "loss": 0.856,
+      "step": 95765
+    },
+    {
+      "epoch": 6.506998233455633,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.331261040902296e-06,
+      "loss": 0.7231,
+      "step": 95770
+    },
+    {
+      "epoch": 6.507337953526294,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.329137790460662e-06,
+      "loss": 0.8168,
+      "step": 95775
+    },
+    {
+      "epoch": 6.507677673596956,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.327014540019024e-06,
+      "loss": 0.8406,
+      "step": 95780
+    },
+    {
+      "epoch": 6.508017393667618,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.32489128957739e-06,
+      "loss": 0.8711,
+      "step": 95785
+    },
+    {
+      "epoch": 6.508357113738279,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.322768039135752e-06,
+      "loss": 0.8371,
+      "step": 95790
+    },
+    {
+      "epoch": 6.5086968338089415,
+      "grad_norm": 8.625,
+      "learning_rate": 9.320644788694116e-06,
+      "loss": 0.8562,
+      "step": 95795
+    },
+    {
+      "epoch": 6.509036553879604,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.318521538252482e-06,
+      "loss": 0.8117,
+      "step": 95800
+    },
+    {
+      "epoch": 6.509376273950265,
+      "grad_norm": 8.625,
+      "learning_rate": 9.316398287810844e-06,
+      "loss": 0.878,
+      "step": 95805
+    },
+    {
+      "epoch": 6.509715994020927,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.314275037369208e-06,
+      "loss": 0.7877,
+      "step": 95810
+    },
+    {
+      "epoch": 6.510055714091589,
+      "grad_norm": 8.375,
+      "learning_rate": 9.312151786927572e-06,
+      "loss": 0.8499,
+      "step": 95815
+    },
+    {
+      "epoch": 6.51039543416225,
+      "grad_norm": 8.0,
+      "learning_rate": 9.310028536485936e-06,
+      "loss": 0.7373,
+      "step": 95820
+    },
+    {
+      "epoch": 6.510735154232912,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.3079052860443e-06,
+      "loss": 0.8197,
+      "step": 95825
+    },
+    {
+      "epoch": 6.511074874303574,
+      "grad_norm": 8.375,
+      "learning_rate": 9.305782035602664e-06,
+      "loss": 0.8776,
+      "step": 95830
+    },
+    {
+      "epoch": 6.511414594374235,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.303658785161028e-06,
+      "loss": 0.7937,
+      "step": 95835
+    },
+    {
+      "epoch": 6.5117543144448975,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.30153553471939e-06,
+      "loss": 0.8212,
+      "step": 95840
+    },
+    {
+      "epoch": 6.51209403451556,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.299412284277756e-06,
+      "loss": 0.8624,
+      "step": 95845
+    },
+    {
+      "epoch": 6.512433754586221,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.29728903383612e-06,
+      "loss": 0.9307,
+      "step": 95850
+    },
+    {
+      "epoch": 6.512773474656883,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.295165783394482e-06,
+      "loss": 0.8126,
+      "step": 95855
+    },
+    {
+      "epoch": 6.513113194727545,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.293042532952848e-06,
+      "loss": 0.8662,
+      "step": 95860
+    },
+    {
+      "epoch": 6.513452914798206,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.29091928251121e-06,
+      "loss": 0.7963,
+      "step": 95865
+    },
+    {
+      "epoch": 6.513792634868868,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.288796032069576e-06,
+      "loss": 0.7741,
+      "step": 95870
+    },
+    {
+      "epoch": 6.51413235493953,
+      "grad_norm": 7.75,
+      "learning_rate": 9.286672781627938e-06,
+      "loss": 0.7563,
+      "step": 95875
+    },
+    {
+      "epoch": 6.514472075010191,
+      "grad_norm": 7.0,
+      "learning_rate": 9.284549531186302e-06,
+      "loss": 0.7817,
+      "step": 95880
+    },
+    {
+      "epoch": 6.5148117950808535,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.282426280744668e-06,
+      "loss": 0.8638,
+      "step": 95885
+    },
+    {
+      "epoch": 6.515151515151516,
+      "grad_norm": 8.25,
+      "learning_rate": 9.28030303030303e-06,
+      "loss": 0.9527,
+      "step": 95890
+    },
+    {
+      "epoch": 6.515491235222177,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.278179779861394e-06,
+      "loss": 0.8443,
+      "step": 95895
+    },
+    {
+      "epoch": 6.515830955292839,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.276056529419758e-06,
+      "loss": 0.9209,
+      "step": 95900
+    },
+    {
+      "epoch": 6.516170675363501,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.273933278978122e-06,
+      "loss": 0.7302,
+      "step": 95905
+    },
+    {
+      "epoch": 6.516510395434162,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.271810028536486e-06,
+      "loss": 0.8119,
+      "step": 95910
+    },
+    {
+      "epoch": 6.516850115504824,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.26968677809485e-06,
+      "loss": 0.8573,
+      "step": 95915
+    },
+    {
+      "epoch": 6.517189835575486,
+      "grad_norm": 8.375,
+      "learning_rate": 9.267563527653214e-06,
+      "loss": 0.9052,
+      "step": 95920
+    },
+    {
+      "epoch": 6.517529555646147,
+      "grad_norm": 8.375,
+      "learning_rate": 9.265440277211577e-06,
+      "loss": 0.8182,
+      "step": 95925
+    },
+    {
+      "epoch": 6.5178692757168095,
+      "grad_norm": 8.5,
+      "learning_rate": 9.263317026769942e-06,
+      "loss": 0.8172,
+      "step": 95930
+    },
+    {
+      "epoch": 6.518208995787472,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.261193776328306e-06,
+      "loss": 0.8131,
+      "step": 95935
+    },
+    {
+      "epoch": 6.518548715858133,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.259070525886669e-06,
+      "loss": 0.858,
+      "step": 95940
+    },
+    {
+      "epoch": 6.518888435928795,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.256947275445034e-06,
+      "loss": 0.8281,
+      "step": 95945
+    },
+    {
+      "epoch": 6.519228155999457,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.254824025003397e-06,
+      "loss": 0.8572,
+      "step": 95950
+    },
+    {
+      "epoch": 6.519567876070118,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.252700774561762e-06,
+      "loss": 0.8481,
+      "step": 95955
+    },
+    {
+      "epoch": 6.51990759614078,
+      "grad_norm": 6.4375,
+      "learning_rate": 9.250577524120126e-06,
+      "loss": 0.8575,
+      "step": 95960
+    },
+    {
+      "epoch": 6.520247316211442,
+      "grad_norm": 8.25,
+      "learning_rate": 9.248454273678489e-06,
+      "loss": 0.7685,
+      "step": 95965
+    },
+    {
+      "epoch": 6.5205870362821035,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.246331023236854e-06,
+      "loss": 0.7705,
+      "step": 95970
+    },
+    {
+      "epoch": 6.5209267563527655,
+      "grad_norm": 8.25,
+      "learning_rate": 9.244207772795217e-06,
+      "loss": 0.8045,
+      "step": 95975
+    },
+    {
+      "epoch": 6.521266476423427,
+      "grad_norm": 6.78125,
+      "learning_rate": 9.24208452235358e-06,
+      "loss": 0.931,
+      "step": 95980
+    },
+    {
+      "epoch": 6.521606196494089,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.239961271911945e-06,
+      "loss": 0.9279,
+      "step": 95985
+    },
+    {
+      "epoch": 6.521945916564751,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.237838021470309e-06,
+      "loss": 0.878,
+      "step": 95990
+    },
+    {
+      "epoch": 6.522285636635412,
+      "grad_norm": 6.9375,
+      "learning_rate": 9.235714771028673e-06,
+      "loss": 0.7803,
+      "step": 95995
+    },
+    {
+      "epoch": 6.522625356706074,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.233591520587037e-06,
+      "loss": 0.8552,
+      "step": 96000
+    },
+    {
+      "epoch": 6.522965076776736,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.2314682701454e-06,
+      "loss": 0.8658,
+      "step": 96005
+    },
+    {
+      "epoch": 6.523304796847397,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.229345019703765e-06,
+      "loss": 0.9211,
+      "step": 96010
+    },
+    {
+      "epoch": 6.5236445169180595,
+      "grad_norm": 8.375,
+      "learning_rate": 9.227221769262129e-06,
+      "loss": 0.8182,
+      "step": 96015
+    },
+    {
+      "epoch": 6.5239842369887215,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.225098518820493e-06,
+      "loss": 0.8654,
+      "step": 96020
+    },
+    {
+      "epoch": 6.524323957059383,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.222975268378855e-06,
+      "loss": 0.8383,
+      "step": 96025
+    },
+    {
+      "epoch": 6.524663677130045,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.22085201793722e-06,
+      "loss": 0.9149,
+      "step": 96030
+    },
+    {
+      "epoch": 6.525003397200707,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.218728767495583e-06,
+      "loss": 0.8397,
+      "step": 96035
+    },
+    {
+      "epoch": 6.525343117271368,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.216605517053949e-06,
+      "loss": 0.7957,
+      "step": 96040
+    },
+    {
+      "epoch": 6.52568283734203,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.214482266612313e-06,
+      "loss": 0.8562,
+      "step": 96045
+    },
+    {
+      "epoch": 6.526022557412692,
+      "grad_norm": 8.875,
+      "learning_rate": 9.212359016170675e-06,
+      "loss": 0.8822,
+      "step": 96050
+    },
+    {
+      "epoch": 6.526362277483353,
+      "grad_norm": 8.25,
+      "learning_rate": 9.21023576572904e-06,
+      "loss": 0.8645,
+      "step": 96055
+    },
+    {
+      "epoch": 6.5267019975540155,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.208112515287403e-06,
+      "loss": 0.9057,
+      "step": 96060
+    },
+    {
+      "epoch": 6.5270417176246776,
+      "grad_norm": 8.25,
+      "learning_rate": 9.205989264845767e-06,
+      "loss": 0.7918,
+      "step": 96065
+    },
+    {
+      "epoch": 6.527381437695339,
+      "grad_norm": 8.0,
+      "learning_rate": 9.203866014404131e-06,
+      "loss": 0.8322,
+      "step": 96070
+    },
+    {
+      "epoch": 6.527721157766001,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.201742763962495e-06,
+      "loss": 0.8724,
+      "step": 96075
+    },
+    {
+      "epoch": 6.528060877836663,
+      "grad_norm": 8.75,
+      "learning_rate": 9.199619513520859e-06,
+      "loss": 0.872,
+      "step": 96080
+    },
+    {
+      "epoch": 6.528400597907324,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.197496263079223e-06,
+      "loss": 0.8895,
+      "step": 96085
+    },
+    {
+      "epoch": 6.528740317977986,
+      "grad_norm": 8.625,
+      "learning_rate": 9.195373012637587e-06,
+      "loss": 0.7752,
+      "step": 96090
+    },
+    {
+      "epoch": 6.529080038048648,
+      "grad_norm": 8.375,
+      "learning_rate": 9.193249762195951e-06,
+      "loss": 0.9067,
+      "step": 96095
+    },
+    {
+      "epoch": 6.529419758119309,
+      "grad_norm": 7.21875,
+      "learning_rate": 9.191126511754315e-06,
+      "loss": 0.8733,
+      "step": 96100
+    },
+    {
+      "epoch": 6.5297594781899715,
+      "grad_norm": 7.875,
+      "learning_rate": 9.189003261312679e-06,
+      "loss": 0.8688,
+      "step": 96105
+    },
+    {
+      "epoch": 6.530099198260634,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.186880010871041e-06,
+      "loss": 0.8343,
+      "step": 96110
+    },
+    {
+      "epoch": 6.530438918331295,
+      "grad_norm": 8.125,
+      "learning_rate": 9.184756760429407e-06,
+      "loss": 0.8958,
+      "step": 96115
+    },
+    {
+      "epoch": 6.530778638401957,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.18263350998777e-06,
+      "loss": 0.9079,
+      "step": 96120
+    },
+    {
+      "epoch": 6.531118358472619,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.180510259546135e-06,
+      "loss": 0.8025,
+      "step": 96125
+    },
+    {
+      "epoch": 6.53145807854328,
+      "grad_norm": 9.25,
+      "learning_rate": 9.1783870091045e-06,
+      "loss": 0.9099,
+      "step": 96130
+    },
+    {
+      "epoch": 6.531797798613942,
+      "grad_norm": 7.5,
+      "learning_rate": 9.176263758662861e-06,
+      "loss": 0.7639,
+      "step": 96135
+    },
+    {
+      "epoch": 6.532137518684604,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.174140508221227e-06,
+      "loss": 0.7053,
+      "step": 96140
+    },
+    {
+      "epoch": 6.532477238755265,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.17201725777959e-06,
+      "loss": 0.8145,
+      "step": 96145
+    },
+    {
+      "epoch": 6.5328169588259275,
+      "grad_norm": 7.75,
+      "learning_rate": 9.169894007337953e-06,
+      "loss": 0.8806,
+      "step": 96150
+    },
+    {
+      "epoch": 6.53315667889659,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.16777075689632e-06,
+      "loss": 0.8154,
+      "step": 96155
+    },
+    {
+      "epoch": 6.533496398967251,
+      "grad_norm": 7.5,
+      "learning_rate": 9.165647506454681e-06,
+      "loss": 0.8199,
+      "step": 96160
+    },
+    {
+      "epoch": 6.533836119037913,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.163524256013045e-06,
+      "loss": 0.8982,
+      "step": 96165
+    },
+    {
+      "epoch": 6.534175839108575,
+      "grad_norm": 8.375,
+      "learning_rate": 9.16140100557141e-06,
+      "loss": 0.9193,
+      "step": 96170
+    },
+    {
+      "epoch": 6.534515559179236,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.159277755129774e-06,
+      "loss": 0.9106,
+      "step": 96175
+    },
+    {
+      "epoch": 6.534855279249898,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.157154504688138e-06,
+      "loss": 0.8339,
+      "step": 96180
+    },
+    {
+      "epoch": 6.535194999320559,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.155031254246502e-06,
+      "loss": 0.861,
+      "step": 96185
+    },
+    {
+      "epoch": 6.535534719391221,
+      "grad_norm": 8.5,
+      "learning_rate": 9.152908003804866e-06,
+      "loss": 0.8074,
+      "step": 96190
+    },
+    {
+      "epoch": 6.5358744394618835,
+      "grad_norm": 7.125,
+      "learning_rate": 9.150784753363228e-06,
+      "loss": 0.7526,
+      "step": 96195
+    },
+    {
+      "epoch": 6.536214159532545,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.148661502921594e-06,
+      "loss": 0.7346,
+      "step": 96200
+    },
+    {
+      "epoch": 6.536553879603207,
+      "grad_norm": 8.0,
+      "learning_rate": 9.146538252479958e-06,
+      "loss": 0.8865,
+      "step": 96205
+    },
+    {
+      "epoch": 6.536893599673869,
+      "grad_norm": 8.125,
+      "learning_rate": 9.144415002038322e-06,
+      "loss": 0.9586,
+      "step": 96210
+    },
+    {
+      "epoch": 6.53723331974453,
+      "grad_norm": 9.0,
+      "learning_rate": 9.142291751596686e-06,
+      "loss": 0.8766,
+      "step": 96215
+    },
+    {
+      "epoch": 6.537573039815192,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.140168501155048e-06,
+      "loss": 0.8385,
+      "step": 96220
+    },
+    {
+      "epoch": 6.537912759885854,
+      "grad_norm": 8.5,
+      "learning_rate": 9.138045250713414e-06,
+      "loss": 0.778,
+      "step": 96225
+    },
+    {
+      "epoch": 6.538252479956515,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.135922000271776e-06,
+      "loss": 0.8215,
+      "step": 96230
+    },
+    {
+      "epoch": 6.5385922000271774,
+      "grad_norm": 7.34375,
+      "learning_rate": 9.13379874983014e-06,
+      "loss": 0.7239,
+      "step": 96235
+    },
+    {
+      "epoch": 6.5389319200978395,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.131675499388506e-06,
+      "loss": 0.8646,
+      "step": 96240
+    },
+    {
+      "epoch": 6.539271640168501,
+      "grad_norm": 8.5,
+      "learning_rate": 9.129552248946868e-06,
+      "loss": 0.8652,
+      "step": 96245
+    },
+    {
+      "epoch": 6.539611360239163,
+      "grad_norm": 7.46875,
+      "learning_rate": 9.127428998505232e-06,
+      "loss": 0.8186,
+      "step": 96250
+    },
+    {
+      "epoch": 6.539951080309825,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.125305748063596e-06,
+      "loss": 0.8819,
+      "step": 96255
+    },
+    {
+      "epoch": 6.540290800380486,
+      "grad_norm": 8.125,
+      "learning_rate": 9.12318249762196e-06,
+      "loss": 0.8171,
+      "step": 96260
+    },
+    {
+      "epoch": 6.540630520451148,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.121059247180324e-06,
+      "loss": 0.8823,
+      "step": 96265
+    },
+    {
+      "epoch": 6.54097024052181,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.118935996738688e-06,
+      "loss": 0.8569,
+      "step": 96270
+    },
+    {
+      "epoch": 6.541309960592471,
+      "grad_norm": 7.65625,
+      "learning_rate": 9.116812746297052e-06,
+      "loss": 0.8653,
+      "step": 96275
+    },
+    {
+      "epoch": 6.5416496806631335,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.114689495855414e-06,
+      "loss": 0.7637,
+      "step": 96280
+    },
+    {
+      "epoch": 6.5419894007337955,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.11256624541378e-06,
+      "loss": 0.7733,
+      "step": 96285
+    },
+    {
+      "epoch": 6.542329120804457,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.110442994972144e-06,
+      "loss": 0.8738,
+      "step": 96290
+    },
+    {
+      "epoch": 6.542668840875119,
+      "grad_norm": 8.625,
+      "learning_rate": 9.108319744530508e-06,
+      "loss": 0.8867,
+      "step": 96295
+    },
+    {
+      "epoch": 6.543008560945781,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.106196494088872e-06,
+      "loss": 0.8544,
+      "step": 96300
+    },
+    {
+      "epoch": 6.543348281016442,
+      "grad_norm": 8.5,
+      "learning_rate": 9.104073243647234e-06,
+      "loss": 0.8685,
+      "step": 96305
+    },
+    {
+      "epoch": 6.543688001087104,
+      "grad_norm": 8.375,
+      "learning_rate": 9.1019499932056e-06,
+      "loss": 0.883,
+      "step": 96310
+    },
+    {
+      "epoch": 6.544027721157766,
+      "grad_norm": 8.625,
+      "learning_rate": 9.099826742763962e-06,
+      "loss": 0.9524,
+      "step": 96315
+    },
+    {
+      "epoch": 6.544367441228427,
+      "grad_norm": 8.375,
+      "learning_rate": 9.097703492322326e-06,
+      "loss": 0.7827,
+      "step": 96320
+    },
+    {
+      "epoch": 6.5447071612990895,
+      "grad_norm": 7.125,
+      "learning_rate": 9.095580241880692e-06,
+      "loss": 0.8039,
+      "step": 96325
+    },
+    {
+      "epoch": 6.5450468813697515,
+      "grad_norm": 7.0,
+      "learning_rate": 9.093456991439054e-06,
+      "loss": 0.825,
+      "step": 96330
+    },
+    {
+      "epoch": 6.545386601440413,
+      "grad_norm": 7.53125,
+      "learning_rate": 9.091333740997418e-06,
+      "loss": 0.8113,
+      "step": 96335
+    },
+    {
+      "epoch": 6.545726321511075,
+      "grad_norm": 8.5,
+      "learning_rate": 9.089210490555782e-06,
+      "loss": 0.9274,
+      "step": 96340
+    },
+    {
+      "epoch": 6.546066041581737,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.087087240114146e-06,
+      "loss": 0.8485,
+      "step": 96345
+    },
+    {
+      "epoch": 6.546405761652398,
+      "grad_norm": 8.25,
+      "learning_rate": 9.08496398967251e-06,
+      "loss": 0.7789,
+      "step": 96350
+    },
+    {
+      "epoch": 6.54674548172306,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.082840739230874e-06,
+      "loss": 0.7836,
+      "step": 96355
+    },
+    {
+      "epoch": 6.547085201793722,
+      "grad_norm": 6.625,
+      "learning_rate": 9.080717488789238e-06,
+      "loss": 0.8624,
+      "step": 96360
+    },
+    {
+      "epoch": 6.547424921864383,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.0785942383476e-06,
+      "loss": 0.8833,
+      "step": 96365
+    },
+    {
+      "epoch": 6.5477646419350455,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.076470987905966e-06,
+      "loss": 0.8907,
+      "step": 96370
+    },
+    {
+      "epoch": 6.548104362005708,
+      "grad_norm": 7.125,
+      "learning_rate": 9.07434773746433e-06,
+      "loss": 0.8146,
+      "step": 96375
+    },
+    {
+      "epoch": 6.548444082076369,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.072224487022694e-06,
+      "loss": 0.8786,
+      "step": 96380
+    },
+    {
+      "epoch": 6.548783802147031,
+      "grad_norm": 8.125,
+      "learning_rate": 9.070101236581058e-06,
+      "loss": 0.8093,
+      "step": 96385
+    },
+    {
+      "epoch": 6.549123522217693,
+      "grad_norm": 8.5625,
+      "learning_rate": 9.06797798613942e-06,
+      "loss": 0.9191,
+      "step": 96390
+    },
+    {
+      "epoch": 6.549463242288354,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.065854735697786e-06,
+      "loss": 0.8938,
+      "step": 96395
+    },
+    {
+      "epoch": 6.549802962359016,
+      "grad_norm": 7.6875,
+      "learning_rate": 9.06373148525615e-06,
+      "loss": 0.7798,
+      "step": 96400
+    },
+    {
+      "epoch": 6.550142682429678,
+      "grad_norm": 8.25,
+      "learning_rate": 9.061608234814513e-06,
+      "loss": 0.8206,
+      "step": 96405
+    },
+    {
+      "epoch": 6.550482402500339,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.059484984372878e-06,
+      "loss": 0.7777,
+      "step": 96410
+    },
+    {
+      "epoch": 6.5508221225710015,
+      "grad_norm": 9.0625,
+      "learning_rate": 9.05736173393124e-06,
+      "loss": 0.9402,
+      "step": 96415
+    },
+    {
+      "epoch": 6.551161842641664,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.055238483489605e-06,
+      "loss": 0.8258,
+      "step": 96420
+    },
+    {
+      "epoch": 6.551501562712325,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.053115233047969e-06,
+      "loss": 0.822,
+      "step": 96425
+    },
+    {
+      "epoch": 6.551841282782987,
+      "grad_norm": 9.5,
+      "learning_rate": 9.050991982606333e-06,
+      "loss": 0.8455,
+      "step": 96430
+    },
+    {
+      "epoch": 6.552181002853649,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.048868732164697e-06,
+      "loss": 0.8403,
+      "step": 96435
+    },
+    {
+      "epoch": 6.55252072292431,
+      "grad_norm": 7.40625,
+      "learning_rate": 9.04674548172306e-06,
+      "loss": 0.769,
+      "step": 96440
+    },
+    {
+      "epoch": 6.552860442994972,
+      "grad_norm": 7.625,
+      "learning_rate": 9.044622231281425e-06,
+      "loss": 0.8045,
+      "step": 96445
+    },
+    {
+      "epoch": 6.553200163065634,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.042498980839789e-06,
+      "loss": 0.8398,
+      "step": 96450
+    },
+    {
+      "epoch": 6.553539883136295,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.040375730398153e-06,
+      "loss": 0.796,
+      "step": 96455
+    },
+    {
+      "epoch": 6.5538796032069575,
+      "grad_norm": 7.09375,
+      "learning_rate": 9.038252479956517e-06,
+      "loss": 0.876,
+      "step": 96460
+    },
+    {
+      "epoch": 6.55421932327762,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.03612922951488e-06,
+      "loss": 0.8209,
+      "step": 96465
+    },
+    {
+      "epoch": 6.554559043348281,
+      "grad_norm": 7.0625,
+      "learning_rate": 9.034005979073245e-06,
+      "loss": 0.8998,
+      "step": 96470
+    },
+    {
+      "epoch": 6.554898763418943,
+      "grad_norm": 7.28125,
+      "learning_rate": 9.031882728631607e-06,
+      "loss": 0.8784,
+      "step": 96475
+    },
+    {
+      "epoch": 6.555238483489605,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.029759478189973e-06,
+      "loss": 0.8585,
+      "step": 96480
+    },
+    {
+      "epoch": 6.555578203560266,
+      "grad_norm": 8.375,
+      "learning_rate": 9.027636227748337e-06,
+      "loss": 0.8172,
+      "step": 96485
+    },
+    {
+      "epoch": 6.555917923630928,
+      "grad_norm": 8.5,
+      "learning_rate": 9.025512977306699e-06,
+      "loss": 0.8872,
+      "step": 96490
+    },
+    {
+      "epoch": 6.55625764370159,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.023389726865065e-06,
+      "loss": 0.7853,
+      "step": 96495
+    },
+    {
+      "epoch": 6.556597363772251,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.021266476423427e-06,
+      "loss": 0.8677,
+      "step": 96500
+    },
+    {
+      "epoch": 6.5569370838429135,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.019143225981791e-06,
+      "loss": 0.9081,
+      "step": 96505
+    },
+    {
+      "epoch": 6.557276803913576,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.017019975540155e-06,
+      "loss": 0.866,
+      "step": 96510
+    },
+    {
+      "epoch": 6.557616523984237,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.014896725098519e-06,
+      "loss": 0.886,
+      "step": 96515
+    },
+    {
+      "epoch": 6.557956244054899,
+      "grad_norm": 7.875,
+      "learning_rate": 9.012773474656883e-06,
+      "loss": 0.8459,
+      "step": 96520
+    },
+    {
+      "epoch": 6.558295964125561,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.010650224215247e-06,
+      "loss": 0.9143,
+      "step": 96525
+    },
+    {
+      "epoch": 6.558635684196222,
+      "grad_norm": 7.9375,
+      "learning_rate": 9.008526973773611e-06,
+      "loss": 0.822,
+      "step": 96530
+    },
+    {
+      "epoch": 6.558975404266884,
+      "grad_norm": 8.0,
+      "learning_rate": 9.006403723331975e-06,
+      "loss": 0.8205,
+      "step": 96535
+    },
+    {
+      "epoch": 6.559315124337546,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.004280472890339e-06,
+      "loss": 0.8317,
+      "step": 96540
+    },
+    {
+      "epoch": 6.5596548444082075,
+      "grad_norm": 7.78125,
+      "learning_rate": 9.002157222448703e-06,
+      "loss": 0.9018,
+      "step": 96545
+    },
+    {
+      "epoch": 6.5599945644788695,
+      "grad_norm": 8.125,
+      "learning_rate": 9.000033972007067e-06,
+      "loss": 0.8817,
+      "step": 96550
+    },
+    {
+      "epoch": 6.560334284549532,
+      "grad_norm": 6.6875,
+      "learning_rate": 8.997910721565431e-06,
+      "loss": 0.8819,
+      "step": 96555
+    },
+    {
+      "epoch": 6.560674004620193,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.995787471123793e-06,
+      "loss": 0.8983,
+      "step": 96560
+    },
+    {
+      "epoch": 6.561013724690855,
+      "grad_norm": 9.0,
+      "learning_rate": 8.993664220682159e-06,
+      "loss": 0.8491,
+      "step": 96565
+    },
+    {
+      "epoch": 6.561353444761517,
+      "grad_norm": 7.1875,
+      "learning_rate": 8.991540970240523e-06,
+      "loss": 0.8663,
+      "step": 96570
+    },
+    {
+      "epoch": 6.561693164832178,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.989417719798885e-06,
+      "loss": 0.7902,
+      "step": 96575
+    },
+    {
+      "epoch": 6.56203288490284,
+      "grad_norm": 8.25,
+      "learning_rate": 8.987294469357251e-06,
+      "loss": 0.9588,
+      "step": 96580
+    },
+    {
+      "epoch": 6.562372604973502,
+      "grad_norm": 8.25,
+      "learning_rate": 8.985171218915613e-06,
+      "loss": 0.8291,
+      "step": 96585
+    },
+    {
+      "epoch": 6.5627123250441635,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.983047968473977e-06,
+      "loss": 0.8658,
+      "step": 96590
+    },
+    {
+      "epoch": 6.5630520451148255,
+      "grad_norm": 7.375,
+      "learning_rate": 8.980924718032341e-06,
+      "loss": 0.7708,
+      "step": 96595
+    },
+    {
+      "epoch": 6.563391765185488,
+      "grad_norm": 8.9375,
+      "learning_rate": 8.978801467590705e-06,
+      "loss": 0.811,
+      "step": 96600
+    },
+    {
+      "epoch": 6.563731485256149,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.97667821714907e-06,
+      "loss": 0.8291,
+      "step": 96605
+    },
+    {
+      "epoch": 6.564071205326811,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.974554966707433e-06,
+      "loss": 0.934,
+      "step": 96610
+    },
+    {
+      "epoch": 6.564410925397473,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.972431716265797e-06,
+      "loss": 0.8644,
+      "step": 96615
+    },
+    {
+      "epoch": 6.564750645468134,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.970308465824161e-06,
+      "loss": 0.7261,
+      "step": 96620
+    },
+    {
+      "epoch": 6.565090365538796,
+      "grad_norm": 8.9375,
+      "learning_rate": 8.968185215382525e-06,
+      "loss": 0.818,
+      "step": 96625
+    },
+    {
+      "epoch": 6.565430085609458,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.96606196494089e-06,
+      "loss": 0.8272,
+      "step": 96630
+    },
+    {
+      "epoch": 6.5657698056801195,
+      "grad_norm": 8.125,
+      "learning_rate": 8.963938714499253e-06,
+      "loss": 0.8401,
+      "step": 96635
+    },
+    {
+      "epoch": 6.5661095257507816,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.961815464057617e-06,
+      "loss": 0.7753,
+      "step": 96640
+    },
+    {
+      "epoch": 6.566449245821444,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.95969221361598e-06,
+      "loss": 0.8088,
+      "step": 96645
+    },
+    {
+      "epoch": 6.566788965892105,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.957568963174345e-06,
+      "loss": 0.9544,
+      "step": 96650
+    },
+    {
+      "epoch": 6.567128685962767,
+      "grad_norm": 7.75,
+      "learning_rate": 8.95544571273271e-06,
+      "loss": 0.8921,
+      "step": 96655
+    },
+    {
+      "epoch": 6.567468406033428,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.953322462291072e-06,
+      "loss": 0.9151,
+      "step": 96660
+    },
+    {
+      "epoch": 6.56780812610409,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.951199211849437e-06,
+      "loss": 0.8216,
+      "step": 96665
+    },
+    {
+      "epoch": 6.568147846174752,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.9490759614078e-06,
+      "loss": 0.8137,
+      "step": 96670
+    },
+    {
+      "epoch": 6.568487566245413,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.946952710966164e-06,
+      "loss": 0.8312,
+      "step": 96675
+    },
+    {
+      "epoch": 6.5688272863160755,
+      "grad_norm": 8.25,
+      "learning_rate": 8.94482946052453e-06,
+      "loss": 0.8599,
+      "step": 96680
+    },
+    {
+      "epoch": 6.569167006386738,
+      "grad_norm": 8.875,
+      "learning_rate": 8.942706210082892e-06,
+      "loss": 0.8246,
+      "step": 96685
+    },
+    {
+      "epoch": 6.569506726457399,
+      "grad_norm": 7.875,
+      "learning_rate": 8.940582959641256e-06,
+      "loss": 0.7718,
+      "step": 96690
+    },
+    {
+      "epoch": 6.569846446528061,
+      "grad_norm": 8.9375,
+      "learning_rate": 8.93845970919962e-06,
+      "loss": 0.8761,
+      "step": 96695
+    },
+    {
+      "epoch": 6.570186166598723,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.936336458757984e-06,
+      "loss": 0.8665,
+      "step": 96700
+    },
+    {
+      "epoch": 6.570525886669384,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.934213208316348e-06,
+      "loss": 0.7248,
+      "step": 96705
+    },
+    {
+      "epoch": 6.570865606740046,
+      "grad_norm": 8.5,
+      "learning_rate": 8.932089957874712e-06,
+      "loss": 0.7706,
+      "step": 96710
+    },
+    {
+      "epoch": 6.571205326810708,
+      "grad_norm": 8.875,
+      "learning_rate": 8.929966707433076e-06,
+      "loss": 0.8404,
+      "step": 96715
+    },
+    {
+      "epoch": 6.571545046881369,
+      "grad_norm": 7.75,
+      "learning_rate": 8.92784345699144e-06,
+      "loss": 0.8213,
+      "step": 96720
+    },
+    {
+      "epoch": 6.5718847669520315,
+      "grad_norm": 8.25,
+      "learning_rate": 8.925720206549804e-06,
+      "loss": 0.8086,
+      "step": 96725
+    },
+    {
+      "epoch": 6.572224487022694,
+      "grad_norm": 7.375,
+      "learning_rate": 8.923596956108168e-06,
+      "loss": 0.7965,
+      "step": 96730
+    },
+    {
+      "epoch": 6.572564207093355,
+      "grad_norm": 9.125,
+      "learning_rate": 8.921473705666532e-06,
+      "loss": 0.8216,
+      "step": 96735
+    },
+    {
+      "epoch": 6.572903927164017,
+      "grad_norm": 7.53125,
+      "learning_rate": 8.919350455224896e-06,
+      "loss": 0.8691,
+      "step": 96740
+    },
+    {
+      "epoch": 6.573243647234679,
+      "grad_norm": 7.75,
+      "learning_rate": 8.917227204783258e-06,
+      "loss": 0.7726,
+      "step": 96745
+    },
+    {
+      "epoch": 6.57358336730534,
+      "grad_norm": 7.625,
+      "learning_rate": 8.915103954341624e-06,
+      "loss": 0.84,
+      "step": 96750
+    },
+    {
+      "epoch": 6.573923087376002,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.912980703899986e-06,
+      "loss": 0.8107,
+      "step": 96755
+    },
+    {
+      "epoch": 6.574262807446664,
+      "grad_norm": 7.875,
+      "learning_rate": 8.91085745345835e-06,
+      "loss": 0.7784,
+      "step": 96760
+    },
+    {
+      "epoch": 6.574602527517325,
+      "grad_norm": 8.875,
+      "learning_rate": 8.908734203016716e-06,
+      "loss": 0.8528,
+      "step": 96765
+    },
+    {
+      "epoch": 6.5749422475879875,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.906610952575078e-06,
+      "loss": 0.7615,
+      "step": 96770
+    },
+    {
+      "epoch": 6.57528196765865,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.904487702133442e-06,
+      "loss": 0.8898,
+      "step": 96775
+    },
+    {
+      "epoch": 6.575621687729311,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.902364451691806e-06,
+      "loss": 0.8264,
+      "step": 96780
+    },
+    {
+      "epoch": 6.575961407799973,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.90024120125017e-06,
+      "loss": 0.9009,
+      "step": 96785
+    },
+    {
+      "epoch": 6.576301127870635,
+      "grad_norm": 7.75,
+      "learning_rate": 8.898117950808534e-06,
+      "loss": 0.82,
+      "step": 96790
+    },
+    {
+      "epoch": 6.576640847941296,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.895994700366898e-06,
+      "loss": 0.8383,
+      "step": 96795
+    },
+    {
+      "epoch": 6.576980568011958,
+      "grad_norm": 8.375,
+      "learning_rate": 8.893871449925262e-06,
+      "loss": 0.8938,
+      "step": 96800
+    },
+    {
+      "epoch": 6.57732028808262,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.891748199483626e-06,
+      "loss": 0.8462,
+      "step": 96805
+    },
+    {
+      "epoch": 6.5776600081532814,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.88962494904199e-06,
+      "loss": 0.9761,
+      "step": 96810
+    },
+    {
+      "epoch": 6.5779997282239435,
+      "grad_norm": 8.125,
+      "learning_rate": 8.887501698600354e-06,
+      "loss": 0.9174,
+      "step": 96815
+    },
+    {
+      "epoch": 6.578339448294606,
+      "grad_norm": 7.625,
+      "learning_rate": 8.885378448158718e-06,
+      "loss": 0.8577,
+      "step": 96820
+    },
+    {
+      "epoch": 6.578679168365267,
+      "grad_norm": 7.75,
+      "learning_rate": 8.883255197717082e-06,
+      "loss": 0.8107,
+      "step": 96825
+    },
+    {
+      "epoch": 6.579018888435929,
+      "grad_norm": 8.125,
+      "learning_rate": 8.881131947275444e-06,
+      "loss": 0.7576,
+      "step": 96830
+    },
+    {
+      "epoch": 6.579358608506591,
+      "grad_norm": 8.0,
+      "learning_rate": 8.87900869683381e-06,
+      "loss": 0.866,
+      "step": 96835
+    },
+    {
+      "epoch": 6.579698328577252,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.876885446392172e-06,
+      "loss": 0.846,
+      "step": 96840
+    },
+    {
+      "epoch": 6.580038048647914,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.874762195950536e-06,
+      "loss": 0.872,
+      "step": 96845
+    },
+    {
+      "epoch": 6.580377768718576,
+      "grad_norm": 7.625,
+      "learning_rate": 8.872638945508902e-06,
+      "loss": 0.8328,
+      "step": 96850
+    },
+    {
+      "epoch": 6.5807174887892375,
+      "grad_norm": 8.375,
+      "learning_rate": 8.870515695067264e-06,
+      "loss": 0.8079,
+      "step": 96855
+    },
+    {
+      "epoch": 6.5810572088598995,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.868392444625628e-06,
+      "loss": 0.8629,
+      "step": 96860
+    },
+    {
+      "epoch": 6.581396928930561,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.866269194183992e-06,
+      "loss": 0.9207,
+      "step": 96865
+    },
+    {
+      "epoch": 6.581736649001223,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.864145943742356e-06,
+      "loss": 0.8392,
+      "step": 96870
+    },
+    {
+      "epoch": 6.582076369071885,
+      "grad_norm": 8.0,
+      "learning_rate": 8.86202269330072e-06,
+      "loss": 0.9742,
+      "step": 96875
+    },
+    {
+      "epoch": 6.582416089142546,
+      "grad_norm": 8.5,
+      "learning_rate": 8.859899442859084e-06,
+      "loss": 0.8024,
+      "step": 96880
+    },
+    {
+      "epoch": 6.582755809213208,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.857776192417448e-06,
+      "loss": 0.8944,
+      "step": 96885
+    },
+    {
+      "epoch": 6.58309552928387,
+      "grad_norm": 7.21875,
+      "learning_rate": 8.855652941975813e-06,
+      "loss": 0.7924,
+      "step": 96890
+    },
+    {
+      "epoch": 6.583435249354531,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.853529691534177e-06,
+      "loss": 0.7792,
+      "step": 96895
+    },
+    {
+      "epoch": 6.5837749694251935,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.85140644109254e-06,
+      "loss": 0.9081,
+      "step": 96900
+    },
+    {
+      "epoch": 6.5841146894958555,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.849283190650905e-06,
+      "loss": 0.8322,
+      "step": 96905
+    },
+    {
+      "epoch": 6.584454409566517,
+      "grad_norm": 6.96875,
+      "learning_rate": 8.847159940209269e-06,
+      "loss": 0.7703,
+      "step": 96910
+    },
+    {
+      "epoch": 6.584794129637179,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.84503668976763e-06,
+      "loss": 0.7851,
+      "step": 96915
+    },
+    {
+      "epoch": 6.585133849707841,
+      "grad_norm": 8.125,
+      "learning_rate": 8.842913439325997e-06,
+      "loss": 0.8111,
+      "step": 96920
+    },
+    {
+      "epoch": 6.585473569778502,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.84079018888436e-06,
+      "loss": 0.8164,
+      "step": 96925
+    },
+    {
+      "epoch": 6.585813289849164,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.838666938442723e-06,
+      "loss": 0.9061,
+      "step": 96930
+    },
+    {
+      "epoch": 6.586153009919826,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.836543688001089e-06,
+      "loss": 0.827,
+      "step": 96935
+    },
+    {
+      "epoch": 6.586492729990487,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.83442043755945e-06,
+      "loss": 0.8381,
+      "step": 96940
+    },
+    {
+      "epoch": 6.5868324500611495,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.832297187117815e-06,
+      "loss": 0.837,
+      "step": 96945
+    },
+    {
+      "epoch": 6.5871721701318116,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.830173936676179e-06,
+      "loss": 0.8116,
+      "step": 96950
+    },
+    {
+      "epoch": 6.587511890202473,
+      "grad_norm": 9.125,
+      "learning_rate": 8.828050686234543e-06,
+      "loss": 0.8227,
+      "step": 96955
+    },
+    {
+      "epoch": 6.587851610273135,
+      "grad_norm": 8.0,
+      "learning_rate": 8.825927435792907e-06,
+      "loss": 0.9126,
+      "step": 96960
+    },
+    {
+      "epoch": 6.588191330343797,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.823804185351271e-06,
+      "loss": 0.8299,
+      "step": 96965
+    },
+    {
+      "epoch": 6.588531050414458,
+      "grad_norm": 8.375,
+      "learning_rate": 8.821680934909635e-06,
+      "loss": 0.9839,
+      "step": 96970
+    },
+    {
+      "epoch": 6.58887077048512,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.819557684467999e-06,
+      "loss": 0.8181,
+      "step": 96975
+    },
+    {
+      "epoch": 6.589210490555782,
+      "grad_norm": 8.75,
+      "learning_rate": 8.817434434026363e-06,
+      "loss": 0.9357,
+      "step": 96980
+    },
+    {
+      "epoch": 6.589550210626443,
+      "grad_norm": 8.625,
+      "learning_rate": 8.815311183584727e-06,
+      "loss": 0.9157,
+      "step": 96985
+    },
+    {
+      "epoch": 6.5898899306971055,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.813187933143091e-06,
+      "loss": 0.7799,
+      "step": 96990
+    },
+    {
+      "epoch": 6.590229650767768,
+      "grad_norm": 8.75,
+      "learning_rate": 8.811064682701455e-06,
+      "loss": 0.8454,
+      "step": 96995
+    },
+    {
+      "epoch": 6.590569370838429,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.808941432259817e-06,
+      "loss": 0.9362,
+      "step": 97000
+    },
+    {
+      "epoch": 6.590909090909091,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.806818181818183e-06,
+      "loss": 0.8421,
+      "step": 97005
+    },
+    {
+      "epoch": 6.591248810979753,
+      "grad_norm": 8.125,
+      "learning_rate": 8.804694931376547e-06,
+      "loss": 0.7804,
+      "step": 97010
+    },
+    {
+      "epoch": 6.591588531050414,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.80257168093491e-06,
+      "loss": 0.7946,
+      "step": 97015
+    },
+    {
+      "epoch": 6.591928251121076,
+      "grad_norm": 7.5,
+      "learning_rate": 8.800448430493275e-06,
+      "loss": 0.8562,
+      "step": 97020
+    },
+    {
+      "epoch": 6.592267971191738,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.798325180051637e-06,
+      "loss": 0.7913,
+      "step": 97025
+    },
+    {
+      "epoch": 6.592607691262399,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.796201929610001e-06,
+      "loss": 0.8788,
+      "step": 97030
+    },
+    {
+      "epoch": 6.5929474113330615,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.794078679168365e-06,
+      "loss": 0.719,
+      "step": 97035
+    },
+    {
+      "epoch": 6.593287131403724,
+      "grad_norm": 8.125,
+      "learning_rate": 8.79195542872673e-06,
+      "loss": 0.8359,
+      "step": 97040
+    },
+    {
+      "epoch": 6.593626851474385,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.789832178285093e-06,
+      "loss": 0.8392,
+      "step": 97045
+    },
+    {
+      "epoch": 6.593966571545047,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.787708927843457e-06,
+      "loss": 0.8616,
+      "step": 97050
+    },
+    {
+      "epoch": 6.594306291615709,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.785585677401821e-06,
+      "loss": 0.7842,
+      "step": 97055
+    },
+    {
+      "epoch": 6.59464601168637,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.783462426960185e-06,
+      "loss": 0.7586,
+      "step": 97060
+    },
+    {
+      "epoch": 6.594985731757032,
+      "grad_norm": 7.40625,
+      "learning_rate": 8.78133917651855e-06,
+      "loss": 0.7605,
+      "step": 97065
+    },
+    {
+      "epoch": 6.595325451827694,
+      "grad_norm": 7.375,
+      "learning_rate": 8.779215926076913e-06,
+      "loss": 0.8354,
+      "step": 97070
+    },
+    {
+      "epoch": 6.595665171898355,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.777092675635277e-06,
+      "loss": 0.8462,
+      "step": 97075
+    },
+    {
+      "epoch": 6.5960048919690175,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.774969425193641e-06,
+      "loss": 0.8131,
+      "step": 97080
+    },
+    {
+      "epoch": 6.59634461203968,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.772846174752004e-06,
+      "loss": 0.9211,
+      "step": 97085
+    },
+    {
+      "epoch": 6.596684332110341,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.77072292431037e-06,
+      "loss": 0.8879,
+      "step": 97090
+    },
+    {
+      "epoch": 6.597024052181003,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.768599673868733e-06,
+      "loss": 0.829,
+      "step": 97095
+    },
+    {
+      "epoch": 6.597363772251665,
+      "grad_norm": 6.96875,
+      "learning_rate": 8.766476423427096e-06,
+      "loss": 0.8419,
+      "step": 97100
+    },
+    {
+      "epoch": 6.597703492322326,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.764353172985461e-06,
+      "loss": 0.9011,
+      "step": 97105
+    },
+    {
+      "epoch": 6.598043212392988,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.762229922543824e-06,
+      "loss": 0.8882,
+      "step": 97110
+    },
+    {
+      "epoch": 6.59838293246365,
+      "grad_norm": 8.0,
+      "learning_rate": 8.760106672102188e-06,
+      "loss": 0.8858,
+      "step": 97115
+    },
+    {
+      "epoch": 6.5987226525343115,
+      "grad_norm": 7.0,
+      "learning_rate": 8.757983421660553e-06,
+      "loss": 0.7622,
+      "step": 97120
+    },
+    {
+      "epoch": 6.5990623726049735,
+      "grad_norm": 8.75,
+      "learning_rate": 8.755860171218916e-06,
+      "loss": 0.7464,
+      "step": 97125
+    },
+    {
+      "epoch": 6.599402092675636,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.75373692077728e-06,
+      "loss": 0.8419,
+      "step": 97130
+    },
+    {
+      "epoch": 6.599741812746297,
+      "grad_norm": 8.0,
+      "learning_rate": 8.751613670335644e-06,
+      "loss": 0.8008,
+      "step": 97135
+    },
+    {
+      "epoch": 6.600081532816959,
+      "grad_norm": 7.40625,
+      "learning_rate": 8.749490419894008e-06,
+      "loss": 0.8567,
+      "step": 97140
+    },
+    {
+      "epoch": 6.600421252887621,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.747367169452372e-06,
+      "loss": 0.787,
+      "step": 97145
+    },
+    {
+      "epoch": 6.600760972958282,
+      "grad_norm": 7.09375,
+      "learning_rate": 8.745243919010736e-06,
+      "loss": 0.8431,
+      "step": 97150
+    },
+    {
+      "epoch": 6.601100693028944,
+      "grad_norm": 8.5,
+      "learning_rate": 8.7431206685691e-06,
+      "loss": 0.8327,
+      "step": 97155
+    },
+    {
+      "epoch": 6.601440413099606,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.740997418127464e-06,
+      "loss": 0.8895,
+      "step": 97160
+    },
+    {
+      "epoch": 6.6017801331702675,
+      "grad_norm": 9.375,
+      "learning_rate": 8.738874167685828e-06,
+      "loss": 0.9878,
+      "step": 97165
+    },
+    {
+      "epoch": 6.6021198532409295,
+      "grad_norm": 7.5,
+      "learning_rate": 8.736750917244192e-06,
+      "loss": 0.842,
+      "step": 97170
+    },
+    {
+      "epoch": 6.602459573311592,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.734627666802556e-06,
+      "loss": 0.9101,
+      "step": 97175
+    },
+    {
+      "epoch": 6.602799293382253,
+      "grad_norm": 8.375,
+      "learning_rate": 8.73250441636092e-06,
+      "loss": 0.8596,
+      "step": 97180
+    },
+    {
+      "epoch": 6.603139013452915,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.730381165919282e-06,
+      "loss": 0.7851,
+      "step": 97185
+    },
+    {
+      "epoch": 6.603478733523577,
+      "grad_norm": 8.25,
+      "learning_rate": 8.728257915477648e-06,
+      "loss": 0.9136,
+      "step": 97190
+    },
+    {
+      "epoch": 6.603818453594238,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.72613466503601e-06,
+      "loss": 0.8374,
+      "step": 97195
+    },
+    {
+      "epoch": 6.6041581736649,
+      "grad_norm": 7.75,
+      "learning_rate": 8.724011414594374e-06,
+      "loss": 0.865,
+      "step": 97200
+    },
+    {
+      "epoch": 6.604497893735562,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.72188816415274e-06,
+      "loss": 0.7577,
+      "step": 97205
+    },
+    {
+      "epoch": 6.6048376138062235,
+      "grad_norm": 8.75,
+      "learning_rate": 8.719764913711102e-06,
+      "loss": 0.9066,
+      "step": 97210
+    },
+    {
+      "epoch": 6.6051773338768855,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.717641663269466e-06,
+      "loss": 0.8744,
+      "step": 97215
+    },
+    {
+      "epoch": 6.605517053947548,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.71551841282783e-06,
+      "loss": 0.7919,
+      "step": 97220
+    },
+    {
+      "epoch": 6.605856774018209,
+      "grad_norm": 7.75,
+      "learning_rate": 8.713395162386194e-06,
+      "loss": 0.8027,
+      "step": 97225
+    },
+    {
+      "epoch": 6.606196494088871,
+      "grad_norm": 8.625,
+      "learning_rate": 8.711271911944558e-06,
+      "loss": 0.8975,
+      "step": 97230
+    },
+    {
+      "epoch": 6.606536214159533,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.709148661502922e-06,
+      "loss": 0.9492,
+      "step": 97235
+    },
+    {
+      "epoch": 6.606875934230194,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.707025411061286e-06,
+      "loss": 0.7437,
+      "step": 97240
+    },
+    {
+      "epoch": 6.607215654300856,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.70490216061965e-06,
+      "loss": 0.8715,
+      "step": 97245
+    },
+    {
+      "epoch": 6.607555374371518,
+      "grad_norm": 8.125,
+      "learning_rate": 8.702778910178014e-06,
+      "loss": 0.8361,
+      "step": 97250
+    },
+    {
+      "epoch": 6.6078950944421795,
+      "grad_norm": 8.125,
+      "learning_rate": 8.700655659736378e-06,
+      "loss": 0.8405,
+      "step": 97255
+    },
+    {
+      "epoch": 6.608234814512842,
+      "grad_norm": 8.0,
+      "learning_rate": 8.698532409294742e-06,
+      "loss": 0.9142,
+      "step": 97260
+    },
+    {
+      "epoch": 6.608574534583504,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.696409158853106e-06,
+      "loss": 0.8871,
+      "step": 97265
+    },
+    {
+      "epoch": 6.608914254654165,
+      "grad_norm": 8.5,
+      "learning_rate": 8.694285908411468e-06,
+      "loss": 0.8547,
+      "step": 97270
+    },
+    {
+      "epoch": 6.609253974724827,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.692162657969834e-06,
+      "loss": 0.8576,
+      "step": 97275
+    },
+    {
+      "epoch": 6.609593694795489,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.690039407528196e-06,
+      "loss": 0.8642,
+      "step": 97280
+    },
+    {
+      "epoch": 6.60993341486615,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.68791615708656e-06,
+      "loss": 0.8152,
+      "step": 97285
+    },
+    {
+      "epoch": 6.610273134936812,
+      "grad_norm": 7.0625,
+      "learning_rate": 8.685792906644926e-06,
+      "loss": 0.7909,
+      "step": 97290
+    },
+    {
+      "epoch": 6.610612855007474,
+      "grad_norm": 7.75,
+      "learning_rate": 8.683669656203288e-06,
+      "loss": 0.8658,
+      "step": 97295
+    },
+    {
+      "epoch": 6.6109525750781355,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.681546405761652e-06,
+      "loss": 0.9334,
+      "step": 97300
+    },
+    {
+      "epoch": 6.611292295148798,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.679423155320016e-06,
+      "loss": 0.8199,
+      "step": 97305
+    },
+    {
+      "epoch": 6.61163201521946,
+      "grad_norm": 8.75,
+      "learning_rate": 8.67729990487838e-06,
+      "loss": 0.793,
+      "step": 97310
+    },
+    {
+      "epoch": 6.611971735290121,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.675176654436746e-06,
+      "loss": 0.88,
+      "step": 97315
+    },
+    {
+      "epoch": 6.612311455360783,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.673053403995108e-06,
+      "loss": 0.8039,
+      "step": 97320
+    },
+    {
+      "epoch": 6.612651175431445,
+      "grad_norm": 8.625,
+      "learning_rate": 8.670930153553472e-06,
+      "loss": 0.8152,
+      "step": 97325
+    },
+    {
+      "epoch": 6.612990895502106,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.668806903111836e-06,
+      "loss": 0.8142,
+      "step": 97330
+    },
+    {
+      "epoch": 6.613330615572768,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.6666836526702e-06,
+      "loss": 0.7328,
+      "step": 97335
+    },
+    {
+      "epoch": 6.613670335643429,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.664560402228564e-06,
+      "loss": 0.7756,
+      "step": 97340
+    },
+    {
+      "epoch": 6.6140100557140915,
+      "grad_norm": 8.25,
+      "learning_rate": 8.662437151786928e-06,
+      "loss": 0.911,
+      "step": 97345
+    },
+    {
+      "epoch": 6.614349775784754,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.660313901345292e-06,
+      "loss": 0.7945,
+      "step": 97350
+    },
+    {
+      "epoch": 6.614689495855415,
+      "grad_norm": 7.75,
+      "learning_rate": 8.658190650903655e-06,
+      "loss": 0.8318,
+      "step": 97355
+    },
+    {
+      "epoch": 6.615029215926077,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.65606740046202e-06,
+      "loss": 0.8013,
+      "step": 97360
+    },
+    {
+      "epoch": 6.615368935996739,
+      "grad_norm": 8.5,
+      "learning_rate": 8.653944150020384e-06,
+      "loss": 0.8601,
+      "step": 97365
+    },
+    {
+      "epoch": 6.6157086560674,
+      "grad_norm": 8.25,
+      "learning_rate": 8.651820899578747e-06,
+      "loss": 0.8172,
+      "step": 97370
+    },
+    {
+      "epoch": 6.616048376138062,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.649697649137112e-06,
+      "loss": 0.8105,
+      "step": 97375
+    },
+    {
+      "epoch": 6.616388096208724,
+      "grad_norm": 8.25,
+      "learning_rate": 8.647574398695475e-06,
+      "loss": 0.7873,
+      "step": 97380
+    },
+    {
+      "epoch": 6.616727816279385,
+      "grad_norm": 7.875,
+      "learning_rate": 8.645451148253839e-06,
+      "loss": 0.931,
+      "step": 97385
+    },
+    {
+      "epoch": 6.6170675363500475,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.643327897812203e-06,
+      "loss": 0.8553,
+      "step": 97390
+    },
+    {
+      "epoch": 6.61740725642071,
+      "grad_norm": 8.375,
+      "learning_rate": 8.641204647370567e-06,
+      "loss": 0.8601,
+      "step": 97395
+    },
+    {
+      "epoch": 6.617746976491371,
+      "grad_norm": 7.875,
+      "learning_rate": 8.639081396928932e-06,
+      "loss": 0.8416,
+      "step": 97400
+    },
+    {
+      "epoch": 6.618086696562033,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.636958146487295e-06,
+      "loss": 0.8141,
+      "step": 97405
+    },
+    {
+      "epoch": 6.618426416632695,
+      "grad_norm": 9.375,
+      "learning_rate": 8.634834896045659e-06,
+      "loss": 0.8869,
+      "step": 97410
+    },
+    {
+      "epoch": 6.618766136703356,
+      "grad_norm": 7.625,
+      "learning_rate": 8.632711645604023e-06,
+      "loss": 0.84,
+      "step": 97415
+    },
+    {
+      "epoch": 6.619105856774018,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.630588395162387e-06,
+      "loss": 0.8336,
+      "step": 97420
+    },
+    {
+      "epoch": 6.61944557684468,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.62846514472075e-06,
+      "loss": 0.8756,
+      "step": 97425
+    },
+    {
+      "epoch": 6.6197852969153415,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.626341894279115e-06,
+      "loss": 0.8623,
+      "step": 97430
+    },
+    {
+      "epoch": 6.6201250169860035,
+      "grad_norm": 8.9375,
+      "learning_rate": 8.624218643837479e-06,
+      "loss": 0.8068,
+      "step": 97435
+    },
+    {
+      "epoch": 6.620464737056666,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.622095393395841e-06,
+      "loss": 0.689,
+      "step": 97440
+    },
+    {
+      "epoch": 6.620804457127327,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.619972142954207e-06,
+      "loss": 0.8027,
+      "step": 97445
+    },
+    {
+      "epoch": 6.621144177197989,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.61784889251257e-06,
+      "loss": 0.8594,
+      "step": 97450
+    },
+    {
+      "epoch": 6.621483897268651,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.615725642070933e-06,
+      "loss": 0.8377,
+      "step": 97455
+    },
+    {
+      "epoch": 6.621823617339312,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.613602391629299e-06,
+      "loss": 0.8956,
+      "step": 97460
+    },
+    {
+      "epoch": 6.622163337409974,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.611479141187661e-06,
+      "loss": 0.8069,
+      "step": 97465
+    },
+    {
+      "epoch": 6.622503057480636,
+      "grad_norm": 8.625,
+      "learning_rate": 8.609355890746025e-06,
+      "loss": 0.756,
+      "step": 97470
+    },
+    {
+      "epoch": 6.6228427775512975,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.607232640304389e-06,
+      "loss": 0.8725,
+      "step": 97475
+    },
+    {
+      "epoch": 6.6231824976219595,
+      "grad_norm": 7.53125,
+      "learning_rate": 8.605109389862753e-06,
+      "loss": 0.8476,
+      "step": 97480
+    },
+    {
+      "epoch": 6.623522217692622,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.602986139421119e-06,
+      "loss": 0.8633,
+      "step": 97485
+    },
+    {
+      "epoch": 6.623861937763283,
+      "grad_norm": 6.78125,
+      "learning_rate": 8.600862888979481e-06,
+      "loss": 0.8742,
+      "step": 97490
+    },
+    {
+      "epoch": 6.624201657833945,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.598739638537845e-06,
+      "loss": 0.8087,
+      "step": 97495
+    },
+    {
+      "epoch": 6.624541377904607,
+      "grad_norm": 7.53125,
+      "learning_rate": 8.596616388096209e-06,
+      "loss": 0.8502,
+      "step": 97500
+    },
+    {
+      "epoch": 6.624881097975268,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.594493137654573e-06,
+      "loss": 0.7886,
+      "step": 97505
+    },
+    {
+      "epoch": 6.62522081804593,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.592369887212937e-06,
+      "loss": 0.92,
+      "step": 97510
+    },
+    {
+      "epoch": 6.625560538116592,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.590246636771301e-06,
+      "loss": 0.8828,
+      "step": 97515
+    },
+    {
+      "epoch": 6.6259002581872535,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.588123386329665e-06,
+      "loss": 0.8873,
+      "step": 97520
+    },
+    {
+      "epoch": 6.6262399782579156,
+      "grad_norm": 7.875,
+      "learning_rate": 8.586000135888027e-06,
+      "loss": 0.7793,
+      "step": 97525
+    },
+    {
+      "epoch": 6.626579698328578,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.583876885446393e-06,
+      "loss": 0.8954,
+      "step": 97530
+    },
+    {
+      "epoch": 6.626919418399239,
+      "grad_norm": 9.875,
+      "learning_rate": 8.581753635004757e-06,
+      "loss": 0.8641,
+      "step": 97535
+    },
+    {
+      "epoch": 6.627259138469901,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.57963038456312e-06,
+      "loss": 0.8489,
+      "step": 97540
+    },
+    {
+      "epoch": 6.627598858540563,
+      "grad_norm": 7.875,
+      "learning_rate": 8.577507134121485e-06,
+      "loss": 0.8131,
+      "step": 97545
+    },
+    {
+      "epoch": 6.627938578611224,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.575383883679847e-06,
+      "loss": 0.976,
+      "step": 97550
+    },
+    {
+      "epoch": 6.628278298681886,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.573260633238211e-06,
+      "loss": 0.8164,
+      "step": 97555
+    },
+    {
+      "epoch": 6.628618018752547,
+      "grad_norm": 7.375,
+      "learning_rate": 8.571137382796575e-06,
+      "loss": 0.9078,
+      "step": 97560
+    },
+    {
+      "epoch": 6.6289577388232095,
+      "grad_norm": 8.375,
+      "learning_rate": 8.56901413235494e-06,
+      "loss": 0.8067,
+      "step": 97565
+    },
+    {
+      "epoch": 6.629297458893872,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.566890881913305e-06,
+      "loss": 0.8553,
+      "step": 97570
+    },
+    {
+      "epoch": 6.629637178964533,
+      "grad_norm": 8.125,
+      "learning_rate": 8.564767631471667e-06,
+      "loss": 0.8118,
+      "step": 97575
+    },
+    {
+      "epoch": 6.629976899035195,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.562644381030031e-06,
+      "loss": 0.8615,
+      "step": 97580
+    },
+    {
+      "epoch": 6.630316619105857,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.560521130588395e-06,
+      "loss": 0.8177,
+      "step": 97585
+    },
+    {
+      "epoch": 6.630656339176518,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.55839788014676e-06,
+      "loss": 0.832,
+      "step": 97590
+    },
+    {
+      "epoch": 6.63099605924718,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.556274629705123e-06,
+      "loss": 0.7861,
+      "step": 97595
+    },
+    {
+      "epoch": 6.631335779317842,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.554151379263487e-06,
+      "loss": 0.8446,
+      "step": 97600
+    },
+    {
+      "epoch": 6.631675499388503,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.552028128821851e-06,
+      "loss": 0.8004,
+      "step": 97605
+    },
+    {
+      "epoch": 6.6320152194591655,
+      "grad_norm": 8.375,
+      "learning_rate": 8.549904878380214e-06,
+      "loss": 0.8442,
+      "step": 97610
+    },
+    {
+      "epoch": 6.632354939529828,
+      "grad_norm": 8.375,
+      "learning_rate": 8.54778162793858e-06,
+      "loss": 0.798,
+      "step": 97615
+    },
+    {
+      "epoch": 6.632694659600489,
+      "grad_norm": 7.875,
+      "learning_rate": 8.545658377496944e-06,
+      "loss": 0.7656,
+      "step": 97620
+    },
+    {
+      "epoch": 6.633034379671151,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.543535127055306e-06,
+      "loss": 0.737,
+      "step": 97625
+    },
+    {
+      "epoch": 6.633374099741813,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.541411876613672e-06,
+      "loss": 0.808,
+      "step": 97630
+    },
+    {
+      "epoch": 6.633713819812474,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.539288626172034e-06,
+      "loss": 0.7904,
+      "step": 97635
+    },
+    {
+      "epoch": 6.634053539883136,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.537165375730398e-06,
+      "loss": 0.7839,
+      "step": 97640
+    },
+    {
+      "epoch": 6.634393259953798,
+      "grad_norm": 8.0,
+      "learning_rate": 8.535042125288764e-06,
+      "loss": 0.9073,
+      "step": 97645
+    },
+    {
+      "epoch": 6.634732980024459,
+      "grad_norm": 7.375,
+      "learning_rate": 8.532918874847126e-06,
+      "loss": 0.8959,
+      "step": 97650
+    },
+    {
+      "epoch": 6.6350727000951215,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.530795624405492e-06,
+      "loss": 0.8125,
+      "step": 97655
+    },
+    {
+      "epoch": 6.635412420165784,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.528672373963854e-06,
+      "loss": 0.892,
+      "step": 97660
+    },
+    {
+      "epoch": 6.635752140236445,
+      "grad_norm": 8.125,
+      "learning_rate": 8.526549123522218e-06,
+      "loss": 0.8875,
+      "step": 97665
+    },
+    {
+      "epoch": 6.636091860307107,
+      "grad_norm": 7.875,
+      "learning_rate": 8.524425873080582e-06,
+      "loss": 0.7999,
+      "step": 97670
+    },
+    {
+      "epoch": 6.636431580377769,
+      "grad_norm": 9.375,
+      "learning_rate": 8.522302622638946e-06,
+      "loss": 0.8777,
+      "step": 97675
+    },
+    {
+      "epoch": 6.63677130044843,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.52017937219731e-06,
+      "loss": 0.8472,
+      "step": 97680
+    },
+    {
+      "epoch": 6.637111020519092,
+      "grad_norm": 7.875,
+      "learning_rate": 8.518056121755674e-06,
+      "loss": 0.8247,
+      "step": 97685
+    },
+    {
+      "epoch": 6.637450740589754,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.515932871314038e-06,
+      "loss": 0.9241,
+      "step": 97690
+    },
+    {
+      "epoch": 6.6377904606604154,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.513809620872402e-06,
+      "loss": 0.8167,
+      "step": 97695
+    },
+    {
+      "epoch": 6.6381301807310775,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.511686370430766e-06,
+      "loss": 0.9205,
+      "step": 97700
+    },
+    {
+      "epoch": 6.63846990080174,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.50956311998913e-06,
+      "loss": 0.8292,
+      "step": 97705
+    },
+    {
+      "epoch": 6.638809620872401,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.507439869547492e-06,
+      "loss": 0.7925,
+      "step": 97710
+    },
+    {
+      "epoch": 6.639149340943063,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.505316619105858e-06,
+      "loss": 0.8609,
+      "step": 97715
+    },
+    {
+      "epoch": 6.639489061013725,
+      "grad_norm": 8.625,
+      "learning_rate": 8.50319336866422e-06,
+      "loss": 0.9365,
+      "step": 97720
+    },
+    {
+      "epoch": 6.639828781084386,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.501070118222584e-06,
+      "loss": 0.8501,
+      "step": 97725
+    },
+    {
+      "epoch": 6.640168501155048,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.49894686778095e-06,
+      "loss": 0.8176,
+      "step": 97730
+    },
+    {
+      "epoch": 6.64050822122571,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.496823617339312e-06,
+      "loss": 0.8021,
+      "step": 97735
+    },
+    {
+      "epoch": 6.6408479412963715,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.494700366897678e-06,
+      "loss": 0.8871,
+      "step": 97740
+    },
+    {
+      "epoch": 6.6411876613670335,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.49257711645604e-06,
+      "loss": 0.8699,
+      "step": 97745
+    },
+    {
+      "epoch": 6.641527381437696,
+      "grad_norm": 8.125,
+      "learning_rate": 8.490453866014404e-06,
+      "loss": 0.8982,
+      "step": 97750
+    },
+    {
+      "epoch": 6.641867101508357,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.488330615572768e-06,
+      "loss": 0.7709,
+      "step": 97755
+    },
+    {
+      "epoch": 6.642206821579019,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.486207365131132e-06,
+      "loss": 0.8321,
+      "step": 97760
+    },
+    {
+      "epoch": 6.642546541649681,
+      "grad_norm": 7.625,
+      "learning_rate": 8.484084114689496e-06,
+      "loss": 0.9597,
+      "step": 97765
+    },
+    {
+      "epoch": 6.642886261720342,
+      "grad_norm": 8.25,
+      "learning_rate": 8.48196086424786e-06,
+      "loss": 0.884,
+      "step": 97770
+    },
+    {
+      "epoch": 6.643225981791004,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.479837613806224e-06,
+      "loss": 0.8793,
+      "step": 97775
+    },
+    {
+      "epoch": 6.643565701861666,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.477714363364588e-06,
+      "loss": 0.8564,
+      "step": 97780
+    },
+    {
+      "epoch": 6.6439054219323275,
+      "grad_norm": 8.5,
+      "learning_rate": 8.475591112922952e-06,
+      "loss": 0.909,
+      "step": 97785
+    },
+    {
+      "epoch": 6.6442451420029895,
+      "grad_norm": 7.5,
+      "learning_rate": 8.473467862481316e-06,
+      "loss": 0.8105,
+      "step": 97790
+    },
+    {
+      "epoch": 6.644584862073652,
+      "grad_norm": 7.75,
+      "learning_rate": 8.471344612039679e-06,
+      "loss": 0.8684,
+      "step": 97795
+    },
+    {
+      "epoch": 6.644924582144313,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.469221361598044e-06,
+      "loss": 0.8242,
+      "step": 97800
+    },
+    {
+      "epoch": 6.645264302214975,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.467098111156407e-06,
+      "loss": 0.8213,
+      "step": 97805
+    },
+    {
+      "epoch": 6.645604022285637,
+      "grad_norm": 6.9375,
+      "learning_rate": 8.46497486071477e-06,
+      "loss": 0.9132,
+      "step": 97810
+    },
+    {
+      "epoch": 6.645943742356298,
+      "grad_norm": 6.71875,
+      "learning_rate": 8.462851610273136e-06,
+      "loss": 0.8302,
+      "step": 97815
+    },
+    {
+      "epoch": 6.64628346242696,
+      "grad_norm": 8.375,
+      "learning_rate": 8.460728359831499e-06,
+      "loss": 0.8274,
+      "step": 97820
+    },
+    {
+      "epoch": 6.646623182497622,
+      "grad_norm": 7.5,
+      "learning_rate": 8.458605109389864e-06,
+      "loss": 0.7607,
+      "step": 97825
+    },
+    {
+      "epoch": 6.6469629025682835,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.456481858948227e-06,
+      "loss": 0.7867,
+      "step": 97830
+    },
+    {
+      "epoch": 6.647302622638946,
+      "grad_norm": 8.9375,
+      "learning_rate": 8.45435860850659e-06,
+      "loss": 0.8231,
+      "step": 97835
+    },
+    {
+      "epoch": 6.647642342709608,
+      "grad_norm": 7.5,
+      "learning_rate": 8.452235358064956e-06,
+      "loss": 0.8297,
+      "step": 97840
+    },
+    {
+      "epoch": 6.647982062780269,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.450112107623319e-06,
+      "loss": 0.7906,
+      "step": 97845
+    },
+    {
+      "epoch": 6.648321782850931,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.447988857181683e-06,
+      "loss": 0.9002,
+      "step": 97850
+    },
+    {
+      "epoch": 6.648661502921593,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.445865606740047e-06,
+      "loss": 0.8085,
+      "step": 97855
+    },
+    {
+      "epoch": 6.649001222992254,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.44374235629841e-06,
+      "loss": 0.9635,
+      "step": 97860
+    },
+    {
+      "epoch": 6.649340943062916,
+      "grad_norm": 7.125,
+      "learning_rate": 8.441619105856775e-06,
+      "loss": 0.8378,
+      "step": 97865
+    },
+    {
+      "epoch": 6.649680663133578,
+      "grad_norm": 7.75,
+      "learning_rate": 8.439495855415139e-06,
+      "loss": 0.8903,
+      "step": 97870
+    },
+    {
+      "epoch": 6.6500203832042395,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.437372604973503e-06,
+      "loss": 0.867,
+      "step": 97875
+    },
+    {
+      "epoch": 6.650360103274902,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.435249354531865e-06,
+      "loss": 0.8225,
+      "step": 97880
+    },
+    {
+      "epoch": 6.650699823345564,
+      "grad_norm": 7.0625,
+      "learning_rate": 8.43312610409023e-06,
+      "loss": 0.7984,
+      "step": 97885
+    },
+    {
+      "epoch": 6.651039543416225,
+      "grad_norm": 6.125,
+      "learning_rate": 8.431002853648595e-06,
+      "loss": 0.8636,
+      "step": 97890
+    },
+    {
+      "epoch": 6.651379263486887,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.428879603206957e-06,
+      "loss": 0.8386,
+      "step": 97895
+    },
+    {
+      "epoch": 6.651718983557549,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.426756352765323e-06,
+      "loss": 0.9042,
+      "step": 97900
+    },
+    {
+      "epoch": 6.65205870362821,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.424633102323685e-06,
+      "loss": 0.8968,
+      "step": 97905
+    },
+    {
+      "epoch": 6.652398423698872,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.42250985188205e-06,
+      "loss": 0.8073,
+      "step": 97910
+    },
+    {
+      "epoch": 6.652738143769534,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.420386601440413e-06,
+      "loss": 0.8016,
+      "step": 97915
+    },
+    {
+      "epoch": 6.6530778638401955,
+      "grad_norm": 6.75,
+      "learning_rate": 8.418263350998777e-06,
+      "loss": 0.8132,
+      "step": 97920
+    },
+    {
+      "epoch": 6.653417583910858,
+      "grad_norm": 8.875,
+      "learning_rate": 8.416140100557143e-06,
+      "loss": 0.9151,
+      "step": 97925
+    },
+    {
+      "epoch": 6.65375730398152,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.414016850115505e-06,
+      "loss": 0.8822,
+      "step": 97930
+    },
+    {
+      "epoch": 6.654097024052181,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.411893599673869e-06,
+      "loss": 0.927,
+      "step": 97935
+    },
+    {
+      "epoch": 6.654436744122843,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.409770349232233e-06,
+      "loss": 0.7882,
+      "step": 97940
+    },
+    {
+      "epoch": 6.654776464193505,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.407647098790597e-06,
+      "loss": 0.8581,
+      "step": 97945
+    },
+    {
+      "epoch": 6.655116184264166,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.405523848348961e-06,
+      "loss": 0.7691,
+      "step": 97950
+    },
+    {
+      "epoch": 6.655455904334828,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.403400597907325e-06,
+      "loss": 0.7584,
+      "step": 97955
+    },
+    {
+      "epoch": 6.65579562440549,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.401277347465689e-06,
+      "loss": 0.7992,
+      "step": 97960
+    },
+    {
+      "epoch": 6.6561353444761515,
+      "grad_norm": 8.25,
+      "learning_rate": 8.399154097024051e-06,
+      "loss": 0.8398,
+      "step": 97965
+    },
+    {
+      "epoch": 6.656475064546814,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.397030846582417e-06,
+      "loss": 0.8955,
+      "step": 97970
+    },
+    {
+      "epoch": 6.656814784617476,
+      "grad_norm": 8.25,
+      "learning_rate": 8.394907596140781e-06,
+      "loss": 0.9098,
+      "step": 97975
+    },
+    {
+      "epoch": 6.657154504688137,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.392784345699143e-06,
+      "loss": 0.8226,
+      "step": 97980
+    },
+    {
+      "epoch": 6.657494224758799,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.390661095257509e-06,
+      "loss": 0.875,
+      "step": 97985
+    },
+    {
+      "epoch": 6.657833944829461,
+      "grad_norm": 7.75,
+      "learning_rate": 8.388537844815871e-06,
+      "loss": 0.8416,
+      "step": 97990
+    },
+    {
+      "epoch": 6.658173664900122,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.386414594374237e-06,
+      "loss": 0.9352,
+      "step": 97995
+    },
+    {
+      "epoch": 6.658513384970784,
+      "grad_norm": 8.375,
+      "learning_rate": 8.3842913439326e-06,
+      "loss": 0.8204,
+      "step": 98000
+    },
+    {
+      "epoch": 6.658853105041446,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.382168093490963e-06,
+      "loss": 0.9182,
+      "step": 98005
+    },
+    {
+      "epoch": 6.6591928251121075,
+      "grad_norm": 8.25,
+      "learning_rate": 8.380044843049329e-06,
+      "loss": 0.8195,
+      "step": 98010
+    },
+    {
+      "epoch": 6.65953254518277,
+      "grad_norm": 7.09375,
+      "learning_rate": 8.377921592607691e-06,
+      "loss": 0.7847,
+      "step": 98015
+    },
+    {
+      "epoch": 6.659872265253431,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.375798342166055e-06,
+      "loss": 0.9131,
+      "step": 98020
+    },
+    {
+      "epoch": 6.660211985324093,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.37367509172442e-06,
+      "loss": 0.8612,
+      "step": 98025
+    },
+    {
+      "epoch": 6.660551705394755,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.371551841282783e-06,
+      "loss": 0.8861,
+      "step": 98030
+    },
+    {
+      "epoch": 6.660891425465416,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.369428590841147e-06,
+      "loss": 0.9932,
+      "step": 98035
+    },
+    {
+      "epoch": 6.661231145536078,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.367305340399511e-06,
+      "loss": 0.8473,
+      "step": 98040
+    },
+    {
+      "epoch": 6.66157086560674,
+      "grad_norm": 8.875,
+      "learning_rate": 8.365182089957875e-06,
+      "loss": 0.946,
+      "step": 98045
+    },
+    {
+      "epoch": 6.6619105856774015,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.363058839516238e-06,
+      "loss": 0.8228,
+      "step": 98050
+    },
+    {
+      "epoch": 6.6622503057480635,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.360935589074603e-06,
+      "loss": 0.8697,
+      "step": 98055
+    },
+    {
+      "epoch": 6.662590025818726,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.358812338632967e-06,
+      "loss": 0.9193,
+      "step": 98060
+    },
+    {
+      "epoch": 6.662929745889387,
+      "grad_norm": 8.25,
+      "learning_rate": 8.35668908819133e-06,
+      "loss": 0.8394,
+      "step": 98065
+    },
+    {
+      "epoch": 6.663269465960049,
+      "grad_norm": 8.125,
+      "learning_rate": 8.354565837749695e-06,
+      "loss": 0.8255,
+      "step": 98070
+    },
+    {
+      "epoch": 6.663609186030711,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.352442587308058e-06,
+      "loss": 0.8584,
+      "step": 98075
+    },
+    {
+      "epoch": 6.663948906101372,
+      "grad_norm": 9.25,
+      "learning_rate": 8.350319336866423e-06,
+      "loss": 0.805,
+      "step": 98080
+    },
+    {
+      "epoch": 6.664288626172034,
+      "grad_norm": 6.9375,
+      "learning_rate": 8.348196086424787e-06,
+      "loss": 0.8396,
+      "step": 98085
+    },
+    {
+      "epoch": 6.664628346242696,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.34607283598315e-06,
+      "loss": 0.923,
+      "step": 98090
+    },
+    {
+      "epoch": 6.6649680663133575,
+      "grad_norm": 9.75,
+      "learning_rate": 8.343949585541515e-06,
+      "loss": 0.8645,
+      "step": 98095
+    },
+    {
+      "epoch": 6.6653077863840196,
+      "grad_norm": 8.125,
+      "learning_rate": 8.341826335099878e-06,
+      "loss": 0.7925,
+      "step": 98100
+    },
+    {
+      "epoch": 6.665647506454682,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.339703084658242e-06,
+      "loss": 0.8399,
+      "step": 98105
+    },
+    {
+      "epoch": 6.665987226525343,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.337579834216606e-06,
+      "loss": 0.8481,
+      "step": 98110
+    },
+    {
+      "epoch": 6.666326946596005,
+      "grad_norm": 7.0,
+      "learning_rate": 8.33545658377497e-06,
+      "loss": 0.8596,
+      "step": 98115
+    },
+    {
+      "epoch": 6.666666666666667,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.854,
+      "step": 98120
+    },
+    {
+      "epoch": 6.667006386737328,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.331210082891698e-06,
+      "loss": 0.8553,
+      "step": 98125
+    },
+    {
+      "epoch": 6.66734610680799,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.329086832450062e-06,
+      "loss": 0.7372,
+      "step": 98130
+    },
+    {
+      "epoch": 6.667685826878652,
+      "grad_norm": 7.09375,
+      "learning_rate": 8.326963582008426e-06,
+      "loss": 0.8099,
+      "step": 98135
+    },
+    {
+      "epoch": 6.6680255469493135,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.32484033156679e-06,
+      "loss": 0.772,
+      "step": 98140
+    },
+    {
+      "epoch": 6.668365267019976,
+      "grad_norm": 7.28125,
+      "learning_rate": 8.322717081125154e-06,
+      "loss": 0.8866,
+      "step": 98145
+    },
+    {
+      "epoch": 6.668704987090638,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.320593830683516e-06,
+      "loss": 0.8772,
+      "step": 98150
+    },
+    {
+      "epoch": 6.669044707161299,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.318470580241882e-06,
+      "loss": 0.8348,
+      "step": 98155
+    },
+    {
+      "epoch": 6.669384427231961,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.316347329800244e-06,
+      "loss": 0.8135,
+      "step": 98160
+    },
+    {
+      "epoch": 6.669724147302623,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.31422407935861e-06,
+      "loss": 1.0675,
+      "step": 98165
+    },
+    {
+      "epoch": 6.670063867373284,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.312100828916974e-06,
+      "loss": 0.9648,
+      "step": 98170
+    },
+    {
+      "epoch": 6.670403587443946,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.309977578475336e-06,
+      "loss": 0.8551,
+      "step": 98175
+    },
+    {
+      "epoch": 6.670743307514608,
+      "grad_norm": 7.5,
+      "learning_rate": 8.307854328033702e-06,
+      "loss": 0.8475,
+      "step": 98180
+    },
+    {
+      "epoch": 6.6710830275852695,
+      "grad_norm": 8.0,
+      "learning_rate": 8.305731077592064e-06,
+      "loss": 0.7913,
+      "step": 98185
+    },
+    {
+      "epoch": 6.671422747655932,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.303607827150428e-06,
+      "loss": 0.8807,
+      "step": 98190
+    },
+    {
+      "epoch": 6.671762467726594,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.301484576708792e-06,
+      "loss": 0.8002,
+      "step": 98195
+    },
+    {
+      "epoch": 6.672102187797255,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.299361326267156e-06,
+      "loss": 0.8485,
+      "step": 98200
+    },
+    {
+      "epoch": 6.672441907867917,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.29723807582552e-06,
+      "loss": 0.8296,
+      "step": 98205
+    },
+    {
+      "epoch": 6.672781627938579,
+      "grad_norm": 7.25,
+      "learning_rate": 8.295114825383884e-06,
+      "loss": 0.911,
+      "step": 98210
+    },
+    {
+      "epoch": 6.67312134800924,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.292991574942248e-06,
+      "loss": 0.7692,
+      "step": 98215
+    },
+    {
+      "epoch": 6.673461068079902,
+      "grad_norm": 8.0,
+      "learning_rate": 8.290868324500612e-06,
+      "loss": 0.8778,
+      "step": 98220
+    },
+    {
+      "epoch": 6.673800788150564,
+      "grad_norm": 7.625,
+      "learning_rate": 8.288745074058976e-06,
+      "loss": 0.8395,
+      "step": 98225
+    },
+    {
+      "epoch": 6.6741405082212255,
+      "grad_norm": 8.25,
+      "learning_rate": 8.28662182361734e-06,
+      "loss": 0.7678,
+      "step": 98230
+    },
+    {
+      "epoch": 6.674480228291888,
+      "grad_norm": 8.125,
+      "learning_rate": 8.284498573175702e-06,
+      "loss": 0.8543,
+      "step": 98235
+    },
+    {
+      "epoch": 6.674819948362549,
+      "grad_norm": 8.25,
+      "learning_rate": 8.282375322734068e-06,
+      "loss": 0.8264,
+      "step": 98240
+    },
+    {
+      "epoch": 6.675159668433211,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.28025207229243e-06,
+      "loss": 0.9102,
+      "step": 98245
+    },
+    {
+      "epoch": 6.675499388503873,
+      "grad_norm": 6.03125,
+      "learning_rate": 8.278128821850796e-06,
+      "loss": 0.7952,
+      "step": 98250
+    },
+    {
+      "epoch": 6.675839108574534,
+      "grad_norm": 8.375,
+      "learning_rate": 8.27600557140916e-06,
+      "loss": 0.8141,
+      "step": 98255
+    },
+    {
+      "epoch": 6.676178828645196,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.273882320967522e-06,
+      "loss": 0.8793,
+      "step": 98260
+    },
+    {
+      "epoch": 6.676518548715858,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.271759070525888e-06,
+      "loss": 0.8714,
+      "step": 98265
+    },
+    {
+      "epoch": 6.6768582687865194,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.26963582008425e-06,
+      "loss": 0.8803,
+      "step": 98270
+    },
+    {
+      "epoch": 6.6771979888571815,
+      "grad_norm": 7.625,
+      "learning_rate": 8.267512569642614e-06,
+      "loss": 0.8137,
+      "step": 98275
+    },
+    {
+      "epoch": 6.677537708927844,
+      "grad_norm": 7.53125,
+      "learning_rate": 8.265389319200978e-06,
+      "loss": 0.8959,
+      "step": 98280
+    },
+    {
+      "epoch": 6.677877428998505,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.263266068759342e-06,
+      "loss": 0.8742,
+      "step": 98285
+    },
+    {
+      "epoch": 6.678217149069167,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.261142818317706e-06,
+      "loss": 0.8604,
+      "step": 98290
+    },
+    {
+      "epoch": 6.678556869139829,
+      "grad_norm": 8.25,
+      "learning_rate": 8.25901956787607e-06,
+      "loss": 0.848,
+      "step": 98295
+    },
+    {
+      "epoch": 6.67889658921049,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.256896317434434e-06,
+      "loss": 0.8864,
+      "step": 98300
+    },
+    {
+      "epoch": 6.679236309281152,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.254773066992798e-06,
+      "loss": 0.8718,
+      "step": 98305
+    },
+    {
+      "epoch": 6.679576029351814,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.252649816551162e-06,
+      "loss": 0.7709,
+      "step": 98310
+    },
+    {
+      "epoch": 6.6799157494224755,
+      "grad_norm": 9.1875,
+      "learning_rate": 8.250526566109526e-06,
+      "loss": 0.8628,
+      "step": 98315
+    },
+    {
+      "epoch": 6.6802554694931375,
+      "grad_norm": 8.25,
+      "learning_rate": 8.248403315667889e-06,
+      "loss": 0.8807,
+      "step": 98320
+    },
+    {
+      "epoch": 6.6805951895638,
+      "grad_norm": 8.0,
+      "learning_rate": 8.246280065226255e-06,
+      "loss": 0.7755,
+      "step": 98325
+    },
+    {
+      "epoch": 6.680934909634461,
+      "grad_norm": 9.1875,
+      "learning_rate": 8.244156814784617e-06,
+      "loss": 0.809,
+      "step": 98330
+    },
+    {
+      "epoch": 6.681274629705123,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.242033564342983e-06,
+      "loss": 0.8002,
+      "step": 98335
+    },
+    {
+      "epoch": 6.681614349775785,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.239910313901347e-06,
+      "loss": 0.9248,
+      "step": 98340
+    },
+    {
+      "epoch": 6.681954069846446,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.237787063459709e-06,
+      "loss": 0.8278,
+      "step": 98345
+    },
+    {
+      "epoch": 6.682293789917108,
+      "grad_norm": 8.375,
+      "learning_rate": 8.235663813018075e-06,
+      "loss": 0.8488,
+      "step": 98350
+    },
+    {
+      "epoch": 6.68263350998777,
+      "grad_norm": 7.46875,
+      "learning_rate": 8.233540562576437e-06,
+      "loss": 0.869,
+      "step": 98355
+    },
+    {
+      "epoch": 6.6829732300584315,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.2314173121348e-06,
+      "loss": 0.8308,
+      "step": 98360
+    },
+    {
+      "epoch": 6.6833129501290935,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.229294061693167e-06,
+      "loss": 0.7979,
+      "step": 98365
+    },
+    {
+      "epoch": 6.683652670199756,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.227170811251529e-06,
+      "loss": 0.8381,
+      "step": 98370
+    },
+    {
+      "epoch": 6.683992390270417,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.225047560809893e-06,
+      "loss": 0.8044,
+      "step": 98375
+    },
+    {
+      "epoch": 6.684332110341079,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.222924310368257e-06,
+      "loss": 0.7701,
+      "step": 98380
+    },
+    {
+      "epoch": 6.684671830411741,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.220801059926621e-06,
+      "loss": 0.8748,
+      "step": 98385
+    },
+    {
+      "epoch": 6.685011550482402,
+      "grad_norm": 8.25,
+      "learning_rate": 8.218677809484985e-06,
+      "loss": 0.7218,
+      "step": 98390
+    },
+    {
+      "epoch": 6.685351270553064,
+      "grad_norm": 7.625,
+      "learning_rate": 8.216554559043349e-06,
+      "loss": 0.9182,
+      "step": 98395
+    },
+    {
+      "epoch": 6.685690990623726,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.214431308601713e-06,
+      "loss": 0.8272,
+      "step": 98400
+    },
+    {
+      "epoch": 6.6860307106943875,
+      "grad_norm": 8.25,
+      "learning_rate": 8.212308058160075e-06,
+      "loss": 0.8699,
+      "step": 98405
+    },
+    {
+      "epoch": 6.6863704307650496,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.210184807718441e-06,
+      "loss": 0.7834,
+      "step": 98410
+    },
+    {
+      "epoch": 6.686710150835712,
+      "grad_norm": 6.5,
+      "learning_rate": 8.208061557276805e-06,
+      "loss": 1.15,
+      "step": 98415
+    },
+    {
+      "epoch": 6.687049870906373,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.205938306835169e-06,
+      "loss": 0.8355,
+      "step": 98420
+    },
+    {
+      "epoch": 6.687389590977035,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.203815056393533e-06,
+      "loss": 0.8622,
+      "step": 98425
+    },
+    {
+      "epoch": 6.687729311047697,
+      "grad_norm": 7.5,
+      "learning_rate": 8.201691805951895e-06,
+      "loss": 0.8119,
+      "step": 98430
+    },
+    {
+      "epoch": 6.688069031118358,
+      "grad_norm": 7.40625,
+      "learning_rate": 8.199568555510261e-06,
+      "loss": 0.7734,
+      "step": 98435
+    },
+    {
+      "epoch": 6.68840875118902,
+      "grad_norm": 7.875,
+      "learning_rate": 8.197445305068623e-06,
+      "loss": 0.8354,
+      "step": 98440
+    },
+    {
+      "epoch": 6.688748471259682,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.195322054626987e-06,
+      "loss": 0.8029,
+      "step": 98445
+    },
+    {
+      "epoch": 6.6890881913303435,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.193198804185353e-06,
+      "loss": 1.0089,
+      "step": 98450
+    },
+    {
+      "epoch": 6.689427911401006,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.191075553743715e-06,
+      "loss": 0.9439,
+      "step": 98455
+    },
+    {
+      "epoch": 6.689767631471668,
+      "grad_norm": 9.1875,
+      "learning_rate": 8.18895230330208e-06,
+      "loss": 0.8519,
+      "step": 98460
+    },
+    {
+      "epoch": 6.690107351542329,
+      "grad_norm": 7.0625,
+      "learning_rate": 8.186829052860443e-06,
+      "loss": 0.8238,
+      "step": 98465
+    },
+    {
+      "epoch": 6.690447071612991,
+      "grad_norm": 7.09375,
+      "learning_rate": 8.184705802418807e-06,
+      "loss": 0.8627,
+      "step": 98470
+    },
+    {
+      "epoch": 6.690786791683653,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.182582551977171e-06,
+      "loss": 0.8816,
+      "step": 98475
+    },
+    {
+      "epoch": 6.691126511754314,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.180459301535535e-06,
+      "loss": 0.934,
+      "step": 98480
+    },
+    {
+      "epoch": 6.691466231824976,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.1783360510939e-06,
+      "loss": 0.8994,
+      "step": 98485
+    },
+    {
+      "epoch": 6.691805951895638,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.176212800652262e-06,
+      "loss": 0.8117,
+      "step": 98490
+    },
+    {
+      "epoch": 6.6921456719662995,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.174089550210627e-06,
+      "loss": 0.8747,
+      "step": 98495
+    },
+    {
+      "epoch": 6.692485392036962,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.171966299768991e-06,
+      "loss": 0.8882,
+      "step": 98500
+    },
+    {
+      "epoch": 6.692825112107624,
+      "grad_norm": 7.75,
+      "learning_rate": 8.169843049327355e-06,
+      "loss": 0.8362,
+      "step": 98505
+    },
+    {
+      "epoch": 6.693164832178285,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.16771979888572e-06,
+      "loss": 0.8412,
+      "step": 98510
+    },
+    {
+      "epoch": 6.693504552248947,
+      "grad_norm": 6.40625,
+      "learning_rate": 8.165596548444082e-06,
+      "loss": 0.8492,
+      "step": 98515
+    },
+    {
+      "epoch": 6.693844272319609,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.163473298002447e-06,
+      "loss": 0.7327,
+      "step": 98520
+    },
+    {
+      "epoch": 6.69418399239027,
+      "grad_norm": 7.0,
+      "learning_rate": 8.16135004756081e-06,
+      "loss": 0.7719,
+      "step": 98525
+    },
+    {
+      "epoch": 6.694523712460932,
+      "grad_norm": 8.625,
+      "learning_rate": 8.159226797119174e-06,
+      "loss": 0.8935,
+      "step": 98530
+    },
+    {
+      "epoch": 6.694863432531594,
+      "grad_norm": 8.25,
+      "learning_rate": 8.15710354667754e-06,
+      "loss": 0.9342,
+      "step": 98535
+    },
+    {
+      "epoch": 6.6952031526022555,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.154980296235902e-06,
+      "loss": 0.839,
+      "step": 98540
+    },
+    {
+      "epoch": 6.695542872672918,
+      "grad_norm": 9.4375,
+      "learning_rate": 8.152857045794266e-06,
+      "loss": 0.6672,
+      "step": 98545
+    },
+    {
+      "epoch": 6.69588259274358,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.15073379535263e-06,
+      "loss": 0.8873,
+      "step": 98550
+    },
+    {
+      "epoch": 6.696222312814241,
+      "grad_norm": 7.34375,
+      "learning_rate": 8.148610544910994e-06,
+      "loss": 0.8287,
+      "step": 98555
+    },
+    {
+      "epoch": 6.696562032884903,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.146487294469358e-06,
+      "loss": 0.9227,
+      "step": 98560
+    },
+    {
+      "epoch": 6.696901752955565,
+      "grad_norm": 7.1875,
+      "learning_rate": 8.144364044027722e-06,
+      "loss": 0.7332,
+      "step": 98565
+    },
+    {
+      "epoch": 6.697241473026226,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.142240793586086e-06,
+      "loss": 0.8455,
+      "step": 98570
+    },
+    {
+      "epoch": 6.697581193096888,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.140117543144448e-06,
+      "loss": 0.9342,
+      "step": 98575
+    },
+    {
+      "epoch": 6.69792091316755,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.137994292702814e-06,
+      "loss": 0.8856,
+      "step": 98580
+    },
+    {
+      "epoch": 6.6982606332382115,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.135871042261178e-06,
+      "loss": 0.8106,
+      "step": 98585
+    },
+    {
+      "epoch": 6.698600353308874,
+      "grad_norm": 8.5,
+      "learning_rate": 8.133747791819542e-06,
+      "loss": 0.7883,
+      "step": 98590
+    },
+    {
+      "epoch": 6.698940073379536,
+      "grad_norm": 7.6875,
+      "learning_rate": 8.131624541377906e-06,
+      "loss": 0.7866,
+      "step": 98595
+    },
+    {
+      "epoch": 6.699279793450197,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.129501290936268e-06,
+      "loss": 0.7171,
+      "step": 98600
+    },
+    {
+      "epoch": 6.699619513520859,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.127378040494634e-06,
+      "loss": 0.9165,
+      "step": 98605
+    },
+    {
+      "epoch": 6.699959233591521,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.125254790052998e-06,
+      "loss": 0.8569,
+      "step": 98610
+    },
+    {
+      "epoch": 6.700298953662182,
+      "grad_norm": 7.875,
+      "learning_rate": 8.12313153961136e-06,
+      "loss": 0.7997,
+      "step": 98615
+    },
+    {
+      "epoch": 6.700638673732844,
+      "grad_norm": 8.25,
+      "learning_rate": 8.121008289169726e-06,
+      "loss": 0.9397,
+      "step": 98620
+    },
+    {
+      "epoch": 6.700978393803506,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.118885038728088e-06,
+      "loss": 0.7822,
+      "step": 98625
+    },
+    {
+      "epoch": 6.7013181138741675,
+      "grad_norm": 7.875,
+      "learning_rate": 8.116761788286452e-06,
+      "loss": 0.8891,
+      "step": 98630
+    },
+    {
+      "epoch": 6.70165783394483,
+      "grad_norm": 8.625,
+      "learning_rate": 8.114638537844816e-06,
+      "loss": 0.8653,
+      "step": 98635
+    },
+    {
+      "epoch": 6.701997554015492,
+      "grad_norm": 7.375,
+      "learning_rate": 8.11251528740318e-06,
+      "loss": 0.7748,
+      "step": 98640
+    },
+    {
+      "epoch": 6.702337274086153,
+      "grad_norm": 8.25,
+      "learning_rate": 8.110392036961544e-06,
+      "loss": 0.8878,
+      "step": 98645
+    },
+    {
+      "epoch": 6.702676994156815,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.108268786519908e-06,
+      "loss": 0.8755,
+      "step": 98650
+    },
+    {
+      "epoch": 6.703016714227477,
+      "grad_norm": 8.0,
+      "learning_rate": 8.106145536078272e-06,
+      "loss": 0.7439,
+      "step": 98655
+    },
+    {
+      "epoch": 6.703356434298138,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.104022285636636e-06,
+      "loss": 0.9238,
+      "step": 98660
+    },
+    {
+      "epoch": 6.7036961543688,
+      "grad_norm": 7.53125,
+      "learning_rate": 8.101899035195e-06,
+      "loss": 0.8356,
+      "step": 98665
+    },
+    {
+      "epoch": 6.704035874439462,
+      "grad_norm": 8.25,
+      "learning_rate": 8.099775784753364e-06,
+      "loss": 0.8094,
+      "step": 98670
+    },
+    {
+      "epoch": 6.7043755945101235,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.097652534311728e-06,
+      "loss": 0.9215,
+      "step": 98675
+    },
+    {
+      "epoch": 6.704715314580786,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.095529283870092e-06,
+      "loss": 0.8476,
+      "step": 98680
+    },
+    {
+      "epoch": 6.705055034651448,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.093406033428454e-06,
+      "loss": 0.8215,
+      "step": 98685
+    },
+    {
+      "epoch": 6.705394754722109,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.09128278298682e-06,
+      "loss": 0.8389,
+      "step": 98690
+    },
+    {
+      "epoch": 6.705734474792771,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.089159532545184e-06,
+      "loss": 0.8032,
+      "step": 98695
+    },
+    {
+      "epoch": 6.706074194863432,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.087036282103546e-06,
+      "loss": 0.8656,
+      "step": 98700
+    },
+    {
+      "epoch": 6.706413914934094,
+      "grad_norm": 7.5,
+      "learning_rate": 8.084913031661912e-06,
+      "loss": 0.8376,
+      "step": 98705
+    },
+    {
+      "epoch": 6.706753635004756,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.082789781220274e-06,
+      "loss": 0.8704,
+      "step": 98710
+    },
+    {
+      "epoch": 6.7070933550754175,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.080666530778638e-06,
+      "loss": 0.8787,
+      "step": 98715
+    },
+    {
+      "epoch": 6.70743307514608,
+      "grad_norm": 7.875,
+      "learning_rate": 8.078543280337002e-06,
+      "loss": 0.9344,
+      "step": 98720
+    },
+    {
+      "epoch": 6.707772795216742,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.076420029895366e-06,
+      "loss": 0.8924,
+      "step": 98725
+    },
+    {
+      "epoch": 6.708112515287403,
+      "grad_norm": 8.125,
+      "learning_rate": 8.07429677945373e-06,
+      "loss": 0.8999,
+      "step": 98730
+    },
+    {
+      "epoch": 6.708452235358065,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.072173529012094e-06,
+      "loss": 0.8076,
+      "step": 98735
+    },
+    {
+      "epoch": 6.708791955428727,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.070050278570458e-06,
+      "loss": 0.9074,
+      "step": 98740
+    },
+    {
+      "epoch": 6.709131675499388,
+      "grad_norm": 8.375,
+      "learning_rate": 8.067927028128822e-06,
+      "loss": 0.8056,
+      "step": 98745
+    },
+    {
+      "epoch": 6.70947139557005,
+      "grad_norm": 8.625,
+      "learning_rate": 8.065803777687186e-06,
+      "loss": 0.8351,
+      "step": 98750
+    },
+    {
+      "epoch": 6.709811115640712,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.06368052724555e-06,
+      "loss": 0.9171,
+      "step": 98755
+    },
+    {
+      "epoch": 6.7101508357113735,
+      "grad_norm": 7.5625,
+      "learning_rate": 8.061557276803914e-06,
+      "loss": 0.8836,
+      "step": 98760
+    },
+    {
+      "epoch": 6.710490555782036,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.059434026362278e-06,
+      "loss": 0.8884,
+      "step": 98765
+    },
+    {
+      "epoch": 6.710830275852698,
+      "grad_norm": 9.125,
+      "learning_rate": 8.05731077592064e-06,
+      "loss": 0.9348,
+      "step": 98770
+    },
+    {
+      "epoch": 6.711169995923359,
+      "grad_norm": 8.9375,
+      "learning_rate": 8.055187525479006e-06,
+      "loss": 0.8277,
+      "step": 98775
+    },
+    {
+      "epoch": 6.711509715994021,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.05306427503737e-06,
+      "loss": 0.8612,
+      "step": 98780
+    },
+    {
+      "epoch": 6.711849436064683,
+      "grad_norm": 7.375,
+      "learning_rate": 8.050941024595733e-06,
+      "loss": 0.802,
+      "step": 98785
+    },
+    {
+      "epoch": 6.712189156135344,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.048817774154098e-06,
+      "loss": 0.8969,
+      "step": 98790
+    },
+    {
+      "epoch": 6.712528876206006,
+      "grad_norm": 8.25,
+      "learning_rate": 8.04669452371246e-06,
+      "loss": 0.836,
+      "step": 98795
+    },
+    {
+      "epoch": 6.712868596276668,
+      "grad_norm": 7.625,
+      "learning_rate": 8.044571273270825e-06,
+      "loss": 0.8701,
+      "step": 98800
+    },
+    {
+      "epoch": 6.7132083163473295,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.04244802282919e-06,
+      "loss": 0.8425,
+      "step": 98805
+    },
+    {
+      "epoch": 6.713548036417992,
+      "grad_norm": 7.75,
+      "learning_rate": 8.040324772387553e-06,
+      "loss": 0.8032,
+      "step": 98810
+    },
+    {
+      "epoch": 6.713887756488654,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.038201521945917e-06,
+      "loss": 0.89,
+      "step": 98815
+    },
+    {
+      "epoch": 6.714227476559315,
+      "grad_norm": 8.375,
+      "learning_rate": 8.03607827150428e-06,
+      "loss": 0.8868,
+      "step": 98820
+    },
+    {
+      "epoch": 6.714567196629977,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.033955021062645e-06,
+      "loss": 0.9384,
+      "step": 98825
+    },
+    {
+      "epoch": 6.714906916700639,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.031831770621009e-06,
+      "loss": 0.8913,
+      "step": 98830
+    },
+    {
+      "epoch": 6.7152466367713,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.029708520179373e-06,
+      "loss": 0.881,
+      "step": 98835
+    },
+    {
+      "epoch": 6.715586356841962,
+      "grad_norm": 6.78125,
+      "learning_rate": 8.027585269737737e-06,
+      "loss": 0.8175,
+      "step": 98840
+    },
+    {
+      "epoch": 6.715926076912624,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.0254620192961e-06,
+      "loss": 0.8402,
+      "step": 98845
+    },
+    {
+      "epoch": 6.7162657969832855,
+      "grad_norm": 9.0,
+      "learning_rate": 8.023338768854465e-06,
+      "loss": 0.8548,
+      "step": 98850
+    },
+    {
+      "epoch": 6.716605517053948,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.021215518412829e-06,
+      "loss": 0.8424,
+      "step": 98855
+    },
+    {
+      "epoch": 6.71694523712461,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.019092267971193e-06,
+      "loss": 0.825,
+      "step": 98860
+    },
+    {
+      "epoch": 6.717284957195271,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.016969017529557e-06,
+      "loss": 0.927,
+      "step": 98865
+    },
+    {
+      "epoch": 6.717624677265933,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.014845767087919e-06,
+      "loss": 0.8754,
+      "step": 98870
+    },
+    {
+      "epoch": 6.717964397336595,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.012722516646285e-06,
+      "loss": 0.8133,
+      "step": 98875
+    },
+    {
+      "epoch": 6.718304117407256,
+      "grad_norm": 7.71875,
+      "learning_rate": 8.010599266204647e-06,
+      "loss": 0.8028,
+      "step": 98880
+    },
+    {
+      "epoch": 6.718643837477918,
+      "grad_norm": 6.65625,
+      "learning_rate": 8.008476015763011e-06,
+      "loss": 0.7163,
+      "step": 98885
+    },
+    {
+      "epoch": 6.71898355754858,
+      "grad_norm": 8.125,
+      "learning_rate": 8.006352765321377e-06,
+      "loss": 0.8766,
+      "step": 98890
+    },
+    {
+      "epoch": 6.7193232776192415,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.004229514879739e-06,
+      "loss": 0.8754,
+      "step": 98895
+    },
+    {
+      "epoch": 6.719662997689904,
+      "grad_norm": 8.125,
+      "learning_rate": 8.002106264438103e-06,
+      "loss": 0.9029,
+      "step": 98900
+    },
+    {
+      "epoch": 6.720002717760566,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.999983013996467e-06,
+      "loss": 0.8228,
+      "step": 98905
+    },
+    {
+      "epoch": 6.720342437831227,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.997859763554831e-06,
+      "loss": 0.8847,
+      "step": 98910
+    },
+    {
+      "epoch": 6.720682157901889,
+      "grad_norm": 8.25,
+      "learning_rate": 7.995736513113195e-06,
+      "loss": 0.8322,
+      "step": 98915
+    },
+    {
+      "epoch": 6.72102187797255,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.993613262671559e-06,
+      "loss": 0.8398,
+      "step": 98920
+    },
+    {
+      "epoch": 6.721361598043212,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.991490012229923e-06,
+      "loss": 0.9175,
+      "step": 98925
+    },
+    {
+      "epoch": 6.721701318113874,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.989366761788287e-06,
+      "loss": 0.9052,
+      "step": 98930
+    },
+    {
+      "epoch": 6.7220410381845355,
+      "grad_norm": 7.75,
+      "learning_rate": 7.987243511346651e-06,
+      "loss": 0.8984,
+      "step": 98935
+    },
+    {
+      "epoch": 6.7223807582551975,
+      "grad_norm": 8.375,
+      "learning_rate": 7.985120260905015e-06,
+      "loss": 0.9068,
+      "step": 98940
+    },
+    {
+      "epoch": 6.72272047832586,
+      "grad_norm": 8.625,
+      "learning_rate": 7.982997010463379e-06,
+      "loss": 0.8781,
+      "step": 98945
+    },
+    {
+      "epoch": 6.723060198396521,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.980873760021743e-06,
+      "loss": 0.8648,
+      "step": 98950
+    },
+    {
+      "epoch": 6.723399918467183,
+      "grad_norm": 6.4375,
+      "learning_rate": 7.978750509580105e-06,
+      "loss": 0.8031,
+      "step": 98955
+    },
+    {
+      "epoch": 6.723739638537845,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.976627259138471e-06,
+      "loss": 0.8101,
+      "step": 98960
+    },
+    {
+      "epoch": 6.724079358608506,
+      "grad_norm": 7.75,
+      "learning_rate": 7.974504008696833e-06,
+      "loss": 0.8394,
+      "step": 98965
+    },
+    {
+      "epoch": 6.724419078679168,
+      "grad_norm": 8.25,
+      "learning_rate": 7.972380758255197e-06,
+      "loss": 0.8524,
+      "step": 98970
+    },
+    {
+      "epoch": 6.72475879874983,
+      "grad_norm": 8.5,
+      "learning_rate": 7.970257507813563e-06,
+      "loss": 0.8168,
+      "step": 98975
+    },
+    {
+      "epoch": 6.7250985188204915,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.968134257371925e-06,
+      "loss": 0.831,
+      "step": 98980
+    },
+    {
+      "epoch": 6.7254382388911536,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.96601100693029e-06,
+      "loss": 0.8422,
+      "step": 98985
+    },
+    {
+      "epoch": 6.725777958961816,
+      "grad_norm": 8.125,
+      "learning_rate": 7.963887756488653e-06,
+      "loss": 0.7918,
+      "step": 98990
+    },
+    {
+      "epoch": 6.726117679032477,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.961764506047017e-06,
+      "loss": 0.9019,
+      "step": 98995
+    },
+    {
+      "epoch": 6.726457399103139,
+      "grad_norm": 8.125,
+      "learning_rate": 7.959641255605381e-06,
+      "loss": 0.8469,
+      "step": 99000
+    },
+    {
+      "epoch": 6.726797119173801,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.957518005163745e-06,
+      "loss": 0.8759,
+      "step": 99005
+    },
+    {
+      "epoch": 6.727136839244462,
+      "grad_norm": 8.375,
+      "learning_rate": 7.95539475472211e-06,
+      "loss": 0.8255,
+      "step": 99010
+    },
+    {
+      "epoch": 6.727476559315124,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.953271504280473e-06,
+      "loss": 0.9181,
+      "step": 99015
+    },
+    {
+      "epoch": 6.727816279385786,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.951148253838837e-06,
+      "loss": 0.8764,
+      "step": 99020
+    },
+    {
+      "epoch": 6.7281559994564475,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.949025003397201e-06,
+      "loss": 0.8977,
+      "step": 99025
+    },
+    {
+      "epoch": 6.72849571952711,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.946901752955565e-06,
+      "loss": 0.81,
+      "step": 99030
+    },
+    {
+      "epoch": 6.728835439597772,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.94477850251393e-06,
+      "loss": 0.8253,
+      "step": 99035
+    },
+    {
+      "epoch": 6.729175159668433,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.942655252072292e-06,
+      "loss": 0.8846,
+      "step": 99040
+    },
+    {
+      "epoch": 6.729514879739095,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.940532001630658e-06,
+      "loss": 0.9385,
+      "step": 99045
+    },
+    {
+      "epoch": 6.729854599809757,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.938408751189022e-06,
+      "loss": 0.8935,
+      "step": 99050
+    },
+    {
+      "epoch": 6.730194319880418,
+      "grad_norm": 9.3125,
+      "learning_rate": 7.936285500747384e-06,
+      "loss": 0.8939,
+      "step": 99055
+    },
+    {
+      "epoch": 6.73053403995108,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.93416225030575e-06,
+      "loss": 0.7847,
+      "step": 99060
+    },
+    {
+      "epoch": 6.730873760021742,
+      "grad_norm": 8.75,
+      "learning_rate": 7.932038999864112e-06,
+      "loss": 0.7882,
+      "step": 99065
+    },
+    {
+      "epoch": 6.7312134800924035,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.929915749422476e-06,
+      "loss": 0.8643,
+      "step": 99070
+    },
+    {
+      "epoch": 6.731553200163066,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.92779249898084e-06,
+      "loss": 0.8948,
+      "step": 99075
+    },
+    {
+      "epoch": 6.731892920233728,
+      "grad_norm": 8.5,
+      "learning_rate": 7.925669248539204e-06,
+      "loss": 0.8741,
+      "step": 99080
+    },
+    {
+      "epoch": 6.732232640304389,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.923545998097568e-06,
+      "loss": 0.8672,
+      "step": 99085
+    },
+    {
+      "epoch": 6.732572360375051,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.921422747655932e-06,
+      "loss": 0.9621,
+      "step": 99090
+    },
+    {
+      "epoch": 6.732912080445713,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.919299497214296e-06,
+      "loss": 0.8542,
+      "step": 99095
+    },
+    {
+      "epoch": 6.733251800516374,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.91717624677266e-06,
+      "loss": 0.8189,
+      "step": 99100
+    },
+    {
+      "epoch": 6.733591520587036,
+      "grad_norm": 7.21875,
+      "learning_rate": 7.915052996331024e-06,
+      "loss": 0.812,
+      "step": 99105
+    },
+    {
+      "epoch": 6.733931240657698,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.912929745889388e-06,
+      "loss": 0.8226,
+      "step": 99110
+    },
+    {
+      "epoch": 6.7342709607283595,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.910806495447752e-06,
+      "loss": 0.7809,
+      "step": 99115
+    },
+    {
+      "epoch": 6.734610680799022,
+      "grad_norm": 8.0,
+      "learning_rate": 7.908683245006116e-06,
+      "loss": 0.8337,
+      "step": 99120
+    },
+    {
+      "epoch": 6.734950400869684,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.906559994564478e-06,
+      "loss": 0.7484,
+      "step": 99125
+    },
+    {
+      "epoch": 6.735290120940345,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.904436744122844e-06,
+      "loss": 0.8577,
+      "step": 99130
+    },
+    {
+      "epoch": 6.735629841011007,
+      "grad_norm": 7.875,
+      "learning_rate": 7.902313493681208e-06,
+      "loss": 0.8246,
+      "step": 99135
+    },
+    {
+      "epoch": 6.735969561081669,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.90019024323957e-06,
+      "loss": 0.8755,
+      "step": 99140
+    },
+    {
+      "epoch": 6.73630928115233,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.898066992797936e-06,
+      "loss": 0.88,
+      "step": 99145
+    },
+    {
+      "epoch": 6.736649001222992,
+      "grad_norm": 8.25,
+      "learning_rate": 7.895943742356298e-06,
+      "loss": 0.949,
+      "step": 99150
+    },
+    {
+      "epoch": 6.736988721293654,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.893820491914662e-06,
+      "loss": 0.807,
+      "step": 99155
+    },
+    {
+      "epoch": 6.7373284413643155,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.891697241473026e-06,
+      "loss": 0.8203,
+      "step": 99160
+    },
+    {
+      "epoch": 6.737668161434978,
+      "grad_norm": 8.125,
+      "learning_rate": 7.88957399103139e-06,
+      "loss": 0.8443,
+      "step": 99165
+    },
+    {
+      "epoch": 6.73800788150564,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.887450740589754e-06,
+      "loss": 0.8902,
+      "step": 99170
+    },
+    {
+      "epoch": 6.738347601576301,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.885327490148118e-06,
+      "loss": 0.8672,
+      "step": 99175
+    },
+    {
+      "epoch": 6.738687321646963,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.883204239706482e-06,
+      "loss": 0.8368,
+      "step": 99180
+    },
+    {
+      "epoch": 6.739027041717625,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.881080989264846e-06,
+      "loss": 0.8035,
+      "step": 99185
+    },
+    {
+      "epoch": 6.739366761788286,
+      "grad_norm": 7.09375,
+      "learning_rate": 7.87895773882321e-06,
+      "loss": 0.8514,
+      "step": 99190
+    },
+    {
+      "epoch": 6.739706481858948,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.876834488381574e-06,
+      "loss": 0.8717,
+      "step": 99195
+    },
+    {
+      "epoch": 6.74004620192961,
+      "grad_norm": 8.25,
+      "learning_rate": 7.874711237939938e-06,
+      "loss": 0.7689,
+      "step": 99200
+    },
+    {
+      "epoch": 6.7403859220002715,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.872587987498302e-06,
+      "loss": 0.8107,
+      "step": 99205
+    },
+    {
+      "epoch": 6.740725642070934,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.870464737056665e-06,
+      "loss": 0.9086,
+      "step": 99210
+    },
+    {
+      "epoch": 6.741065362141596,
+      "grad_norm": 7.25,
+      "learning_rate": 7.86834148661503e-06,
+      "loss": 0.8099,
+      "step": 99215
+    },
+    {
+      "epoch": 6.741405082212257,
+      "grad_norm": 8.375,
+      "learning_rate": 7.866218236173394e-06,
+      "loss": 0.8429,
+      "step": 99220
+    },
+    {
+      "epoch": 6.741744802282919,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.864094985731757e-06,
+      "loss": 0.8424,
+      "step": 99225
+    },
+    {
+      "epoch": 6.742084522353581,
+      "grad_norm": 9.0,
+      "learning_rate": 7.861971735290122e-06,
+      "loss": 0.8793,
+      "step": 99230
+    },
+    {
+      "epoch": 6.742424242424242,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.859848484848485e-06,
+      "loss": 0.9439,
+      "step": 99235
+    },
+    {
+      "epoch": 6.742763962494904,
+      "grad_norm": 8.75,
+      "learning_rate": 7.857725234406849e-06,
+      "loss": 0.9408,
+      "step": 99240
+    },
+    {
+      "epoch": 6.743103682565566,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.855601983965213e-06,
+      "loss": 0.8045,
+      "step": 99245
+    },
+    {
+      "epoch": 6.7434434026362275,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.853478733523577e-06,
+      "loss": 0.8627,
+      "step": 99250
+    },
+    {
+      "epoch": 6.74378312270689,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.85135548308194e-06,
+      "loss": 0.8145,
+      "step": 99255
+    },
+    {
+      "epoch": 6.744122842777552,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.849232232640305e-06,
+      "loss": 0.8371,
+      "step": 99260
+    },
+    {
+      "epoch": 6.744462562848213,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.847108982198669e-06,
+      "loss": 0.7984,
+      "step": 99265
+    },
+    {
+      "epoch": 6.744802282918875,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.844985731757033e-06,
+      "loss": 0.773,
+      "step": 99270
+    },
+    {
+      "epoch": 6.745142002989537,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.842862481315397e-06,
+      "loss": 0.7667,
+      "step": 99275
+    },
+    {
+      "epoch": 6.745481723060198,
+      "grad_norm": 7.75,
+      "learning_rate": 7.84073923087376e-06,
+      "loss": 0.8298,
+      "step": 99280
+    },
+    {
+      "epoch": 6.74582144313086,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.838615980432125e-06,
+      "loss": 0.8054,
+      "step": 99285
+    },
+    {
+      "epoch": 6.746161163201522,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.836492729990489e-06,
+      "loss": 0.8125,
+      "step": 99290
+    },
+    {
+      "epoch": 6.746500883272184,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.834369479548851e-06,
+      "loss": 0.8312,
+      "step": 99295
+    },
+    {
+      "epoch": 6.746840603342846,
+      "grad_norm": 7.75,
+      "learning_rate": 7.832246229107217e-06,
+      "loss": 0.8889,
+      "step": 99300
+    },
+    {
+      "epoch": 6.747180323413508,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.83012297866558e-06,
+      "loss": 0.8567,
+      "step": 99305
+    },
+    {
+      "epoch": 6.747520043484169,
+      "grad_norm": 7.28125,
+      "learning_rate": 7.827999728223943e-06,
+      "loss": 0.8297,
+      "step": 99310
+    },
+    {
+      "epoch": 6.747859763554831,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.825876477782309e-06,
+      "loss": 0.8458,
+      "step": 99315
+    },
+    {
+      "epoch": 6.748199483625493,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.823753227340671e-06,
+      "loss": 0.8688,
+      "step": 99320
+    },
+    {
+      "epoch": 6.748539203696154,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.821629976899035e-06,
+      "loss": 0.9056,
+      "step": 99325
+    },
+    {
+      "epoch": 6.748878923766816,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.8195067264574e-06,
+      "loss": 0.8169,
+      "step": 99330
+    },
+    {
+      "epoch": 6.749218643837478,
+      "grad_norm": 9.0,
+      "learning_rate": 7.817383476015763e-06,
+      "loss": 0.8839,
+      "step": 99335
+    },
+    {
+      "epoch": 6.74955836390814,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.815260225574127e-06,
+      "loss": 0.8044,
+      "step": 99340
+    },
+    {
+      "epoch": 6.749898083978802,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.813136975132491e-06,
+      "loss": 0.9335,
+      "step": 99345
+    },
+    {
+      "epoch": 6.750237804049464,
+      "grad_norm": 7.5,
+      "learning_rate": 7.811013724690855e-06,
+      "loss": 0.6873,
+      "step": 99350
+    },
+    {
+      "epoch": 6.750577524120125,
+      "grad_norm": 7.75,
+      "learning_rate": 7.808890474249219e-06,
+      "loss": 0.7817,
+      "step": 99355
+    },
+    {
+      "epoch": 6.750917244190787,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.806767223807583e-06,
+      "loss": 0.8174,
+      "step": 99360
+    },
+    {
+      "epoch": 6.751256964261449,
+      "grad_norm": 7.21875,
+      "learning_rate": 7.804643973365947e-06,
+      "loss": 0.8606,
+      "step": 99365
+    },
+    {
+      "epoch": 6.75159668433211,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.802520722924311e-06,
+      "loss": 0.8773,
+      "step": 99370
+    },
+    {
+      "epoch": 6.751936404402772,
+      "grad_norm": 8.125,
+      "learning_rate": 7.800397472482675e-06,
+      "loss": 0.8509,
+      "step": 99375
+    },
+    {
+      "epoch": 6.7522761244734335,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.798274222041039e-06,
+      "loss": 0.8355,
+      "step": 99380
+    },
+    {
+      "epoch": 6.752615844544096,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.796150971599403e-06,
+      "loss": 0.8078,
+      "step": 99385
+    },
+    {
+      "epoch": 6.752955564614758,
+      "grad_norm": 9.25,
+      "learning_rate": 7.794027721157767e-06,
+      "loss": 0.8675,
+      "step": 99390
+    },
+    {
+      "epoch": 6.753295284685419,
+      "grad_norm": 8.25,
+      "learning_rate": 7.79190447071613e-06,
+      "loss": 0.8632,
+      "step": 99395
+    },
+    {
+      "epoch": 6.753635004756081,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.789781220274495e-06,
+      "loss": 0.9583,
+      "step": 99400
+    },
+    {
+      "epoch": 6.753974724826743,
+      "grad_norm": 8.125,
+      "learning_rate": 7.787657969832857e-06,
+      "loss": 0.7554,
+      "step": 99405
+    },
+    {
+      "epoch": 6.754314444897404,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.785534719391221e-06,
+      "loss": 0.7801,
+      "step": 99410
+    },
+    {
+      "epoch": 6.754654164968066,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.783411468949587e-06,
+      "loss": 0.7374,
+      "step": 99415
+    },
+    {
+      "epoch": 6.754993885038728,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.78128821850795e-06,
+      "loss": 0.865,
+      "step": 99420
+    },
+    {
+      "epoch": 6.7553336051093895,
+      "grad_norm": 8.25,
+      "learning_rate": 7.779164968066313e-06,
+      "loss": 0.91,
+      "step": 99425
+    },
+    {
+      "epoch": 6.755673325180052,
+      "grad_norm": 7.0,
+      "learning_rate": 7.777041717624677e-06,
+      "loss": 0.8816,
+      "step": 99430
+    },
+    {
+      "epoch": 6.756013045250714,
+      "grad_norm": 8.375,
+      "learning_rate": 7.774918467183041e-06,
+      "loss": 0.8766,
+      "step": 99435
+    },
+    {
+      "epoch": 6.756352765321375,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.772795216741405e-06,
+      "loss": 0.8848,
+      "step": 99440
+    },
+    {
+      "epoch": 6.756692485392037,
+      "grad_norm": 9.5,
+      "learning_rate": 7.77067196629977e-06,
+      "loss": 0.8486,
+      "step": 99445
+    },
+    {
+      "epoch": 6.757032205462699,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.768548715858133e-06,
+      "loss": 0.8036,
+      "step": 99450
+    },
+    {
+      "epoch": 6.75737192553336,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.766425465416497e-06,
+      "loss": 0.9207,
+      "step": 99455
+    },
+    {
+      "epoch": 6.757711645604022,
+      "grad_norm": 7.5,
+      "learning_rate": 7.764302214974861e-06,
+      "loss": 0.9255,
+      "step": 99460
+    },
+    {
+      "epoch": 6.758051365674684,
+      "grad_norm": 7.625,
+      "learning_rate": 7.762178964533225e-06,
+      "loss": 0.8286,
+      "step": 99465
+    },
+    {
+      "epoch": 6.7583910857453455,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.76005571409159e-06,
+      "loss": 0.8269,
+      "step": 99470
+    },
+    {
+      "epoch": 6.758730805816008,
+      "grad_norm": 8.875,
+      "learning_rate": 7.757932463649953e-06,
+      "loss": 0.7506,
+      "step": 99475
+    },
+    {
+      "epoch": 6.75907052588667,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.755809213208316e-06,
+      "loss": 0.7878,
+      "step": 99480
+    },
+    {
+      "epoch": 6.759410245957331,
+      "grad_norm": 7.5,
+      "learning_rate": 7.753685962766681e-06,
+      "loss": 0.9204,
+      "step": 99485
+    },
+    {
+      "epoch": 6.759749966027993,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.751562712325044e-06,
+      "loss": 0.8816,
+      "step": 99490
+    },
+    {
+      "epoch": 6.760089686098655,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.749439461883408e-06,
+      "loss": 0.8844,
+      "step": 99495
+    },
+    {
+      "epoch": 6.760429406169316,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.747316211441773e-06,
+      "loss": 0.8598,
+      "step": 99500
+    },
+    {
+      "epoch": 6.760769126239978,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.745192961000136e-06,
+      "loss": 0.8545,
+      "step": 99505
+    },
+    {
+      "epoch": 6.76110884631064,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.7430697105585e-06,
+      "loss": 0.8761,
+      "step": 99510
+    },
+    {
+      "epoch": 6.7614485663813015,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.740946460116864e-06,
+      "loss": 0.8766,
+      "step": 99515
+    },
+    {
+      "epoch": 6.761788286451964,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.738823209675228e-06,
+      "loss": 0.8074,
+      "step": 99520
+    },
+    {
+      "epoch": 6.762128006522626,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.736699959233592e-06,
+      "loss": 1.0042,
+      "step": 99525
+    },
+    {
+      "epoch": 6.762467726593287,
+      "grad_norm": 7.4375,
+      "learning_rate": 7.734576708791956e-06,
+      "loss": 0.8133,
+      "step": 99530
+    },
+    {
+      "epoch": 6.762807446663949,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.73245345835032e-06,
+      "loss": 0.7809,
+      "step": 99535
+    },
+    {
+      "epoch": 6.763147166734611,
+      "grad_norm": 8.5,
+      "learning_rate": 7.730330207908684e-06,
+      "loss": 0.8379,
+      "step": 99540
+    },
+    {
+      "epoch": 6.763486886805272,
+      "grad_norm": 8.25,
+      "learning_rate": 7.728206957467048e-06,
+      "loss": 0.8259,
+      "step": 99545
+    },
+    {
+      "epoch": 6.763826606875934,
+      "grad_norm": 9.125,
+      "learning_rate": 7.726083707025412e-06,
+      "loss": 0.9004,
+      "step": 99550
+    },
+    {
+      "epoch": 6.764166326946596,
+      "grad_norm": 8.0,
+      "learning_rate": 7.723960456583776e-06,
+      "loss": 0.8755,
+      "step": 99555
+    },
+    {
+      "epoch": 6.7645060470172576,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.72183720614214e-06,
+      "loss": 0.8125,
+      "step": 99560
+    },
+    {
+      "epoch": 6.76484576708792,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.719713955700502e-06,
+      "loss": 0.9475,
+      "step": 99565
+    },
+    {
+      "epoch": 6.765185487158582,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.717590705258868e-06,
+      "loss": 0.8512,
+      "step": 99570
+    },
+    {
+      "epoch": 6.765525207229243,
+      "grad_norm": 8.0,
+      "learning_rate": 7.715467454817232e-06,
+      "loss": 0.9127,
+      "step": 99575
+    },
+    {
+      "epoch": 6.765864927299905,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.713344204375594e-06,
+      "loss": 0.8441,
+      "step": 99580
+    },
+    {
+      "epoch": 6.766204647370567,
+      "grad_norm": 7.28125,
+      "learning_rate": 7.71122095393396e-06,
+      "loss": 0.8943,
+      "step": 99585
+    },
+    {
+      "epoch": 6.766544367441228,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.709097703492322e-06,
+      "loss": 0.7996,
+      "step": 99590
+    },
+    {
+      "epoch": 6.76688408751189,
+      "grad_norm": 8.125,
+      "learning_rate": 7.706974453050686e-06,
+      "loss": 0.805,
+      "step": 99595
+    },
+    {
+      "epoch": 6.7672238075825515,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.70485120260905e-06,
+      "loss": 0.7766,
+      "step": 99600
+    },
+    {
+      "epoch": 6.767563527653214,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.702727952167414e-06,
+      "loss": 0.9067,
+      "step": 99605
+    },
+    {
+      "epoch": 6.767903247723876,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.700604701725778e-06,
+      "loss": 0.9156,
+      "step": 99610
+    },
+    {
+      "epoch": 6.768242967794537,
+      "grad_norm": 8.875,
+      "learning_rate": 7.698481451284142e-06,
+      "loss": 0.8665,
+      "step": 99615
+    },
+    {
+      "epoch": 6.768582687865199,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.696358200842506e-06,
+      "loss": 0.7833,
+      "step": 99620
+    },
+    {
+      "epoch": 6.768922407935861,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.69423495040087e-06,
+      "loss": 0.8934,
+      "step": 99625
+    },
+    {
+      "epoch": 6.769262128006522,
+      "grad_norm": 8.125,
+      "learning_rate": 7.692111699959234e-06,
+      "loss": 0.7609,
+      "step": 99630
+    },
+    {
+      "epoch": 6.769601848077184,
+      "grad_norm": 8.125,
+      "learning_rate": 7.689988449517598e-06,
+      "loss": 0.8978,
+      "step": 99635
+    },
+    {
+      "epoch": 6.769941568147846,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.687865199075962e-06,
+      "loss": 0.7287,
+      "step": 99640
+    },
+    {
+      "epoch": 6.7702812882185075,
+      "grad_norm": 6.65625,
+      "learning_rate": 7.685741948634326e-06,
+      "loss": 0.8424,
+      "step": 99645
+    },
+    {
+      "epoch": 6.77062100828917,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.683618698192688e-06,
+      "loss": 0.804,
+      "step": 99650
+    },
+    {
+      "epoch": 6.770960728359832,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.681495447751054e-06,
+      "loss": 0.8305,
+      "step": 99655
+    },
+    {
+      "epoch": 6.771300448430493,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.679372197309418e-06,
+      "loss": 0.8545,
+      "step": 99660
+    },
+    {
+      "epoch": 6.771640168501155,
+      "grad_norm": 8.125,
+      "learning_rate": 7.67724894686778e-06,
+      "loss": 0.8513,
+      "step": 99665
+    },
+    {
+      "epoch": 6.771979888571817,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.675125696426146e-06,
+      "loss": 0.8524,
+      "step": 99670
+    },
+    {
+      "epoch": 6.772319608642478,
+      "grad_norm": 8.625,
+      "learning_rate": 7.673002445984508e-06,
+      "loss": 0.7787,
+      "step": 99675
+    },
+    {
+      "epoch": 6.77265932871314,
+      "grad_norm": 7.03125,
+      "learning_rate": 7.670879195542872e-06,
+      "loss": 0.7917,
+      "step": 99680
+    },
+    {
+      "epoch": 6.772999048783802,
+      "grad_norm": 9.0,
+      "learning_rate": 7.668755945101236e-06,
+      "loss": 0.851,
+      "step": 99685
+    },
+    {
+      "epoch": 6.7733387688544635,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.6666326946596e-06,
+      "loss": 0.9048,
+      "step": 99690
+    },
+    {
+      "epoch": 6.773678488925126,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.664509444217964e-06,
+      "loss": 0.9556,
+      "step": 99695
+    },
+    {
+      "epoch": 6.774018208995788,
+      "grad_norm": 7.375,
+      "learning_rate": 7.662386193776328e-06,
+      "loss": 0.9201,
+      "step": 99700
+    },
+    {
+      "epoch": 6.774357929066449,
+      "grad_norm": 8.375,
+      "learning_rate": 7.660262943334692e-06,
+      "loss": 0.8557,
+      "step": 99705
+    },
+    {
+      "epoch": 6.774697649137111,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.658139692893056e-06,
+      "loss": 0.7775,
+      "step": 99710
+    },
+    {
+      "epoch": 6.775037369207773,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.65601644245142e-06,
+      "loss": 0.7755,
+      "step": 99715
+    },
+    {
+      "epoch": 6.775377089278434,
+      "grad_norm": 6.71875,
+      "learning_rate": 7.653893192009784e-06,
+      "loss": 0.7725,
+      "step": 99720
+    },
+    {
+      "epoch": 6.775716809349096,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.651769941568148e-06,
+      "loss": 0.8397,
+      "step": 99725
+    },
+    {
+      "epoch": 6.776056529419758,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.649646691126512e-06,
+      "loss": 0.8198,
+      "step": 99730
+    },
+    {
+      "epoch": 6.7763962494904195,
+      "grad_norm": 8.875,
+      "learning_rate": 7.647523440684875e-06,
+      "loss": 0.8158,
+      "step": 99735
+    },
+    {
+      "epoch": 6.776735969561082,
+      "grad_norm": 8.0,
+      "learning_rate": 7.64540019024324e-06,
+      "loss": 0.8146,
+      "step": 99740
+    },
+    {
+      "epoch": 6.777075689631744,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.643276939801604e-06,
+      "loss": 0.8321,
+      "step": 99745
+    },
+    {
+      "epoch": 6.777415409702405,
+      "grad_norm": 8.125,
+      "learning_rate": 7.641153689359967e-06,
+      "loss": 0.8663,
+      "step": 99750
+    },
+    {
+      "epoch": 6.777755129773067,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.639030438918333e-06,
+      "loss": 0.9396,
+      "step": 99755
+    },
+    {
+      "epoch": 6.778094849843729,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.636907188476695e-06,
+      "loss": 0.7209,
+      "step": 99760
+    },
+    {
+      "epoch": 6.77843456991439,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.634783938035059e-06,
+      "loss": 0.8447,
+      "step": 99765
+    },
+    {
+      "epoch": 6.778774289985052,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.632660687593425e-06,
+      "loss": 0.756,
+      "step": 99770
+    },
+    {
+      "epoch": 6.779114010055714,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.630537437151787e-06,
+      "loss": 0.8645,
+      "step": 99775
+    },
+    {
+      "epoch": 6.7794537301263755,
+      "grad_norm": 6.1875,
+      "learning_rate": 7.628414186710151e-06,
+      "loss": 0.8453,
+      "step": 99780
+    },
+    {
+      "epoch": 6.779793450197038,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.626290936268516e-06,
+      "loss": 0.8402,
+      "step": 99785
+    },
+    {
+      "epoch": 6.7801331702677,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.624167685826879e-06,
+      "loss": 0.8857,
+      "step": 99790
+    },
+    {
+      "epoch": 6.780472890338361,
+      "grad_norm": 8.25,
+      "learning_rate": 7.622044435385244e-06,
+      "loss": 0.888,
+      "step": 99795
+    },
+    {
+      "epoch": 6.780812610409023,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.619921184943607e-06,
+      "loss": 0.849,
+      "step": 99800
+    },
+    {
+      "epoch": 6.781152330479685,
+      "grad_norm": 7.5,
+      "learning_rate": 7.61779793450197e-06,
+      "loss": 0.8771,
+      "step": 99805
+    },
+    {
+      "epoch": 6.781492050550346,
+      "grad_norm": 8.625,
+      "learning_rate": 7.615674684060335e-06,
+      "loss": 0.8249,
+      "step": 99810
+    },
+    {
+      "epoch": 6.781831770621008,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.613551433618699e-06,
+      "loss": 0.8342,
+      "step": 99815
+    },
+    {
+      "epoch": 6.78217149069167,
+      "grad_norm": 6.84375,
+      "learning_rate": 7.611428183177062e-06,
+      "loss": 0.8856,
+      "step": 99820
+    },
+    {
+      "epoch": 6.7825112107623315,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.609304932735427e-06,
+      "loss": 0.7872,
+      "step": 99825
+    },
+    {
+      "epoch": 6.782850930832994,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.60718168229379e-06,
+      "loss": 0.7806,
+      "step": 99830
+    },
+    {
+      "epoch": 6.783190650903656,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.605058431852154e-06,
+      "loss": 0.8128,
+      "step": 99835
+    },
+    {
+      "epoch": 6.783530370974317,
+      "grad_norm": 8.75,
+      "learning_rate": 7.602935181410519e-06,
+      "loss": 0.9872,
+      "step": 99840
+    },
+    {
+      "epoch": 6.783870091044979,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.600811930968882e-06,
+      "loss": 0.8366,
+      "step": 99845
+    },
+    {
+      "epoch": 6.784209811115641,
+      "grad_norm": 7.75,
+      "learning_rate": 7.598688680527245e-06,
+      "loss": 0.8906,
+      "step": 99850
+    },
+    {
+      "epoch": 6.784549531186302,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.59656543008561e-06,
+      "loss": 0.8894,
+      "step": 99855
+    },
+    {
+      "epoch": 6.784889251256964,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.594442179643973e-06,
+      "loss": 0.8528,
+      "step": 99860
+    },
+    {
+      "epoch": 6.785228971327626,
+      "grad_norm": 7.75,
+      "learning_rate": 7.592318929202337e-06,
+      "loss": 0.8525,
+      "step": 99865
+    },
+    {
+      "epoch": 6.7855686913982876,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.590195678760702e-06,
+      "loss": 0.8472,
+      "step": 99870
+    },
+    {
+      "epoch": 6.78590841146895,
+      "grad_norm": 8.625,
+      "learning_rate": 7.588072428319065e-06,
+      "loss": 0.8195,
+      "step": 99875
+    },
+    {
+      "epoch": 6.786248131539612,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.58594917787743e-06,
+      "loss": 0.9407,
+      "step": 99880
+    },
+    {
+      "epoch": 6.786587851610273,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.583825927435793e-06,
+      "loss": 0.781,
+      "step": 99885
+    },
+    {
+      "epoch": 6.786927571680935,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.581702676994157e-06,
+      "loss": 0.9998,
+      "step": 99890
+    },
+    {
+      "epoch": 6.787267291751597,
+      "grad_norm": 7.3125,
+      "learning_rate": 7.579579426552521e-06,
+      "loss": 0.89,
+      "step": 99895
+    },
+    {
+      "epoch": 6.787607011822258,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.577456176110885e-06,
+      "loss": 0.8495,
+      "step": 99900
+    },
+    {
+      "epoch": 6.78794673189292,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.575332925669248e-06,
+      "loss": 0.8696,
+      "step": 99905
+    },
+    {
+      "epoch": 6.788286451963582,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.573209675227613e-06,
+      "loss": 0.8324,
+      "step": 99910
+    },
+    {
+      "epoch": 6.788626172034244,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.571086424785976e-06,
+      "loss": 0.8165,
+      "step": 99915
+    },
+    {
+      "epoch": 6.788965892104906,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.56896317434434e-06,
+      "loss": 0.8189,
+      "step": 99920
+    },
+    {
+      "epoch": 6.789305612175568,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.566839923902705e-06,
+      "loss": 0.7641,
+      "step": 99925
+    },
+    {
+      "epoch": 6.789645332246229,
+      "grad_norm": 8.25,
+      "learning_rate": 7.564716673461068e-06,
+      "loss": 0.9089,
+      "step": 99930
+    },
+    {
+      "epoch": 6.789985052316891,
+      "grad_norm": 8.0,
+      "learning_rate": 7.5625934230194316e-06,
+      "loss": 0.8787,
+      "step": 99935
+    },
+    {
+      "epoch": 6.790324772387553,
+      "grad_norm": 6.78125,
+      "learning_rate": 7.560470172577796e-06,
+      "loss": 0.8775,
+      "step": 99940
+    },
+    {
+      "epoch": 6.790664492458214,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.5583469221361596e-06,
+      "loss": 0.7922,
+      "step": 99945
+    },
+    {
+      "epoch": 6.791004212528876,
+      "grad_norm": 8.625,
+      "learning_rate": 7.5562236716945236e-06,
+      "loss": 0.7812,
+      "step": 99950
+    },
+    {
+      "epoch": 6.791343932599538,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.554100421252888e-06,
+      "loss": 0.9048,
+      "step": 99955
+    },
+    {
+      "epoch": 6.7916836526702,
+      "grad_norm": 6.625,
+      "learning_rate": 7.551977170811252e-06,
+      "loss": 0.86,
+      "step": 99960
+    },
+    {
+      "epoch": 6.792023372740862,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.5498539203696164e-06,
+      "loss": 0.7929,
+      "step": 99965
+    },
+    {
+      "epoch": 6.792363092811524,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.54773066992798e-06,
+      "loss": 0.8276,
+      "step": 99970
+    },
+    {
+      "epoch": 6.792702812882185,
+      "grad_norm": 7.375,
+      "learning_rate": 7.545607419486344e-06,
+      "loss": 0.8594,
+      "step": 99975
+    },
+    {
+      "epoch": 6.793042532952847,
+      "grad_norm": 9.4375,
+      "learning_rate": 7.5434841690447084e-06,
+      "loss": 0.8134,
+      "step": 99980
+    },
+    {
+      "epoch": 6.793382253023509,
+      "grad_norm": 8.25,
+      "learning_rate": 7.541360918603072e-06,
+      "loss": 0.9688,
+      "step": 99985
+    },
+    {
+      "epoch": 6.79372197309417,
+      "grad_norm": 6.8125,
+      "learning_rate": 7.539237668161435e-06,
+      "loss": 0.7189,
+      "step": 99990
+    },
+    {
+      "epoch": 6.794061693164832,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.5371144177198e-06,
+      "loss": 0.785,
+      "step": 99995
+    },
+    {
+      "epoch": 6.794401413235494,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.534991167278163e-06,
+      "loss": 0.9031,
+      "step": 100000
+    },
+    {
+      "epoch": 6.794741133306156,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.532867916836527e-06,
+      "loss": 0.8258,
+      "step": 100005
+    },
+    {
+      "epoch": 6.795080853376818,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.530744666394892e-06,
+      "loss": 0.7964,
+      "step": 100010
+    },
+    {
+      "epoch": 6.79542057344748,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.528621415953255e-06,
+      "loss": 0.9004,
+      "step": 100015
+    },
+    {
+      "epoch": 6.795760293518141,
+      "grad_norm": 8.0,
+      "learning_rate": 7.526498165511618e-06,
+      "loss": 0.8509,
+      "step": 100020
+    },
+    {
+      "epoch": 6.796100013588803,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.524374915069983e-06,
+      "loss": 0.8445,
+      "step": 100025
+    },
+    {
+      "epoch": 6.796439733659465,
+      "grad_norm": 8.25,
+      "learning_rate": 7.522251664628347e-06,
+      "loss": 0.8215,
+      "step": 100030
+    },
+    {
+      "epoch": 6.796779453730126,
+      "grad_norm": 8.875,
+      "learning_rate": 7.52012841418671e-06,
+      "loss": 0.87,
+      "step": 100035
+    },
+    {
+      "epoch": 6.797119173800788,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.518005163745075e-06,
+      "loss": 0.7968,
+      "step": 100040
+    },
+    {
+      "epoch": 6.79745889387145,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.515881913303438e-06,
+      "loss": 0.8497,
+      "step": 100045
+    },
+    {
+      "epoch": 6.797798613942112,
+      "grad_norm": 7.875,
+      "learning_rate": 7.513758662861803e-06,
+      "loss": 0.8028,
+      "step": 100050
+    },
+    {
+      "epoch": 6.798138334012774,
+      "grad_norm": 8.5,
+      "learning_rate": 7.511635412420166e-06,
+      "loss": 0.8392,
+      "step": 100055
+    },
+    {
+      "epoch": 6.798478054083435,
+      "grad_norm": 6.90625,
+      "learning_rate": 7.50951216197853e-06,
+      "loss": 0.8805,
+      "step": 100060
+    },
+    {
+      "epoch": 6.798817774154097,
+      "grad_norm": 8.25,
+      "learning_rate": 7.507388911536895e-06,
+      "loss": 0.9367,
+      "step": 100065
+    },
+    {
+      "epoch": 6.799157494224759,
+      "grad_norm": 8.25,
+      "learning_rate": 7.505265661095258e-06,
+      "loss": 0.9024,
+      "step": 100070
+    },
+    {
+      "epoch": 6.79949721429542,
+      "grad_norm": 8.25,
+      "learning_rate": 7.503142410653621e-06,
+      "loss": 0.7736,
+      "step": 100075
+    },
+    {
+      "epoch": 6.799836934366082,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.501019160211986e-06,
+      "loss": 0.9894,
+      "step": 100080
+    },
+    {
+      "epoch": 6.800176654436744,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.49889590977035e-06,
+      "loss": 0.8223,
+      "step": 100085
+    },
+    {
+      "epoch": 6.8005163745074055,
+      "grad_norm": 8.0,
+      "learning_rate": 7.496772659328713e-06,
+      "loss": 0.829,
+      "step": 100090
+    },
+    {
+      "epoch": 6.800856094578068,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.494649408887078e-06,
+      "loss": 0.9202,
+      "step": 100095
+    },
+    {
+      "epoch": 6.80119581464873,
+      "grad_norm": 7.5,
+      "learning_rate": 7.492526158445441e-06,
+      "loss": 0.8126,
+      "step": 100100
+    },
+    {
+      "epoch": 6.801535534719391,
+      "grad_norm": 7.28125,
+      "learning_rate": 7.490402908003804e-06,
+      "loss": 0.8288,
+      "step": 100105
+    },
+    {
+      "epoch": 6.801875254790053,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.488279657562169e-06,
+      "loss": 0.7296,
+      "step": 100110
+    },
+    {
+      "epoch": 6.802214974860715,
+      "grad_norm": 8.5,
+      "learning_rate": 7.486156407120533e-06,
+      "loss": 0.9363,
+      "step": 100115
+    },
+    {
+      "epoch": 6.802554694931376,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.484033156678896e-06,
+      "loss": 0.7248,
+      "step": 100120
+    },
+    {
+      "epoch": 6.802894415002038,
+      "grad_norm": 7.125,
+      "learning_rate": 7.481909906237261e-06,
+      "loss": 0.9448,
+      "step": 100125
+    },
+    {
+      "epoch": 6.8032341350727,
+      "grad_norm": 6.75,
+      "learning_rate": 7.479786655795624e-06,
+      "loss": 0.8479,
+      "step": 100130
+    },
+    {
+      "epoch": 6.8035738551433615,
+      "grad_norm": 6.6875,
+      "learning_rate": 7.477663405353989e-06,
+      "loss": 0.8987,
+      "step": 100135
+    },
+    {
+      "epoch": 6.803913575214024,
+      "grad_norm": 8.875,
+      "learning_rate": 7.475540154912352e-06,
+      "loss": 0.9443,
+      "step": 100140
+    },
+    {
+      "epoch": 6.804253295284686,
+      "grad_norm": 7.625,
+      "learning_rate": 7.473416904470716e-06,
+      "loss": 0.7991,
+      "step": 100145
+    },
+    {
+      "epoch": 6.804593015355347,
+      "grad_norm": 7.28125,
+      "learning_rate": 7.471293654029081e-06,
+      "loss": 0.8903,
+      "step": 100150
+    },
+    {
+      "epoch": 6.804932735426009,
+      "grad_norm": 6.75,
+      "learning_rate": 7.469170403587444e-06,
+      "loss": 0.8693,
+      "step": 100155
+    },
+    {
+      "epoch": 6.805272455496671,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.4670471531458075e-06,
+      "loss": 0.7422,
+      "step": 100160
+    },
+    {
+      "epoch": 6.805612175567332,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.464923902704172e-06,
+      "loss": 0.8381,
+      "step": 100165
+    },
+    {
+      "epoch": 6.805951895637994,
+      "grad_norm": 8.125,
+      "learning_rate": 7.462800652262536e-06,
+      "loss": 0.7631,
+      "step": 100170
+    },
+    {
+      "epoch": 6.806291615708656,
+      "grad_norm": 6.84375,
+      "learning_rate": 7.4606774018208995e-06,
+      "loss": 0.8337,
+      "step": 100175
+    },
+    {
+      "epoch": 6.806631335779318,
+      "grad_norm": 7.3125,
+      "learning_rate": 7.458554151379264e-06,
+      "loss": 0.846,
+      "step": 100180
+    },
+    {
+      "epoch": 6.80697105584998,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.4564309009376275e-06,
+      "loss": 0.8354,
+      "step": 100185
+    },
+    {
+      "epoch": 6.807310775920642,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.454307650495991e-06,
+      "loss": 0.8773,
+      "step": 100190
+    },
+    {
+      "epoch": 6.807650495991303,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.4521844000543555e-06,
+      "loss": 0.8293,
+      "step": 100195
+    },
+    {
+      "epoch": 6.807990216061965,
+      "grad_norm": 8.0,
+      "learning_rate": 7.4500611496127195e-06,
+      "loss": 0.9024,
+      "step": 100200
+    },
+    {
+      "epoch": 6.808329936132627,
+      "grad_norm": 7.0625,
+      "learning_rate": 7.447937899171083e-06,
+      "loss": 0.7379,
+      "step": 100205
+    },
+    {
+      "epoch": 6.808669656203288,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.4458146487294475e-06,
+      "loss": 0.7981,
+      "step": 100210
+    },
+    {
+      "epoch": 6.80900937627395,
+      "grad_norm": 9.25,
+      "learning_rate": 7.443691398287811e-06,
+      "loss": 0.9439,
+      "step": 100215
+    },
+    {
+      "epoch": 6.809349096344612,
+      "grad_norm": 6.53125,
+      "learning_rate": 7.4415681478461755e-06,
+      "loss": 0.8476,
+      "step": 100220
+    },
+    {
+      "epoch": 6.809688816415274,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.4394448974045395e-06,
+      "loss": 0.8734,
+      "step": 100225
+    },
+    {
+      "epoch": 6.810028536485936,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.437321646962903e-06,
+      "loss": 0.8341,
+      "step": 100230
+    },
+    {
+      "epoch": 6.810368256556598,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.4351983965212676e-06,
+      "loss": 0.8037,
+      "step": 100235
+    },
+    {
+      "epoch": 6.810707976627259,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.433075146079631e-06,
+      "loss": 0.9017,
+      "step": 100240
+    },
+    {
+      "epoch": 6.811047696697921,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.430951895637994e-06,
+      "loss": 0.9753,
+      "step": 100245
+    },
+    {
+      "epoch": 6.811387416768583,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.428828645196359e-06,
+      "loss": 0.8031,
+      "step": 100250
+    },
+    {
+      "epoch": 6.811727136839244,
+      "grad_norm": 8.125,
+      "learning_rate": 7.426705394754723e-06,
+      "loss": 0.7665,
+      "step": 100255
+    },
+    {
+      "epoch": 6.812066856909906,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.424582144313086e-06,
+      "loss": 0.8205,
+      "step": 100260
+    },
+    {
+      "epoch": 6.812406576980568,
+      "grad_norm": 8.125,
+      "learning_rate": 7.422458893871451e-06,
+      "loss": 0.8325,
+      "step": 100265
+    },
+    {
+      "epoch": 6.81274629705123,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.420335643429814e-06,
+      "loss": 0.803,
+      "step": 100270
+    },
+    {
+      "epoch": 6.813086017121892,
+      "grad_norm": 8.125,
+      "learning_rate": 7.418212392988178e-06,
+      "loss": 0.8138,
+      "step": 100275
+    },
+    {
+      "epoch": 6.813425737192553,
+      "grad_norm": 7.375,
+      "learning_rate": 7.416089142546542e-06,
+      "loss": 0.7163,
+      "step": 100280
+    },
+    {
+      "epoch": 6.813765457263215,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.413965892104906e-06,
+      "loss": 0.8746,
+      "step": 100285
+    },
+    {
+      "epoch": 6.814105177333877,
+      "grad_norm": 6.25,
+      "learning_rate": 7.411842641663269e-06,
+      "loss": 0.7778,
+      "step": 100290
+    },
+    {
+      "epoch": 6.814444897404538,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.409719391221634e-06,
+      "loss": 0.8051,
+      "step": 100295
+    },
+    {
+      "epoch": 6.8147846174752,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.407596140779997e-06,
+      "loss": 0.7877,
+      "step": 100300
+    },
+    {
+      "epoch": 6.815124337545862,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.405472890338362e-06,
+      "loss": 0.7844,
+      "step": 100305
+    },
+    {
+      "epoch": 6.8154640576165235,
+      "grad_norm": 8.25,
+      "learning_rate": 7.403349639896726e-06,
+      "loss": 0.8838,
+      "step": 100310
+    },
+    {
+      "epoch": 6.815803777687186,
+      "grad_norm": 7.375,
+      "learning_rate": 7.401226389455089e-06,
+      "loss": 0.8363,
+      "step": 100315
+    },
+    {
+      "epoch": 6.816143497757848,
+      "grad_norm": 6.53125,
+      "learning_rate": 7.399103139013454e-06,
+      "loss": 0.896,
+      "step": 100320
+    },
+    {
+      "epoch": 6.816483217828509,
+      "grad_norm": 8.25,
+      "learning_rate": 7.396979888571817e-06,
+      "loss": 0.8432,
+      "step": 100325
+    },
+    {
+      "epoch": 6.816822937899171,
+      "grad_norm": 8.25,
+      "learning_rate": 7.39485663813018e-06,
+      "loss": 0.9311,
+      "step": 100330
+    },
+    {
+      "epoch": 6.817162657969833,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.392733387688545e-06,
+      "loss": 0.8444,
+      "step": 100335
+    },
+    {
+      "epoch": 6.817502378040494,
+      "grad_norm": 8.875,
+      "learning_rate": 7.390610137246909e-06,
+      "loss": 0.8329,
+      "step": 100340
+    },
+    {
+      "epoch": 6.817842098111156,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.388486886805272e-06,
+      "loss": 0.8499,
+      "step": 100345
+    },
+    {
+      "epoch": 6.818181818181818,
+      "grad_norm": 8.0,
+      "learning_rate": 7.386363636363637e-06,
+      "loss": 0.7726,
+      "step": 100350
+    },
+    {
+      "epoch": 6.8185215382524795,
+      "grad_norm": 7.03125,
+      "learning_rate": 7.384240385922e-06,
+      "loss": 0.918,
+      "step": 100355
+    },
+    {
+      "epoch": 6.818861258323142,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.382117135480364e-06,
+      "loss": 0.8773,
+      "step": 100360
+    },
+    {
+      "epoch": 6.819200978393804,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.379993885038729e-06,
+      "loss": 1.0136,
+      "step": 100365
+    },
+    {
+      "epoch": 6.819540698464465,
+      "grad_norm": 8.625,
+      "learning_rate": 7.377870634597092e-06,
+      "loss": 0.8457,
+      "step": 100370
+    },
+    {
+      "epoch": 6.819880418535127,
+      "grad_norm": 8.25,
+      "learning_rate": 7.3757473841554554e-06,
+      "loss": 0.8639,
+      "step": 100375
+    },
+    {
+      "epoch": 6.820220138605789,
+      "grad_norm": 8.625,
+      "learning_rate": 7.37362413371382e-06,
+      "loss": 0.8597,
+      "step": 100380
+    },
+    {
+      "epoch": 6.82055985867645,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.3715008832721834e-06,
+      "loss": 0.821,
+      "step": 100385
+    },
+    {
+      "epoch": 6.820899578747112,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.369377632830548e-06,
+      "loss": 0.8216,
+      "step": 100390
+    },
+    {
+      "epoch": 6.821239298817774,
+      "grad_norm": 8.125,
+      "learning_rate": 7.367254382388912e-06,
+      "loss": 0.9101,
+      "step": 100395
+    },
+    {
+      "epoch": 6.8215790188884355,
+      "grad_norm": 8.0,
+      "learning_rate": 7.3651311319472754e-06,
+      "loss": 0.8692,
+      "step": 100400
+    },
+    {
+      "epoch": 6.821918738959098,
+      "grad_norm": 8.0,
+      "learning_rate": 7.36300788150564e-06,
+      "loss": 0.8258,
+      "step": 100405
+    },
+    {
+      "epoch": 6.82225845902976,
+      "grad_norm": 8.0,
+      "learning_rate": 7.3608846310640035e-06,
+      "loss": 0.8188,
+      "step": 100410
+    },
+    {
+      "epoch": 6.822598179100421,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.3587613806223675e-06,
+      "loss": 0.8608,
+      "step": 100415
+    },
+    {
+      "epoch": 6.822937899171083,
+      "grad_norm": 8.75,
+      "learning_rate": 7.356638130180732e-06,
+      "loss": 0.846,
+      "step": 100420
+    },
+    {
+      "epoch": 6.823277619241745,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.3545148797390955e-06,
+      "loss": 0.8651,
+      "step": 100425
+    },
+    {
+      "epoch": 6.823617339312406,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.352391629297459e-06,
+      "loss": 0.7925,
+      "step": 100430
+    },
+    {
+      "epoch": 6.823957059383068,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.3502683788558235e-06,
+      "loss": 0.8113,
+      "step": 100435
+    },
+    {
+      "epoch": 6.82429677945373,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.348145128414187e-06,
+      "loss": 0.7993,
+      "step": 100440
+    },
+    {
+      "epoch": 6.8246364995243916,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.346021877972551e-06,
+      "loss": 0.8751,
+      "step": 100445
+    },
+    {
+      "epoch": 6.824976219595054,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.3438986275309155e-06,
+      "loss": 0.7874,
+      "step": 100450
+    },
+    {
+      "epoch": 6.825315939665716,
+      "grad_norm": 7.75,
+      "learning_rate": 7.341775377089279e-06,
+      "loss": 0.8521,
+      "step": 100455
+    },
+    {
+      "epoch": 6.825655659736377,
+      "grad_norm": 7.875,
+      "learning_rate": 7.339652126647642e-06,
+      "loss": 0.9122,
+      "step": 100460
+    },
+    {
+      "epoch": 6.825995379807039,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.337528876206007e-06,
+      "loss": 0.9101,
+      "step": 100465
+    },
+    {
+      "epoch": 6.826335099877701,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.335405625764371e-06,
+      "loss": 0.7779,
+      "step": 100470
+    },
+    {
+      "epoch": 6.826674819948362,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.333282375322735e-06,
+      "loss": 0.8801,
+      "step": 100475
+    },
+    {
+      "epoch": 6.827014540019024,
+      "grad_norm": 7.15625,
+      "learning_rate": 7.331159124881099e-06,
+      "loss": 0.8622,
+      "step": 100480
+    },
+    {
+      "epoch": 6.827354260089686,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.329035874439462e-06,
+      "loss": 0.8921,
+      "step": 100485
+    },
+    {
+      "epoch": 6.827693980160348,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.326912623997827e-06,
+      "loss": 0.7351,
+      "step": 100490
+    },
+    {
+      "epoch": 6.82803370023101,
+      "grad_norm": 8.9375,
+      "learning_rate": 7.32478937355619e-06,
+      "loss": 0.9177,
+      "step": 100495
+    },
+    {
+      "epoch": 6.828373420301672,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.322666123114554e-06,
+      "loss": 0.7799,
+      "step": 100500
+    },
+    {
+      "epoch": 6.828713140372333,
+      "grad_norm": 8.125,
+      "learning_rate": 7.320542872672919e-06,
+      "loss": 0.8121,
+      "step": 100505
+    },
+    {
+      "epoch": 6.829052860442995,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.318419622231282e-06,
+      "loss": 0.8293,
+      "step": 100510
+    },
+    {
+      "epoch": 6.829392580513657,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.316296371789645e-06,
+      "loss": 0.9604,
+      "step": 100515
+    },
+    {
+      "epoch": 6.829732300584318,
+      "grad_norm": 8.625,
+      "learning_rate": 7.31417312134801e-06,
+      "loss": 0.7515,
+      "step": 100520
+    },
+    {
+      "epoch": 6.83007202065498,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.312049870906373e-06,
+      "loss": 0.8553,
+      "step": 100525
+    },
+    {
+      "epoch": 6.830411740725642,
+      "grad_norm": 8.375,
+      "learning_rate": 7.309926620464737e-06,
+      "loss": 0.8615,
+      "step": 100530
+    },
+    {
+      "epoch": 6.830751460796304,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.307803370023102e-06,
+      "loss": 0.8521,
+      "step": 100535
+    },
+    {
+      "epoch": 6.831091180866966,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.305680119581465e-06,
+      "loss": 0.8166,
+      "step": 100540
+    },
+    {
+      "epoch": 6.831430900937628,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.303556869139828e-06,
+      "loss": 0.8316,
+      "step": 100545
+    },
+    {
+      "epoch": 6.831770621008289,
+      "grad_norm": 8.75,
+      "learning_rate": 7.301433618698193e-06,
+      "loss": 0.89,
+      "step": 100550
+    },
+    {
+      "epoch": 6.832110341078951,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.299310368256557e-06,
+      "loss": 0.8885,
+      "step": 100555
+    },
+    {
+      "epoch": 6.832450061149613,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.297187117814922e-06,
+      "loss": 0.8711,
+      "step": 100560
+    },
+    {
+      "epoch": 6.832789781220274,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.295063867373285e-06,
+      "loss": 0.8626,
+      "step": 100565
+    },
+    {
+      "epoch": 6.833129501290936,
+      "grad_norm": 8.0,
+      "learning_rate": 7.292940616931648e-06,
+      "loss": 0.8699,
+      "step": 100570
+    },
+    {
+      "epoch": 6.833469221361598,
+      "grad_norm": 8.125,
+      "learning_rate": 7.290817366490013e-06,
+      "loss": 0.8944,
+      "step": 100575
+    },
+    {
+      "epoch": 6.83380894143226,
+      "grad_norm": 8.375,
+      "learning_rate": 7.288694116048376e-06,
+      "loss": 0.8314,
+      "step": 100580
+    },
+    {
+      "epoch": 6.834148661502922,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.28657086560674e-06,
+      "loss": 0.7651,
+      "step": 100585
+    },
+    {
+      "epoch": 6.834488381573584,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.284447615165105e-06,
+      "loss": 0.9195,
+      "step": 100590
+    },
+    {
+      "epoch": 6.834828101644245,
+      "grad_norm": 7.875,
+      "learning_rate": 7.282324364723468e-06,
+      "loss": 0.7986,
+      "step": 100595
+    },
+    {
+      "epoch": 6.835167821714907,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.280201114281831e-06,
+      "loss": 0.7957,
+      "step": 100600
+    },
+    {
+      "epoch": 6.835507541785569,
+      "grad_norm": 8.125,
+      "learning_rate": 7.278077863840196e-06,
+      "loss": 0.9015,
+      "step": 100605
+    },
+    {
+      "epoch": 6.83584726185623,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.27595461339856e-06,
+      "loss": 0.8178,
+      "step": 100610
+    },
+    {
+      "epoch": 6.836186981926892,
+      "grad_norm": 8.75,
+      "learning_rate": 7.273831362956923e-06,
+      "loss": 0.81,
+      "step": 100615
+    },
+    {
+      "epoch": 6.836526701997554,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.271708112515288e-06,
+      "loss": 0.7925,
+      "step": 100620
+    },
+    {
+      "epoch": 6.836866422068216,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.269584862073651e-06,
+      "loss": 0.6923,
+      "step": 100625
+    },
+    {
+      "epoch": 6.837206142138878,
+      "grad_norm": 7.625,
+      "learning_rate": 7.2674616116320145e-06,
+      "loss": 0.8301,
+      "step": 100630
+    },
+    {
+      "epoch": 6.83754586220954,
+      "grad_norm": 8.0,
+      "learning_rate": 7.265338361190379e-06,
+      "loss": 0.8345,
+      "step": 100635
+    },
+    {
+      "epoch": 6.837885582280201,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.263215110748743e-06,
+      "loss": 0.8519,
+      "step": 100640
+    },
+    {
+      "epoch": 6.838225302350863,
+      "grad_norm": 8.375,
+      "learning_rate": 7.261091860307108e-06,
+      "loss": 0.8252,
+      "step": 100645
+    },
+    {
+      "epoch": 6.838565022421525,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.258968609865471e-06,
+      "loss": 0.8419,
+      "step": 100650
+    },
+    {
+      "epoch": 6.838904742492186,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.2568453594238346e-06,
+      "loss": 0.8501,
+      "step": 100655
+    },
+    {
+      "epoch": 6.839244462562848,
+      "grad_norm": 8.25,
+      "learning_rate": 7.254722108982199e-06,
+      "loss": 0.936,
+      "step": 100660
+    },
+    {
+      "epoch": 6.83958418263351,
+      "grad_norm": 9.0625,
+      "learning_rate": 7.2525988585405626e-06,
+      "loss": 0.951,
+      "step": 100665
+    },
+    {
+      "epoch": 6.839923902704172,
+      "grad_norm": 8.25,
+      "learning_rate": 7.2504756080989266e-06,
+      "loss": 0.8134,
+      "step": 100670
+    },
+    {
+      "epoch": 6.840263622774834,
+      "grad_norm": 8.75,
+      "learning_rate": 7.2483523576572914e-06,
+      "loss": 0.8139,
+      "step": 100675
+    },
+    {
+      "epoch": 6.840603342845496,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.246229107215655e-06,
+      "loss": 0.823,
+      "step": 100680
+    },
+    {
+      "epoch": 6.840943062916157,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.244105856774018e-06,
+      "loss": 0.8823,
+      "step": 100685
+    },
+    {
+      "epoch": 6.841282782986819,
+      "grad_norm": 9.0,
+      "learning_rate": 7.241982606332383e-06,
+      "loss": 0.8537,
+      "step": 100690
+    },
+    {
+      "epoch": 6.841622503057481,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.239859355890747e-06,
+      "loss": 0.8322,
+      "step": 100695
+    },
+    {
+      "epoch": 6.841962223128142,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.23773610544911e-06,
+      "loss": 0.8821,
+      "step": 100700
+    },
+    {
+      "epoch": 6.842301943198804,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.235612855007475e-06,
+      "loss": 0.8637,
+      "step": 100705
+    },
+    {
+      "epoch": 6.842641663269466,
+      "grad_norm": 7.875,
+      "learning_rate": 7.233489604565838e-06,
+      "loss": 0.7215,
+      "step": 100710
+    },
+    {
+      "epoch": 6.842981383340128,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.231366354124201e-06,
+      "loss": 0.881,
+      "step": 100715
+    },
+    {
+      "epoch": 6.84332110341079,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.229243103682566e-06,
+      "loss": 0.81,
+      "step": 100720
+    },
+    {
+      "epoch": 6.843660823481452,
+      "grad_norm": 8.375,
+      "learning_rate": 7.22711985324093e-06,
+      "loss": 0.8227,
+      "step": 100725
+    },
+    {
+      "epoch": 6.844000543552113,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.224996602799295e-06,
+      "loss": 0.8086,
+      "step": 100730
+    },
+    {
+      "epoch": 6.844340263622775,
+      "grad_norm": 8.375,
+      "learning_rate": 7.222873352357658e-06,
+      "loss": 0.8539,
+      "step": 100735
+    },
+    {
+      "epoch": 6.844679983693436,
+      "grad_norm": 8.5,
+      "learning_rate": 7.220750101916021e-06,
+      "loss": 0.9168,
+      "step": 100740
+    },
+    {
+      "epoch": 6.845019703764098,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.218626851474386e-06,
+      "loss": 0.8231,
+      "step": 100745
+    },
+    {
+      "epoch": 6.84535942383476,
+      "grad_norm": 8.375,
+      "learning_rate": 7.21650360103275e-06,
+      "loss": 0.845,
+      "step": 100750
+    },
+    {
+      "epoch": 6.845699143905422,
+      "grad_norm": 7.59375,
+      "learning_rate": 7.214380350591113e-06,
+      "loss": 0.7971,
+      "step": 100755
+    },
+    {
+      "epoch": 6.846038863976084,
+      "grad_norm": 8.125,
+      "learning_rate": 7.212257100149478e-06,
+      "loss": 0.8561,
+      "step": 100760
+    },
+    {
+      "epoch": 6.846378584046746,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.210133849707841e-06,
+      "loss": 0.8423,
+      "step": 100765
+    },
+    {
+      "epoch": 6.846718304117407,
+      "grad_norm": 7.09375,
+      "learning_rate": 7.208010599266204e-06,
+      "loss": 0.7836,
+      "step": 100770
+    },
+    {
+      "epoch": 6.847058024188069,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.205887348824569e-06,
+      "loss": 0.7599,
+      "step": 100775
+    },
+    {
+      "epoch": 6.847397744258731,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.203764098382933e-06,
+      "loss": 0.7533,
+      "step": 100780
+    },
+    {
+      "epoch": 6.847737464329392,
+      "grad_norm": 8.375,
+      "learning_rate": 7.201640847941296e-06,
+      "loss": 0.8204,
+      "step": 100785
+    },
+    {
+      "epoch": 6.848077184400054,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.199517597499661e-06,
+      "loss": 0.9623,
+      "step": 100790
+    },
+    {
+      "epoch": 6.848416904470716,
+      "grad_norm": 7.4375,
+      "learning_rate": 7.197394347058024e-06,
+      "loss": 0.764,
+      "step": 100795
+    },
+    {
+      "epoch": 6.848756624541378,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.195271096616388e-06,
+      "loss": 0.8043,
+      "step": 100800
+    },
+    {
+      "epoch": 6.84909634461204,
+      "grad_norm": 7.375,
+      "learning_rate": 7.193147846174753e-06,
+      "loss": 0.8008,
+      "step": 100805
+    },
+    {
+      "epoch": 6.849436064682702,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.191024595733116e-06,
+      "loss": 0.8825,
+      "step": 100810
+    },
+    {
+      "epoch": 6.849775784753363,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.188901345291481e-06,
+      "loss": 0.9322,
+      "step": 100815
+    },
+    {
+      "epoch": 6.850115504824025,
+      "grad_norm": 7.625,
+      "learning_rate": 7.186778094849844e-06,
+      "loss": 0.8635,
+      "step": 100820
+    },
+    {
+      "epoch": 6.850455224894687,
+      "grad_norm": 7.25,
+      "learning_rate": 7.184654844408207e-06,
+      "loss": 0.8331,
+      "step": 100825
+    },
+    {
+      "epoch": 6.850794944965348,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.182531593966572e-06,
+      "loss": 0.7907,
+      "step": 100830
+    },
+    {
+      "epoch": 6.85113466503601,
+      "grad_norm": 7.4375,
+      "learning_rate": 7.180408343524936e-06,
+      "loss": 0.8444,
+      "step": 100835
+    },
+    {
+      "epoch": 6.851474385106672,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.178285093083299e-06,
+      "loss": 0.853,
+      "step": 100840
+    },
+    {
+      "epoch": 6.851814105177334,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.176161842641664e-06,
+      "loss": 0.8644,
+      "step": 100845
+    },
+    {
+      "epoch": 6.852153825247996,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.174038592200027e-06,
+      "loss": 0.8039,
+      "step": 100850
+    },
+    {
+      "epoch": 6.852493545318658,
+      "grad_norm": 7.875,
+      "learning_rate": 7.171915341758391e-06,
+      "loss": 0.8748,
+      "step": 100855
+    },
+    {
+      "epoch": 6.852833265389319,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.169792091316755e-06,
+      "loss": 0.7802,
+      "step": 100860
+    },
+    {
+      "epoch": 6.853172985459981,
+      "grad_norm": 8.125,
+      "learning_rate": 7.167668840875119e-06,
+      "loss": 0.893,
+      "step": 100865
+    },
+    {
+      "epoch": 6.853512705530643,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.1655455904334825e-06,
+      "loss": 0.8213,
+      "step": 100870
+    },
+    {
+      "epoch": 6.853852425601304,
+      "grad_norm": 9.8125,
+      "learning_rate": 7.163422339991847e-06,
+      "loss": 0.948,
+      "step": 100875
+    },
+    {
+      "epoch": 6.854192145671966,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.1612990895502105e-06,
+      "loss": 0.8688,
+      "step": 100880
+    },
+    {
+      "epoch": 6.854531865742628,
+      "grad_norm": 8.875,
+      "learning_rate": 7.1591758391085745e-06,
+      "loss": 0.8689,
+      "step": 100885
+    },
+    {
+      "epoch": 6.85487158581329,
+      "grad_norm": 8.625,
+      "learning_rate": 7.157052588666939e-06,
+      "loss": 0.8397,
+      "step": 100890
+    },
+    {
+      "epoch": 6.855211305883952,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.1549293382253025e-06,
+      "loss": 0.7644,
+      "step": 100895
+    },
+    {
+      "epoch": 6.855551025954614,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.152806087783667e-06,
+      "loss": 0.8527,
+      "step": 100900
+    },
+    {
+      "epoch": 6.855890746025275,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.1506828373420305e-06,
+      "loss": 0.7486,
+      "step": 100905
+    },
+    {
+      "epoch": 6.856230466095937,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.148559586900394e-06,
+      "loss": 0.7826,
+      "step": 100910
+    },
+    {
+      "epoch": 6.856570186166599,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.1464363364587585e-06,
+      "loss": 0.9593,
+      "step": 100915
+    },
+    {
+      "epoch": 6.85690990623726,
+      "grad_norm": 7.53125,
+      "learning_rate": 7.1443130860171225e-06,
+      "loss": 0.9189,
+      "step": 100920
+    },
+    {
+      "epoch": 6.857249626307922,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.142189835575486e-06,
+      "loss": 0.9282,
+      "step": 100925
+    },
+    {
+      "epoch": 6.857589346378584,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.1400665851338505e-06,
+      "loss": 0.8978,
+      "step": 100930
+    },
+    {
+      "epoch": 6.857929066449246,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.137943334692214e-06,
+      "loss": 0.8296,
+      "step": 100935
+    },
+    {
+      "epoch": 6.858268786519908,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.135820084250578e-06,
+      "loss": 0.8326,
+      "step": 100940
+    },
+    {
+      "epoch": 6.85860850659057,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.1336968338089426e-06,
+      "loss": 0.8294,
+      "step": 100945
+    },
+    {
+      "epoch": 6.858948226661231,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.131573583367306e-06,
+      "loss": 0.9366,
+      "step": 100950
+    },
+    {
+      "epoch": 6.859287946731893,
+      "grad_norm": 8.25,
+      "learning_rate": 7.129450332925669e-06,
+      "loss": 0.8014,
+      "step": 100955
+    },
+    {
+      "epoch": 6.859627666802554,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.127327082484034e-06,
+      "loss": 0.8216,
+      "step": 100960
+    },
+    {
+      "epoch": 6.859967386873216,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.125203832042397e-06,
+      "loss": 0.7528,
+      "step": 100965
+    },
+    {
+      "epoch": 6.860307106943878,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.123080581600761e-06,
+      "loss": 0.8442,
+      "step": 100970
+    },
+    {
+      "epoch": 6.8606468270145395,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.120957331159126e-06,
+      "loss": 0.7897,
+      "step": 100975
+    },
+    {
+      "epoch": 6.860986547085202,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.118834080717489e-06,
+      "loss": 0.8784,
+      "step": 100980
+    },
+    {
+      "epoch": 6.861326267155864,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.116710830275854e-06,
+      "loss": 0.8388,
+      "step": 100985
+    },
+    {
+      "epoch": 6.861665987226525,
+      "grad_norm": 7.25,
+      "learning_rate": 7.114587579834217e-06,
+      "loss": 0.8571,
+      "step": 100990
+    },
+    {
+      "epoch": 6.862005707297187,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.112464329392581e-06,
+      "loss": 0.8662,
+      "step": 100995
+    },
+    {
+      "epoch": 6.862345427367849,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.110341078950945e-06,
+      "loss": 0.7837,
+      "step": 101000
+    },
+    {
+      "epoch": 6.86268514743851,
+      "grad_norm": 7.375,
+      "learning_rate": 7.108217828509309e-06,
+      "loss": 0.7723,
+      "step": 101005
+    },
+    {
+      "epoch": 6.863024867509172,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.106094578067672e-06,
+      "loss": 0.9038,
+      "step": 101010
+    },
+    {
+      "epoch": 6.863364587579834,
+      "grad_norm": 7.375,
+      "learning_rate": 7.103971327626037e-06,
+      "loss": 0.8035,
+      "step": 101015
+    },
+    {
+      "epoch": 6.8637043076504956,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.1018480771844e-06,
+      "loss": 0.8307,
+      "step": 101020
+    },
+    {
+      "epoch": 6.864044027721158,
+      "grad_norm": 9.0,
+      "learning_rate": 7.099724826742764e-06,
+      "loss": 0.7917,
+      "step": 101025
+    },
+    {
+      "epoch": 6.86438374779182,
+      "grad_norm": 7.625,
+      "learning_rate": 7.097601576301129e-06,
+      "loss": 0.8026,
+      "step": 101030
+    },
+    {
+      "epoch": 6.864723467862481,
+      "grad_norm": 8.5,
+      "learning_rate": 7.095478325859492e-06,
+      "loss": 0.8942,
+      "step": 101035
+    },
+    {
+      "epoch": 6.865063187933143,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.093355075417855e-06,
+      "loss": 0.8548,
+      "step": 101040
+    },
+    {
+      "epoch": 6.865402908003805,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.09123182497622e-06,
+      "loss": 0.8651,
+      "step": 101045
+    },
+    {
+      "epoch": 6.865742628074466,
+      "grad_norm": 8.0,
+      "learning_rate": 7.089108574534583e-06,
+      "loss": 0.8385,
+      "step": 101050
+    },
+    {
+      "epoch": 6.866082348145128,
+      "grad_norm": 7.03125,
+      "learning_rate": 7.086985324092947e-06,
+      "loss": 0.7697,
+      "step": 101055
+    },
+    {
+      "epoch": 6.86642206821579,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.084862073651312e-06,
+      "loss": 0.8644,
+      "step": 101060
+    },
+    {
+      "epoch": 6.866761788286452,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.082738823209675e-06,
+      "loss": 0.8689,
+      "step": 101065
+    },
+    {
+      "epoch": 6.867101508357114,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.08061557276804e-06,
+      "loss": 0.833,
+      "step": 101070
+    },
+    {
+      "epoch": 6.867441228427776,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.078492322326403e-06,
+      "loss": 0.7646,
+      "step": 101075
+    },
+    {
+      "epoch": 6.867780948498437,
+      "grad_norm": 8.25,
+      "learning_rate": 7.076369071884767e-06,
+      "loss": 0.8388,
+      "step": 101080
+    },
+    {
+      "epoch": 6.868120668569099,
+      "grad_norm": 8.25,
+      "learning_rate": 7.074245821443132e-06,
+      "loss": 0.77,
+      "step": 101085
+    },
+    {
+      "epoch": 6.868460388639761,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.072122571001495e-06,
+      "loss": 0.8102,
+      "step": 101090
+    },
+    {
+      "epoch": 6.868800108710422,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.0699993205598584e-06,
+      "loss": 0.8434,
+      "step": 101095
+    },
+    {
+      "epoch": 6.869139828781084,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.067876070118223e-06,
+      "loss": 0.7978,
+      "step": 101100
+    },
+    {
+      "epoch": 6.869479548851746,
+      "grad_norm": 8.5,
+      "learning_rate": 7.0657528196765864e-06,
+      "loss": 0.8675,
+      "step": 101105
+    },
+    {
+      "epoch": 6.869819268922408,
+      "grad_norm": 7.34375,
+      "learning_rate": 7.0636295692349504e-06,
+      "loss": 0.8676,
+      "step": 101110
+    },
+    {
+      "epoch": 6.87015898899307,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.061506318793315e-06,
+      "loss": 0.7389,
+      "step": 101115
+    },
+    {
+      "epoch": 6.870498709063732,
+      "grad_norm": 7.6875,
+      "learning_rate": 7.0593830683516785e-06,
+      "loss": 0.8423,
+      "step": 101120
+    },
+    {
+      "epoch": 6.870838429134393,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.057259817910042e-06,
+      "loss": 0.7869,
+      "step": 101125
+    },
+    {
+      "epoch": 6.871178149205055,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.0551365674684065e-06,
+      "loss": 0.9012,
+      "step": 101130
+    },
+    {
+      "epoch": 6.871517869275717,
+      "grad_norm": 7.875,
+      "learning_rate": 7.0530133170267705e-06,
+      "loss": 0.8954,
+      "step": 101135
+    },
+    {
+      "epoch": 6.871857589346378,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.050890066585134e-06,
+      "loss": 0.9532,
+      "step": 101140
+    },
+    {
+      "epoch": 6.87219730941704,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.0487668161434985e-06,
+      "loss": 0.8609,
+      "step": 101145
+    },
+    {
+      "epoch": 6.872537029487702,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.046643565701862e-06,
+      "loss": 0.9394,
+      "step": 101150
+    },
+    {
+      "epoch": 6.872876749558364,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.0445203152602265e-06,
+      "loss": 0.8434,
+      "step": 101155
+    },
+    {
+      "epoch": 6.873216469629026,
+      "grad_norm": 6.90625,
+      "learning_rate": 7.04239706481859e-06,
+      "loss": 0.7906,
+      "step": 101160
+    },
+    {
+      "epoch": 6.873556189699688,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.040273814376954e-06,
+      "loss": 0.7998,
+      "step": 101165
+    },
+    {
+      "epoch": 6.873895909770349,
+      "grad_norm": 8.125,
+      "learning_rate": 7.0381505639353185e-06,
+      "loss": 0.9051,
+      "step": 101170
+    },
+    {
+      "epoch": 6.874235629841011,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.036027313493682e-06,
+      "loss": 0.8274,
+      "step": 101175
+    },
+    {
+      "epoch": 6.874575349911673,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.033904063052045e-06,
+      "loss": 0.873,
+      "step": 101180
+    },
+    {
+      "epoch": 6.874915069982334,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.03178081261041e-06,
+      "loss": 0.8522,
+      "step": 101185
+    },
+    {
+      "epoch": 6.875254790052996,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.029657562168774e-06,
+      "loss": 0.7702,
+      "step": 101190
+    },
+    {
+      "epoch": 6.875594510123658,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.027534311727137e-06,
+      "loss": 0.8102,
+      "step": 101195
+    },
+    {
+      "epoch": 6.87593423019432,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.025411061285502e-06,
+      "loss": 0.8227,
+      "step": 101200
+    },
+    {
+      "epoch": 6.876273950264982,
+      "grad_norm": 6.875,
+      "learning_rate": 7.023287810843865e-06,
+      "loss": 0.9173,
+      "step": 101205
+    },
+    {
+      "epoch": 6.876613670335644,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.021164560402228e-06,
+      "loss": 0.7579,
+      "step": 101210
+    },
+    {
+      "epoch": 6.876953390406305,
+      "grad_norm": 8.5,
+      "learning_rate": 7.019041309960593e-06,
+      "loss": 0.813,
+      "step": 101215
+    },
+    {
+      "epoch": 6.877293110476967,
+      "grad_norm": 7.5,
+      "learning_rate": 7.016918059518957e-06,
+      "loss": 0.8384,
+      "step": 101220
+    },
+    {
+      "epoch": 6.877632830547629,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.01479480907732e-06,
+      "loss": 0.8286,
+      "step": 101225
+    },
+    {
+      "epoch": 6.87797255061829,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.012671558635685e-06,
+      "loss": 0.865,
+      "step": 101230
+    },
+    {
+      "epoch": 6.878312270688952,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.010548308194048e-06,
+      "loss": 0.8478,
+      "step": 101235
+    },
+    {
+      "epoch": 6.878651990759614,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.008425057752413e-06,
+      "loss": 0.8827,
+      "step": 101240
+    },
+    {
+      "epoch": 6.878991710830276,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.006301807310776e-06,
+      "loss": 0.8215,
+      "step": 101245
+    },
+    {
+      "epoch": 6.879331430900938,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.00417855686914e-06,
+      "loss": 0.7559,
+      "step": 101250
+    },
+    {
+      "epoch": 6.8796711509716,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.002055306427505e-06,
+      "loss": 0.8723,
+      "step": 101255
+    },
+    {
+      "epoch": 6.880010871042261,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.999932055985868e-06,
+      "loss": 0.8192,
+      "step": 101260
+    },
+    {
+      "epoch": 6.880350591112923,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.997808805544231e-06,
+      "loss": 0.8266,
+      "step": 101265
+    },
+    {
+      "epoch": 6.880690311183585,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.995685555102596e-06,
+      "loss": 0.7793,
+      "step": 101270
+    },
+    {
+      "epoch": 6.881030031254246,
+      "grad_norm": 8.625,
+      "learning_rate": 6.99356230466096e-06,
+      "loss": 0.9016,
+      "step": 101275
+    },
+    {
+      "epoch": 6.881369751324908,
+      "grad_norm": 8.375,
+      "learning_rate": 6.991439054219323e-06,
+      "loss": 0.819,
+      "step": 101280
+    },
+    {
+      "epoch": 6.88170947139557,
+      "grad_norm": 8.125,
+      "learning_rate": 6.989315803777688e-06,
+      "loss": 0.885,
+      "step": 101285
+    },
+    {
+      "epoch": 6.882049191466232,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.987192553336051e-06,
+      "loss": 0.8715,
+      "step": 101290
+    },
+    {
+      "epoch": 6.882388911536894,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.985069302894414e-06,
+      "loss": 0.814,
+      "step": 101295
+    },
+    {
+      "epoch": 6.882728631607556,
+      "grad_norm": 7.40625,
+      "learning_rate": 6.982946052452779e-06,
+      "loss": 0.8576,
+      "step": 101300
+    },
+    {
+      "epoch": 6.883068351678217,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.980822802011143e-06,
+      "loss": 0.8078,
+      "step": 101305
+    },
+    {
+      "epoch": 6.883408071748879,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.978699551569506e-06,
+      "loss": 0.8153,
+      "step": 101310
+    },
+    {
+      "epoch": 6.883747791819541,
+      "grad_norm": 7.375,
+      "learning_rate": 6.976576301127871e-06,
+      "loss": 0.8373,
+      "step": 101315
+    },
+    {
+      "epoch": 6.884087511890202,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.974453050686234e-06,
+      "loss": 0.8269,
+      "step": 101320
+    },
+    {
+      "epoch": 6.884427231960864,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.972329800244599e-06,
+      "loss": 0.84,
+      "step": 101325
+    },
+    {
+      "epoch": 6.884766952031526,
+      "grad_norm": 7.125,
+      "learning_rate": 6.970206549802963e-06,
+      "loss": 0.8296,
+      "step": 101330
+    },
+    {
+      "epoch": 6.885106672102188,
+      "grad_norm": 9.0,
+      "learning_rate": 6.968083299361326e-06,
+      "loss": 0.8731,
+      "step": 101335
+    },
+    {
+      "epoch": 6.88544639217285,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.965960048919691e-06,
+      "loss": 0.8422,
+      "step": 101340
+    },
+    {
+      "epoch": 6.885786112243512,
+      "grad_norm": 8.0,
+      "learning_rate": 6.963836798478054e-06,
+      "loss": 0.913,
+      "step": 101345
+    },
+    {
+      "epoch": 6.886125832314173,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.9617135480364175e-06,
+      "loss": 0.7999,
+      "step": 101350
+    },
+    {
+      "epoch": 6.886465552384835,
+      "grad_norm": 9.0,
+      "learning_rate": 6.959590297594782e-06,
+      "loss": 0.8325,
+      "step": 101355
+    },
+    {
+      "epoch": 6.886805272455497,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.957467047153146e-06,
+      "loss": 0.8238,
+      "step": 101360
+    },
+    {
+      "epoch": 6.887144992526158,
+      "grad_norm": 8.125,
+      "learning_rate": 6.9553437967115096e-06,
+      "loss": 0.8876,
+      "step": 101365
+    },
+    {
+      "epoch": 6.88748471259682,
+      "grad_norm": 7.375,
+      "learning_rate": 6.953220546269874e-06,
+      "loss": 0.8783,
+      "step": 101370
+    },
+    {
+      "epoch": 6.8878244326674825,
+      "grad_norm": 6.96875,
+      "learning_rate": 6.9510972958282376e-06,
+      "loss": 0.7388,
+      "step": 101375
+    },
+    {
+      "epoch": 6.888164152738144,
+      "grad_norm": 7.40625,
+      "learning_rate": 6.9489740453866016e-06,
+      "loss": 0.8213,
+      "step": 101380
+    },
+    {
+      "epoch": 6.888503872808806,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.9468507949449656e-06,
+      "loss": 0.8396,
+      "step": 101385
+    },
+    {
+      "epoch": 6.888843592879468,
+      "grad_norm": 8.375,
+      "learning_rate": 6.94472754450333e-06,
+      "loss": 0.8089,
+      "step": 101390
+    },
+    {
+      "epoch": 6.889183312950129,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.942604294061693e-06,
+      "loss": 0.9277,
+      "step": 101395
+    },
+    {
+      "epoch": 6.889523033020791,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.940481043620058e-06,
+      "loss": 0.769,
+      "step": 101400
+    },
+    {
+      "epoch": 6.889862753091453,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.938357793178421e-06,
+      "loss": 0.8342,
+      "step": 101405
+    },
+    {
+      "epoch": 6.890202473162114,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.936234542736786e-06,
+      "loss": 0.7692,
+      "step": 101410
+    },
+    {
+      "epoch": 6.890542193232776,
+      "grad_norm": 8.0,
+      "learning_rate": 6.93411129229515e-06,
+      "loss": 0.8465,
+      "step": 101415
+    },
+    {
+      "epoch": 6.8908819133034385,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.931988041853513e-06,
+      "loss": 0.8393,
+      "step": 101420
+    },
+    {
+      "epoch": 6.8912216333741,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.929864791411878e-06,
+      "loss": 0.8837,
+      "step": 101425
+    },
+    {
+      "epoch": 6.891561353444762,
+      "grad_norm": 8.5,
+      "learning_rate": 6.927741540970241e-06,
+      "loss": 0.8777,
+      "step": 101430
+    },
+    {
+      "epoch": 6.891901073515423,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.925618290528604e-06,
+      "loss": 0.8208,
+      "step": 101435
+    },
+    {
+      "epoch": 6.892240793586085,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.923495040086969e-06,
+      "loss": 0.8359,
+      "step": 101440
+    },
+    {
+      "epoch": 6.892580513656747,
+      "grad_norm": 8.625,
+      "learning_rate": 6.921371789645333e-06,
+      "loss": 0.9525,
+      "step": 101445
+    },
+    {
+      "epoch": 6.892920233727408,
+      "grad_norm": 8.5,
+      "learning_rate": 6.919248539203696e-06,
+      "loss": 0.8504,
+      "step": 101450
+    },
+    {
+      "epoch": 6.89325995379807,
+      "grad_norm": 8.125,
+      "learning_rate": 6.917125288762061e-06,
+      "loss": 0.8359,
+      "step": 101455
+    },
+    {
+      "epoch": 6.893599673868732,
+      "grad_norm": 8.875,
+      "learning_rate": 6.915002038320424e-06,
+      "loss": 0.8826,
+      "step": 101460
+    },
+    {
+      "epoch": 6.893939393939394,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.912878787878788e-06,
+      "loss": 0.7975,
+      "step": 101465
+    },
+    {
+      "epoch": 6.894279114010056,
+      "grad_norm": 8.125,
+      "learning_rate": 6.910755537437153e-06,
+      "loss": 0.8678,
+      "step": 101470
+    },
+    {
+      "epoch": 6.894618834080718,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.908632286995516e-06,
+      "loss": 0.8922,
+      "step": 101475
+    },
+    {
+      "epoch": 6.894958554151379,
+      "grad_norm": 7.40625,
+      "learning_rate": 6.906509036553879e-06,
+      "loss": 0.7991,
+      "step": 101480
+    },
+    {
+      "epoch": 6.895298274222041,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.904385786112244e-06,
+      "loss": 0.7983,
+      "step": 101485
+    },
+    {
+      "epoch": 6.895637994292703,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.902262535670607e-06,
+      "loss": 0.8874,
+      "step": 101490
+    },
+    {
+      "epoch": 6.895977714363364,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.900139285228972e-06,
+      "loss": 0.8282,
+      "step": 101495
+    },
+    {
+      "epoch": 6.896317434434026,
+      "grad_norm": 7.375,
+      "learning_rate": 6.898016034787336e-06,
+      "loss": 0.7791,
+      "step": 101500
+    },
+    {
+      "epoch": 6.896657154504688,
+      "grad_norm": 7.5,
+      "learning_rate": 6.895892784345699e-06,
+      "loss": 0.8543,
+      "step": 101505
+    },
+    {
+      "epoch": 6.89699687457535,
+      "grad_norm": 9.125,
+      "learning_rate": 6.893769533904064e-06,
+      "loss": 0.8672,
+      "step": 101510
+    },
+    {
+      "epoch": 6.897336594646012,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.891646283462427e-06,
+      "loss": 0.9498,
+      "step": 101515
+    },
+    {
+      "epoch": 6.897676314716674,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.889523033020791e-06,
+      "loss": 0.8084,
+      "step": 101520
+    },
+    {
+      "epoch": 6.898016034787335,
+      "grad_norm": 8.25,
+      "learning_rate": 6.887399782579156e-06,
+      "loss": 0.9197,
+      "step": 101525
+    },
+    {
+      "epoch": 6.898355754857997,
+      "grad_norm": 8.625,
+      "learning_rate": 6.885276532137519e-06,
+      "loss": 0.8575,
+      "step": 101530
+    },
+    {
+      "epoch": 6.898695474928659,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.883153281695882e-06,
+      "loss": 0.7394,
+      "step": 101535
+    },
+    {
+      "epoch": 6.89903519499932,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.881030031254247e-06,
+      "loss": 0.8652,
+      "step": 101540
+    },
+    {
+      "epoch": 6.899374915069982,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.87890678081261e-06,
+      "loss": 0.7744,
+      "step": 101545
+    },
+    {
+      "epoch": 6.899714635140644,
+      "grad_norm": 7.625,
+      "learning_rate": 6.876783530370974e-06,
+      "loss": 0.8519,
+      "step": 101550
+    },
+    {
+      "epoch": 6.900054355211306,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.874660279929339e-06,
+      "loss": 0.8529,
+      "step": 101555
+    },
+    {
+      "epoch": 6.900394075281968,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.872537029487702e-06,
+      "loss": 0.9163,
+      "step": 101560
+    },
+    {
+      "epoch": 6.90073379535263,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.8704137790460655e-06,
+      "loss": 0.8136,
+      "step": 101565
+    },
+    {
+      "epoch": 6.901073515423291,
+      "grad_norm": 8.375,
+      "learning_rate": 6.86829052860443e-06,
+      "loss": 0.8169,
+      "step": 101570
+    },
+    {
+      "epoch": 6.901413235493953,
+      "grad_norm": 7.625,
+      "learning_rate": 6.866167278162794e-06,
+      "loss": 0.8538,
+      "step": 101575
+    },
+    {
+      "epoch": 6.901752955564615,
+      "grad_norm": 8.375,
+      "learning_rate": 6.864044027721158e-06,
+      "loss": 0.7799,
+      "step": 101580
+    },
+    {
+      "epoch": 6.902092675635276,
+      "grad_norm": 9.25,
+      "learning_rate": 6.861920777279522e-06,
+      "loss": 0.8566,
+      "step": 101585
+    },
+    {
+      "epoch": 6.902432395705938,
+      "grad_norm": 7.5,
+      "learning_rate": 6.8597975268378855e-06,
+      "loss": 0.7565,
+      "step": 101590
+    },
+    {
+      "epoch": 6.9027721157766,
+      "grad_norm": 7.75,
+      "learning_rate": 6.85767427639625e-06,
+      "loss": 0.8252,
+      "step": 101595
+    },
+    {
+      "epoch": 6.903111835847262,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.8555510259546135e-06,
+      "loss": 0.7901,
+      "step": 101600
+    },
+    {
+      "epoch": 6.903451555917924,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.8534277755129775e-06,
+      "loss": 0.9434,
+      "step": 101605
+    },
+    {
+      "epoch": 6.903791275988586,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.851304525071342e-06,
+      "loss": 0.7604,
+      "step": 101610
+    },
+    {
+      "epoch": 6.904130996059247,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.8491812746297055e-06,
+      "loss": 0.8241,
+      "step": 101615
+    },
+    {
+      "epoch": 6.904470716129909,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.847058024188069e-06,
+      "loss": 0.8353,
+      "step": 101620
+    },
+    {
+      "epoch": 6.904810436200571,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.8449347737464335e-06,
+      "loss": 0.8452,
+      "step": 101625
+    },
+    {
+      "epoch": 6.905150156271232,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.842811523304797e-06,
+      "loss": 0.853,
+      "step": 101630
+    },
+    {
+      "epoch": 6.905489876341894,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.840688272863161e-06,
+      "loss": 0.8257,
+      "step": 101635
+    },
+    {
+      "epoch": 6.905829596412556,
+      "grad_norm": 7.625,
+      "learning_rate": 6.8385650224215255e-06,
+      "loss": 0.8706,
+      "step": 101640
+    },
+    {
+      "epoch": 6.906169316483218,
+      "grad_norm": 8.125,
+      "learning_rate": 6.836441771979889e-06,
+      "loss": 0.8377,
+      "step": 101645
+    },
+    {
+      "epoch": 6.90650903655388,
+      "grad_norm": 9.0,
+      "learning_rate": 6.834318521538252e-06,
+      "loss": 0.9785,
+      "step": 101650
+    },
+    {
+      "epoch": 6.906848756624541,
+      "grad_norm": 8.5,
+      "learning_rate": 6.832195271096617e-06,
+      "loss": 0.8875,
+      "step": 101655
+    },
+    {
+      "epoch": 6.907188476695203,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.830072020654981e-06,
+      "loss": 0.8723,
+      "step": 101660
+    },
+    {
+      "epoch": 6.907528196765865,
+      "grad_norm": 8.125,
+      "learning_rate": 6.8279487702133456e-06,
+      "loss": 0.8516,
+      "step": 101665
+    },
+    {
+      "epoch": 6.907867916836526,
+      "grad_norm": 8.0,
+      "learning_rate": 6.825825519771709e-06,
+      "loss": 0.8098,
+      "step": 101670
+    },
+    {
+      "epoch": 6.908207636907188,
+      "grad_norm": 8.875,
+      "learning_rate": 6.823702269330072e-06,
+      "loss": 0.8171,
+      "step": 101675
+    },
+    {
+      "epoch": 6.90854735697785,
+      "grad_norm": 8.875,
+      "learning_rate": 6.821579018888437e-06,
+      "loss": 0.7975,
+      "step": 101680
+    },
+    {
+      "epoch": 6.908887077048512,
+      "grad_norm": 8.8125,
+      "learning_rate": 6.8194557684468e-06,
+      "loss": 0.8154,
+      "step": 101685
+    },
+    {
+      "epoch": 6.909226797119174,
+      "grad_norm": 7.09375,
+      "learning_rate": 6.817332518005164e-06,
+      "loss": 0.8434,
+      "step": 101690
+    },
+    {
+      "epoch": 6.909566517189836,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.815209267563529e-06,
+      "loss": 0.7361,
+      "step": 101695
+    },
+    {
+      "epoch": 6.909906237260497,
+      "grad_norm": 8.875,
+      "learning_rate": 6.813086017121892e-06,
+      "loss": 0.7732,
+      "step": 101700
+    },
+    {
+      "epoch": 6.910245957331159,
+      "grad_norm": 8.375,
+      "learning_rate": 6.810962766680255e-06,
+      "loss": 0.8296,
+      "step": 101705
+    },
+    {
+      "epoch": 6.910585677401821,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.80883951623862e-06,
+      "loss": 0.904,
+      "step": 101710
+    },
+    {
+      "epoch": 6.910925397472482,
+      "grad_norm": 8.5,
+      "learning_rate": 6.806716265796984e-06,
+      "loss": 0.8859,
+      "step": 101715
+    },
+    {
+      "epoch": 6.911265117543144,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.804593015355347e-06,
+      "loss": 0.8893,
+      "step": 101720
+    },
+    {
+      "epoch": 6.911604837613806,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.802469764913712e-06,
+      "loss": 0.8644,
+      "step": 101725
+    },
+    {
+      "epoch": 6.911944557684468,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.800346514472075e-06,
+      "loss": 0.894,
+      "step": 101730
+    },
+    {
+      "epoch": 6.91228427775513,
+      "grad_norm": 8.125,
+      "learning_rate": 6.798223264030438e-06,
+      "loss": 0.8024,
+      "step": 101735
+    },
+    {
+      "epoch": 6.912623997825792,
+      "grad_norm": 8.75,
+      "learning_rate": 6.796100013588803e-06,
+      "loss": 0.9132,
+      "step": 101740
+    },
+    {
+      "epoch": 6.912963717896453,
+      "grad_norm": 7.59375,
+      "learning_rate": 6.793976763147167e-06,
+      "loss": 0.9034,
+      "step": 101745
+    },
+    {
+      "epoch": 6.913303437967115,
+      "grad_norm": 8.5,
+      "learning_rate": 6.791853512705532e-06,
+      "loss": 0.7562,
+      "step": 101750
+    },
+    {
+      "epoch": 6.913643158037777,
+      "grad_norm": 8.375,
+      "learning_rate": 6.789730262263895e-06,
+      "loss": 0.8225,
+      "step": 101755
+    },
+    {
+      "epoch": 6.913982878108438,
+      "grad_norm": 8.625,
+      "learning_rate": 6.787607011822258e-06,
+      "loss": 0.7414,
+      "step": 101760
+    },
+    {
+      "epoch": 6.9143225981791,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.785483761380623e-06,
+      "loss": 0.8536,
+      "step": 101765
+    },
+    {
+      "epoch": 6.914662318249762,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.783360510938986e-06,
+      "loss": 0.8841,
+      "step": 101770
+    },
+    {
+      "epoch": 6.915002038320424,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.78123726049735e-06,
+      "loss": 0.7082,
+      "step": 101775
+    },
+    {
+      "epoch": 6.915341758391086,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.779114010055715e-06,
+      "loss": 0.8449,
+      "step": 101780
+    },
+    {
+      "epoch": 6.915681478461748,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.776990759614078e-06,
+      "loss": 0.7922,
+      "step": 101785
+    },
+    {
+      "epoch": 6.916021198532409,
+      "grad_norm": 7.03125,
+      "learning_rate": 6.774867509172441e-06,
+      "loss": 0.7816,
+      "step": 101790
+    },
+    {
+      "epoch": 6.916360918603071,
+      "grad_norm": 6.65625,
+      "learning_rate": 6.772744258730806e-06,
+      "loss": 0.8276,
+      "step": 101795
+    },
+    {
+      "epoch": 6.916700638673733,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.77062100828917e-06,
+      "loss": 0.8506,
+      "step": 101800
+    },
+    {
+      "epoch": 6.917040358744394,
+      "grad_norm": 8.875,
+      "learning_rate": 6.7684977578475334e-06,
+      "loss": 0.804,
+      "step": 101805
+    },
+    {
+      "epoch": 6.917380078815056,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.766374507405898e-06,
+      "loss": 0.8823,
+      "step": 101810
+    },
+    {
+      "epoch": 6.917719798885718,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.7642512569642614e-06,
+      "loss": 0.8283,
+      "step": 101815
+    },
+    {
+      "epoch": 6.91805951895638,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.762128006522625e-06,
+      "loss": 0.8788,
+      "step": 101820
+    },
+    {
+      "epoch": 6.918399239027042,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.7600047560809894e-06,
+      "loss": 0.8808,
+      "step": 101825
+    },
+    {
+      "epoch": 6.918738959097704,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.7578815056393534e-06,
+      "loss": 0.8814,
+      "step": 101830
+    },
+    {
+      "epoch": 6.919078679168365,
+      "grad_norm": 7.5,
+      "learning_rate": 6.755758255197718e-06,
+      "loss": 0.8052,
+      "step": 101835
+    },
+    {
+      "epoch": 6.919418399239027,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.7536350047560815e-06,
+      "loss": 0.7782,
+      "step": 101840
+    },
+    {
+      "epoch": 6.919758119309689,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.751511754314445e-06,
+      "loss": 0.9287,
+      "step": 101845
+    },
+    {
+      "epoch": 6.92009783938035,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.7493885038728095e-06,
+      "loss": 0.8073,
+      "step": 101850
+    },
+    {
+      "epoch": 6.920437559451012,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.7472652534311735e-06,
+      "loss": 0.9866,
+      "step": 101855
+    },
+    {
+      "epoch": 6.920777279521674,
+      "grad_norm": 8.5,
+      "learning_rate": 6.745142002989537e-06,
+      "loss": 0.854,
+      "step": 101860
+    },
+    {
+      "epoch": 6.921116999592336,
+      "grad_norm": 8.125,
+      "learning_rate": 6.7430187525479015e-06,
+      "loss": 0.7548,
+      "step": 101865
+    },
+    {
+      "epoch": 6.921456719662998,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.740895502106265e-06,
+      "loss": 0.742,
+      "step": 101870
+    },
+    {
+      "epoch": 6.92179643973366,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.738772251664628e-06,
+      "loss": 0.7978,
+      "step": 101875
+    },
+    {
+      "epoch": 6.922136159804321,
+      "grad_norm": 6.6875,
+      "learning_rate": 6.736649001222993e-06,
+      "loss": 0.872,
+      "step": 101880
+    },
+    {
+      "epoch": 6.922475879874983,
+      "grad_norm": 7.28125,
+      "learning_rate": 6.734525750781357e-06,
+      "loss": 0.8195,
+      "step": 101885
+    },
+    {
+      "epoch": 6.922815599945645,
+      "grad_norm": 8.125,
+      "learning_rate": 6.73240250033972e-06,
+      "loss": 0.884,
+      "step": 101890
+    },
+    {
+      "epoch": 6.923155320016306,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.730279249898085e-06,
+      "loss": 0.8887,
+      "step": 101895
+    },
+    {
+      "epoch": 6.923495040086968,
+      "grad_norm": 7.375,
+      "learning_rate": 6.728155999456448e-06,
+      "loss": 0.8404,
+      "step": 101900
+    },
+    {
+      "epoch": 6.92383476015763,
+      "grad_norm": 7.875,
+      "learning_rate": 6.726032749014812e-06,
+      "loss": 0.8376,
+      "step": 101905
+    },
+    {
+      "epoch": 6.924174480228292,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.723909498573177e-06,
+      "loss": 0.9245,
+      "step": 101910
+    },
+    {
+      "epoch": 6.924514200298954,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.72178624813154e-06,
+      "loss": 0.7276,
+      "step": 101915
+    },
+    {
+      "epoch": 6.924853920369616,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.719662997689905e-06,
+      "loss": 0.7871,
+      "step": 101920
+    },
+    {
+      "epoch": 6.925193640440277,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.717539747248268e-06,
+      "loss": 0.865,
+      "step": 101925
+    },
+    {
+      "epoch": 6.925533360510939,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.715416496806631e-06,
+      "loss": 0.8206,
+      "step": 101930
+    },
+    {
+      "epoch": 6.925873080581601,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.713293246364996e-06,
+      "loss": 0.9277,
+      "step": 101935
+    },
+    {
+      "epoch": 6.926212800652262,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.71116999592336e-06,
+      "loss": 1.0965,
+      "step": 101940
+    },
+    {
+      "epoch": 6.926552520722924,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.709046745481723e-06,
+      "loss": 0.8403,
+      "step": 101945
+    },
+    {
+      "epoch": 6.9268922407935865,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.706923495040088e-06,
+      "loss": 0.8336,
+      "step": 101950
+    },
+    {
+      "epoch": 6.927231960864248,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.704800244598451e-06,
+      "loss": 0.9059,
+      "step": 101955
+    },
+    {
+      "epoch": 6.92757168093491,
+      "grad_norm": 8.0,
+      "learning_rate": 6.702676994156815e-06,
+      "loss": 0.8586,
+      "step": 101960
+    },
+    {
+      "epoch": 6.927911401005572,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.700553743715179e-06,
+      "loss": 0.8014,
+      "step": 101965
+    },
+    {
+      "epoch": 6.928251121076233,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.698430493273543e-06,
+      "loss": 0.8972,
+      "step": 101970
+    },
+    {
+      "epoch": 6.928590841146895,
+      "grad_norm": 8.5,
+      "learning_rate": 6.696307242831906e-06,
+      "loss": 0.8283,
+      "step": 101975
+    },
+    {
+      "epoch": 6.928930561217557,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.694183992390271e-06,
+      "loss": 0.7438,
+      "step": 101980
+    },
+    {
+      "epoch": 6.929270281288218,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.692060741948634e-06,
+      "loss": 0.7846,
+      "step": 101985
+    },
+    {
+      "epoch": 6.92961000135888,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.689937491506998e-06,
+      "loss": 0.8525,
+      "step": 101990
+    },
+    {
+      "epoch": 6.9299497214295425,
+      "grad_norm": 8.125,
+      "learning_rate": 6.687814241065363e-06,
+      "loss": 0.8596,
+      "step": 101995
+    },
+    {
+      "epoch": 6.930289441500204,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.685690990623726e-06,
+      "loss": 0.93,
+      "step": 102000
+    },
+    {
+      "epoch": 6.930629161570866,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.683567740182091e-06,
+      "loss": 0.933,
+      "step": 102005
+    },
+    {
+      "epoch": 6.930968881641528,
+      "grad_norm": 6.28125,
+      "learning_rate": 6.681444489740454e-06,
+      "loss": 0.8501,
+      "step": 102010
+    },
+    {
+      "epoch": 6.931308601712189,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.679321239298817e-06,
+      "loss": 0.8941,
+      "step": 102015
+    },
+    {
+      "epoch": 6.931648321782851,
+      "grad_norm": 8.25,
+      "learning_rate": 6.677197988857182e-06,
+      "loss": 0.92,
+      "step": 102020
+    },
+    {
+      "epoch": 6.931988041853513,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.675074738415546e-06,
+      "loss": 0.8147,
+      "step": 102025
+    },
+    {
+      "epoch": 6.932327761924174,
+      "grad_norm": 8.625,
+      "learning_rate": 6.672951487973909e-06,
+      "loss": 0.8086,
+      "step": 102030
+    },
+    {
+      "epoch": 6.932667481994836,
+      "grad_norm": 7.625,
+      "learning_rate": 6.670828237532274e-06,
+      "loss": 0.8434,
+      "step": 102035
+    },
+    {
+      "epoch": 6.9330072020654985,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.668704987090637e-06,
+      "loss": 0.8663,
+      "step": 102040
+    },
+    {
+      "epoch": 6.93334692213616,
+      "grad_norm": 8.375,
+      "learning_rate": 6.666581736649001e-06,
+      "loss": 0.9016,
+      "step": 102045
+    },
+    {
+      "epoch": 6.933686642206822,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.664458486207366e-06,
+      "loss": 0.8375,
+      "step": 102050
+    },
+    {
+      "epoch": 6.934026362277484,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.662335235765729e-06,
+      "loss": 0.8834,
+      "step": 102055
+    },
+    {
+      "epoch": 6.934366082348145,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.6602119853240925e-06,
+      "loss": 0.9021,
+      "step": 102060
+    },
+    {
+      "epoch": 6.934705802418807,
+      "grad_norm": 6.59375,
+      "learning_rate": 6.658088734882457e-06,
+      "loss": 0.8582,
+      "step": 102065
+    },
+    {
+      "epoch": 6.935045522489469,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.6559654844408205e-06,
+      "loss": 0.8861,
+      "step": 102070
+    },
+    {
+      "epoch": 6.93538524256013,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.6538422339991846e-06,
+      "loss": 0.8948,
+      "step": 102075
+    },
+    {
+      "epoch": 6.935724962630792,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.651718983557549e-06,
+      "loss": 0.8406,
+      "step": 102080
+    },
+    {
+      "epoch": 6.9360646827014545,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.6495957331159126e-06,
+      "loss": 0.8581,
+      "step": 102085
+    },
+    {
+      "epoch": 6.936404402772116,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.647472482674277e-06,
+      "loss": 0.8249,
+      "step": 102090
+    },
+    {
+      "epoch": 6.936744122842778,
+      "grad_norm": 9.125,
+      "learning_rate": 6.6453492322326406e-06,
+      "loss": 0.8761,
+      "step": 102095
+    },
+    {
+      "epoch": 6.93708384291344,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.6432259817910046e-06,
+      "loss": 0.7581,
+      "step": 102100
+    },
+    {
+      "epoch": 6.937423562984101,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.641102731349369e-06,
+      "loss": 0.8419,
+      "step": 102105
+    },
+    {
+      "epoch": 6.937763283054763,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.638979480907733e-06,
+      "loss": 0.8152,
+      "step": 102110
+    },
+    {
+      "epoch": 6.938103003125424,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.636856230466096e-06,
+      "loss": 0.7738,
+      "step": 102115
+    },
+    {
+      "epoch": 6.938442723196086,
+      "grad_norm": 8.875,
+      "learning_rate": 6.634732980024461e-06,
+      "loss": 0.9085,
+      "step": 102120
+    },
+    {
+      "epoch": 6.938782443266748,
+      "grad_norm": 8.375,
+      "learning_rate": 6.632609729582824e-06,
+      "loss": 0.8684,
+      "step": 102125
+    },
+    {
+      "epoch": 6.93912216333741,
+      "grad_norm": 7.0,
+      "learning_rate": 6.630486479141188e-06,
+      "loss": 0.8853,
+      "step": 102130
+    },
+    {
+      "epoch": 6.939461883408072,
+      "grad_norm": 8.25,
+      "learning_rate": 6.628363228699553e-06,
+      "loss": 0.7514,
+      "step": 102135
+    },
+    {
+      "epoch": 6.939801603478734,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.626239978257916e-06,
+      "loss": 0.8606,
+      "step": 102140
+    },
+    {
+      "epoch": 6.940141323549395,
+      "grad_norm": 8.125,
+      "learning_rate": 6.624116727816279e-06,
+      "loss": 0.882,
+      "step": 102145
+    },
+    {
+      "epoch": 6.940481043620057,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.621993477374644e-06,
+      "loss": 0.9168,
+      "step": 102150
+    },
+    {
+      "epoch": 6.940820763690719,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.619870226933008e-06,
+      "loss": 0.76,
+      "step": 102155
+    },
+    {
+      "epoch": 6.94116048376138,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.617746976491371e-06,
+      "loss": 0.784,
+      "step": 102160
+    },
+    {
+      "epoch": 6.941500203832042,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.615623726049736e-06,
+      "loss": 0.9039,
+      "step": 102165
+    },
+    {
+      "epoch": 6.941839923902704,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.613500475608099e-06,
+      "loss": 0.8404,
+      "step": 102170
+    },
+    {
+      "epoch": 6.942179643973366,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.611377225166464e-06,
+      "loss": 0.8899,
+      "step": 102175
+    },
+    {
+      "epoch": 6.942519364044028,
+      "grad_norm": 6.8125,
+      "learning_rate": 6.609253974724827e-06,
+      "loss": 0.7076,
+      "step": 102180
+    },
+    {
+      "epoch": 6.94285908411469,
+      "grad_norm": 8.0,
+      "learning_rate": 6.607130724283191e-06,
+      "loss": 0.8639,
+      "step": 102185
+    },
+    {
+      "epoch": 6.943198804185351,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.605007473841556e-06,
+      "loss": 0.9754,
+      "step": 102190
+    },
+    {
+      "epoch": 6.943538524256013,
+      "grad_norm": 8.875,
+      "learning_rate": 6.602884223399919e-06,
+      "loss": 0.8125,
+      "step": 102195
+    },
+    {
+      "epoch": 6.943878244326675,
+      "grad_norm": 8.5,
+      "learning_rate": 6.600760972958282e-06,
+      "loss": 0.7943,
+      "step": 102200
+    },
+    {
+      "epoch": 6.944217964397336,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.598637722516647e-06,
+      "loss": 0.9032,
+      "step": 102205
+    },
+    {
+      "epoch": 6.944557684467998,
+      "grad_norm": 7.875,
+      "learning_rate": 6.59651447207501e-06,
+      "loss": 0.7538,
+      "step": 102210
+    },
+    {
+      "epoch": 6.9448974045386604,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.594391221633374e-06,
+      "loss": 0.8215,
+      "step": 102215
+    },
+    {
+      "epoch": 6.945237124609322,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.592267971191739e-06,
+      "loss": 0.8272,
+      "step": 102220
+    },
+    {
+      "epoch": 6.945576844679984,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.590144720750102e-06,
+      "loss": 0.7505,
+      "step": 102225
+    },
+    {
+      "epoch": 6.945916564750646,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.588021470308465e-06,
+      "loss": 0.8643,
+      "step": 102230
+    },
+    {
+      "epoch": 6.946256284821307,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.58589821986683e-06,
+      "loss": 0.8726,
+      "step": 102235
+    },
+    {
+      "epoch": 6.946596004891969,
+      "grad_norm": 8.75,
+      "learning_rate": 6.583774969425194e-06,
+      "loss": 0.915,
+      "step": 102240
+    },
+    {
+      "epoch": 6.946935724962631,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.581651718983557e-06,
+      "loss": 0.9218,
+      "step": 102245
+    },
+    {
+      "epoch": 6.947275445033292,
+      "grad_norm": 8.375,
+      "learning_rate": 6.579528468541922e-06,
+      "loss": 0.8216,
+      "step": 102250
+    },
+    {
+      "epoch": 6.947615165103954,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.577405218100285e-06,
+      "loss": 0.7987,
+      "step": 102255
+    },
+    {
+      "epoch": 6.9479548851746165,
+      "grad_norm": 7.46875,
+      "learning_rate": 6.57528196765865e-06,
+      "loss": 0.8386,
+      "step": 102260
+    },
+    {
+      "epoch": 6.948294605245278,
+      "grad_norm": 7.5,
+      "learning_rate": 6.573158717217013e-06,
+      "loss": 0.8206,
+      "step": 102265
+    },
+    {
+      "epoch": 6.94863432531594,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.571035466775377e-06,
+      "loss": 0.8869,
+      "step": 102270
+    },
+    {
+      "epoch": 6.948974045386602,
+      "grad_norm": 6.875,
+      "learning_rate": 6.568912216333742e-06,
+      "loss": 0.811,
+      "step": 102275
+    },
+    {
+      "epoch": 6.949313765457263,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.566788965892105e-06,
+      "loss": 0.7668,
+      "step": 102280
+    },
+    {
+      "epoch": 6.949653485527925,
+      "grad_norm": 8.625,
+      "learning_rate": 6.5646657154504685e-06,
+      "loss": 0.8474,
+      "step": 102285
+    },
+    {
+      "epoch": 6.949993205598587,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.562542465008833e-06,
+      "loss": 0.8608,
+      "step": 102290
+    },
+    {
+      "epoch": 6.950332925669248,
+      "grad_norm": 9.0,
+      "learning_rate": 6.560419214567197e-06,
+      "loss": 0.7906,
+      "step": 102295
+    },
+    {
+      "epoch": 6.95067264573991,
+      "grad_norm": 7.625,
+      "learning_rate": 6.5582959641255605e-06,
+      "loss": 0.7415,
+      "step": 102300
+    },
+    {
+      "epoch": 6.9510123658105725,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.556172713683925e-06,
+      "loss": 0.8968,
+      "step": 102305
+    },
+    {
+      "epoch": 6.951352085881234,
+      "grad_norm": 7.375,
+      "learning_rate": 6.5540494632422885e-06,
+      "loss": 0.9222,
+      "step": 102310
+    },
+    {
+      "epoch": 6.951691805951896,
+      "grad_norm": 8.25,
+      "learning_rate": 6.551926212800652e-06,
+      "loss": 0.8005,
+      "step": 102315
+    },
+    {
+      "epoch": 6.952031526022557,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.5498029623590165e-06,
+      "loss": 0.8924,
+      "step": 102320
+    },
+    {
+      "epoch": 6.952371246093219,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.5476797119173805e-06,
+      "loss": 0.8169,
+      "step": 102325
+    },
+    {
+      "epoch": 6.952710966163881,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.545556461475744e-06,
+      "loss": 0.8568,
+      "step": 102330
+    },
+    {
+      "epoch": 6.953050686234542,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.5434332110341085e-06,
+      "loss": 0.7669,
+      "step": 102335
+    },
+    {
+      "epoch": 6.953390406305204,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.541309960592472e-06,
+      "loss": 0.8371,
+      "step": 102340
+    },
+    {
+      "epoch": 6.953730126375866,
+      "grad_norm": 8.375,
+      "learning_rate": 6.5391867101508365e-06,
+      "loss": 0.8925,
+      "step": 102345
+    },
+    {
+      "epoch": 6.954069846446528,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.5370634597092e-06,
+      "loss": 0.9685,
+      "step": 102350
+    },
+    {
+      "epoch": 6.95440956651719,
+      "grad_norm": 7.0625,
+      "learning_rate": 6.534940209267564e-06,
+      "loss": 0.9775,
+      "step": 102355
+    },
+    {
+      "epoch": 6.954749286587852,
+      "grad_norm": 8.0,
+      "learning_rate": 6.5328169588259285e-06,
+      "loss": 0.7867,
+      "step": 102360
+    },
+    {
+      "epoch": 6.955089006658513,
+      "grad_norm": 7.875,
+      "learning_rate": 6.530693708384292e-06,
+      "loss": 0.8433,
+      "step": 102365
+    },
+    {
+      "epoch": 6.955428726729175,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.528570457942655e-06,
+      "loss": 0.9523,
+      "step": 102370
+    },
+    {
+      "epoch": 6.955768446799837,
+      "grad_norm": 7.21875,
+      "learning_rate": 6.52644720750102e-06,
+      "loss": 0.8622,
+      "step": 102375
+    },
+    {
+      "epoch": 6.956108166870498,
+      "grad_norm": 8.125,
+      "learning_rate": 6.524323957059384e-06,
+      "loss": 0.8038,
+      "step": 102380
+    },
+    {
+      "epoch": 6.95644788694116,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.522200706617747e-06,
+      "loss": 0.8699,
+      "step": 102385
+    },
+    {
+      "epoch": 6.956787607011822,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.520077456176112e-06,
+      "loss": 0.8615,
+      "step": 102390
+    },
+    {
+      "epoch": 6.957127327082484,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.517954205734475e-06,
+      "loss": 0.7493,
+      "step": 102395
+    },
+    {
+      "epoch": 6.957467047153146,
+      "grad_norm": 8.5,
+      "learning_rate": 6.515830955292838e-06,
+      "loss": 0.9388,
+      "step": 102400
+    },
+    {
+      "epoch": 6.957806767223808,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.513707704851203e-06,
+      "loss": 0.8953,
+      "step": 102405
+    },
+    {
+      "epoch": 6.958146487294469,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.511584454409567e-06,
+      "loss": 0.8996,
+      "step": 102410
+    },
+    {
+      "epoch": 6.958486207365131,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.50946120396793e-06,
+      "loss": 0.7918,
+      "step": 102415
+    },
+    {
+      "epoch": 6.958825927435793,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.507337953526295e-06,
+      "loss": 0.8757,
+      "step": 102420
+    },
+    {
+      "epoch": 6.959165647506454,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.505214703084658e-06,
+      "loss": 0.8584,
+      "step": 102425
+    },
+    {
+      "epoch": 6.959505367577116,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.503091452643023e-06,
+      "loss": 0.8289,
+      "step": 102430
+    },
+    {
+      "epoch": 6.959845087647778,
+      "grad_norm": 8.5,
+      "learning_rate": 6.500968202201387e-06,
+      "loss": 0.8682,
+      "step": 102435
+    },
+    {
+      "epoch": 6.96018480771844,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.49884495175975e-06,
+      "loss": 0.8946,
+      "step": 102440
+    },
+    {
+      "epoch": 6.960524527789102,
+      "grad_norm": 8.125,
+      "learning_rate": 6.496721701318115e-06,
+      "loss": 0.8465,
+      "step": 102445
+    },
+    {
+      "epoch": 6.960864247859764,
+      "grad_norm": 8.5,
+      "learning_rate": 6.494598450876478e-06,
+      "loss": 0.9261,
+      "step": 102450
+    },
+    {
+      "epoch": 6.961203967930425,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.492475200434841e-06,
+      "loss": 0.7829,
+      "step": 102455
+    },
+    {
+      "epoch": 6.961543688001087,
+      "grad_norm": 6.28125,
+      "learning_rate": 6.490351949993206e-06,
+      "loss": 0.8144,
+      "step": 102460
+    },
+    {
+      "epoch": 6.961883408071749,
+      "grad_norm": 7.40625,
+      "learning_rate": 6.48822869955157e-06,
+      "loss": 0.818,
+      "step": 102465
+    },
+    {
+      "epoch": 6.96222312814241,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.486105449109933e-06,
+      "loss": 0.84,
+      "step": 102470
+    },
+    {
+      "epoch": 6.962562848213072,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.483982198668298e-06,
+      "loss": 0.7421,
+      "step": 102475
+    },
+    {
+      "epoch": 6.962902568283734,
+      "grad_norm": 7.78125,
+      "learning_rate": 6.481858948226661e-06,
+      "loss": 0.8191,
+      "step": 102480
+    },
+    {
+      "epoch": 6.963242288354396,
+      "grad_norm": 7.78125,
+      "learning_rate": 6.479735697785025e-06,
+      "loss": 0.9228,
+      "step": 102485
+    },
+    {
+      "epoch": 6.963582008425058,
+      "grad_norm": 8.0,
+      "learning_rate": 6.47761244734339e-06,
+      "loss": 0.9145,
+      "step": 102490
+    },
+    {
+      "epoch": 6.96392172849572,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.475489196901753e-06,
+      "loss": 0.8782,
+      "step": 102495
+    },
+    {
+      "epoch": 6.964261448566381,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.473365946460116e-06,
+      "loss": 0.8148,
+      "step": 102500
+    },
+    {
+      "epoch": 6.964601168637043,
+      "grad_norm": 7.375,
+      "learning_rate": 6.471242696018481e-06,
+      "loss": 0.8229,
+      "step": 102505
+    },
+    {
+      "epoch": 6.964940888707705,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.469119445576844e-06,
+      "loss": 0.8004,
+      "step": 102510
+    },
+    {
+      "epoch": 6.965280608778366,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.466996195135209e-06,
+      "loss": 0.8232,
+      "step": 102515
+    },
+    {
+      "epoch": 6.965620328849028,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.464872944693573e-06,
+      "loss": 0.8584,
+      "step": 102520
+    },
+    {
+      "epoch": 6.9659600489196905,
+      "grad_norm": 8.125,
+      "learning_rate": 6.4627496942519364e-06,
+      "loss": 0.8526,
+      "step": 102525
+    },
+    {
+      "epoch": 6.966299768990352,
+      "grad_norm": 8.75,
+      "learning_rate": 6.460626443810301e-06,
+      "loss": 0.913,
+      "step": 102530
+    },
+    {
+      "epoch": 6.966639489061014,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.4585031933686644e-06,
+      "loss": 0.9426,
+      "step": 102535
+    },
+    {
+      "epoch": 6.966979209131676,
+      "grad_norm": 8.125,
+      "learning_rate": 6.4563799429270284e-06,
+      "loss": 0.9486,
+      "step": 102540
+    },
+    {
+      "epoch": 6.967318929202337,
+      "grad_norm": 8.375,
+      "learning_rate": 6.4542566924853924e-06,
+      "loss": 0.8419,
+      "step": 102545
+    },
+    {
+      "epoch": 6.967658649272999,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.4521334420437564e-06,
+      "loss": 0.8,
+      "step": 102550
+    },
+    {
+      "epoch": 6.967998369343661,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.45001019160212e-06,
+      "loss": 0.8586,
+      "step": 102555
+    },
+    {
+      "epoch": 6.968338089414322,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.4478869411604845e-06,
+      "loss": 0.8741,
+      "step": 102560
+    },
+    {
+      "epoch": 6.968677809484984,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.445763690718848e-06,
+      "loss": 0.9389,
+      "step": 102565
+    },
+    {
+      "epoch": 6.9690175295556465,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.443640440277212e-06,
+      "loss": 0.921,
+      "step": 102570
+    },
+    {
+      "epoch": 6.969357249626308,
+      "grad_norm": 7.125,
+      "learning_rate": 6.4415171898355765e-06,
+      "loss": 0.8488,
+      "step": 102575
+    },
+    {
+      "epoch": 6.96969696969697,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.43939393939394e-06,
+      "loss": 0.886,
+      "step": 102580
+    },
+    {
+      "epoch": 6.970036689767632,
+      "grad_norm": 9.125,
+      "learning_rate": 6.437270688952303e-06,
+      "loss": 0.8374,
+      "step": 102585
+    },
+    {
+      "epoch": 6.970376409838293,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.435147438510668e-06,
+      "loss": 0.798,
+      "step": 102590
+    },
+    {
+      "epoch": 6.970716129908955,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.433024188069031e-06,
+      "loss": 0.8057,
+      "step": 102595
+    },
+    {
+      "epoch": 6.971055849979617,
+      "grad_norm": 7.59375,
+      "learning_rate": 6.430900937627396e-06,
+      "loss": 0.8905,
+      "step": 102600
+    },
+    {
+      "epoch": 6.971395570050278,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.42877768718576e-06,
+      "loss": 0.8832,
+      "step": 102605
+    },
+    {
+      "epoch": 6.97173529012094,
+      "grad_norm": 7.375,
+      "learning_rate": 6.426654436744123e-06,
+      "loss": 0.85,
+      "step": 102610
+    },
+    {
+      "epoch": 6.9720750101916025,
+      "grad_norm": 7.125,
+      "learning_rate": 6.424531186302488e-06,
+      "loss": 0.7929,
+      "step": 102615
+    },
+    {
+      "epoch": 6.972414730262264,
+      "grad_norm": 8.5,
+      "learning_rate": 6.422407935860851e-06,
+      "loss": 0.8251,
+      "step": 102620
+    },
+    {
+      "epoch": 6.972754450332926,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.420284685419215e-06,
+      "loss": 0.8484,
+      "step": 102625
+    },
+    {
+      "epoch": 6.973094170403588,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.41816143497758e-06,
+      "loss": 0.8334,
+      "step": 102630
+    },
+    {
+      "epoch": 6.973433890474249,
+      "grad_norm": 8.375,
+      "learning_rate": 6.416038184535943e-06,
+      "loss": 0.8312,
+      "step": 102635
+    },
+    {
+      "epoch": 6.973773610544911,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.413914934094306e-06,
+      "loss": 0.9145,
+      "step": 102640
+    },
+    {
+      "epoch": 6.974113330615573,
+      "grad_norm": 7.46875,
+      "learning_rate": 6.411791683652671e-06,
+      "loss": 0.7233,
+      "step": 102645
+    },
+    {
+      "epoch": 6.974453050686234,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.409668433211034e-06,
+      "loss": 0.7568,
+      "step": 102650
+    },
+    {
+      "epoch": 6.974792770756896,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.407545182769398e-06,
+      "loss": 0.8612,
+      "step": 102655
+    },
+    {
+      "epoch": 6.9751324908275585,
+      "grad_norm": 7.75,
+      "learning_rate": 6.405421932327763e-06,
+      "loss": 0.799,
+      "step": 102660
+    },
+    {
+      "epoch": 6.97547221089822,
+      "grad_norm": 8.25,
+      "learning_rate": 6.403298681886126e-06,
+      "loss": 0.8147,
+      "step": 102665
+    },
+    {
+      "epoch": 6.975811930968882,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.401175431444489e-06,
+      "loss": 0.8393,
+      "step": 102670
+    },
+    {
+      "epoch": 6.976151651039544,
+      "grad_norm": 8.125,
+      "learning_rate": 6.399052181002854e-06,
+      "loss": 0.7715,
+      "step": 102675
+    },
+    {
+      "epoch": 6.976491371110205,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.396928930561218e-06,
+      "loss": 0.8773,
+      "step": 102680
+    },
+    {
+      "epoch": 6.976831091180867,
+      "grad_norm": 8.0,
+      "learning_rate": 6.394805680119582e-06,
+      "loss": 0.7821,
+      "step": 102685
+    },
+    {
+      "epoch": 6.977170811251529,
+      "grad_norm": 8.25,
+      "learning_rate": 6.392682429677946e-06,
+      "loss": 0.7624,
+      "step": 102690
+    },
+    {
+      "epoch": 6.97751053132219,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.390559179236309e-06,
+      "loss": 0.8377,
+      "step": 102695
+    },
+    {
+      "epoch": 6.977850251392852,
+      "grad_norm": 9.75,
+      "learning_rate": 6.388435928794674e-06,
+      "loss": 0.862,
+      "step": 102700
+    },
+    {
+      "epoch": 6.9781899714635145,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.386312678353037e-06,
+      "loss": 0.9218,
+      "step": 102705
+    },
+    {
+      "epoch": 6.978529691534176,
+      "grad_norm": 8.875,
+      "learning_rate": 6.384189427911401e-06,
+      "loss": 0.8378,
+      "step": 102710
+    },
+    {
+      "epoch": 6.978869411604838,
+      "grad_norm": 6.96875,
+      "learning_rate": 6.382066177469766e-06,
+      "loss": 0.8309,
+      "step": 102715
+    },
+    {
+      "epoch": 6.9792091316755,
+      "grad_norm": 8.125,
+      "learning_rate": 6.379942927028129e-06,
+      "loss": 0.8462,
+      "step": 102720
+    },
+    {
+      "epoch": 6.979548851746161,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.377819676586492e-06,
+      "loss": 0.9126,
+      "step": 102725
+    },
+    {
+      "epoch": 6.979888571816823,
+      "grad_norm": 7.59375,
+      "learning_rate": 6.375696426144857e-06,
+      "loss": 0.8991,
+      "step": 102730
+    },
+    {
+      "epoch": 6.980228291887485,
+      "grad_norm": 8.125,
+      "learning_rate": 6.37357317570322e-06,
+      "loss": 0.8626,
+      "step": 102735
+    },
+    {
+      "epoch": 6.980568011958146,
+      "grad_norm": 8.0,
+      "learning_rate": 6.371449925261584e-06,
+      "loss": 0.7791,
+      "step": 102740
+    },
+    {
+      "epoch": 6.980907732028808,
+      "grad_norm": 8.5,
+      "learning_rate": 6.369326674819949e-06,
+      "loss": 0.819,
+      "step": 102745
+    },
+    {
+      "epoch": 6.9812474520994705,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.367203424378312e-06,
+      "loss": 0.9267,
+      "step": 102750
+    },
+    {
+      "epoch": 6.981587172170132,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.3650801739366755e-06,
+      "loss": 0.7986,
+      "step": 102755
+    },
+    {
+      "epoch": 6.981926892240794,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.36295692349504e-06,
+      "loss": 0.8219,
+      "step": 102760
+    },
+    {
+      "epoch": 6.982266612311456,
+      "grad_norm": 8.625,
+      "learning_rate": 6.360833673053404e-06,
+      "loss": 0.8185,
+      "step": 102765
+    },
+    {
+      "epoch": 6.982606332382117,
+      "grad_norm": 6.75,
+      "learning_rate": 6.358710422611769e-06,
+      "loss": 0.9115,
+      "step": 102770
+    },
+    {
+      "epoch": 6.982946052452779,
+      "grad_norm": 7.875,
+      "learning_rate": 6.356587172170132e-06,
+      "loss": 0.8594,
+      "step": 102775
+    },
+    {
+      "epoch": 6.983285772523441,
+      "grad_norm": 7.09375,
+      "learning_rate": 6.3544639217284955e-06,
+      "loss": 0.9008,
+      "step": 102780
+    },
+    {
+      "epoch": 6.983625492594102,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.35234067128686e-06,
+      "loss": 0.8578,
+      "step": 102785
+    },
+    {
+      "epoch": 6.983965212664764,
+      "grad_norm": 7.78125,
+      "learning_rate": 6.3502174208452236e-06,
+      "loss": 0.8495,
+      "step": 102790
+    },
+    {
+      "epoch": 6.984304932735426,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.3480941704035876e-06,
+      "loss": 0.889,
+      "step": 102795
+    },
+    {
+      "epoch": 6.984644652806088,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.345970919961952e-06,
+      "loss": 0.8237,
+      "step": 102800
+    },
+    {
+      "epoch": 6.98498437287675,
+      "grad_norm": 7.0625,
+      "learning_rate": 6.3438476695203156e-06,
+      "loss": 0.8671,
+      "step": 102805
+    },
+    {
+      "epoch": 6.985324092947411,
+      "grad_norm": 7.3125,
+      "learning_rate": 6.341724419078679e-06,
+      "loss": 0.8834,
+      "step": 102810
+    },
+    {
+      "epoch": 6.985663813018073,
+      "grad_norm": 8.625,
+      "learning_rate": 6.3396011686370436e-06,
+      "loss": 0.8485,
+      "step": 102815
+    },
+    {
+      "epoch": 6.986003533088735,
+      "grad_norm": 7.5,
+      "learning_rate": 6.3374779181954076e-06,
+      "loss": 0.8517,
+      "step": 102820
+    },
+    {
+      "epoch": 6.986343253159396,
+      "grad_norm": 8.625,
+      "learning_rate": 6.335354667753771e-06,
+      "loss": 0.84,
+      "step": 102825
+    },
+    {
+      "epoch": 6.986682973230058,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.333231417312136e-06,
+      "loss": 0.8452,
+      "step": 102830
+    },
+    {
+      "epoch": 6.9870226933007205,
+      "grad_norm": 9.4375,
+      "learning_rate": 6.331108166870499e-06,
+      "loss": 0.8666,
+      "step": 102835
+    },
+    {
+      "epoch": 6.987362413371382,
+      "grad_norm": 8.0,
+      "learning_rate": 6.328984916428862e-06,
+      "loss": 0.8733,
+      "step": 102840
+    },
+    {
+      "epoch": 6.987702133442044,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.326861665987227e-06,
+      "loss": 0.8632,
+      "step": 102845
+    },
+    {
+      "epoch": 6.988041853512706,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.324738415545591e-06,
+      "loss": 0.815,
+      "step": 102850
+    },
+    {
+      "epoch": 6.988381573583367,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.322615165103956e-06,
+      "loss": 0.7143,
+      "step": 102855
+    },
+    {
+      "epoch": 6.988721293654029,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.320491914662319e-06,
+      "loss": 0.8552,
+      "step": 102860
+    },
+    {
+      "epoch": 6.989061013724691,
+      "grad_norm": 8.5,
+      "learning_rate": 6.318368664220682e-06,
+      "loss": 0.7777,
+      "step": 102865
+    },
+    {
+      "epoch": 6.989400733795352,
+      "grad_norm": 8.375,
+      "learning_rate": 6.316245413779047e-06,
+      "loss": 0.9007,
+      "step": 102870
+    },
+    {
+      "epoch": 6.989740453866014,
+      "grad_norm": 8.0,
+      "learning_rate": 6.314122163337411e-06,
+      "loss": 0.799,
+      "step": 102875
+    },
+    {
+      "epoch": 6.9900801739366765,
+      "grad_norm": 8.5,
+      "learning_rate": 6.311998912895774e-06,
+      "loss": 0.9168,
+      "step": 102880
+    },
+    {
+      "epoch": 6.990419894007338,
+      "grad_norm": 9.0,
+      "learning_rate": 6.309875662454139e-06,
+      "loss": 0.8544,
+      "step": 102885
+    },
+    {
+      "epoch": 6.990759614078,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.307752412012502e-06,
+      "loss": 0.8913,
+      "step": 102890
+    },
+    {
+      "epoch": 6.991099334148662,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.305629161570865e-06,
+      "loss": 0.771,
+      "step": 102895
+    },
+    {
+      "epoch": 6.991439054219323,
+      "grad_norm": 7.28125,
+      "learning_rate": 6.30350591112923e-06,
+      "loss": 0.9354,
+      "step": 102900
+    },
+    {
+      "epoch": 6.991778774289985,
+      "grad_norm": 8.625,
+      "learning_rate": 6.301382660687594e-06,
+      "loss": 0.7765,
+      "step": 102905
+    },
+    {
+      "epoch": 6.992118494360647,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.299259410245957e-06,
+      "loss": 0.8556,
+      "step": 102910
+    },
+    {
+      "epoch": 6.992458214431308,
+      "grad_norm": 7.78125,
+      "learning_rate": 6.297136159804322e-06,
+      "loss": 0.8573,
+      "step": 102915
+    },
+    {
+      "epoch": 6.99279793450197,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.295012909362685e-06,
+      "loss": 0.7687,
+      "step": 102920
+    },
+    {
+      "epoch": 6.9931376545726325,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.292889658921049e-06,
+      "loss": 0.8822,
+      "step": 102925
+    },
+    {
+      "epoch": 6.993477374643294,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.290766408479413e-06,
+      "loss": 0.8448,
+      "step": 102930
+    },
+    {
+      "epoch": 6.993817094713956,
+      "grad_norm": 9.0,
+      "learning_rate": 6.288643158037777e-06,
+      "loss": 0.9129,
+      "step": 102935
+    },
+    {
+      "epoch": 6.994156814784618,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.286519907596142e-06,
+      "loss": 0.9058,
+      "step": 102940
+    },
+    {
+      "epoch": 6.994496534855279,
+      "grad_norm": 8.0,
+      "learning_rate": 6.284396657154505e-06,
+      "loss": 0.8999,
+      "step": 102945
+    },
+    {
+      "epoch": 6.994836254925941,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.282273406712868e-06,
+      "loss": 0.922,
+      "step": 102950
+    },
+    {
+      "epoch": 6.995175974996603,
+      "grad_norm": 8.375,
+      "learning_rate": 6.280150156271233e-06,
+      "loss": 0.8914,
+      "step": 102955
+    },
+    {
+      "epoch": 6.995515695067264,
+      "grad_norm": 8.375,
+      "learning_rate": 6.278026905829597e-06,
+      "loss": 0.8397,
+      "step": 102960
+    },
+    {
+      "epoch": 6.995855415137926,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.27590365538796e-06,
+      "loss": 0.7446,
+      "step": 102965
+    },
+    {
+      "epoch": 6.9961951352085885,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.273780404946325e-06,
+      "loss": 0.7854,
+      "step": 102970
+    },
+    {
+      "epoch": 6.99653485527925,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.271657154504688e-06,
+      "loss": 0.8764,
+      "step": 102975
+    },
+    {
+      "epoch": 6.996874575349912,
+      "grad_norm": 7.59375,
+      "learning_rate": 6.2695339040630515e-06,
+      "loss": 0.8829,
+      "step": 102980
+    },
+    {
+      "epoch": 6.997214295420574,
+      "grad_norm": 7.625,
+      "learning_rate": 6.267410653621416e-06,
+      "loss": 0.841,
+      "step": 102985
+    },
+    {
+      "epoch": 6.997554015491235,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.26528740317978e-06,
+      "loss": 0.8968,
+      "step": 102990
+    },
+    {
+      "epoch": 6.997893735561897,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.2631641527381435e-06,
+      "loss": 0.8104,
+      "step": 102995
+    },
+    {
+      "epoch": 6.998233455632558,
+      "grad_norm": 7.375,
+      "learning_rate": 6.261040902296508e-06,
+      "loss": 0.7932,
+      "step": 103000
+    },
+    {
+      "epoch": 6.99857317570322,
+      "grad_norm": 7.25,
+      "learning_rate": 6.2589176518548715e-06,
+      "loss": 0.7751,
+      "step": 103005
+    },
+    {
+      "epoch": 6.998912895773882,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.2567944014132355e-06,
+      "loss": 0.7868,
+      "step": 103010
+    },
+    {
+      "epoch": 6.999252615844544,
+      "grad_norm": 7.03125,
+      "learning_rate": 6.2546711509716e-06,
+      "loss": 0.8611,
+      "step": 103015
+    },
+    {
+      "epoch": 6.999592335915206,
+      "grad_norm": 7.75,
+      "learning_rate": 6.2525479005299635e-06,
+      "loss": 0.8114,
+      "step": 103020
+    },
+    {
+      "epoch": 6.999932055985868,
+      "grad_norm": 7.875,
+      "learning_rate": 6.250424650088328e-06,
+      "loss": 0.7613,
+      "step": 103025
+    },
+    {
+      "epoch": 7.0,
+      "eval_loss": 1.7656691074371338,
+      "eval_runtime": 146.9077,
+      "eval_samples_per_second": 70.241,
+      "eval_steps_per_second": 8.781,
+      "step": 103026
+    },
+    {
+      "epoch": 7.00027177605653,
+      "grad_norm": 7.25,
+      "learning_rate": 6.2483013996466915e-06,
+      "loss": 0.7451,
+      "step": 103030
+    },
+    {
+      "epoch": 7.000611496127191,
+      "grad_norm": 7.78125,
+      "learning_rate": 6.246178149205055e-06,
+      "loss": 0.8399,
+      "step": 103035
+    },
+    {
+      "epoch": 7.000951216197853,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.2440548987634195e-06,
+      "loss": 0.8483,
+      "step": 103040
+    },
+    {
+      "epoch": 7.001290936268515,
+      "grad_norm": 7.46875,
+      "learning_rate": 6.2419316483217835e-06,
+      "loss": 0.9035,
+      "step": 103045
+    },
+    {
+      "epoch": 7.001630656339176,
+      "grad_norm": 8.625,
+      "learning_rate": 6.2398083978801475e-06,
+      "loss": 0.8896,
+      "step": 103050
+    },
+    {
+      "epoch": 7.001970376409838,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.237685147438511e-06,
+      "loss": 0.8434,
+      "step": 103055
+    },
+    {
+      "epoch": 7.0023100964805,
+      "grad_norm": 8.375,
+      "learning_rate": 6.235561896996875e-06,
+      "loss": 0.8971,
+      "step": 103060
+    },
+    {
+      "epoch": 7.002649816551162,
+      "grad_norm": 8.125,
+      "learning_rate": 6.233438646555239e-06,
+      "loss": 0.8811,
+      "step": 103065
+    },
+    {
+      "epoch": 7.002989536621824,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.231315396113603e-06,
+      "loss": 0.8756,
+      "step": 103070
+    },
+    {
+      "epoch": 7.003329256692485,
+      "grad_norm": 8.25,
+      "learning_rate": 6.229192145671967e-06,
+      "loss": 0.7957,
+      "step": 103075
+    },
+    {
+      "epoch": 7.003668976763147,
+      "grad_norm": 7.34375,
+      "learning_rate": 6.227068895230331e-06,
+      "loss": 0.7774,
+      "step": 103080
+    },
+    {
+      "epoch": 7.004008696833809,
+      "grad_norm": 8.5,
+      "learning_rate": 6.224945644788695e-06,
+      "loss": 0.882,
+      "step": 103085
+    },
+    {
+      "epoch": 7.00434841690447,
+      "grad_norm": 8.0,
+      "learning_rate": 6.222822394347058e-06,
+      "loss": 0.7783,
+      "step": 103090
+    },
+    {
+      "epoch": 7.004688136975132,
+      "grad_norm": 8.75,
+      "learning_rate": 6.220699143905422e-06,
+      "loss": 0.8568,
+      "step": 103095
+    },
+    {
+      "epoch": 7.0050278570457944,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.218575893463787e-06,
+      "loss": 0.8352,
+      "step": 103100
+    },
+    {
+      "epoch": 7.005367577116456,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.21645264302215e-06,
+      "loss": 0.8488,
+      "step": 103105
+    },
+    {
+      "epoch": 7.005707297187118,
+      "grad_norm": 7.5,
+      "learning_rate": 6.214329392580514e-06,
+      "loss": 0.7922,
+      "step": 103110
+    },
+    {
+      "epoch": 7.00604701725778,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.212206142138878e-06,
+      "loss": 0.8658,
+      "step": 103115
+    },
+    {
+      "epoch": 7.006386737328441,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.210082891697241e-06,
+      "loss": 0.8079,
+      "step": 103120
+    },
+    {
+      "epoch": 7.006726457399103,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.207959641255606e-06,
+      "loss": 0.8219,
+      "step": 103125
+    },
+    {
+      "epoch": 7.007066177469765,
+      "grad_norm": 8.0,
+      "learning_rate": 6.20583639081397e-06,
+      "loss": 0.9223,
+      "step": 103130
+    },
+    {
+      "epoch": 7.007405897540426,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.203713140372334e-06,
+      "loss": 0.8739,
+      "step": 103135
+    },
+    {
+      "epoch": 7.007745617611088,
+      "grad_norm": 7.625,
+      "learning_rate": 6.201589889930697e-06,
+      "loss": 0.8471,
+      "step": 103140
+    },
+    {
+      "epoch": 7.0080853376817505,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.199466639489061e-06,
+      "loss": 0.8118,
+      "step": 103145
+    },
+    {
+      "epoch": 7.008425057752412,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.197343389047425e-06,
+      "loss": 0.8452,
+      "step": 103150
+    },
+    {
+      "epoch": 7.008764777823074,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.195220138605789e-06,
+      "loss": 0.8623,
+      "step": 103155
+    },
+    {
+      "epoch": 7.009104497893736,
+      "grad_norm": 7.0625,
+      "learning_rate": 6.193096888164153e-06,
+      "loss": 0.8808,
+      "step": 103160
+    },
+    {
+      "epoch": 7.009444217964397,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.190973637722517e-06,
+      "loss": 0.7965,
+      "step": 103165
+    },
+    {
+      "epoch": 7.009783938035059,
+      "grad_norm": 7.5,
+      "learning_rate": 6.188850387280881e-06,
+      "loss": 0.7983,
+      "step": 103170
+    },
+    {
+      "epoch": 7.010123658105721,
+      "grad_norm": 8.5,
+      "learning_rate": 6.186727136839244e-06,
+      "loss": 0.8276,
+      "step": 103175
+    },
+    {
+      "epoch": 7.010463378176382,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.184603886397609e-06,
+      "loss": 0.9193,
+      "step": 103180
+    },
+    {
+      "epoch": 7.010803098247044,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.182480635955973e-06,
+      "loss": 0.7752,
+      "step": 103185
+    },
+    {
+      "epoch": 7.0111428183177065,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.180357385514336e-06,
+      "loss": 0.8088,
+      "step": 103190
+    },
+    {
+      "epoch": 7.011482538388368,
+      "grad_norm": 8.125,
+      "learning_rate": 6.1782341350727e-06,
+      "loss": 0.8468,
+      "step": 103195
+    },
+    {
+      "epoch": 7.01182225845903,
+      "grad_norm": 7.15625,
+      "learning_rate": 6.176110884631064e-06,
+      "loss": 0.8299,
+      "step": 103200
+    },
+    {
+      "epoch": 7.012161978529692,
+      "grad_norm": 8.0,
+      "learning_rate": 6.173987634189428e-06,
+      "loss": 0.7721,
+      "step": 103205
+    },
+    {
+      "epoch": 7.012501698600353,
+      "grad_norm": 7.59375,
+      "learning_rate": 6.171864383747792e-06,
+      "loss": 0.8605,
+      "step": 103210
+    },
+    {
+      "epoch": 7.012841418671015,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.169741133306156e-06,
+      "loss": 0.8431,
+      "step": 103215
+    },
+    {
+      "epoch": 7.013181138741677,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.16761788286452e-06,
+      "loss": 0.886,
+      "step": 103220
+    },
+    {
+      "epoch": 7.013520858812338,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.165494632422883e-06,
+      "loss": 0.8728,
+      "step": 103225
+    },
+    {
+      "epoch": 7.013860578883,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.163371381981247e-06,
+      "loss": 0.8632,
+      "step": 103230
+    },
+    {
+      "epoch": 7.0142002989536625,
+      "grad_norm": 7.875,
+      "learning_rate": 6.161248131539612e-06,
+      "loss": 0.8646,
+      "step": 103235
+    },
+    {
+      "epoch": 7.014540019024324,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.1591248810979754e-06,
+      "loss": 0.885,
+      "step": 103240
+    },
+    {
+      "epoch": 7.014879739094986,
+      "grad_norm": 7.1875,
+      "learning_rate": 6.1570016306563394e-06,
+      "loss": 0.8031,
+      "step": 103245
+    },
+    {
+      "epoch": 7.015219459165648,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.1548783802147034e-06,
+      "loss": 0.838,
+      "step": 103250
+    },
+    {
+      "epoch": 7.015559179236309,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.1527551297730674e-06,
+      "loss": 0.8352,
+      "step": 103255
+    },
+    {
+      "epoch": 7.015898899306971,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.1506318793314314e-06,
+      "loss": 0.9397,
+      "step": 103260
+    },
+    {
+      "epoch": 7.016238619377633,
+      "grad_norm": 6.84375,
+      "learning_rate": 6.1485086288897954e-06,
+      "loss": 0.8675,
+      "step": 103265
+    },
+    {
+      "epoch": 7.016578339448294,
+      "grad_norm": 8.25,
+      "learning_rate": 6.1463853784481595e-06,
+      "loss": 0.8148,
+      "step": 103270
+    },
+    {
+      "epoch": 7.016918059518956,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.144262128006523e-06,
+      "loss": 0.8501,
+      "step": 103275
+    },
+    {
+      "epoch": 7.0172577795896185,
+      "grad_norm": 8.375,
+      "learning_rate": 6.142138877564887e-06,
+      "loss": 0.8447,
+      "step": 103280
+    },
+    {
+      "epoch": 7.01759749966028,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.140015627123251e-06,
+      "loss": 0.8564,
+      "step": 103285
+    },
+    {
+      "epoch": 7.017937219730942,
+      "grad_norm": 8.375,
+      "learning_rate": 6.137892376681615e-06,
+      "loss": 0.8195,
+      "step": 103290
+    },
+    {
+      "epoch": 7.018276939801604,
+      "grad_norm": 8.375,
+      "learning_rate": 6.135769126239979e-06,
+      "loss": 0.8381,
+      "step": 103295
+    },
+    {
+      "epoch": 7.018616659872265,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.133645875798343e-06,
+      "loss": 0.9194,
+      "step": 103300
+    },
+    {
+      "epoch": 7.018956379942927,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.131522625356707e-06,
+      "loss": 0.8589,
+      "step": 103305
+    },
+    {
+      "epoch": 7.019296100013589,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.12939937491507e-06,
+      "loss": 0.7461,
+      "step": 103310
+    },
+    {
+      "epoch": 7.01963582008425,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.127276124473434e-06,
+      "loss": 0.8638,
+      "step": 103315
+    },
+    {
+      "epoch": 7.019975540154912,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.125152874031799e-06,
+      "loss": 0.8389,
+      "step": 103320
+    },
+    {
+      "epoch": 7.0203152602255745,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.123029623590162e-06,
+      "loss": 0.9107,
+      "step": 103325
+    },
+    {
+      "epoch": 7.020654980296236,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.120906373148526e-06,
+      "loss": 0.8455,
+      "step": 103330
+    },
+    {
+      "epoch": 7.020994700366898,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.11878312270689e-06,
+      "loss": 0.8143,
+      "step": 103335
+    },
+    {
+      "epoch": 7.02133442043756,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.116659872265254e-06,
+      "loss": 0.7574,
+      "step": 103340
+    },
+    {
+      "epoch": 7.021674140508221,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.114536621823618e-06,
+      "loss": 0.9324,
+      "step": 103345
+    },
+    {
+      "epoch": 7.022013860578883,
+      "grad_norm": 6.71875,
+      "learning_rate": 6.112413371381982e-06,
+      "loss": 0.9427,
+      "step": 103350
+    },
+    {
+      "epoch": 7.022353580649545,
+      "grad_norm": 7.75,
+      "learning_rate": 6.110290120940346e-06,
+      "loss": 0.7844,
+      "step": 103355
+    },
+    {
+      "epoch": 7.022693300720206,
+      "grad_norm": 7.53125,
+      "learning_rate": 6.108166870498709e-06,
+      "loss": 0.8169,
+      "step": 103360
+    },
+    {
+      "epoch": 7.023033020790868,
+      "grad_norm": 8.5,
+      "learning_rate": 6.106043620057073e-06,
+      "loss": 0.7798,
+      "step": 103365
+    },
+    {
+      "epoch": 7.0233727408615305,
+      "grad_norm": 8.375,
+      "learning_rate": 6.103920369615437e-06,
+      "loss": 0.7961,
+      "step": 103370
+    },
+    {
+      "epoch": 7.023712460932192,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.101797119173801e-06,
+      "loss": 0.7531,
+      "step": 103375
+    },
+    {
+      "epoch": 7.024052181002854,
+      "grad_norm": 7.75,
+      "learning_rate": 6.099673868732165e-06,
+      "loss": 0.7976,
+      "step": 103380
+    },
+    {
+      "epoch": 7.024391901073516,
+      "grad_norm": 7.9375,
+      "learning_rate": 6.097550618290529e-06,
+      "loss": 0.811,
+      "step": 103385
+    },
+    {
+      "epoch": 7.024731621144177,
+      "grad_norm": 8.8125,
+      "learning_rate": 6.095427367848893e-06,
+      "loss": 0.7973,
+      "step": 103390
+    },
+    {
+      "epoch": 7.025071341214839,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.093304117407256e-06,
+      "loss": 0.8162,
+      "step": 103395
+    },
+    {
+      "epoch": 7.025411061285501,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.091180866965621e-06,
+      "loss": 0.8786,
+      "step": 103400
+    },
+    {
+      "epoch": 7.025750781356162,
+      "grad_norm": 7.8125,
+      "learning_rate": 6.089057616523985e-06,
+      "loss": 0.8984,
+      "step": 103405
+    },
+    {
+      "epoch": 7.0260905014268245,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.086934366082348e-06,
+      "loss": 0.762,
+      "step": 103410
+    },
+    {
+      "epoch": 7.026430221497486,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.084811115640712e-06,
+      "loss": 0.7628,
+      "step": 103415
+    },
+    {
+      "epoch": 7.026769941568148,
+      "grad_norm": 9.1875,
+      "learning_rate": 6.082687865199076e-06,
+      "loss": 0.826,
+      "step": 103420
+    },
+    {
+      "epoch": 7.02710966163881,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.08056461475744e-06,
+      "loss": 0.7958,
+      "step": 103425
+    },
+    {
+      "epoch": 7.027449381709471,
+      "grad_norm": 8.375,
+      "learning_rate": 6.078441364315804e-06,
+      "loss": 0.8088,
+      "step": 103430
+    },
+    {
+      "epoch": 7.027789101780133,
+      "grad_norm": 8.5,
+      "learning_rate": 6.076318113874168e-06,
+      "loss": 0.8088,
+      "step": 103435
+    },
+    {
+      "epoch": 7.028128821850795,
+      "grad_norm": 7.625,
+      "learning_rate": 6.074194863432532e-06,
+      "loss": 0.8605,
+      "step": 103440
+    },
+    {
+      "epoch": 7.028468541921456,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.072071612990895e-06,
+      "loss": 0.9131,
+      "step": 103445
+    },
+    {
+      "epoch": 7.028808261992118,
+      "grad_norm": 8.0,
+      "learning_rate": 6.069948362549259e-06,
+      "loss": 0.9407,
+      "step": 103450
+    },
+    {
+      "epoch": 7.0291479820627805,
+      "grad_norm": 9.0625,
+      "learning_rate": 6.067825112107623e-06,
+      "loss": 0.8098,
+      "step": 103455
+    },
+    {
+      "epoch": 7.029487702133442,
+      "grad_norm": 7.0625,
+      "learning_rate": 6.065701861665987e-06,
+      "loss": 0.795,
+      "step": 103460
+    },
+    {
+      "epoch": 7.029827422204104,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.063578611224351e-06,
+      "loss": 0.9382,
+      "step": 103465
+    },
+    {
+      "epoch": 7.030167142274766,
+      "grad_norm": 7.78125,
+      "learning_rate": 6.061455360782715e-06,
+      "loss": 0.844,
+      "step": 103470
+    },
+    {
+      "epoch": 7.030506862345427,
+      "grad_norm": 7.28125,
+      "learning_rate": 6.059332110341079e-06,
+      "loss": 0.9008,
+      "step": 103475
+    },
+    {
+      "epoch": 7.030846582416089,
+      "grad_norm": 7.75,
+      "learning_rate": 6.0572088598994425e-06,
+      "loss": 0.8214,
+      "step": 103480
+    },
+    {
+      "epoch": 7.031186302486751,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.055085609457807e-06,
+      "loss": 0.8046,
+      "step": 103485
+    },
+    {
+      "epoch": 7.031526022557412,
+      "grad_norm": 8.25,
+      "learning_rate": 6.052962359016171e-06,
+      "loss": 0.8301,
+      "step": 103490
+    },
+    {
+      "epoch": 7.031865742628074,
+      "grad_norm": 8.5,
+      "learning_rate": 6.0508391085745345e-06,
+      "loss": 0.83,
+      "step": 103495
+    },
+    {
+      "epoch": 7.0322054626987365,
+      "grad_norm": 7.46875,
+      "learning_rate": 6.0487158581328985e-06,
+      "loss": 0.9048,
+      "step": 103500
+    },
+    {
+      "epoch": 7.032545182769398,
+      "grad_norm": 8.125,
+      "learning_rate": 6.0465926076912626e-06,
+      "loss": 0.8765,
+      "step": 103505
+    },
+    {
+      "epoch": 7.03288490284006,
+      "grad_norm": 9.0,
+      "learning_rate": 6.0444693572496266e-06,
+      "loss": 0.8149,
+      "step": 103510
+    },
+    {
+      "epoch": 7.033224622910722,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.0423461068079906e-06,
+      "loss": 0.8363,
+      "step": 103515
+    },
+    {
+      "epoch": 7.033564342981383,
+      "grad_norm": 7.875,
+      "learning_rate": 6.0402228563663546e-06,
+      "loss": 0.8707,
+      "step": 103520
+    },
+    {
+      "epoch": 7.033904063052045,
+      "grad_norm": 7.0625,
+      "learning_rate": 6.0380996059247186e-06,
+      "loss": 0.7311,
+      "step": 103525
+    },
+    {
+      "epoch": 7.034243783122707,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.035976355483082e-06,
+      "loss": 0.8313,
+      "step": 103530
+    },
+    {
+      "epoch": 7.034583503193368,
+      "grad_norm": 9.3125,
+      "learning_rate": 6.033853105041446e-06,
+      "loss": 0.956,
+      "step": 103535
+    },
+    {
+      "epoch": 7.03492322326403,
+      "grad_norm": 8.375,
+      "learning_rate": 6.031729854599811e-06,
+      "loss": 0.8408,
+      "step": 103540
+    },
+    {
+      "epoch": 7.0352629433346925,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.029606604158174e-06,
+      "loss": 0.8751,
+      "step": 103545
+    },
+    {
+      "epoch": 7.035602663405354,
+      "grad_norm": 8.0,
+      "learning_rate": 6.027483353716538e-06,
+      "loss": 0.8497,
+      "step": 103550
+    },
+    {
+      "epoch": 7.035942383476016,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.025360103274902e-06,
+      "loss": 0.9615,
+      "step": 103555
+    },
+    {
+      "epoch": 7.036282103546678,
+      "grad_norm": 8.625,
+      "learning_rate": 6.023236852833266e-06,
+      "loss": 0.8707,
+      "step": 103560
+    },
+    {
+      "epoch": 7.036621823617339,
+      "grad_norm": 7.625,
+      "learning_rate": 6.02111360239163e-06,
+      "loss": 0.8953,
+      "step": 103565
+    },
+    {
+      "epoch": 7.036961543688001,
+      "grad_norm": 8.625,
+      "learning_rate": 6.018990351949994e-06,
+      "loss": 0.8264,
+      "step": 103570
+    },
+    {
+      "epoch": 7.037301263758663,
+      "grad_norm": 8.625,
+      "learning_rate": 6.016867101508358e-06,
+      "loss": 0.8531,
+      "step": 103575
+    },
+    {
+      "epoch": 7.037640983829324,
+      "grad_norm": 8.5,
+      "learning_rate": 6.014743851066721e-06,
+      "loss": 0.8347,
+      "step": 103580
+    },
+    {
+      "epoch": 7.037980703899986,
+      "grad_norm": 7.46875,
+      "learning_rate": 6.012620600625085e-06,
+      "loss": 0.9234,
+      "step": 103585
+    },
+    {
+      "epoch": 7.0383204239706485,
+      "grad_norm": 6.8125,
+      "learning_rate": 6.010497350183449e-06,
+      "loss": 0.8882,
+      "step": 103590
+    },
+    {
+      "epoch": 7.03866014404131,
+      "grad_norm": 8.75,
+      "learning_rate": 6.008374099741814e-06,
+      "loss": 0.8569,
+      "step": 103595
+    },
+    {
+      "epoch": 7.038999864111972,
+      "grad_norm": 7.6875,
+      "learning_rate": 6.006250849300177e-06,
+      "loss": 0.7939,
+      "step": 103600
+    },
+    {
+      "epoch": 7.039339584182634,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.004127598858541e-06,
+      "loss": 0.9306,
+      "step": 103605
+    },
+    {
+      "epoch": 7.039679304253295,
+      "grad_norm": 7.90625,
+      "learning_rate": 6.002004348416905e-06,
+      "loss": 0.9182,
+      "step": 103610
+    },
+    {
+      "epoch": 7.040019024323957,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.999881097975268e-06,
+      "loss": 0.8177,
+      "step": 103615
+    },
+    {
+      "epoch": 7.040358744394619,
+      "grad_norm": 8.25,
+      "learning_rate": 5.997757847533633e-06,
+      "loss": 0.8777,
+      "step": 103620
+    },
+    {
+      "epoch": 7.04069846446528,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.995634597091997e-06,
+      "loss": 0.837,
+      "step": 103625
+    },
+    {
+      "epoch": 7.041038184535942,
+      "grad_norm": 8.125,
+      "learning_rate": 5.99351134665036e-06,
+      "loss": 0.7999,
+      "step": 103630
+    },
+    {
+      "epoch": 7.0413779046066045,
+      "grad_norm": 8.375,
+      "learning_rate": 5.991388096208724e-06,
+      "loss": 0.907,
+      "step": 103635
+    },
+    {
+      "epoch": 7.041717624677266,
+      "grad_norm": 7.375,
+      "learning_rate": 5.989264845767088e-06,
+      "loss": 0.8742,
+      "step": 103640
+    },
+    {
+      "epoch": 7.042057344747928,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.987141595325452e-06,
+      "loss": 0.8364,
+      "step": 103645
+    },
+    {
+      "epoch": 7.04239706481859,
+      "grad_norm": 8.875,
+      "learning_rate": 5.985018344883816e-06,
+      "loss": 0.8588,
+      "step": 103650
+    },
+    {
+      "epoch": 7.042736784889251,
+      "grad_norm": 7.375,
+      "learning_rate": 5.98289509444218e-06,
+      "loss": 0.8327,
+      "step": 103655
+    },
+    {
+      "epoch": 7.043076504959913,
+      "grad_norm": 7.34375,
+      "learning_rate": 5.980771844000544e-06,
+      "loss": 0.9229,
+      "step": 103660
+    },
+    {
+      "epoch": 7.043416225030575,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.978648593558907e-06,
+      "loss": 0.8916,
+      "step": 103665
+    },
+    {
+      "epoch": 7.043755945101236,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.976525343117271e-06,
+      "loss": 0.8307,
+      "step": 103670
+    },
+    {
+      "epoch": 7.0440956651718984,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.974402092675635e-06,
+      "loss": 0.8645,
+      "step": 103675
+    },
+    {
+      "epoch": 7.0444353852425605,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.972278842234e-06,
+      "loss": 0.8916,
+      "step": 103680
+    },
+    {
+      "epoch": 7.044775105313222,
+      "grad_norm": 7.375,
+      "learning_rate": 5.970155591792363e-06,
+      "loss": 0.8307,
+      "step": 103685
+    },
+    {
+      "epoch": 7.045114825383884,
+      "grad_norm": 7.09375,
+      "learning_rate": 5.968032341350727e-06,
+      "loss": 0.693,
+      "step": 103690
+    },
+    {
+      "epoch": 7.045454545454546,
+      "grad_norm": 7.0625,
+      "learning_rate": 5.965909090909091e-06,
+      "loss": 0.8764,
+      "step": 103695
+    },
+    {
+      "epoch": 7.045794265525207,
+      "grad_norm": 8.5,
+      "learning_rate": 5.9637858404674545e-06,
+      "loss": 0.8174,
+      "step": 103700
+    },
+    {
+      "epoch": 7.046133985595869,
+      "grad_norm": 9.125,
+      "learning_rate": 5.961662590025819e-06,
+      "loss": 0.8298,
+      "step": 103705
+    },
+    {
+      "epoch": 7.046473705666531,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.959539339584183e-06,
+      "loss": 0.8605,
+      "step": 103710
+    },
+    {
+      "epoch": 7.046813425737192,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.9574160891425465e-06,
+      "loss": 0.8163,
+      "step": 103715
+    },
+    {
+      "epoch": 7.0471531458078545,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.9552928387009105e-06,
+      "loss": 0.8287,
+      "step": 103720
+    },
+    {
+      "epoch": 7.0474928658785165,
+      "grad_norm": 8.125,
+      "learning_rate": 5.9531695882592745e-06,
+      "loss": 0.856,
+      "step": 103725
+    },
+    {
+      "epoch": 7.047832585949178,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.9510463378176385e-06,
+      "loss": 0.7891,
+      "step": 103730
+    },
+    {
+      "epoch": 7.04817230601984,
+      "grad_norm": 7.03125,
+      "learning_rate": 5.9489230873760025e-06,
+      "loss": 0.7542,
+      "step": 103735
+    },
+    {
+      "epoch": 7.048512026090501,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.9467998369343665e-06,
+      "loss": 0.7859,
+      "step": 103740
+    },
+    {
+      "epoch": 7.048851746161163,
+      "grad_norm": 8.625,
+      "learning_rate": 5.9446765864927305e-06,
+      "loss": 0.863,
+      "step": 103745
+    },
+    {
+      "epoch": 7.049191466231825,
+      "grad_norm": 6.90625,
+      "learning_rate": 5.942553336051094e-06,
+      "loss": 0.8501,
+      "step": 103750
+    },
+    {
+      "epoch": 7.049531186302486,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.940430085609458e-06,
+      "loss": 0.8634,
+      "step": 103755
+    },
+    {
+      "epoch": 7.049870906373148,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.9383068351678225e-06,
+      "loss": 0.9015,
+      "step": 103760
+    },
+    {
+      "epoch": 7.0502106264438105,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.9361835847261865e-06,
+      "loss": 0.7481,
+      "step": 103765
+    },
+    {
+      "epoch": 7.050550346514472,
+      "grad_norm": 7.125,
+      "learning_rate": 5.93406033428455e-06,
+      "loss": 0.8499,
+      "step": 103770
+    },
+    {
+      "epoch": 7.050890066585134,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.931937083842914e-06,
+      "loss": 0.8467,
+      "step": 103775
+    },
+    {
+      "epoch": 7.051229786655796,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.929813833401278e-06,
+      "loss": 0.8992,
+      "step": 103780
+    },
+    {
+      "epoch": 7.051569506726457,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.927690582959642e-06,
+      "loss": 0.8573,
+      "step": 103785
+    },
+    {
+      "epoch": 7.051909226797119,
+      "grad_norm": 7.25,
+      "learning_rate": 5.925567332518006e-06,
+      "loss": 0.8754,
+      "step": 103790
+    },
+    {
+      "epoch": 7.052248946867781,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.92344408207637e-06,
+      "loss": 0.8889,
+      "step": 103795
+    },
+    {
+      "epoch": 7.052588666938442,
+      "grad_norm": 8.375,
+      "learning_rate": 5.921320831634733e-06,
+      "loss": 0.7447,
+      "step": 103800
+    },
+    {
+      "epoch": 7.052928387009104,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.919197581193097e-06,
+      "loss": 0.8174,
+      "step": 103805
+    },
+    {
+      "epoch": 7.0532681070797665,
+      "grad_norm": 6.96875,
+      "learning_rate": 5.917074330751461e-06,
+      "loss": 0.806,
+      "step": 103810
+    },
+    {
+      "epoch": 7.053607827150428,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.914951080309825e-06,
+      "loss": 0.85,
+      "step": 103815
+    },
+    {
+      "epoch": 7.05394754722109,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.912827829868189e-06,
+      "loss": 0.7765,
+      "step": 103820
+    },
+    {
+      "epoch": 7.054287267291752,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.910704579426553e-06,
+      "loss": 0.8995,
+      "step": 103825
+    },
+    {
+      "epoch": 7.054626987362413,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.908581328984917e-06,
+      "loss": 0.8403,
+      "step": 103830
+    },
+    {
+      "epoch": 7.054966707433075,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.90645807854328e-06,
+      "loss": 0.8054,
+      "step": 103835
+    },
+    {
+      "epoch": 7.055306427503737,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.904334828101644e-06,
+      "loss": 0.8627,
+      "step": 103840
+    },
+    {
+      "epoch": 7.055646147574398,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.902211577660009e-06,
+      "loss": 0.887,
+      "step": 103845
+    },
+    {
+      "epoch": 7.05598586764506,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.900088327218373e-06,
+      "loss": 0.865,
+      "step": 103850
+    },
+    {
+      "epoch": 7.0563255877157225,
+      "grad_norm": 8.375,
+      "learning_rate": 5.897965076776736e-06,
+      "loss": 0.8613,
+      "step": 103855
+    },
+    {
+      "epoch": 7.056665307786384,
+      "grad_norm": 8.0,
+      "learning_rate": 5.8958418263351e-06,
+      "loss": 0.8875,
+      "step": 103860
+    },
+    {
+      "epoch": 7.057005027857046,
+      "grad_norm": 7.0,
+      "learning_rate": 5.893718575893464e-06,
+      "loss": 0.8088,
+      "step": 103865
+    },
+    {
+      "epoch": 7.057344747927708,
+      "grad_norm": 8.125,
+      "learning_rate": 5.891595325451828e-06,
+      "loss": 0.8014,
+      "step": 103870
+    },
+    {
+      "epoch": 7.057684467998369,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.889472075010192e-06,
+      "loss": 0.833,
+      "step": 103875
+    },
+    {
+      "epoch": 7.058024188069031,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.887348824568556e-06,
+      "loss": 0.7473,
+      "step": 103880
+    },
+    {
+      "epoch": 7.058363908139693,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.885225574126919e-06,
+      "loss": 0.8642,
+      "step": 103885
+    },
+    {
+      "epoch": 7.058703628210354,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.883102323685283e-06,
+      "loss": 0.8041,
+      "step": 103890
+    },
+    {
+      "epoch": 7.059043348281016,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.880979073243647e-06,
+      "loss": 0.9113,
+      "step": 103895
+    },
+    {
+      "epoch": 7.0593830683516785,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.878855822802012e-06,
+      "loss": 0.8397,
+      "step": 103900
+    },
+    {
+      "epoch": 7.05972278842234,
+      "grad_norm": 7.21875,
+      "learning_rate": 5.876732572360375e-06,
+      "loss": 0.9283,
+      "step": 103905
+    },
+    {
+      "epoch": 7.060062508493002,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.874609321918739e-06,
+      "loss": 0.8275,
+      "step": 103910
+    },
+    {
+      "epoch": 7.060402228563664,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.872486071477103e-06,
+      "loss": 0.8352,
+      "step": 103915
+    },
+    {
+      "epoch": 7.060741948634325,
+      "grad_norm": 8.375,
+      "learning_rate": 5.870362821035466e-06,
+      "loss": 0.8824,
+      "step": 103920
+    },
+    {
+      "epoch": 7.061081668704987,
+      "grad_norm": 7.4375,
+      "learning_rate": 5.868239570593831e-06,
+      "loss": 0.8057,
+      "step": 103925
+    },
+    {
+      "epoch": 7.061421388775649,
+      "grad_norm": 8.25,
+      "learning_rate": 5.866116320152195e-06,
+      "loss": 0.7965,
+      "step": 103930
+    },
+    {
+      "epoch": 7.06176110884631,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.863993069710559e-06,
+      "loss": 0.9646,
+      "step": 103935
+    },
+    {
+      "epoch": 7.062100828916972,
+      "grad_norm": 7.375,
+      "learning_rate": 5.861869819268922e-06,
+      "loss": 0.8139,
+      "step": 103940
+    },
+    {
+      "epoch": 7.0624405489876345,
+      "grad_norm": 6.53125,
+      "learning_rate": 5.859746568827286e-06,
+      "loss": 0.84,
+      "step": 103945
+    },
+    {
+      "epoch": 7.062780269058296,
+      "grad_norm": 8.75,
+      "learning_rate": 5.8576233183856504e-06,
+      "loss": 0.7842,
+      "step": 103950
+    },
+    {
+      "epoch": 7.063119989128958,
+      "grad_norm": 7.15625,
+      "learning_rate": 5.8555000679440144e-06,
+      "loss": 0.9457,
+      "step": 103955
+    },
+    {
+      "epoch": 7.06345970919962,
+      "grad_norm": 7.28125,
+      "learning_rate": 5.8533768175023784e-06,
+      "loss": 0.8311,
+      "step": 103960
+    },
+    {
+      "epoch": 7.063799429270281,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.8512535670607424e-06,
+      "loss": 0.9193,
+      "step": 103965
+    },
+    {
+      "epoch": 7.064139149340943,
+      "grad_norm": 8.375,
+      "learning_rate": 5.849130316619106e-06,
+      "loss": 0.9015,
+      "step": 103970
+    },
+    {
+      "epoch": 7.064478869411605,
+      "grad_norm": 7.28125,
+      "learning_rate": 5.84700706617747e-06,
+      "loss": 0.8502,
+      "step": 103975
+    },
+    {
+      "epoch": 7.064818589482266,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.8448838157358344e-06,
+      "loss": 0.8348,
+      "step": 103980
+    },
+    {
+      "epoch": 7.0651583095529285,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.8427605652941985e-06,
+      "loss": 0.8214,
+      "step": 103985
+    },
+    {
+      "epoch": 7.0654980296235905,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.840637314852562e-06,
+      "loss": 0.8037,
+      "step": 103990
+    },
+    {
+      "epoch": 7.065837749694252,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.838514064410926e-06,
+      "loss": 0.7983,
+      "step": 103995
+    },
+    {
+      "epoch": 7.066177469764914,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.83639081396929e-06,
+      "loss": 0.8929,
+      "step": 104000
+    },
+    {
+      "epoch": 7.066517189835576,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.834267563527654e-06,
+      "loss": 0.8662,
+      "step": 104005
+    },
+    {
+      "epoch": 7.066856909906237,
+      "grad_norm": 6.625,
+      "learning_rate": 5.832144313086018e-06,
+      "loss": 0.9218,
+      "step": 104010
+    },
+    {
+      "epoch": 7.067196629976899,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.830021062644382e-06,
+      "loss": 0.8927,
+      "step": 104015
+    },
+    {
+      "epoch": 7.067536350047561,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.827897812202746e-06,
+      "loss": 0.8186,
+      "step": 104020
+    },
+    {
+      "epoch": 7.067876070118222,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.825774561761109e-06,
+      "loss": 0.8381,
+      "step": 104025
+    },
+    {
+      "epoch": 7.0682157901888845,
+      "grad_norm": 8.625,
+      "learning_rate": 5.823651311319473e-06,
+      "loss": 0.7917,
+      "step": 104030
+    },
+    {
+      "epoch": 7.0685555102595465,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.821528060877837e-06,
+      "loss": 0.8381,
+      "step": 104035
+    },
+    {
+      "epoch": 7.068895230330208,
+      "grad_norm": 7.21875,
+      "learning_rate": 5.819404810436201e-06,
+      "loss": 0.929,
+      "step": 104040
+    },
+    {
+      "epoch": 7.06923495040087,
+      "grad_norm": 8.25,
+      "learning_rate": 5.817281559994565e-06,
+      "loss": 0.7786,
+      "step": 104045
+    },
+    {
+      "epoch": 7.069574670471532,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.815158309552929e-06,
+      "loss": 0.8511,
+      "step": 104050
+    },
+    {
+      "epoch": 7.069914390542193,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.813035059111292e-06,
+      "loss": 0.893,
+      "step": 104055
+    },
+    {
+      "epoch": 7.070254110612855,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.810911808669656e-06,
+      "loss": 0.9496,
+      "step": 104060
+    },
+    {
+      "epoch": 7.070593830683517,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.808788558228021e-06,
+      "loss": 0.761,
+      "step": 104065
+    },
+    {
+      "epoch": 7.070933550754178,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.806665307786385e-06,
+      "loss": 0.8659,
+      "step": 104070
+    },
+    {
+      "epoch": 7.0712732708248405,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.804542057344748e-06,
+      "loss": 0.8454,
+      "step": 104075
+    },
+    {
+      "epoch": 7.0716129908955025,
+      "grad_norm": 7.28125,
+      "learning_rate": 5.802418806903112e-06,
+      "loss": 0.8705,
+      "step": 104080
+    },
+    {
+      "epoch": 7.071952710966164,
+      "grad_norm": 8.75,
+      "learning_rate": 5.800295556461476e-06,
+      "loss": 0.9148,
+      "step": 104085
+    },
+    {
+      "epoch": 7.072292431036826,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.79817230601984e-06,
+      "loss": 0.9265,
+      "step": 104090
+    },
+    {
+      "epoch": 7.072632151107487,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.796049055578204e-06,
+      "loss": 0.8909,
+      "step": 104095
+    },
+    {
+      "epoch": 7.072971871178149,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.793925805136568e-06,
+      "loss": 0.7966,
+      "step": 104100
+    },
+    {
+      "epoch": 7.073311591248811,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.791802554694932e-06,
+      "loss": 0.8174,
+      "step": 104105
+    },
+    {
+      "epoch": 7.073651311319472,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.789679304253295e-06,
+      "loss": 0.8956,
+      "step": 104110
+    },
+    {
+      "epoch": 7.073991031390134,
+      "grad_norm": 8.375,
+      "learning_rate": 5.787556053811659e-06,
+      "loss": 0.9588,
+      "step": 104115
+    },
+    {
+      "epoch": 7.0743307514607965,
+      "grad_norm": 8.125,
+      "learning_rate": 5.785432803370024e-06,
+      "loss": 0.779,
+      "step": 104120
+    },
+    {
+      "epoch": 7.074670471531458,
+      "grad_norm": 6.9375,
+      "learning_rate": 5.783309552928387e-06,
+      "loss": 0.765,
+      "step": 104125
+    },
+    {
+      "epoch": 7.07501019160212,
+      "grad_norm": 8.5,
+      "learning_rate": 5.781186302486751e-06,
+      "loss": 0.8156,
+      "step": 104130
+    },
+    {
+      "epoch": 7.075349911672782,
+      "grad_norm": 8.125,
+      "learning_rate": 5.779063052045115e-06,
+      "loss": 0.8182,
+      "step": 104135
+    },
+    {
+      "epoch": 7.075689631743443,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.776939801603478e-06,
+      "loss": 0.9201,
+      "step": 104140
+    },
+    {
+      "epoch": 7.076029351814105,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.774816551161843e-06,
+      "loss": 0.8541,
+      "step": 104145
+    },
+    {
+      "epoch": 7.076369071884767,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.772693300720207e-06,
+      "loss": 0.8103,
+      "step": 104150
+    },
+    {
+      "epoch": 7.076708791955428,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.770570050278571e-06,
+      "loss": 0.8968,
+      "step": 104155
+    },
+    {
+      "epoch": 7.07704851202609,
+      "grad_norm": 9.625,
+      "learning_rate": 5.768446799836934e-06,
+      "loss": 0.9412,
+      "step": 104160
+    },
+    {
+      "epoch": 7.0773882320967525,
+      "grad_norm": 7.09375,
+      "learning_rate": 5.766323549395298e-06,
+      "loss": 0.8967,
+      "step": 104165
+    },
+    {
+      "epoch": 7.077727952167414,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.764200298953662e-06,
+      "loss": 0.8778,
+      "step": 104170
+    },
+    {
+      "epoch": 7.078067672238076,
+      "grad_norm": 9.0,
+      "learning_rate": 5.762077048512026e-06,
+      "loss": 0.914,
+      "step": 104175
+    },
+    {
+      "epoch": 7.078407392308738,
+      "grad_norm": 7.75,
+      "learning_rate": 5.75995379807039e-06,
+      "loss": 0.9253,
+      "step": 104180
+    },
+    {
+      "epoch": 7.078747112379399,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.757830547628754e-06,
+      "loss": 0.7826,
+      "step": 104185
+    },
+    {
+      "epoch": 7.079086832450061,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.755707297187118e-06,
+      "loss": 0.8489,
+      "step": 104190
+    },
+    {
+      "epoch": 7.079426552520723,
+      "grad_norm": 6.75,
+      "learning_rate": 5.7535840467454815e-06,
+      "loss": 0.8868,
+      "step": 104195
+    },
+    {
+      "epoch": 7.079766272591384,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.751460796303846e-06,
+      "loss": 0.8802,
+      "step": 104200
+    },
+    {
+      "epoch": 7.080105992662046,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.74933754586221e-06,
+      "loss": 0.8415,
+      "step": 104205
+    },
+    {
+      "epoch": 7.0804457127327085,
+      "grad_norm": 9.1875,
+      "learning_rate": 5.7472142954205735e-06,
+      "loss": 0.8279,
+      "step": 104210
+    },
+    {
+      "epoch": 7.08078543280337,
+      "grad_norm": 8.0,
+      "learning_rate": 5.7450910449789375e-06,
+      "loss": 0.9302,
+      "step": 104215
+    },
+    {
+      "epoch": 7.081125152874032,
+      "grad_norm": 7.75,
+      "learning_rate": 5.7429677945373016e-06,
+      "loss": 0.779,
+      "step": 104220
+    },
+    {
+      "epoch": 7.081464872944694,
+      "grad_norm": 8.375,
+      "learning_rate": 5.7408445440956656e-06,
+      "loss": 0.8083,
+      "step": 104225
+    },
+    {
+      "epoch": 7.081804593015355,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.7387212936540296e-06,
+      "loss": 0.8169,
+      "step": 104230
+    },
+    {
+      "epoch": 7.082144313086017,
+      "grad_norm": 8.5,
+      "learning_rate": 5.7365980432123936e-06,
+      "loss": 0.7908,
+      "step": 104235
+    },
+    {
+      "epoch": 7.082484033156679,
+      "grad_norm": 5.90625,
+      "learning_rate": 5.7344747927707576e-06,
+      "loss": 1.0885,
+      "step": 104240
+    },
+    {
+      "epoch": 7.08282375322734,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.732351542329121e-06,
+      "loss": 0.8673,
+      "step": 104245
+    },
+    {
+      "epoch": 7.083163473298002,
+      "grad_norm": 8.9375,
+      "learning_rate": 5.730228291887485e-06,
+      "loss": 0.8695,
+      "step": 104250
+    },
+    {
+      "epoch": 7.0835031933686645,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.728105041445849e-06,
+      "loss": 0.8564,
+      "step": 104255
+    },
+    {
+      "epoch": 7.083842913439326,
+      "grad_norm": 8.625,
+      "learning_rate": 5.725981791004213e-06,
+      "loss": 0.8442,
+      "step": 104260
+    },
+    {
+      "epoch": 7.084182633509988,
+      "grad_norm": 6.5,
+      "learning_rate": 5.723858540562577e-06,
+      "loss": 0.7878,
+      "step": 104265
+    },
+    {
+      "epoch": 7.08452235358065,
+      "grad_norm": 8.5,
+      "learning_rate": 5.721735290120941e-06,
+      "loss": 0.7969,
+      "step": 104270
+    },
+    {
+      "epoch": 7.084862073651311,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.719612039679305e-06,
+      "loss": 0.858,
+      "step": 104275
+    },
+    {
+      "epoch": 7.085201793721973,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.717488789237668e-06,
+      "loss": 0.8287,
+      "step": 104280
+    },
+    {
+      "epoch": 7.085541513792635,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.715365538796033e-06,
+      "loss": 0.8507,
+      "step": 104285
+    },
+    {
+      "epoch": 7.085881233863296,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.713242288354397e-06,
+      "loss": 0.9888,
+      "step": 104290
+    },
+    {
+      "epoch": 7.0862209539339585,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.71111903791276e-06,
+      "loss": 0.8331,
+      "step": 104295
+    },
+    {
+      "epoch": 7.0865606740046205,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.708995787471124e-06,
+      "loss": 0.7458,
+      "step": 104300
+    },
+    {
+      "epoch": 7.086900394075282,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.706872537029488e-06,
+      "loss": 0.809,
+      "step": 104305
+    },
+    {
+      "epoch": 7.087240114145944,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.704749286587852e-06,
+      "loss": 0.775,
+      "step": 104310
+    },
+    {
+      "epoch": 7.087579834216606,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.702626036146216e-06,
+      "loss": 0.8591,
+      "step": 104315
+    },
+    {
+      "epoch": 7.087919554287267,
+      "grad_norm": 7.75,
+      "learning_rate": 5.70050278570458e-06,
+      "loss": 0.8734,
+      "step": 104320
+    },
+    {
+      "epoch": 7.088259274357929,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.698379535262944e-06,
+      "loss": 0.8894,
+      "step": 104325
+    },
+    {
+      "epoch": 7.088598994428591,
+      "grad_norm": 7.375,
+      "learning_rate": 5.696256284821307e-06,
+      "loss": 0.7882,
+      "step": 104330
+    },
+    {
+      "epoch": 7.088938714499252,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.694133034379671e-06,
+      "loss": 0.9733,
+      "step": 104335
+    },
+    {
+      "epoch": 7.0892784345699145,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.692009783938036e-06,
+      "loss": 0.9181,
+      "step": 104340
+    },
+    {
+      "epoch": 7.0896181546405765,
+      "grad_norm": 8.9375,
+      "learning_rate": 5.689886533496399e-06,
+      "loss": 0.8163,
+      "step": 104345
+    },
+    {
+      "epoch": 7.089957874711238,
+      "grad_norm": 8.875,
+      "learning_rate": 5.687763283054763e-06,
+      "loss": 0.878,
+      "step": 104350
+    },
+    {
+      "epoch": 7.0902975947819,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.685640032613127e-06,
+      "loss": 0.889,
+      "step": 104355
+    },
+    {
+      "epoch": 7.090637314852562,
+      "grad_norm": 7.875,
+      "learning_rate": 5.683516782171491e-06,
+      "loss": 0.8018,
+      "step": 104360
+    },
+    {
+      "epoch": 7.090977034923223,
+      "grad_norm": 7.625,
+      "learning_rate": 5.681393531729855e-06,
+      "loss": 0.8261,
+      "step": 104365
+    },
+    {
+      "epoch": 7.091316754993885,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.679270281288219e-06,
+      "loss": 0.8378,
+      "step": 104370
+    },
+    {
+      "epoch": 7.091656475064547,
+      "grad_norm": 7.875,
+      "learning_rate": 5.677147030846583e-06,
+      "loss": 0.8741,
+      "step": 104375
+    },
+    {
+      "epoch": 7.091996195135208,
+      "grad_norm": 8.5,
+      "learning_rate": 5.675023780404946e-06,
+      "loss": 0.8819,
+      "step": 104380
+    },
+    {
+      "epoch": 7.0923359152058705,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.67290052996331e-06,
+      "loss": 0.8347,
+      "step": 104385
+    },
+    {
+      "epoch": 7.0926756352765326,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.670777279521674e-06,
+      "loss": 0.886,
+      "step": 104390
+    },
+    {
+      "epoch": 7.093015355347194,
+      "grad_norm": 8.75,
+      "learning_rate": 5.668654029080038e-06,
+      "loss": 0.9011,
+      "step": 104395
+    },
+    {
+      "epoch": 7.093355075417856,
+      "grad_norm": 6.65625,
+      "learning_rate": 5.666530778638402e-06,
+      "loss": 0.8439,
+      "step": 104400
+    },
+    {
+      "epoch": 7.093694795488518,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.664407528196766e-06,
+      "loss": 0.7803,
+      "step": 104405
+    },
+    {
+      "epoch": 7.094034515559179,
+      "grad_norm": 8.25,
+      "learning_rate": 5.66228427775513e-06,
+      "loss": 0.8955,
+      "step": 104410
+    },
+    {
+      "epoch": 7.094374235629841,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.6601610273134935e-06,
+      "loss": 0.8275,
+      "step": 104415
+    },
+    {
+      "epoch": 7.094713955700502,
+      "grad_norm": 8.625,
+      "learning_rate": 5.6580377768718575e-06,
+      "loss": 0.8282,
+      "step": 104420
+    },
+    {
+      "epoch": 7.095053675771164,
+      "grad_norm": 7.625,
+      "learning_rate": 5.655914526430222e-06,
+      "loss": 0.8314,
+      "step": 104425
+    },
+    {
+      "epoch": 7.0953933958418265,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.6537912759885855e-06,
+      "loss": 0.8522,
+      "step": 104430
+    },
+    {
+      "epoch": 7.095733115912488,
+      "grad_norm": 7.75,
+      "learning_rate": 5.6516680255469495e-06,
+      "loss": 0.8644,
+      "step": 104435
+    },
+    {
+      "epoch": 7.09607283598315,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.6495447751053135e-06,
+      "loss": 0.8659,
+      "step": 104440
+    },
+    {
+      "epoch": 7.096412556053812,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.6474215246636775e-06,
+      "loss": 0.8373,
+      "step": 104445
+    },
+    {
+      "epoch": 7.096752276124473,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.6452982742220415e-06,
+      "loss": 0.8275,
+      "step": 104450
+    },
+    {
+      "epoch": 7.097091996195135,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.6431750237804055e-06,
+      "loss": 0.7717,
+      "step": 104455
+    },
+    {
+      "epoch": 7.097431716265797,
+      "grad_norm": 8.25,
+      "learning_rate": 5.6410517733387695e-06,
+      "loss": 0.7539,
+      "step": 104460
+    },
+    {
+      "epoch": 7.097771436336458,
+      "grad_norm": 8.5,
+      "learning_rate": 5.638928522897133e-06,
+      "loss": 0.9164,
+      "step": 104465
+    },
+    {
+      "epoch": 7.09811115640712,
+      "grad_norm": 7.625,
+      "learning_rate": 5.636805272455497e-06,
+      "loss": 0.7977,
+      "step": 104470
+    },
+    {
+      "epoch": 7.0984508764777825,
+      "grad_norm": 8.9375,
+      "learning_rate": 5.634682022013861e-06,
+      "loss": 0.8119,
+      "step": 104475
+    },
+    {
+      "epoch": 7.098790596548444,
+      "grad_norm": 6.875,
+      "learning_rate": 5.632558771572225e-06,
+      "loss": 0.8403,
+      "step": 104480
+    },
+    {
+      "epoch": 7.099130316619106,
+      "grad_norm": 8.625,
+      "learning_rate": 5.630435521130589e-06,
+      "loss": 0.8818,
+      "step": 104485
+    },
+    {
+      "epoch": 7.099470036689768,
+      "grad_norm": 8.125,
+      "learning_rate": 5.628312270688953e-06,
+      "loss": 0.9275,
+      "step": 104490
+    },
+    {
+      "epoch": 7.099809756760429,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.626189020247317e-06,
+      "loss": 0.8334,
+      "step": 104495
+    },
+    {
+      "epoch": 7.100149476831091,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.62406576980568e-06,
+      "loss": 0.8199,
+      "step": 104500
+    },
+    {
+      "epoch": 7.100489196901753,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.621942519364045e-06,
+      "loss": 0.9264,
+      "step": 104505
+    },
+    {
+      "epoch": 7.100828916972414,
+      "grad_norm": 8.5,
+      "learning_rate": 5.619819268922409e-06,
+      "loss": 0.8438,
+      "step": 104510
+    },
+    {
+      "epoch": 7.101168637043076,
+      "grad_norm": 8.5,
+      "learning_rate": 5.617696018480772e-06,
+      "loss": 0.9565,
+      "step": 104515
+    },
+    {
+      "epoch": 7.1015083571137385,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.615572768039136e-06,
+      "loss": 0.822,
+      "step": 104520
+    },
+    {
+      "epoch": 7.1018480771844,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.6134495175975e-06,
+      "loss": 0.7897,
+      "step": 104525
+    },
+    {
+      "epoch": 7.102187797255062,
+      "grad_norm": 7.21875,
+      "learning_rate": 5.611326267155864e-06,
+      "loss": 0.9093,
+      "step": 104530
+    },
+    {
+      "epoch": 7.102527517325724,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.609203016714228e-06,
+      "loss": 0.8485,
+      "step": 104535
+    },
+    {
+      "epoch": 7.102867237396385,
+      "grad_norm": 7.75,
+      "learning_rate": 5.607079766272592e-06,
+      "loss": 0.7977,
+      "step": 104540
+    },
+    {
+      "epoch": 7.103206957467047,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.604956515830956e-06,
+      "loss": 0.9166,
+      "step": 104545
+    },
+    {
+      "epoch": 7.103546677537709,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.602833265389319e-06,
+      "loss": 0.8471,
+      "step": 104550
+    },
+    {
+      "epoch": 7.10388639760837,
+      "grad_norm": 7.4375,
+      "learning_rate": 5.600710014947683e-06,
+      "loss": 0.8092,
+      "step": 104555
+    },
+    {
+      "epoch": 7.1042261176790324,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.598586764506048e-06,
+      "loss": 0.8101,
+      "step": 104560
+    },
+    {
+      "epoch": 7.1045658377496945,
+      "grad_norm": 8.0,
+      "learning_rate": 5.596463514064411e-06,
+      "loss": 0.8472,
+      "step": 104565
+    },
+    {
+      "epoch": 7.104905557820356,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.594340263622775e-06,
+      "loss": 0.7473,
+      "step": 104570
+    },
+    {
+      "epoch": 7.105245277891018,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.592217013181139e-06,
+      "loss": 0.8443,
+      "step": 104575
+    },
+    {
+      "epoch": 7.10558499796168,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.590093762739503e-06,
+      "loss": 0.7782,
+      "step": 104580
+    },
+    {
+      "epoch": 7.105924718032341,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.587970512297867e-06,
+      "loss": 0.8846,
+      "step": 104585
+    },
+    {
+      "epoch": 7.106264438103003,
+      "grad_norm": 8.375,
+      "learning_rate": 5.585847261856231e-06,
+      "loss": 0.8999,
+      "step": 104590
+    },
+    {
+      "epoch": 7.106604158173665,
+      "grad_norm": 8.375,
+      "learning_rate": 5.583724011414595e-06,
+      "loss": 0.7382,
+      "step": 104595
+    },
+    {
+      "epoch": 7.106943878244326,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.581600760972958e-06,
+      "loss": 0.842,
+      "step": 104600
+    },
+    {
+      "epoch": 7.1072835983149885,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.579477510531322e-06,
+      "loss": 0.8465,
+      "step": 104605
+    },
+    {
+      "epoch": 7.1076233183856505,
+      "grad_norm": 7.09375,
+      "learning_rate": 5.577354260089686e-06,
+      "loss": 0.7808,
+      "step": 104610
+    },
+    {
+      "epoch": 7.107963038456312,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.57523100964805e-06,
+      "loss": 0.8805,
+      "step": 104615
+    },
+    {
+      "epoch": 7.108302758526974,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.573107759206414e-06,
+      "loss": 0.814,
+      "step": 104620
+    },
+    {
+      "epoch": 7.108642478597636,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.570984508764778e-06,
+      "loss": 0.8146,
+      "step": 104625
+    },
+    {
+      "epoch": 7.108982198668297,
+      "grad_norm": 8.125,
+      "learning_rate": 5.568861258323142e-06,
+      "loss": 0.8323,
+      "step": 104630
+    },
+    {
+      "epoch": 7.109321918738959,
+      "grad_norm": 7.5,
+      "learning_rate": 5.566738007881505e-06,
+      "loss": 0.8226,
+      "step": 104635
+    },
+    {
+      "epoch": 7.109661638809621,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.564614757439869e-06,
+      "loss": 0.8352,
+      "step": 104640
+    },
+    {
+      "epoch": 7.110001358880282,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.562491506998234e-06,
+      "loss": 0.8169,
+      "step": 104645
+    },
+    {
+      "epoch": 7.1103410789509445,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.560368256556597e-06,
+      "loss": 0.8614,
+      "step": 104650
+    },
+    {
+      "epoch": 7.1106807990216065,
+      "grad_norm": 8.375,
+      "learning_rate": 5.558245006114961e-06,
+      "loss": 0.7383,
+      "step": 104655
+    },
+    {
+      "epoch": 7.111020519092268,
+      "grad_norm": 8.25,
+      "learning_rate": 5.556121755673325e-06,
+      "loss": 0.8762,
+      "step": 104660
+    },
+    {
+      "epoch": 7.11136023916293,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.5539985052316894e-06,
+      "loss": 0.6925,
+      "step": 104665
+    },
+    {
+      "epoch": 7.111699959233592,
+      "grad_norm": 8.25,
+      "learning_rate": 5.5518752547900534e-06,
+      "loss": 0.8166,
+      "step": 104670
+    },
+    {
+      "epoch": 7.112039679304253,
+      "grad_norm": 8.375,
+      "learning_rate": 5.5497520043484174e-06,
+      "loss": 0.7989,
+      "step": 104675
+    },
+    {
+      "epoch": 7.112379399374915,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.5476287539067814e-06,
+      "loss": 0.7911,
+      "step": 104680
+    },
+    {
+      "epoch": 7.112719119445577,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.545505503465145e-06,
+      "loss": 0.8977,
+      "step": 104685
+    },
+    {
+      "epoch": 7.113058839516238,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.543382253023509e-06,
+      "loss": 0.8681,
+      "step": 104690
+    },
+    {
+      "epoch": 7.1133985595869005,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.541259002581873e-06,
+      "loss": 0.7901,
+      "step": 104695
+    },
+    {
+      "epoch": 7.113738279657563,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.5391357521402375e-06,
+      "loss": 0.9657,
+      "step": 104700
+    },
+    {
+      "epoch": 7.114077999728224,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.537012501698601e-06,
+      "loss": 0.9018,
+      "step": 104705
+    },
+    {
+      "epoch": 7.114417719798886,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.534889251256965e-06,
+      "loss": 0.7958,
+      "step": 104710
+    },
+    {
+      "epoch": 7.114757439869548,
+      "grad_norm": 7.625,
+      "learning_rate": 5.532766000815329e-06,
+      "loss": 0.8439,
+      "step": 104715
+    },
+    {
+      "epoch": 7.115097159940209,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.530642750373692e-06,
+      "loss": 0.8639,
+      "step": 104720
+    },
+    {
+      "epoch": 7.115436880010871,
+      "grad_norm": 9.3125,
+      "learning_rate": 5.528519499932057e-06,
+      "loss": 0.8051,
+      "step": 104725
+    },
+    {
+      "epoch": 7.115776600081533,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.526396249490421e-06,
+      "loss": 0.9106,
+      "step": 104730
+    },
+    {
+      "epoch": 7.116116320152194,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.524272999048784e-06,
+      "loss": 0.9161,
+      "step": 104735
+    },
+    {
+      "epoch": 7.1164560402228565,
+      "grad_norm": 7.75,
+      "learning_rate": 5.522149748607148e-06,
+      "loss": 0.8436,
+      "step": 104740
+    },
+    {
+      "epoch": 7.116795760293519,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.520026498165512e-06,
+      "loss": 0.7973,
+      "step": 104745
+    },
+    {
+      "epoch": 7.11713548036418,
+      "grad_norm": 7.34375,
+      "learning_rate": 5.517903247723876e-06,
+      "loss": 0.7944,
+      "step": 104750
+    },
+    {
+      "epoch": 7.117475200434842,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.51577999728224e-06,
+      "loss": 0.8698,
+      "step": 104755
+    },
+    {
+      "epoch": 7.117814920505504,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.513656746840604e-06,
+      "loss": 0.7554,
+      "step": 104760
+    },
+    {
+      "epoch": 7.118154640576165,
+      "grad_norm": 7.875,
+      "learning_rate": 5.511533496398968e-06,
+      "loss": 0.8206,
+      "step": 104765
+    },
+    {
+      "epoch": 7.118494360646827,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.509410245957331e-06,
+      "loss": 0.9282,
+      "step": 104770
+    },
+    {
+      "epoch": 7.118834080717488,
+      "grad_norm": 8.75,
+      "learning_rate": 5.507286995515695e-06,
+      "loss": 0.7726,
+      "step": 104775
+    },
+    {
+      "epoch": 7.11917380078815,
+      "grad_norm": 7.15625,
+      "learning_rate": 5.505163745074059e-06,
+      "loss": 0.911,
+      "step": 104780
+    },
+    {
+      "epoch": 7.1195135208588125,
+      "grad_norm": 8.0,
+      "learning_rate": 5.503040494632424e-06,
+      "loss": 0.9003,
+      "step": 104785
+    },
+    {
+      "epoch": 7.119853240929474,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.500917244190787e-06,
+      "loss": 0.9553,
+      "step": 104790
+    },
+    {
+      "epoch": 7.120192961000136,
+      "grad_norm": 8.25,
+      "learning_rate": 5.498793993749151e-06,
+      "loss": 0.841,
+      "step": 104795
+    },
+    {
+      "epoch": 7.120532681070798,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.496670743307515e-06,
+      "loss": 0.8656,
+      "step": 104800
+    },
+    {
+      "epoch": 7.120872401141459,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.494547492865878e-06,
+      "loss": 0.8844,
+      "step": 104805
+    },
+    {
+      "epoch": 7.121212121212121,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.492424242424243e-06,
+      "loss": 0.7925,
+      "step": 104810
+    },
+    {
+      "epoch": 7.121551841282783,
+      "grad_norm": 9.125,
+      "learning_rate": 5.490300991982607e-06,
+      "loss": 0.9026,
+      "step": 104815
+    },
+    {
+      "epoch": 7.121891561353444,
+      "grad_norm": 8.25,
+      "learning_rate": 5.48817774154097e-06,
+      "loss": 1.0697,
+      "step": 104820
+    },
+    {
+      "epoch": 7.122231281424106,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.486054491099334e-06,
+      "loss": 0.8688,
+      "step": 104825
+    },
+    {
+      "epoch": 7.1225710014947685,
+      "grad_norm": 7.75,
+      "learning_rate": 5.483931240657698e-06,
+      "loss": 0.8407,
+      "step": 104830
+    },
+    {
+      "epoch": 7.12291072156543,
+      "grad_norm": 8.25,
+      "learning_rate": 5.481807990216062e-06,
+      "loss": 0.8987,
+      "step": 104835
+    },
+    {
+      "epoch": 7.123250441636092,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.479684739774426e-06,
+      "loss": 0.8114,
+      "step": 104840
+    },
+    {
+      "epoch": 7.123590161706754,
+      "grad_norm": 8.625,
+      "learning_rate": 5.47756148933279e-06,
+      "loss": 0.877,
+      "step": 104845
+    },
+    {
+      "epoch": 7.123929881777415,
+      "grad_norm": 9.625,
+      "learning_rate": 5.475438238891154e-06,
+      "loss": 0.9241,
+      "step": 104850
+    },
+    {
+      "epoch": 7.124269601848077,
+      "grad_norm": 6.875,
+      "learning_rate": 5.473314988449517e-06,
+      "loss": 0.8487,
+      "step": 104855
+    },
+    {
+      "epoch": 7.124609321918739,
+      "grad_norm": 6.65625,
+      "learning_rate": 5.471191738007881e-06,
+      "loss": 0.8326,
+      "step": 104860
+    },
+    {
+      "epoch": 7.1249490419894,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.469068487566246e-06,
+      "loss": 0.8593,
+      "step": 104865
+    },
+    {
+      "epoch": 7.1252887620600625,
+      "grad_norm": 8.125,
+      "learning_rate": 5.466945237124609e-06,
+      "loss": 0.841,
+      "step": 104870
+    },
+    {
+      "epoch": 7.1256284821307245,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.464821986682973e-06,
+      "loss": 0.9131,
+      "step": 104875
+    },
+    {
+      "epoch": 7.125968202201386,
+      "grad_norm": 7.15625,
+      "learning_rate": 5.462698736241337e-06,
+      "loss": 0.8612,
+      "step": 104880
+    },
+    {
+      "epoch": 7.126307922272048,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.460575485799701e-06,
+      "loss": 0.7921,
+      "step": 104885
+    },
+    {
+      "epoch": 7.12664764234271,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.458452235358065e-06,
+      "loss": 0.8462,
+      "step": 104890
+    },
+    {
+      "epoch": 7.126987362413371,
+      "grad_norm": 7.40625,
+      "learning_rate": 5.456328984916429e-06,
+      "loss": 0.7886,
+      "step": 104895
+    },
+    {
+      "epoch": 7.127327082484033,
+      "grad_norm": 8.5,
+      "learning_rate": 5.454205734474793e-06,
+      "loss": 0.9525,
+      "step": 104900
+    },
+    {
+      "epoch": 7.127666802554695,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.4520824840331565e-06,
+      "loss": 0.8517,
+      "step": 104905
+    },
+    {
+      "epoch": 7.128006522625356,
+      "grad_norm": 8.25,
+      "learning_rate": 5.4499592335915205e-06,
+      "loss": 0.8873,
+      "step": 104910
+    },
+    {
+      "epoch": 7.1283462426960185,
+      "grad_norm": 8.375,
+      "learning_rate": 5.4478359831498845e-06,
+      "loss": 0.8568,
+      "step": 104915
+    },
+    {
+      "epoch": 7.1286859627666805,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.445712732708249e-06,
+      "loss": 0.9299,
+      "step": 104920
+    },
+    {
+      "epoch": 7.129025682837342,
+      "grad_norm": 7.09375,
+      "learning_rate": 5.4435894822666125e-06,
+      "loss": 0.8784,
+      "step": 104925
+    },
+    {
+      "epoch": 7.129365402908004,
+      "grad_norm": 8.125,
+      "learning_rate": 5.4414662318249765e-06,
+      "loss": 0.8533,
+      "step": 104930
+    },
+    {
+      "epoch": 7.129705122978666,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.4393429813833406e-06,
+      "loss": 0.8425,
+      "step": 104935
+    },
+    {
+      "epoch": 7.130044843049327,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.437219730941704e-06,
+      "loss": 0.8455,
+      "step": 104940
+    },
+    {
+      "epoch": 7.130384563119989,
+      "grad_norm": 8.375,
+      "learning_rate": 5.4350964805000686e-06,
+      "loss": 0.9149,
+      "step": 104945
+    },
+    {
+      "epoch": 7.130724283190651,
+      "grad_norm": 8.125,
+      "learning_rate": 5.4329732300584326e-06,
+      "loss": 0.7783,
+      "step": 104950
+    },
+    {
+      "epoch": 7.131064003261312,
+      "grad_norm": 6.34375,
+      "learning_rate": 5.430849979616796e-06,
+      "loss": 0.8105,
+      "step": 104955
+    },
+    {
+      "epoch": 7.1314037233319745,
+      "grad_norm": 8.375,
+      "learning_rate": 5.42872672917516e-06,
+      "loss": 0.7448,
+      "step": 104960
+    },
+    {
+      "epoch": 7.1317434434026366,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.426603478733524e-06,
+      "loss": 0.7583,
+      "step": 104965
+    },
+    {
+      "epoch": 7.132083163473298,
+      "grad_norm": 8.125,
+      "learning_rate": 5.424480228291888e-06,
+      "loss": 0.8718,
+      "step": 104970
+    },
+    {
+      "epoch": 7.13242288354396,
+      "grad_norm": 8.375,
+      "learning_rate": 5.422356977850252e-06,
+      "loss": 0.9091,
+      "step": 104975
+    },
+    {
+      "epoch": 7.132762603614622,
+      "grad_norm": 8.125,
+      "learning_rate": 5.420233727408616e-06,
+      "loss": 0.8145,
+      "step": 104980
+    },
+    {
+      "epoch": 7.133102323685283,
+      "grad_norm": 8.0,
+      "learning_rate": 5.41811047696698e-06,
+      "loss": 0.9055,
+      "step": 104985
+    },
+    {
+      "epoch": 7.133442043755945,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.415987226525343e-06,
+      "loss": 0.7974,
+      "step": 104990
+    },
+    {
+      "epoch": 7.133781763826607,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.413863976083707e-06,
+      "loss": 0.8568,
+      "step": 104995
+    },
+    {
+      "epoch": 7.134121483897268,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.411740725642071e-06,
+      "loss": 0.7439,
+      "step": 105000
+    },
+    {
+      "epoch": 7.1344612039679305,
+      "grad_norm": 8.125,
+      "learning_rate": 5.409617475200436e-06,
+      "loss": 0.8692,
+      "step": 105005
+    },
+    {
+      "epoch": 7.134800924038593,
+      "grad_norm": 8.625,
+      "learning_rate": 5.407494224758799e-06,
+      "loss": 0.8136,
+      "step": 105010
+    },
+    {
+      "epoch": 7.135140644109254,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.405370974317163e-06,
+      "loss": 0.8053,
+      "step": 105015
+    },
+    {
+      "epoch": 7.135480364179916,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.403247723875527e-06,
+      "loss": 0.8086,
+      "step": 105020
+    },
+    {
+      "epoch": 7.135820084250578,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.40112447343389e-06,
+      "loss": 0.7984,
+      "step": 105025
+    },
+    {
+      "epoch": 7.136159804321239,
+      "grad_norm": 8.375,
+      "learning_rate": 5.399001222992255e-06,
+      "loss": 0.8787,
+      "step": 105030
+    },
+    {
+      "epoch": 7.136499524391901,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.396877972550619e-06,
+      "loss": 0.8065,
+      "step": 105035
+    },
+    {
+      "epoch": 7.136839244462563,
+      "grad_norm": 8.125,
+      "learning_rate": 5.394754722108982e-06,
+      "loss": 0.8366,
+      "step": 105040
+    },
+    {
+      "epoch": 7.137178964533224,
+      "grad_norm": 9.1875,
+      "learning_rate": 5.392631471667346e-06,
+      "loss": 0.8984,
+      "step": 105045
+    },
+    {
+      "epoch": 7.1375186846038865,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.39050822122571e-06,
+      "loss": 0.9088,
+      "step": 105050
+    },
+    {
+      "epoch": 7.137858404674549,
+      "grad_norm": 8.75,
+      "learning_rate": 5.388384970784074e-06,
+      "loss": 0.7747,
+      "step": 105055
+    },
+    {
+      "epoch": 7.13819812474521,
+      "grad_norm": 7.375,
+      "learning_rate": 5.386261720342438e-06,
+      "loss": 0.8122,
+      "step": 105060
+    },
+    {
+      "epoch": 7.138537844815872,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.384138469900802e-06,
+      "loss": 0.8143,
+      "step": 105065
+    },
+    {
+      "epoch": 7.138877564886534,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.382015219459166e-06,
+      "loss": 0.8038,
+      "step": 105070
+    },
+    {
+      "epoch": 7.139217284957195,
+      "grad_norm": 8.25,
+      "learning_rate": 5.379891969017529e-06,
+      "loss": 0.7953,
+      "step": 105075
+    },
+    {
+      "epoch": 7.139557005027857,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.377768718575893e-06,
+      "loss": 0.7994,
+      "step": 105080
+    },
+    {
+      "epoch": 7.139896725098519,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.375645468134258e-06,
+      "loss": 0.7782,
+      "step": 105085
+    },
+    {
+      "epoch": 7.14023644516918,
+      "grad_norm": 8.125,
+      "learning_rate": 5.373522217692622e-06,
+      "loss": 0.7951,
+      "step": 105090
+    },
+    {
+      "epoch": 7.1405761652398425,
+      "grad_norm": 8.25,
+      "learning_rate": 5.371398967250985e-06,
+      "loss": 0.8026,
+      "step": 105095
+    },
+    {
+      "epoch": 7.140915885310504,
+      "grad_norm": 8.375,
+      "learning_rate": 5.369275716809349e-06,
+      "loss": 0.8878,
+      "step": 105100
+    },
+    {
+      "epoch": 7.141255605381166,
+      "grad_norm": 9.125,
+      "learning_rate": 5.367152466367713e-06,
+      "loss": 0.8643,
+      "step": 105105
+    },
+    {
+      "epoch": 7.141595325451828,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.365029215926077e-06,
+      "loss": 0.868,
+      "step": 105110
+    },
+    {
+      "epoch": 7.141935045522489,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.362905965484441e-06,
+      "loss": 0.7881,
+      "step": 105115
+    },
+    {
+      "epoch": 7.142274765593151,
+      "grad_norm": 8.75,
+      "learning_rate": 5.360782715042805e-06,
+      "loss": 0.8045,
+      "step": 105120
+    },
+    {
+      "epoch": 7.142614485663813,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.3586594646011685e-06,
+      "loss": 0.9377,
+      "step": 105125
+    },
+    {
+      "epoch": 7.142954205734474,
+      "grad_norm": 6.375,
+      "learning_rate": 5.3565362141595325e-06,
+      "loss": 0.8189,
+      "step": 105130
+    },
+    {
+      "epoch": 7.1432939258051364,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.3544129637178965e-06,
+      "loss": 0.8294,
+      "step": 105135
+    },
+    {
+      "epoch": 7.1436336458757985,
+      "grad_norm": 8.375,
+      "learning_rate": 5.3522897132762605e-06,
+      "loss": 0.882,
+      "step": 105140
+    },
+    {
+      "epoch": 7.14397336594646,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.3501664628346245e-06,
+      "loss": 0.7882,
+      "step": 105145
+    },
+    {
+      "epoch": 7.144313086017122,
+      "grad_norm": 7.125,
+      "learning_rate": 5.3480432123929885e-06,
+      "loss": 0.8563,
+      "step": 105150
+    },
+    {
+      "epoch": 7.144652806087784,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.3459199619513525e-06,
+      "loss": 0.9094,
+      "step": 105155
+    },
+    {
+      "epoch": 7.144992526158445,
+      "grad_norm": 8.375,
+      "learning_rate": 5.343796711509716e-06,
+      "loss": 0.8399,
+      "step": 105160
+    },
+    {
+      "epoch": 7.145332246229107,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.34167346106808e-06,
+      "loss": 0.8962,
+      "step": 105165
+    },
+    {
+      "epoch": 7.145671966299769,
+      "grad_norm": 8.125,
+      "learning_rate": 5.3395502106264445e-06,
+      "loss": 0.8825,
+      "step": 105170
+    },
+    {
+      "epoch": 7.14601168637043,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.3374269601848085e-06,
+      "loss": 0.8931,
+      "step": 105175
+    },
+    {
+      "epoch": 7.1463514064410925,
+      "grad_norm": 8.125,
+      "learning_rate": 5.335303709743172e-06,
+      "loss": 0.7977,
+      "step": 105180
+    },
+    {
+      "epoch": 7.1466911265117545,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.333180459301536e-06,
+      "loss": 0.8453,
+      "step": 105185
+    },
+    {
+      "epoch": 7.147030846582416,
+      "grad_norm": 8.25,
+      "learning_rate": 5.3310572088599e-06,
+      "loss": 0.8095,
+      "step": 105190
+    },
+    {
+      "epoch": 7.147370566653078,
+      "grad_norm": 8.0,
+      "learning_rate": 5.328933958418264e-06,
+      "loss": 0.8372,
+      "step": 105195
+    },
+    {
+      "epoch": 7.14771028672374,
+      "grad_norm": 8.5,
+      "learning_rate": 5.326810707976628e-06,
+      "loss": 0.8133,
+      "step": 105200
+    },
+    {
+      "epoch": 7.148050006794401,
+      "grad_norm": 7.0625,
+      "learning_rate": 5.324687457534992e-06,
+      "loss": 0.7614,
+      "step": 105205
+    },
+    {
+      "epoch": 7.148389726865063,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.322564207093355e-06,
+      "loss": 0.7963,
+      "step": 105210
+    },
+    {
+      "epoch": 7.148729446935725,
+      "grad_norm": 7.40625,
+      "learning_rate": 5.320440956651719e-06,
+      "loss": 0.8301,
+      "step": 105215
+    },
+    {
+      "epoch": 7.149069167006386,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.318317706210083e-06,
+      "loss": 0.8471,
+      "step": 105220
+    },
+    {
+      "epoch": 7.1494088870770485,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.316194455768448e-06,
+      "loss": 0.8284,
+      "step": 105225
+    },
+    {
+      "epoch": 7.1497486071477105,
+      "grad_norm": 8.9375,
+      "learning_rate": 5.314071205326811e-06,
+      "loss": 0.8618,
+      "step": 105230
+    },
+    {
+      "epoch": 7.150088327218372,
+      "grad_norm": 7.25,
+      "learning_rate": 5.311947954885175e-06,
+      "loss": 0.7945,
+      "step": 105235
+    },
+    {
+      "epoch": 7.150428047289034,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.309824704443539e-06,
+      "loss": 0.8486,
+      "step": 105240
+    },
+    {
+      "epoch": 7.150767767359696,
+      "grad_norm": 8.25,
+      "learning_rate": 5.307701454001902e-06,
+      "loss": 0.9363,
+      "step": 105245
+    },
+    {
+      "epoch": 7.151107487430357,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.305578203560267e-06,
+      "loss": 0.8654,
+      "step": 105250
+    },
+    {
+      "epoch": 7.151447207501019,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.303454953118631e-06,
+      "loss": 0.873,
+      "step": 105255
+    },
+    {
+      "epoch": 7.151786927571681,
+      "grad_norm": 8.125,
+      "learning_rate": 5.301331702676995e-06,
+      "loss": 0.8594,
+      "step": 105260
+    },
+    {
+      "epoch": 7.152126647642342,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.299208452235358e-06,
+      "loss": 0.8308,
+      "step": 105265
+    },
+    {
+      "epoch": 7.1524663677130045,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.297085201793722e-06,
+      "loss": 0.9757,
+      "step": 105270
+    },
+    {
+      "epoch": 7.1528060877836666,
+      "grad_norm": 6.78125,
+      "learning_rate": 5.294961951352086e-06,
+      "loss": 0.7722,
+      "step": 105275
+    },
+    {
+      "epoch": 7.153145807854328,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.29283870091045e-06,
+      "loss": 0.832,
+      "step": 105280
+    },
+    {
+      "epoch": 7.15348552792499,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.290715450468814e-06,
+      "loss": 0.7859,
+      "step": 105285
+    },
+    {
+      "epoch": 7.153825247995652,
+      "grad_norm": 6.78125,
+      "learning_rate": 5.288592200027178e-06,
+      "loss": 0.887,
+      "step": 105290
+    },
+    {
+      "epoch": 7.154164968066313,
+      "grad_norm": 8.125,
+      "learning_rate": 5.286468949585541e-06,
+      "loss": 0.8775,
+      "step": 105295
+    },
+    {
+      "epoch": 7.154504688136975,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.284345699143905e-06,
+      "loss": 0.9228,
+      "step": 105300
+    },
+    {
+      "epoch": 7.154844408207637,
+      "grad_norm": 8.125,
+      "learning_rate": 5.28222244870227e-06,
+      "loss": 0.8991,
+      "step": 105305
+    },
+    {
+      "epoch": 7.155184128278298,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.280099198260634e-06,
+      "loss": 0.8618,
+      "step": 105310
+    },
+    {
+      "epoch": 7.1555238483489605,
+      "grad_norm": 8.25,
+      "learning_rate": 5.277975947818997e-06,
+      "loss": 0.7434,
+      "step": 105315
+    },
+    {
+      "epoch": 7.155863568419623,
+      "grad_norm": 8.375,
+      "learning_rate": 5.275852697377361e-06,
+      "loss": 0.9059,
+      "step": 105320
+    },
+    {
+      "epoch": 7.156203288490284,
+      "grad_norm": 6.90625,
+      "learning_rate": 5.273729446935725e-06,
+      "loss": 0.8627,
+      "step": 105325
+    },
+    {
+      "epoch": 7.156543008560946,
+      "grad_norm": 7.40625,
+      "learning_rate": 5.271606196494089e-06,
+      "loss": 0.7666,
+      "step": 105330
+    },
+    {
+      "epoch": 7.156882728631608,
+      "grad_norm": 8.625,
+      "learning_rate": 5.269482946052453e-06,
+      "loss": 0.8395,
+      "step": 105335
+    },
+    {
+      "epoch": 7.157222448702269,
+      "grad_norm": 7.34375,
+      "learning_rate": 5.267359695610817e-06,
+      "loss": 0.883,
+      "step": 105340
+    },
+    {
+      "epoch": 7.157562168772931,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.265236445169181e-06,
+      "loss": 0.9257,
+      "step": 105345
+    },
+    {
+      "epoch": 7.157901888843593,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.263113194727544e-06,
+      "loss": 0.9216,
+      "step": 105350
+    },
+    {
+      "epoch": 7.158241608914254,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.260989944285908e-06,
+      "loss": 0.8545,
+      "step": 105355
+    },
+    {
+      "epoch": 7.1585813289849165,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.258866693844272e-06,
+      "loss": 0.8764,
+      "step": 105360
+    },
+    {
+      "epoch": 7.158921049055579,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.256743443402636e-06,
+      "loss": 0.8998,
+      "step": 105365
+    },
+    {
+      "epoch": 7.15926076912624,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.254620192961e-06,
+      "loss": 0.7493,
+      "step": 105370
+    },
+    {
+      "epoch": 7.159600489196902,
+      "grad_norm": 7.125,
+      "learning_rate": 5.252496942519364e-06,
+      "loss": 0.9002,
+      "step": 105375
+    },
+    {
+      "epoch": 7.159940209267564,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.250373692077728e-06,
+      "loss": 0.7998,
+      "step": 105380
+    },
+    {
+      "epoch": 7.160279929338225,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.248250441636092e-06,
+      "loss": 0.8449,
+      "step": 105385
+    },
+    {
+      "epoch": 7.160619649408887,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.2461271911944564e-06,
+      "loss": 0.869,
+      "step": 105390
+    },
+    {
+      "epoch": 7.160959369479549,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.2440039407528204e-06,
+      "loss": 0.8495,
+      "step": 105395
+    },
+    {
+      "epoch": 7.16129908955021,
+      "grad_norm": 7.4375,
+      "learning_rate": 5.241880690311184e-06,
+      "loss": 0.76,
+      "step": 105400
+    },
+    {
+      "epoch": 7.1616388096208725,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.239757439869548e-06,
+      "loss": 0.825,
+      "step": 105405
+    },
+    {
+      "epoch": 7.161978529691535,
+      "grad_norm": 6.28125,
+      "learning_rate": 5.237634189427912e-06,
+      "loss": 0.8337,
+      "step": 105410
+    },
+    {
+      "epoch": 7.162318249762196,
+      "grad_norm": 7.875,
+      "learning_rate": 5.235510938986276e-06,
+      "loss": 0.8214,
+      "step": 105415
+    },
+    {
+      "epoch": 7.162657969832858,
+      "grad_norm": 7.46875,
+      "learning_rate": 5.23338768854464e-06,
+      "loss": 0.8317,
+      "step": 105420
+    },
+    {
+      "epoch": 7.16299768990352,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.231264438103004e-06,
+      "loss": 0.8037,
+      "step": 105425
+    },
+    {
+      "epoch": 7.163337409974181,
+      "grad_norm": 7.21875,
+      "learning_rate": 5.229141187661368e-06,
+      "loss": 0.852,
+      "step": 105430
+    },
+    {
+      "epoch": 7.163677130044843,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.227017937219731e-06,
+      "loss": 0.9313,
+      "step": 105435
+    },
+    {
+      "epoch": 7.164016850115505,
+      "grad_norm": 7.5,
+      "learning_rate": 5.224894686778095e-06,
+      "loss": 0.78,
+      "step": 105440
+    },
+    {
+      "epoch": 7.1643565701861665,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.22277143633646e-06,
+      "loss": 0.9157,
+      "step": 105445
+    },
+    {
+      "epoch": 7.1646962902568285,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.220648185894823e-06,
+      "loss": 0.8417,
+      "step": 105450
+    },
+    {
+      "epoch": 7.165036010327491,
+      "grad_norm": 8.25,
+      "learning_rate": 5.218524935453187e-06,
+      "loss": 0.7737,
+      "step": 105455
+    },
+    {
+      "epoch": 7.165375730398152,
+      "grad_norm": 9.625,
+      "learning_rate": 5.216401685011551e-06,
+      "loss": 0.8965,
+      "step": 105460
+    },
+    {
+      "epoch": 7.165715450468814,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.214278434569914e-06,
+      "loss": 0.8336,
+      "step": 105465
+    },
+    {
+      "epoch": 7.166055170539475,
+      "grad_norm": 8.75,
+      "learning_rate": 5.212155184128279e-06,
+      "loss": 0.7851,
+      "step": 105470
+    },
+    {
+      "epoch": 7.166394890610137,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.210031933686643e-06,
+      "loss": 0.9238,
+      "step": 105475
+    },
+    {
+      "epoch": 7.166734610680799,
+      "grad_norm": 7.75,
+      "learning_rate": 5.207908683245007e-06,
+      "loss": 0.8543,
+      "step": 105480
+    },
+    {
+      "epoch": 7.16707433075146,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.20578543280337e-06,
+      "loss": 0.7383,
+      "step": 105485
+    },
+    {
+      "epoch": 7.1674140508221225,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.203662182361734e-06,
+      "loss": 0.7983,
+      "step": 105490
+    },
+    {
+      "epoch": 7.1677537708927845,
+      "grad_norm": 8.5,
+      "learning_rate": 5.201538931920098e-06,
+      "loss": 0.863,
+      "step": 105495
+    },
+    {
+      "epoch": 7.168093490963446,
+      "grad_norm": 9.25,
+      "learning_rate": 5.199415681478462e-06,
+      "loss": 0.8341,
+      "step": 105500
+    },
+    {
+      "epoch": 7.168433211034108,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.197292431036826e-06,
+      "loss": 0.8111,
+      "step": 105505
+    },
+    {
+      "epoch": 7.16877293110477,
+      "grad_norm": 7.0625,
+      "learning_rate": 5.19516918059519e-06,
+      "loss": 0.7678,
+      "step": 105510
+    },
+    {
+      "epoch": 7.169112651175431,
+      "grad_norm": 7.375,
+      "learning_rate": 5.193045930153554e-06,
+      "loss": 0.8354,
+      "step": 105515
+    },
+    {
+      "epoch": 7.169452371246093,
+      "grad_norm": 8.5625,
+      "learning_rate": 5.190922679711917e-06,
+      "loss": 0.8447,
+      "step": 105520
+    },
+    {
+      "epoch": 7.169792091316755,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.188799429270281e-06,
+      "loss": 0.995,
+      "step": 105525
+    },
+    {
+      "epoch": 7.170131811387416,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.186676178828646e-06,
+      "loss": 0.7589,
+      "step": 105530
+    },
+    {
+      "epoch": 7.1704715314580785,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.184552928387009e-06,
+      "loss": 0.8581,
+      "step": 105535
+    },
+    {
+      "epoch": 7.1708112515287405,
+      "grad_norm": 8.125,
+      "learning_rate": 5.182429677945373e-06,
+      "loss": 0.9292,
+      "step": 105540
+    },
+    {
+      "epoch": 7.171150971599402,
+      "grad_norm": 8.625,
+      "learning_rate": 5.180306427503737e-06,
+      "loss": 0.8333,
+      "step": 105545
+    },
+    {
+      "epoch": 7.171490691670064,
+      "grad_norm": 8.125,
+      "learning_rate": 5.1781831770621e-06,
+      "loss": 0.8256,
+      "step": 105550
+    },
+    {
+      "epoch": 7.171830411740726,
+      "grad_norm": 8.0,
+      "learning_rate": 5.176059926620465e-06,
+      "loss": 0.8161,
+      "step": 105555
+    },
+    {
+      "epoch": 7.172170131811387,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.173936676178829e-06,
+      "loss": 0.8452,
+      "step": 105560
+    },
+    {
+      "epoch": 7.172509851882049,
+      "grad_norm": 8.0,
+      "learning_rate": 5.171813425737193e-06,
+      "loss": 0.8319,
+      "step": 105565
+    },
+    {
+      "epoch": 7.172849571952711,
+      "grad_norm": 7.75,
+      "learning_rate": 5.169690175295556e-06,
+      "loss": 0.9177,
+      "step": 105570
+    },
+    {
+      "epoch": 7.173189292023372,
+      "grad_norm": 7.625,
+      "learning_rate": 5.16756692485392e-06,
+      "loss": 0.7343,
+      "step": 105575
+    },
+    {
+      "epoch": 7.1735290120940345,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.165443674412284e-06,
+      "loss": 0.8346,
+      "step": 105580
+    },
+    {
+      "epoch": 7.173868732164697,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.163320423970648e-06,
+      "loss": 0.8107,
+      "step": 105585
+    },
+    {
+      "epoch": 7.174208452235358,
+      "grad_norm": 6.375,
+      "learning_rate": 5.161197173529012e-06,
+      "loss": 0.7986,
+      "step": 105590
+    },
+    {
+      "epoch": 7.17454817230602,
+      "grad_norm": 8.5,
+      "learning_rate": 5.159073923087376e-06,
+      "loss": 0.8808,
+      "step": 105595
+    },
+    {
+      "epoch": 7.174887892376682,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.15695067264574e-06,
+      "loss": 0.883,
+      "step": 105600
+    },
+    {
+      "epoch": 7.175227612447343,
+      "grad_norm": 6.5625,
+      "learning_rate": 5.1548274222041035e-06,
+      "loss": 0.8179,
+      "step": 105605
+    },
+    {
+      "epoch": 7.175567332518005,
+      "grad_norm": 7.875,
+      "learning_rate": 5.152704171762468e-06,
+      "loss": 0.7729,
+      "step": 105610
+    },
+    {
+      "epoch": 7.175907052588667,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.150580921320832e-06,
+      "loss": 0.7322,
+      "step": 105615
+    },
+    {
+      "epoch": 7.176246772659328,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.1484576708791955e-06,
+      "loss": 0.7456,
+      "step": 105620
+    },
+    {
+      "epoch": 7.1765864927299905,
+      "grad_norm": 8.5,
+      "learning_rate": 5.1463344204375595e-06,
+      "loss": 0.7366,
+      "step": 105625
+    },
+    {
+      "epoch": 7.176926212800653,
+      "grad_norm": 6.96875,
+      "learning_rate": 5.1442111699959235e-06,
+      "loss": 0.8229,
+      "step": 105630
+    },
+    {
+      "epoch": 7.177265932871314,
+      "grad_norm": 7.4375,
+      "learning_rate": 5.1420879195542875e-06,
+      "loss": 0.8034,
+      "step": 105635
+    },
+    {
+      "epoch": 7.177605652941976,
+      "grad_norm": 9.25,
+      "learning_rate": 5.1399646691126515e-06,
+      "loss": 0.82,
+      "step": 105640
+    },
+    {
+      "epoch": 7.177945373012638,
+      "grad_norm": 7.21875,
+      "learning_rate": 5.1378414186710155e-06,
+      "loss": 0.8448,
+      "step": 105645
+    },
+    {
+      "epoch": 7.178285093083299,
+      "grad_norm": 8.375,
+      "learning_rate": 5.1357181682293796e-06,
+      "loss": 0.9558,
+      "step": 105650
+    },
+    {
+      "epoch": 7.178624813153961,
+      "grad_norm": 9.0625,
+      "learning_rate": 5.133594917787743e-06,
+      "loss": 0.8707,
+      "step": 105655
+    },
+    {
+      "epoch": 7.178964533224623,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.131471667346107e-06,
+      "loss": 0.8842,
+      "step": 105660
+    },
+    {
+      "epoch": 7.179304253295284,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.1293484169044716e-06,
+      "loss": 0.8731,
+      "step": 105665
+    },
+    {
+      "epoch": 7.1796439733659465,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.127225166462835e-06,
+      "loss": 0.8509,
+      "step": 105670
+    },
+    {
+      "epoch": 7.179983693436609,
+      "grad_norm": 8.875,
+      "learning_rate": 5.125101916021199e-06,
+      "loss": 0.8614,
+      "step": 105675
+    },
+    {
+      "epoch": 7.18032341350727,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.122978665579563e-06,
+      "loss": 0.8604,
+      "step": 105680
+    },
+    {
+      "epoch": 7.180663133577932,
+      "grad_norm": 9.4375,
+      "learning_rate": 5.120855415137927e-06,
+      "loss": 0.9472,
+      "step": 105685
+    },
+    {
+      "epoch": 7.181002853648594,
+      "grad_norm": 7.34375,
+      "learning_rate": 5.118732164696291e-06,
+      "loss": 0.8298,
+      "step": 105690
+    },
+    {
+      "epoch": 7.181342573719255,
+      "grad_norm": 8.8125,
+      "learning_rate": 5.116608914254655e-06,
+      "loss": 0.8505,
+      "step": 105695
+    },
+    {
+      "epoch": 7.181682293789917,
+      "grad_norm": 7.75,
+      "learning_rate": 5.114485663813019e-06,
+      "loss": 0.9335,
+      "step": 105700
+    },
+    {
+      "epoch": 7.182022013860579,
+      "grad_norm": 8.0,
+      "learning_rate": 5.112362413371382e-06,
+      "loss": 0.8054,
+      "step": 105705
+    },
+    {
+      "epoch": 7.18236173393124,
+      "grad_norm": 8.125,
+      "learning_rate": 5.110239162929746e-06,
+      "loss": 0.8895,
+      "step": 105710
+    },
+    {
+      "epoch": 7.1827014540019025,
+      "grad_norm": 7.5,
+      "learning_rate": 5.10811591248811e-06,
+      "loss": 0.8136,
+      "step": 105715
+    },
+    {
+      "epoch": 7.183041174072565,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.105992662046474e-06,
+      "loss": 0.8329,
+      "step": 105720
+    },
+    {
+      "epoch": 7.183380894143226,
+      "grad_norm": 7.0,
+      "learning_rate": 5.103869411604838e-06,
+      "loss": 0.8124,
+      "step": 105725
+    },
+    {
+      "epoch": 7.183720614213888,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.101746161163202e-06,
+      "loss": 0.8181,
+      "step": 105730
+    },
+    {
+      "epoch": 7.18406033428455,
+      "grad_norm": 7.90625,
+      "learning_rate": 5.099622910721566e-06,
+      "loss": 0.8038,
+      "step": 105735
+    },
+    {
+      "epoch": 7.184400054355211,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.097499660279929e-06,
+      "loss": 0.887,
+      "step": 105740
+    },
+    {
+      "epoch": 7.184739774425873,
+      "grad_norm": 8.5,
+      "learning_rate": 5.095376409838293e-06,
+      "loss": 0.8323,
+      "step": 105745
+    },
+    {
+      "epoch": 7.185079494496535,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.093253159396658e-06,
+      "loss": 0.7699,
+      "step": 105750
+    },
+    {
+      "epoch": 7.1854192145671965,
+      "grad_norm": 8.625,
+      "learning_rate": 5.091129908955021e-06,
+      "loss": 0.8485,
+      "step": 105755
+    },
+    {
+      "epoch": 7.1857589346378585,
+      "grad_norm": 7.34375,
+      "learning_rate": 5.089006658513385e-06,
+      "loss": 0.8541,
+      "step": 105760
+    },
+    {
+      "epoch": 7.186098654708521,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.086883408071749e-06,
+      "loss": 0.7824,
+      "step": 105765
+    },
+    {
+      "epoch": 7.186438374779182,
+      "grad_norm": 7.03125,
+      "learning_rate": 5.084760157630113e-06,
+      "loss": 0.8207,
+      "step": 105770
+    },
+    {
+      "epoch": 7.186778094849844,
+      "grad_norm": 6.84375,
+      "learning_rate": 5.082636907188477e-06,
+      "loss": 0.8573,
+      "step": 105775
+    },
+    {
+      "epoch": 7.187117814920505,
+      "grad_norm": 8.125,
+      "learning_rate": 5.080513656746841e-06,
+      "loss": 0.869,
+      "step": 105780
+    },
+    {
+      "epoch": 7.187457534991167,
+      "grad_norm": 8.3125,
+      "learning_rate": 5.078390406305205e-06,
+      "loss": 0.8735,
+      "step": 105785
+    },
+    {
+      "epoch": 7.187797255061829,
+      "grad_norm": 7.6875,
+      "learning_rate": 5.076267155863568e-06,
+      "loss": 0.8713,
+      "step": 105790
+    },
+    {
+      "epoch": 7.18813697513249,
+      "grad_norm": 7.75,
+      "learning_rate": 5.074143905421932e-06,
+      "loss": 0.8794,
+      "step": 105795
+    },
+    {
+      "epoch": 7.1884766952031525,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.072020654980296e-06,
+      "loss": 0.793,
+      "step": 105800
+    },
+    {
+      "epoch": 7.1888164152738145,
+      "grad_norm": 8.25,
+      "learning_rate": 5.06989740453866e-06,
+      "loss": 0.8533,
+      "step": 105805
+    },
+    {
+      "epoch": 7.189156135344476,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.067774154097024e-06,
+      "loss": 0.8562,
+      "step": 105810
+    },
+    {
+      "epoch": 7.189495855415138,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.065650903655388e-06,
+      "loss": 1.0031,
+      "step": 105815
+    },
+    {
+      "epoch": 7.1898355754858,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.063527653213752e-06,
+      "loss": 0.8058,
+      "step": 105820
+    },
+    {
+      "epoch": 7.190175295556461,
+      "grad_norm": 8.125,
+      "learning_rate": 5.0614044027721155e-06,
+      "loss": 0.8496,
+      "step": 105825
+    },
+    {
+      "epoch": 7.190515015627123,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.05928115233048e-06,
+      "loss": 0.8505,
+      "step": 105830
+    },
+    {
+      "epoch": 7.190854735697785,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.057157901888844e-06,
+      "loss": 0.814,
+      "step": 105835
+    },
+    {
+      "epoch": 7.191194455768446,
+      "grad_norm": 8.625,
+      "learning_rate": 5.0550346514472075e-06,
+      "loss": 0.8292,
+      "step": 105840
+    },
+    {
+      "epoch": 7.1915341758391085,
+      "grad_norm": 8.25,
+      "learning_rate": 5.0529114010055715e-06,
+      "loss": 0.8199,
+      "step": 105845
+    },
+    {
+      "epoch": 7.1918738959097706,
+      "grad_norm": 8.25,
+      "learning_rate": 5.0507881505639355e-06,
+      "loss": 0.7999,
+      "step": 105850
+    },
+    {
+      "epoch": 7.192213615980432,
+      "grad_norm": 7.96875,
+      "learning_rate": 5.0486649001222995e-06,
+      "loss": 0.8617,
+      "step": 105855
+    },
+    {
+      "epoch": 7.192553336051094,
+      "grad_norm": 8.125,
+      "learning_rate": 5.0465416496806635e-06,
+      "loss": 0.8822,
+      "step": 105860
+    },
+    {
+      "epoch": 7.192893056121756,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.0444183992390275e-06,
+      "loss": 0.8753,
+      "step": 105865
+    },
+    {
+      "epoch": 7.193232776192417,
+      "grad_norm": 8.25,
+      "learning_rate": 5.0422951487973915e-06,
+      "loss": 0.7771,
+      "step": 105870
+    },
+    {
+      "epoch": 7.193572496263079,
+      "grad_norm": 7.75,
+      "learning_rate": 5.040171898355755e-06,
+      "loss": 0.7458,
+      "step": 105875
+    },
+    {
+      "epoch": 7.193912216333741,
+      "grad_norm": 7.84375,
+      "learning_rate": 5.038048647914119e-06,
+      "loss": 0.7919,
+      "step": 105880
+    },
+    {
+      "epoch": 7.194251936404402,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.035925397472483e-06,
+      "loss": 0.7049,
+      "step": 105885
+    },
+    {
+      "epoch": 7.1945916564750645,
+      "grad_norm": 6.9375,
+      "learning_rate": 5.033802147030847e-06,
+      "loss": 0.9031,
+      "step": 105890
+    },
+    {
+      "epoch": 7.194931376545727,
+      "grad_norm": 8.9375,
+      "learning_rate": 5.031678896589211e-06,
+      "loss": 0.8763,
+      "step": 105895
+    },
+    {
+      "epoch": 7.195271096616388,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.029555646147575e-06,
+      "loss": 0.9476,
+      "step": 105900
+    },
+    {
+      "epoch": 7.19561081668705,
+      "grad_norm": 7.5625,
+      "learning_rate": 5.027432395705939e-06,
+      "loss": 0.7566,
+      "step": 105905
+    },
+    {
+      "epoch": 7.195950536757712,
+      "grad_norm": 8.125,
+      "learning_rate": 5.025309145264302e-06,
+      "loss": 0.7348,
+      "step": 105910
+    },
+    {
+      "epoch": 7.196290256828373,
+      "grad_norm": 7.8125,
+      "learning_rate": 5.023185894822667e-06,
+      "loss": 0.9246,
+      "step": 105915
+    },
+    {
+      "epoch": 7.196629976899035,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.021062644381031e-06,
+      "loss": 0.8713,
+      "step": 105920
+    },
+    {
+      "epoch": 7.196969696969697,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.018939393939394e-06,
+      "loss": 0.8859,
+      "step": 105925
+    },
+    {
+      "epoch": 7.197309417040358,
+      "grad_norm": 9.25,
+      "learning_rate": 5.016816143497758e-06,
+      "loss": 0.8031,
+      "step": 105930
+    },
+    {
+      "epoch": 7.1976491371110205,
+      "grad_norm": 7.5,
+      "learning_rate": 5.014692893056122e-06,
+      "loss": 0.7941,
+      "step": 105935
+    },
+    {
+      "epoch": 7.197988857181683,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.012569642614486e-06,
+      "loss": 1.0233,
+      "step": 105940
+    },
+    {
+      "epoch": 7.198328577252344,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.01044639217285e-06,
+      "loss": 0.9353,
+      "step": 105945
+    },
+    {
+      "epoch": 7.198668297323006,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.008323141731214e-06,
+      "loss": 0.8334,
+      "step": 105950
+    },
+    {
+      "epoch": 7.199008017393668,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.006199891289578e-06,
+      "loss": 0.8634,
+      "step": 105955
+    },
+    {
+      "epoch": 7.199347737464329,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.004076640847941e-06,
+      "loss": 0.8491,
+      "step": 105960
+    },
+    {
+      "epoch": 7.199687457534991,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.001953390406305e-06,
+      "loss": 0.7597,
+      "step": 105965
+    },
+    {
+      "epoch": 7.200027177605653,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.99983013996467e-06,
+      "loss": 0.8168,
+      "step": 105970
+    },
+    {
+      "epoch": 7.200366897676314,
+      "grad_norm": 8.125,
+      "learning_rate": 4.997706889523033e-06,
+      "loss": 0.8465,
+      "step": 105975
+    },
+    {
+      "epoch": 7.2007066177469765,
+      "grad_norm": 8.75,
+      "learning_rate": 4.995583639081397e-06,
+      "loss": 0.8359,
+      "step": 105980
+    },
+    {
+      "epoch": 7.201046337817639,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.993460388639761e-06,
+      "loss": 0.8141,
+      "step": 105985
+    },
+    {
+      "epoch": 7.2013860578883,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.991337138198125e-06,
+      "loss": 0.8378,
+      "step": 105990
+    },
+    {
+      "epoch": 7.201725777958962,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.989213887756489e-06,
+      "loss": 0.9225,
+      "step": 105995
+    },
+    {
+      "epoch": 7.202065498029624,
+      "grad_norm": 8.0,
+      "learning_rate": 4.987090637314853e-06,
+      "loss": 0.7627,
+      "step": 106000
+    },
+    {
+      "epoch": 7.202405218100285,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.984967386873217e-06,
+      "loss": 0.8961,
+      "step": 106005
+    },
+    {
+      "epoch": 7.202744938170947,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.98284413643158e-06,
+      "loss": 0.8128,
+      "step": 106010
+    },
+    {
+      "epoch": 7.203084658241609,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.980720885989944e-06,
+      "loss": 0.8228,
+      "step": 106015
+    },
+    {
+      "epoch": 7.2034243783122704,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.978597635548308e-06,
+      "loss": 0.87,
+      "step": 106020
+    },
+    {
+      "epoch": 7.2037640983829325,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.976474385106673e-06,
+      "loss": 0.8542,
+      "step": 106025
+    },
+    {
+      "epoch": 7.204103818453595,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.974351134665036e-06,
+      "loss": 0.8967,
+      "step": 106030
+    },
+    {
+      "epoch": 7.204443538524256,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.9722278842234e-06,
+      "loss": 0.8225,
+      "step": 106035
+    },
+    {
+      "epoch": 7.204783258594918,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.970104633781764e-06,
+      "loss": 0.7315,
+      "step": 106040
+    },
+    {
+      "epoch": 7.20512297866558,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.967981383340127e-06,
+      "loss": 0.8099,
+      "step": 106045
+    },
+    {
+      "epoch": 7.205462698736241,
+      "grad_norm": 8.0,
+      "learning_rate": 4.965858132898492e-06,
+      "loss": 0.8639,
+      "step": 106050
+    },
+    {
+      "epoch": 7.205802418806903,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.963734882456856e-06,
+      "loss": 0.7915,
+      "step": 106055
+    },
+    {
+      "epoch": 7.206142138877565,
+      "grad_norm": 8.375,
+      "learning_rate": 4.961611632015219e-06,
+      "loss": 0.9573,
+      "step": 106060
+    },
+    {
+      "epoch": 7.2064818589482265,
+      "grad_norm": 8.625,
+      "learning_rate": 4.959488381573583e-06,
+      "loss": 0.835,
+      "step": 106065
+    },
+    {
+      "epoch": 7.2068215790188885,
+      "grad_norm": 8.625,
+      "learning_rate": 4.957365131131947e-06,
+      "loss": 0.9,
+      "step": 106070
+    },
+    {
+      "epoch": 7.207161299089551,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.955241880690311e-06,
+      "loss": 0.9071,
+      "step": 106075
+    },
+    {
+      "epoch": 7.207501019160212,
+      "grad_norm": 7.625,
+      "learning_rate": 4.953118630248675e-06,
+      "loss": 0.7735,
+      "step": 106080
+    },
+    {
+      "epoch": 7.207840739230874,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.950995379807039e-06,
+      "loss": 0.8483,
+      "step": 106085
+    },
+    {
+      "epoch": 7.208180459301536,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.948872129365403e-06,
+      "loss": 0.8118,
+      "step": 106090
+    },
+    {
+      "epoch": 7.208520179372197,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.946748878923767e-06,
+      "loss": 0.7388,
+      "step": 106095
+    },
+    {
+      "epoch": 7.208859899442859,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.944625628482131e-06,
+      "loss": 0.9193,
+      "step": 106100
+    },
+    {
+      "epoch": 7.209199619513521,
+      "grad_norm": 8.25,
+      "learning_rate": 4.942502378040495e-06,
+      "loss": 0.852,
+      "step": 106105
+    },
+    {
+      "epoch": 7.2095393395841825,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.9403791275988594e-06,
+      "loss": 0.8223,
+      "step": 106110
+    },
+    {
+      "epoch": 7.2098790596548445,
+      "grad_norm": 7.875,
+      "learning_rate": 4.938255877157223e-06,
+      "loss": 0.8328,
+      "step": 106115
+    },
+    {
+      "epoch": 7.210218779725507,
+      "grad_norm": 8.375,
+      "learning_rate": 4.936132626715587e-06,
+      "loss": 0.8389,
+      "step": 106120
+    },
+    {
+      "epoch": 7.210558499796168,
+      "grad_norm": 7.875,
+      "learning_rate": 4.934009376273951e-06,
+      "loss": 0.826,
+      "step": 106125
+    },
+    {
+      "epoch": 7.21089821986683,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.931886125832314e-06,
+      "loss": 0.7933,
+      "step": 106130
+    },
+    {
+      "epoch": 7.211237939937492,
+      "grad_norm": 8.5,
+      "learning_rate": 4.929762875390679e-06,
+      "loss": 0.7882,
+      "step": 106135
+    },
+    {
+      "epoch": 7.211577660008153,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.927639624949043e-06,
+      "loss": 0.8173,
+      "step": 106140
+    },
+    {
+      "epoch": 7.211917380078815,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.925516374507406e-06,
+      "loss": 0.8734,
+      "step": 106145
+    },
+    {
+      "epoch": 7.212257100149476,
+      "grad_norm": 8.125,
+      "learning_rate": 4.92339312406577e-06,
+      "loss": 0.8976,
+      "step": 106150
+    },
+    {
+      "epoch": 7.2125968202201385,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.921269873624134e-06,
+      "loss": 0.7831,
+      "step": 106155
+    },
+    {
+      "epoch": 7.212936540290801,
+      "grad_norm": 8.5,
+      "learning_rate": 4.919146623182498e-06,
+      "loss": 0.8746,
+      "step": 106160
+    },
+    {
+      "epoch": 7.213276260361462,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.917023372740862e-06,
+      "loss": 0.8422,
+      "step": 106165
+    },
+    {
+      "epoch": 7.213615980432124,
+      "grad_norm": 6.6875,
+      "learning_rate": 4.914900122299226e-06,
+      "loss": 0.8573,
+      "step": 106170
+    },
+    {
+      "epoch": 7.213955700502786,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.91277687185759e-06,
+      "loss": 0.7989,
+      "step": 106175
+    },
+    {
+      "epoch": 7.214295420573447,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.910653621415953e-06,
+      "loss": 0.8318,
+      "step": 106180
+    },
+    {
+      "epoch": 7.214635140644109,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.908530370974317e-06,
+      "loss": 0.8685,
+      "step": 106185
+    },
+    {
+      "epoch": 7.214974860714771,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.906407120532682e-06,
+      "loss": 0.7598,
+      "step": 106190
+    },
+    {
+      "epoch": 7.215314580785432,
+      "grad_norm": 8.625,
+      "learning_rate": 4.904283870091046e-06,
+      "loss": 0.8507,
+      "step": 106195
+    },
+    {
+      "epoch": 7.2156543008560945,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.902160619649409e-06,
+      "loss": 0.8427,
+      "step": 106200
+    },
+    {
+      "epoch": 7.215994020926757,
+      "grad_norm": 8.625,
+      "learning_rate": 4.900037369207773e-06,
+      "loss": 0.9272,
+      "step": 106205
+    },
+    {
+      "epoch": 7.216333740997418,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.897914118766137e-06,
+      "loss": 0.8707,
+      "step": 106210
+    },
+    {
+      "epoch": 7.21667346106808,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.895790868324501e-06,
+      "loss": 0.8049,
+      "step": 106215
+    },
+    {
+      "epoch": 7.217013181138742,
+      "grad_norm": 8.0,
+      "learning_rate": 4.893667617882865e-06,
+      "loss": 0.7607,
+      "step": 106220
+    },
+    {
+      "epoch": 7.217352901209403,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.891544367441229e-06,
+      "loss": 0.7835,
+      "step": 106225
+    },
+    {
+      "epoch": 7.217692621280065,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.889421116999592e-06,
+      "loss": 0.8481,
+      "step": 106230
+    },
+    {
+      "epoch": 7.218032341350727,
+      "grad_norm": 8.0,
+      "learning_rate": 4.887297866557956e-06,
+      "loss": 0.8567,
+      "step": 106235
+    },
+    {
+      "epoch": 7.218372061421388,
+      "grad_norm": 8.75,
+      "learning_rate": 4.88517461611632e-06,
+      "loss": 0.7878,
+      "step": 106240
+    },
+    {
+      "epoch": 7.2187117814920505,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.883051365674685e-06,
+      "loss": 0.7631,
+      "step": 106245
+    },
+    {
+      "epoch": 7.219051501562713,
+      "grad_norm": 8.25,
+      "learning_rate": 4.880928115233048e-06,
+      "loss": 0.7462,
+      "step": 106250
+    },
+    {
+      "epoch": 7.219391221633374,
+      "grad_norm": 7.875,
+      "learning_rate": 4.878804864791412e-06,
+      "loss": 0.8061,
+      "step": 106255
+    },
+    {
+      "epoch": 7.219730941704036,
+      "grad_norm": 8.625,
+      "learning_rate": 4.876681614349776e-06,
+      "loss": 0.8218,
+      "step": 106260
+    },
+    {
+      "epoch": 7.220070661774698,
+      "grad_norm": 7.5,
+      "learning_rate": 4.874558363908139e-06,
+      "loss": 0.8585,
+      "step": 106265
+    },
+    {
+      "epoch": 7.220410381845359,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.872435113466504e-06,
+      "loss": 0.884,
+      "step": 106270
+    },
+    {
+      "epoch": 7.220750101916021,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.870311863024868e-06,
+      "loss": 0.8499,
+      "step": 106275
+    },
+    {
+      "epoch": 7.221089821986683,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.868188612583232e-06,
+      "loss": 1.0492,
+      "step": 106280
+    },
+    {
+      "epoch": 7.221429542057344,
+      "grad_norm": 8.0,
+      "learning_rate": 4.866065362141595e-06,
+      "loss": 0.9358,
+      "step": 106285
+    },
+    {
+      "epoch": 7.2217692621280065,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.863942111699959e-06,
+      "loss": 0.9455,
+      "step": 106290
+    },
+    {
+      "epoch": 7.222108982198669,
+      "grad_norm": 8.75,
+      "learning_rate": 4.861818861258323e-06,
+      "loss": 0.9075,
+      "step": 106295
+    },
+    {
+      "epoch": 7.22244870226933,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.859695610816687e-06,
+      "loss": 0.8636,
+      "step": 106300
+    },
+    {
+      "epoch": 7.222788422339992,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.857572360375051e-06,
+      "loss": 0.8328,
+      "step": 106305
+    },
+    {
+      "epoch": 7.223128142410654,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.855449109933415e-06,
+      "loss": 0.8244,
+      "step": 106310
+    },
+    {
+      "epoch": 7.223467862481315,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.8533258594917785e-06,
+      "loss": 0.7824,
+      "step": 106315
+    },
+    {
+      "epoch": 7.223807582551977,
+      "grad_norm": 8.125,
+      "learning_rate": 4.8512026090501425e-06,
+      "loss": 0.9566,
+      "step": 106320
+    },
+    {
+      "epoch": 7.224147302622639,
+      "grad_norm": 8.375,
+      "learning_rate": 4.8490793586085065e-06,
+      "loss": 0.8563,
+      "step": 106325
+    },
+    {
+      "epoch": 7.2244870226933005,
+      "grad_norm": 9.25,
+      "learning_rate": 4.846956108166871e-06,
+      "loss": 0.8278,
+      "step": 106330
+    },
+    {
+      "epoch": 7.2248267427639625,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.8448328577252345e-06,
+      "loss": 0.7466,
+      "step": 106335
+    },
+    {
+      "epoch": 7.225166462834625,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.8427096072835985e-06,
+      "loss": 0.897,
+      "step": 106340
+    },
+    {
+      "epoch": 7.225506182905286,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.8405863568419625e-06,
+      "loss": 0.8682,
+      "step": 106345
+    },
+    {
+      "epoch": 7.225845902975948,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.838463106400326e-06,
+      "loss": 0.8742,
+      "step": 106350
+    },
+    {
+      "epoch": 7.22618562304661,
+      "grad_norm": 8.625,
+      "learning_rate": 4.8363398559586905e-06,
+      "loss": 0.8371,
+      "step": 106355
+    },
+    {
+      "epoch": 7.226525343117271,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.8342166055170545e-06,
+      "loss": 0.755,
+      "step": 106360
+    },
+    {
+      "epoch": 7.226865063187933,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.8320933550754185e-06,
+      "loss": 0.8113,
+      "step": 106365
+    },
+    {
+      "epoch": 7.227204783258595,
+      "grad_norm": 8.25,
+      "learning_rate": 4.829970104633782e-06,
+      "loss": 0.8991,
+      "step": 106370
+    },
+    {
+      "epoch": 7.2275445033292565,
+      "grad_norm": 8.0,
+      "learning_rate": 4.827846854192146e-06,
+      "loss": 0.8089,
+      "step": 106375
+    },
+    {
+      "epoch": 7.2278842233999185,
+      "grad_norm": 8.5,
+      "learning_rate": 4.82572360375051e-06,
+      "loss": 0.9634,
+      "step": 106380
+    },
+    {
+      "epoch": 7.228223943470581,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.823600353308874e-06,
+      "loss": 0.6893,
+      "step": 106385
+    },
+    {
+      "epoch": 7.228563663541242,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.821477102867238e-06,
+      "loss": 0.8534,
+      "step": 106390
+    },
+    {
+      "epoch": 7.228903383611904,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.819353852425602e-06,
+      "loss": 0.8281,
+      "step": 106395
+    },
+    {
+      "epoch": 7.229243103682566,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.817230601983965e-06,
+      "loss": 0.8843,
+      "step": 106400
+    },
+    {
+      "epoch": 7.229582823753227,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.815107351542329e-06,
+      "loss": 0.8555,
+      "step": 106405
+    },
+    {
+      "epoch": 7.229922543823889,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.812984101100694e-06,
+      "loss": 0.8667,
+      "step": 106410
+    },
+    {
+      "epoch": 7.230262263894551,
+      "grad_norm": 7.875,
+      "learning_rate": 4.810860850659058e-06,
+      "loss": 0.8686,
+      "step": 106415
+    },
+    {
+      "epoch": 7.2306019839652125,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.808737600217421e-06,
+      "loss": 0.8493,
+      "step": 106420
+    },
+    {
+      "epoch": 7.2309417040358746,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.806614349775785e-06,
+      "loss": 0.7855,
+      "step": 106425
+    },
+    {
+      "epoch": 7.231281424106537,
+      "grad_norm": 8.125,
+      "learning_rate": 4.804491099334149e-06,
+      "loss": 0.851,
+      "step": 106430
+    },
+    {
+      "epoch": 7.231621144177198,
+      "grad_norm": 8.375,
+      "learning_rate": 4.802367848892513e-06,
+      "loss": 0.7831,
+      "step": 106435
+    },
+    {
+      "epoch": 7.23196086424786,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.800244598450877e-06,
+      "loss": 0.8405,
+      "step": 106440
+    },
+    {
+      "epoch": 7.232300584318522,
+      "grad_norm": 9.0,
+      "learning_rate": 4.798121348009241e-06,
+      "loss": 0.8416,
+      "step": 106445
+    },
+    {
+      "epoch": 7.232640304389183,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.795998097567605e-06,
+      "loss": 0.8309,
+      "step": 106450
+    },
+    {
+      "epoch": 7.232980024459845,
+      "grad_norm": 8.125,
+      "learning_rate": 4.793874847125968e-06,
+      "loss": 0.8003,
+      "step": 106455
+    },
+    {
+      "epoch": 7.233319744530506,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.791751596684332e-06,
+      "loss": 0.782,
+      "step": 106460
+    },
+    {
+      "epoch": 7.2336594646011685,
+      "grad_norm": 7.0,
+      "learning_rate": 4.789628346242696e-06,
+      "loss": 0.8139,
+      "step": 106465
+    },
+    {
+      "epoch": 7.233999184671831,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.78750509580106e-06,
+      "loss": 0.8368,
+      "step": 106470
+    },
+    {
+      "epoch": 7.234338904742492,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.785381845359424e-06,
+      "loss": 0.7579,
+      "step": 106475
+    },
+    {
+      "epoch": 7.234678624813154,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.783258594917788e-06,
+      "loss": 0.8411,
+      "step": 106480
+    },
+    {
+      "epoch": 7.235018344883816,
+      "grad_norm": 8.625,
+      "learning_rate": 4.781135344476151e-06,
+      "loss": 0.9103,
+      "step": 106485
+    },
+    {
+      "epoch": 7.235358064954477,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.779012094034515e-06,
+      "loss": 0.7477,
+      "step": 106490
+    },
+    {
+      "epoch": 7.235697785025139,
+      "grad_norm": 7.0,
+      "learning_rate": 4.77688884359288e-06,
+      "loss": 0.8627,
+      "step": 106495
+    },
+    {
+      "epoch": 7.236037505095801,
+      "grad_norm": 7.625,
+      "learning_rate": 4.774765593151244e-06,
+      "loss": 0.8421,
+      "step": 106500
+    },
+    {
+      "epoch": 7.236377225166462,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.772642342709607e-06,
+      "loss": 0.8296,
+      "step": 106505
+    },
+    {
+      "epoch": 7.2367169452371245,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.770519092267971e-06,
+      "loss": 0.8443,
+      "step": 106510
+    },
+    {
+      "epoch": 7.237056665307787,
+      "grad_norm": 9.125,
+      "learning_rate": 4.768395841826335e-06,
+      "loss": 0.8471,
+      "step": 106515
+    },
+    {
+      "epoch": 7.237396385378448,
+      "grad_norm": 8.25,
+      "learning_rate": 4.766272591384699e-06,
+      "loss": 0.8688,
+      "step": 106520
+    },
+    {
+      "epoch": 7.23773610544911,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.764149340943063e-06,
+      "loss": 0.8815,
+      "step": 106525
+    },
+    {
+      "epoch": 7.238075825519772,
+      "grad_norm": 7.875,
+      "learning_rate": 4.762026090501427e-06,
+      "loss": 0.7913,
+      "step": 106530
+    },
+    {
+      "epoch": 7.238415545590433,
+      "grad_norm": 8.375,
+      "learning_rate": 4.759902840059791e-06,
+      "loss": 0.8785,
+      "step": 106535
+    },
+    {
+      "epoch": 7.238755265661095,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.7577795896181544e-06,
+      "loss": 0.8544,
+      "step": 106540
+    },
+    {
+      "epoch": 7.239094985731757,
+      "grad_norm": 8.875,
+      "learning_rate": 4.7556563391765185e-06,
+      "loss": 0.8455,
+      "step": 106545
+    },
+    {
+      "epoch": 7.239434705802418,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.753533088734883e-06,
+      "loss": 0.8486,
+      "step": 106550
+    },
+    {
+      "epoch": 7.2397744258730805,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.7514098382932465e-06,
+      "loss": 0.8437,
+      "step": 106555
+    },
+    {
+      "epoch": 7.240114145943743,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.7492865878516105e-06,
+      "loss": 0.8854,
+      "step": 106560
+    },
+    {
+      "epoch": 7.240453866014404,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.7471633374099745e-06,
+      "loss": 0.8607,
+      "step": 106565
+    },
+    {
+      "epoch": 7.240793586085066,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.745040086968338e-06,
+      "loss": 0.9932,
+      "step": 106570
+    },
+    {
+      "epoch": 7.241133306155728,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.7429168365267025e-06,
+      "loss": 0.8633,
+      "step": 106575
+    },
+    {
+      "epoch": 7.241473026226389,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.7407935860850665e-06,
+      "loss": 0.797,
+      "step": 106580
+    },
+    {
+      "epoch": 7.241812746297051,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.7386703356434305e-06,
+      "loss": 0.7656,
+      "step": 106585
+    },
+    {
+      "epoch": 7.242152466367713,
+      "grad_norm": 8.375,
+      "learning_rate": 4.736547085201794e-06,
+      "loss": 0.8365,
+      "step": 106590
+    },
+    {
+      "epoch": 7.2424921864383744,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.734423834760158e-06,
+      "loss": 0.7657,
+      "step": 106595
+    },
+    {
+      "epoch": 7.2428319065090365,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.732300584318522e-06,
+      "loss": 0.8366,
+      "step": 106600
+    },
+    {
+      "epoch": 7.243171626579699,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.730177333876886e-06,
+      "loss": 0.9171,
+      "step": 106605
+    },
+    {
+      "epoch": 7.24351134665036,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.72805408343525e-06,
+      "loss": 0.906,
+      "step": 106610
+    },
+    {
+      "epoch": 7.243851066721022,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.725930832993614e-06,
+      "loss": 0.8338,
+      "step": 106615
+    },
+    {
+      "epoch": 7.244190786791684,
+      "grad_norm": 7.875,
+      "learning_rate": 4.723807582551978e-06,
+      "loss": 0.7538,
+      "step": 106620
+    },
+    {
+      "epoch": 7.244530506862345,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.721684332110341e-06,
+      "loss": 0.867,
+      "step": 106625
+    },
+    {
+      "epoch": 7.244870226933007,
+      "grad_norm": 7.125,
+      "learning_rate": 4.719561081668706e-06,
+      "loss": 0.9867,
+      "step": 106630
+    },
+    {
+      "epoch": 7.245209947003669,
+      "grad_norm": 8.125,
+      "learning_rate": 4.71743783122707e-06,
+      "loss": 0.88,
+      "step": 106635
+    },
+    {
+      "epoch": 7.2455496670743305,
+      "grad_norm": 6.875,
+      "learning_rate": 4.715314580785433e-06,
+      "loss": 0.8352,
+      "step": 106640
+    },
+    {
+      "epoch": 7.2458893871449925,
+      "grad_norm": 9.0,
+      "learning_rate": 4.713191330343797e-06,
+      "loss": 0.8628,
+      "step": 106645
+    },
+    {
+      "epoch": 7.246229107215655,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.711068079902161e-06,
+      "loss": 0.9431,
+      "step": 106650
+    },
+    {
+      "epoch": 7.246568827286316,
+      "grad_norm": 7.375,
+      "learning_rate": 4.708944829460525e-06,
+      "loss": 0.6881,
+      "step": 106655
+    },
+    {
+      "epoch": 7.246908547356978,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.706821579018889e-06,
+      "loss": 0.8065,
+      "step": 106660
+    },
+    {
+      "epoch": 7.24724826742764,
+      "grad_norm": 8.25,
+      "learning_rate": 4.704698328577253e-06,
+      "loss": 0.8633,
+      "step": 106665
+    },
+    {
+      "epoch": 7.247587987498301,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.702575078135617e-06,
+      "loss": 0.7336,
+      "step": 106670
+    },
+    {
+      "epoch": 7.247927707568963,
+      "grad_norm": 8.375,
+      "learning_rate": 4.70045182769398e-06,
+      "loss": 0.8498,
+      "step": 106675
+    },
+    {
+      "epoch": 7.248267427639625,
+      "grad_norm": 8.125,
+      "learning_rate": 4.698328577252344e-06,
+      "loss": 0.7638,
+      "step": 106680
+    },
+    {
+      "epoch": 7.2486071477102865,
+      "grad_norm": 8.625,
+      "learning_rate": 4.696205326810708e-06,
+      "loss": 0.8291,
+      "step": 106685
+    },
+    {
+      "epoch": 7.2489468677809485,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.694082076369072e-06,
+      "loss": 0.7748,
+      "step": 106690
+    },
+    {
+      "epoch": 7.249286587851611,
+      "grad_norm": 8.625,
+      "learning_rate": 4.691958825927436e-06,
+      "loss": 0.8414,
+      "step": 106695
+    },
+    {
+      "epoch": 7.249626307922272,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.6898355754858e-06,
+      "loss": 0.8713,
+      "step": 106700
+    },
+    {
+      "epoch": 7.249966027992934,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.687712325044164e-06,
+      "loss": 0.8024,
+      "step": 106705
+    },
+    {
+      "epoch": 7.250305748063596,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.685589074602527e-06,
+      "loss": 0.9082,
+      "step": 106710
+    },
+    {
+      "epoch": 7.250645468134257,
+      "grad_norm": 7.625,
+      "learning_rate": 4.683465824160892e-06,
+      "loss": 0.8377,
+      "step": 106715
+    },
+    {
+      "epoch": 7.250985188204919,
+      "grad_norm": 7.0,
+      "learning_rate": 4.681342573719256e-06,
+      "loss": 0.8563,
+      "step": 106720
+    },
+    {
+      "epoch": 7.251324908275581,
+      "grad_norm": 8.0,
+      "learning_rate": 4.679219323277619e-06,
+      "loss": 0.8781,
+      "step": 106725
+    },
+    {
+      "epoch": 7.2516646283462425,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.677096072835983e-06,
+      "loss": 0.7809,
+      "step": 106730
+    },
+    {
+      "epoch": 7.2520043484169046,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.674972822394347e-06,
+      "loss": 0.7046,
+      "step": 106735
+    },
+    {
+      "epoch": 7.252344068487567,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.672849571952711e-06,
+      "loss": 0.8843,
+      "step": 106740
+    },
+    {
+      "epoch": 7.252683788558228,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.670726321511075e-06,
+      "loss": 0.7387,
+      "step": 106745
+    },
+    {
+      "epoch": 7.25302350862889,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.668603071069439e-06,
+      "loss": 0.8849,
+      "step": 106750
+    },
+    {
+      "epoch": 7.253363228699552,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.666479820627803e-06,
+      "loss": 0.7519,
+      "step": 106755
+    },
+    {
+      "epoch": 7.253702948770213,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.664356570186166e-06,
+      "loss": 0.8575,
+      "step": 106760
+    },
+    {
+      "epoch": 7.254042668840875,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.66223331974453e-06,
+      "loss": 0.7852,
+      "step": 106765
+    },
+    {
+      "epoch": 7.254382388911537,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.660110069302895e-06,
+      "loss": 0.7975,
+      "step": 106770
+    },
+    {
+      "epoch": 7.2547221089821985,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.657986818861258e-06,
+      "loss": 0.8578,
+      "step": 106775
+    },
+    {
+      "epoch": 7.255061829052861,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.655863568419622e-06,
+      "loss": 0.7863,
+      "step": 106780
+    },
+    {
+      "epoch": 7.255401549123523,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.653740317977986e-06,
+      "loss": 0.7893,
+      "step": 106785
+    },
+    {
+      "epoch": 7.255741269194184,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.65161706753635e-06,
+      "loss": 0.8513,
+      "step": 106790
+    },
+    {
+      "epoch": 7.256080989264846,
+      "grad_norm": 8.125,
+      "learning_rate": 4.649493817094714e-06,
+      "loss": 0.8564,
+      "step": 106795
+    },
+    {
+      "epoch": 7.256420709335508,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.647370566653078e-06,
+      "loss": 0.8138,
+      "step": 106800
+    },
+    {
+      "epoch": 7.256760429406169,
+      "grad_norm": 6.625,
+      "learning_rate": 4.645247316211442e-06,
+      "loss": 0.7706,
+      "step": 106805
+    },
+    {
+      "epoch": 7.257100149476831,
+      "grad_norm": 9.125,
+      "learning_rate": 4.6431240657698056e-06,
+      "loss": 0.8979,
+      "step": 106810
+    },
+    {
+      "epoch": 7.257439869547493,
+      "grad_norm": 7.21875,
+      "learning_rate": 4.64100081532817e-06,
+      "loss": 0.8272,
+      "step": 106815
+    },
+    {
+      "epoch": 7.2577795896181545,
+      "grad_norm": 8.0,
+      "learning_rate": 4.638877564886534e-06,
+      "loss": 0.8126,
+      "step": 106820
+    },
+    {
+      "epoch": 7.258119309688817,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.636754314444898e-06,
+      "loss": 0.8267,
+      "step": 106825
+    },
+    {
+      "epoch": 7.258459029759479,
+      "grad_norm": 8.875,
+      "learning_rate": 4.634631064003262e-06,
+      "loss": 0.8345,
+      "step": 106830
+    },
+    {
+      "epoch": 7.25879874983014,
+      "grad_norm": 7.125,
+      "learning_rate": 4.632507813561626e-06,
+      "loss": 0.9552,
+      "step": 106835
+    },
+    {
+      "epoch": 7.259138469900802,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.63038456311999e-06,
+      "loss": 0.8497,
+      "step": 106840
+    },
+    {
+      "epoch": 7.259478189971463,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.628261312678353e-06,
+      "loss": 0.887,
+      "step": 106845
+    },
+    {
+      "epoch": 7.259817910042125,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.626138062236717e-06,
+      "loss": 0.8633,
+      "step": 106850
+    },
+    {
+      "epoch": 7.260157630112787,
+      "grad_norm": 8.25,
+      "learning_rate": 4.624014811795082e-06,
+      "loss": 0.9075,
+      "step": 106855
+    },
+    {
+      "epoch": 7.260497350183448,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.621891561353445e-06,
+      "loss": 0.8025,
+      "step": 106860
+    },
+    {
+      "epoch": 7.2608370702541105,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.619768310911809e-06,
+      "loss": 0.7251,
+      "step": 106865
+    },
+    {
+      "epoch": 7.261176790324773,
+      "grad_norm": 8.5,
+      "learning_rate": 4.617645060470173e-06,
+      "loss": 0.8682,
+      "step": 106870
+    },
+    {
+      "epoch": 7.261516510395434,
+      "grad_norm": 8.375,
+      "learning_rate": 4.615521810028537e-06,
+      "loss": 0.8799,
+      "step": 106875
+    },
+    {
+      "epoch": 7.261856230466096,
+      "grad_norm": 7.875,
+      "learning_rate": 4.613398559586901e-06,
+      "loss": 0.9347,
+      "step": 106880
+    },
+    {
+      "epoch": 7.262195950536758,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.611275309145265e-06,
+      "loss": 0.9662,
+      "step": 106885
+    },
+    {
+      "epoch": 7.262535670607419,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.609152058703629e-06,
+      "loss": 0.7376,
+      "step": 106890
+    },
+    {
+      "epoch": 7.262875390678081,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.607028808261992e-06,
+      "loss": 0.8651,
+      "step": 106895
+    },
+    {
+      "epoch": 7.263215110748743,
+      "grad_norm": 8.625,
+      "learning_rate": 4.604905557820356e-06,
+      "loss": 0.8391,
+      "step": 106900
+    },
+    {
+      "epoch": 7.2635548308194045,
+      "grad_norm": 8.625,
+      "learning_rate": 4.60278230737872e-06,
+      "loss": 0.7974,
+      "step": 106905
+    },
+    {
+      "epoch": 7.2638945508900665,
+      "grad_norm": 8.125,
+      "learning_rate": 4.600659056937084e-06,
+      "loss": 0.8634,
+      "step": 106910
+    },
+    {
+      "epoch": 7.264234270960729,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.598535806495448e-06,
+      "loss": 0.8272,
+      "step": 106915
+    },
+    {
+      "epoch": 7.26457399103139,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.596412556053812e-06,
+      "loss": 0.7652,
+      "step": 106920
+    },
+    {
+      "epoch": 7.264913711102052,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.594289305612176e-06,
+      "loss": 0.8399,
+      "step": 106925
+    },
+    {
+      "epoch": 7.265253431172714,
+      "grad_norm": 8.0,
+      "learning_rate": 4.592166055170539e-06,
+      "loss": 0.7481,
+      "step": 106930
+    },
+    {
+      "epoch": 7.265593151243375,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.590042804728904e-06,
+      "loss": 0.8203,
+      "step": 106935
+    },
+    {
+      "epoch": 7.265932871314037,
+      "grad_norm": 8.625,
+      "learning_rate": 4.587919554287268e-06,
+      "loss": 0.8738,
+      "step": 106940
+    },
+    {
+      "epoch": 7.266272591384699,
+      "grad_norm": 8.25,
+      "learning_rate": 4.585796303845631e-06,
+      "loss": 0.8468,
+      "step": 106945
+    },
+    {
+      "epoch": 7.2666123114553605,
+      "grad_norm": 8.0,
+      "learning_rate": 4.583673053403995e-06,
+      "loss": 0.8665,
+      "step": 106950
+    },
+    {
+      "epoch": 7.2669520315260225,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.581549802962359e-06,
+      "loss": 0.8308,
+      "step": 106955
+    },
+    {
+      "epoch": 7.267291751596685,
+      "grad_norm": 8.125,
+      "learning_rate": 4.579426552520723e-06,
+      "loss": 0.88,
+      "step": 106960
+    },
+    {
+      "epoch": 7.267631471667346,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.577303302079087e-06,
+      "loss": 0.7709,
+      "step": 106965
+    },
+    {
+      "epoch": 7.267971191738008,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.575180051637451e-06,
+      "loss": 0.8448,
+      "step": 106970
+    },
+    {
+      "epoch": 7.26831091180867,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.573056801195815e-06,
+      "loss": 0.8761,
+      "step": 106975
+    },
+    {
+      "epoch": 7.268650631879331,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.570933550754178e-06,
+      "loss": 0.8357,
+      "step": 106980
+    },
+    {
+      "epoch": 7.268990351949993,
+      "grad_norm": 7.21875,
+      "learning_rate": 4.568810300312542e-06,
+      "loss": 0.7904,
+      "step": 106985
+    },
+    {
+      "epoch": 7.269330072020655,
+      "grad_norm": 8.375,
+      "learning_rate": 4.566687049870907e-06,
+      "loss": 0.8919,
+      "step": 106990
+    },
+    {
+      "epoch": 7.2696697920913165,
+      "grad_norm": 8.25,
+      "learning_rate": 4.56456379942927e-06,
+      "loss": 0.85,
+      "step": 106995
+    },
+    {
+      "epoch": 7.2700095121619785,
+      "grad_norm": 8.0,
+      "learning_rate": 4.562440548987634e-06,
+      "loss": 0.8386,
+      "step": 107000
+    },
+    {
+      "epoch": 7.270349232232641,
+      "grad_norm": 8.5,
+      "learning_rate": 4.560317298545998e-06,
+      "loss": 0.8123,
+      "step": 107005
+    },
+    {
+      "epoch": 7.270688952303302,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.558194048104362e-06,
+      "loss": 0.9054,
+      "step": 107010
+    },
+    {
+      "epoch": 7.271028672373964,
+      "grad_norm": 7.5,
+      "learning_rate": 4.556070797662726e-06,
+      "loss": 1.035,
+      "step": 107015
+    },
+    {
+      "epoch": 7.271368392444626,
+      "grad_norm": 8.875,
+      "learning_rate": 4.55394754722109e-06,
+      "loss": 0.8199,
+      "step": 107020
+    },
+    {
+      "epoch": 7.271708112515287,
+      "grad_norm": 7.25,
+      "learning_rate": 4.551824296779454e-06,
+      "loss": 0.823,
+      "step": 107025
+    },
+    {
+      "epoch": 7.272047832585949,
+      "grad_norm": 7.875,
+      "learning_rate": 4.5497010463378175e-06,
+      "loss": 0.7949,
+      "step": 107030
+    },
+    {
+      "epoch": 7.272387552656611,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.5475777958961815e-06,
+      "loss": 0.8077,
+      "step": 107035
+    },
+    {
+      "epoch": 7.2727272727272725,
+      "grad_norm": 8.625,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 0.8704,
+      "step": 107040
+    },
+    {
+      "epoch": 7.273066992797935,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.5433312950129095e-06,
+      "loss": 0.8615,
+      "step": 107045
+    },
+    {
+      "epoch": 7.273406712868597,
+      "grad_norm": 7.625,
+      "learning_rate": 4.5412080445712735e-06,
+      "loss": 0.8266,
+      "step": 107050
+    },
+    {
+      "epoch": 7.273746432939258,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.5390847941296375e-06,
+      "loss": 0.8393,
+      "step": 107055
+    },
+    {
+      "epoch": 7.27408615300992,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.5369615436880015e-06,
+      "loss": 0.8206,
+      "step": 107060
+    },
+    {
+      "epoch": 7.274425873080582,
+      "grad_norm": 8.125,
+      "learning_rate": 4.534838293246365e-06,
+      "loss": 0.8077,
+      "step": 107065
+    },
+    {
+      "epoch": 7.274765593151243,
+      "grad_norm": 8.125,
+      "learning_rate": 4.532715042804729e-06,
+      "loss": 0.8357,
+      "step": 107070
+    },
+    {
+      "epoch": 7.275105313221905,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.5305917923630935e-06,
+      "loss": 0.8291,
+      "step": 107075
+    },
+    {
+      "epoch": 7.275445033292567,
+      "grad_norm": 8.25,
+      "learning_rate": 4.528468541921457e-06,
+      "loss": 0.9503,
+      "step": 107080
+    },
+    {
+      "epoch": 7.2757847533632285,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.526345291479821e-06,
+      "loss": 0.868,
+      "step": 107085
+    },
+    {
+      "epoch": 7.276124473433891,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.524222041038185e-06,
+      "loss": 0.8651,
+      "step": 107090
+    },
+    {
+      "epoch": 7.276464193504553,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.522098790596549e-06,
+      "loss": 0.9309,
+      "step": 107095
+    },
+    {
+      "epoch": 7.276803913575214,
+      "grad_norm": 6.59375,
+      "learning_rate": 4.519975540154913e-06,
+      "loss": 0.897,
+      "step": 107100
+    },
+    {
+      "epoch": 7.277143633645876,
+      "grad_norm": 8.375,
+      "learning_rate": 4.517852289713277e-06,
+      "loss": 0.9054,
+      "step": 107105
+    },
+    {
+      "epoch": 7.277483353716538,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.515729039271641e-06,
+      "loss": 0.8296,
+      "step": 107110
+    },
+    {
+      "epoch": 7.277823073787199,
+      "grad_norm": 8.625,
+      "learning_rate": 4.513605788830004e-06,
+      "loss": 0.9506,
+      "step": 107115
+    },
+    {
+      "epoch": 7.278162793857861,
+      "grad_norm": 6.8125,
+      "learning_rate": 4.511482538388368e-06,
+      "loss": 0.8042,
+      "step": 107120
+    },
+    {
+      "epoch": 7.278502513928522,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.509359287946732e-06,
+      "loss": 0.8753,
+      "step": 107125
+    },
+    {
+      "epoch": 7.2788422339991845,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.507236037505097e-06,
+      "loss": 0.826,
+      "step": 107130
+    },
+    {
+      "epoch": 7.279181954069847,
+      "grad_norm": 7.5,
+      "learning_rate": 4.50511278706346e-06,
+      "loss": 0.7565,
+      "step": 107135
+    },
+    {
+      "epoch": 7.279521674140508,
+      "grad_norm": 8.0,
+      "learning_rate": 4.502989536621824e-06,
+      "loss": 0.8415,
+      "step": 107140
+    },
+    {
+      "epoch": 7.27986139421117,
+      "grad_norm": 6.21875,
+      "learning_rate": 4.500866286180188e-06,
+      "loss": 0.8479,
+      "step": 107145
+    },
+    {
+      "epoch": 7.280201114281832,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.498743035738551e-06,
+      "loss": 0.8907,
+      "step": 107150
+    },
+    {
+      "epoch": 7.280540834352493,
+      "grad_norm": 8.0,
+      "learning_rate": 4.496619785296916e-06,
+      "loss": 0.8135,
+      "step": 107155
+    },
+    {
+      "epoch": 7.280880554423155,
+      "grad_norm": 8.5,
+      "learning_rate": 4.49449653485528e-06,
+      "loss": 0.8137,
+      "step": 107160
+    },
+    {
+      "epoch": 7.281220274493817,
+      "grad_norm": 7.25,
+      "learning_rate": 4.492373284413643e-06,
+      "loss": 0.8589,
+      "step": 107165
+    },
+    {
+      "epoch": 7.2815599945644784,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.490250033972007e-06,
+      "loss": 0.787,
+      "step": 107170
+    },
+    {
+      "epoch": 7.2818997146351405,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.488126783530371e-06,
+      "loss": 0.8677,
+      "step": 107175
+    },
+    {
+      "epoch": 7.282239434705803,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.486003533088735e-06,
+      "loss": 0.8422,
+      "step": 107180
+    },
+    {
+      "epoch": 7.282579154776464,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.483880282647099e-06,
+      "loss": 0.8389,
+      "step": 107185
+    },
+    {
+      "epoch": 7.282918874847126,
+      "grad_norm": 7.875,
+      "learning_rate": 4.481757032205463e-06,
+      "loss": 0.8012,
+      "step": 107190
+    },
+    {
+      "epoch": 7.283258594917788,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.479633781763827e-06,
+      "loss": 0.8451,
+      "step": 107195
+    },
+    {
+      "epoch": 7.283598314988449,
+      "grad_norm": 8.25,
+      "learning_rate": 4.47751053132219e-06,
+      "loss": 0.8586,
+      "step": 107200
+    },
+    {
+      "epoch": 7.283938035059111,
+      "grad_norm": 7.75,
+      "learning_rate": 4.475387280880554e-06,
+      "loss": 0.8416,
+      "step": 107205
+    },
+    {
+      "epoch": 7.284277755129773,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.473264030438918e-06,
+      "loss": 0.781,
+      "step": 107210
+    },
+    {
+      "epoch": 7.2846174752004345,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.471140779997283e-06,
+      "loss": 0.8314,
+      "step": 107215
+    },
+    {
+      "epoch": 7.2849571952710965,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.469017529555646e-06,
+      "loss": 0.8355,
+      "step": 107220
+    },
+    {
+      "epoch": 7.285296915341759,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.46689427911401e-06,
+      "loss": 0.775,
+      "step": 107225
+    },
+    {
+      "epoch": 7.28563663541242,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.464771028672374e-06,
+      "loss": 0.9102,
+      "step": 107230
+    },
+    {
+      "epoch": 7.285976355483082,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.4626477782307374e-06,
+      "loss": 0.7593,
+      "step": 107235
+    },
+    {
+      "epoch": 7.286316075553744,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.460524527789102e-06,
+      "loss": 0.7851,
+      "step": 107240
+    },
+    {
+      "epoch": 7.286655795624405,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.458401277347466e-06,
+      "loss": 0.7055,
+      "step": 107245
+    },
+    {
+      "epoch": 7.286995515695067,
+      "grad_norm": 8.875,
+      "learning_rate": 4.4562780269058294e-06,
+      "loss": 0.8686,
+      "step": 107250
+    },
+    {
+      "epoch": 7.287335235765729,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.4541547764641934e-06,
+      "loss": 0.8338,
+      "step": 107255
+    },
+    {
+      "epoch": 7.2876749558363905,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.4520315260225575e-06,
+      "loss": 0.9106,
+      "step": 107260
+    },
+    {
+      "epoch": 7.2880146759070525,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.4499082755809215e-06,
+      "loss": 0.7249,
+      "step": 107265
+    },
+    {
+      "epoch": 7.288354395977715,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.4477850251392855e-06,
+      "loss": 0.8746,
+      "step": 107270
+    },
+    {
+      "epoch": 7.288694116048376,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.4456617746976495e-06,
+      "loss": 0.9182,
+      "step": 107275
+    },
+    {
+      "epoch": 7.289033836119038,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.4435385242560135e-06,
+      "loss": 0.8324,
+      "step": 107280
+    },
+    {
+      "epoch": 7.2893735561897,
+      "grad_norm": 6.875,
+      "learning_rate": 4.441415273814377e-06,
+      "loss": 0.8234,
+      "step": 107285
+    },
+    {
+      "epoch": 7.289713276260361,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.439292023372741e-06,
+      "loss": 0.9106,
+      "step": 107290
+    },
+    {
+      "epoch": 7.290052996331023,
+      "grad_norm": 7.875,
+      "learning_rate": 4.4371687729311055e-06,
+      "loss": 0.8607,
+      "step": 107295
+    },
+    {
+      "epoch": 7.290392716401685,
+      "grad_norm": 8.5,
+      "learning_rate": 4.4350455224894695e-06,
+      "loss": 0.8926,
+      "step": 107300
+    },
+    {
+      "epoch": 7.2907324364723465,
+      "grad_norm": 8.625,
+      "learning_rate": 4.432922272047833e-06,
+      "loss": 0.8647,
+      "step": 107305
+    },
+    {
+      "epoch": 7.2910721565430086,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.430799021606197e-06,
+      "loss": 0.869,
+      "step": 107310
+    },
+    {
+      "epoch": 7.291411876613671,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.428675771164561e-06,
+      "loss": 0.8271,
+      "step": 107315
+    },
+    {
+      "epoch": 7.291751596684332,
+      "grad_norm": 6.9375,
+      "learning_rate": 4.426552520722925e-06,
+      "loss": 0.8323,
+      "step": 107320
+    },
+    {
+      "epoch": 7.292091316754994,
+      "grad_norm": 8.125,
+      "learning_rate": 4.424429270281289e-06,
+      "loss": 0.8885,
+      "step": 107325
+    },
+    {
+      "epoch": 7.292431036825656,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.422306019839653e-06,
+      "loss": 0.8404,
+      "step": 107330
+    },
+    {
+      "epoch": 7.292770756896317,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.420182769398016e-06,
+      "loss": 0.8394,
+      "step": 107335
+    },
+    {
+      "epoch": 7.293110476966979,
+      "grad_norm": 8.75,
+      "learning_rate": 4.41805951895638e-06,
+      "loss": 0.8842,
+      "step": 107340
+    },
+    {
+      "epoch": 7.293450197037641,
+      "grad_norm": 8.125,
+      "learning_rate": 4.415936268514744e-06,
+      "loss": 0.8704,
+      "step": 107345
+    },
+    {
+      "epoch": 7.2937899171083025,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.413813018073109e-06,
+      "loss": 0.918,
+      "step": 107350
+    },
+    {
+      "epoch": 7.294129637178965,
+      "grad_norm": 7.75,
+      "learning_rate": 4.411689767631472e-06,
+      "loss": 0.8062,
+      "step": 107355
+    },
+    {
+      "epoch": 7.294469357249627,
+      "grad_norm": 8.875,
+      "learning_rate": 4.409566517189836e-06,
+      "loss": 0.8931,
+      "step": 107360
+    },
+    {
+      "epoch": 7.294809077320288,
+      "grad_norm": 8.0,
+      "learning_rate": 4.4074432667482e-06,
+      "loss": 0.8396,
+      "step": 107365
+    },
+    {
+      "epoch": 7.29514879739095,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.405320016306563e-06,
+      "loss": 0.781,
+      "step": 107370
+    },
+    {
+      "epoch": 7.295488517461612,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.403196765864928e-06,
+      "loss": 0.8272,
+      "step": 107375
+    },
+    {
+      "epoch": 7.295828237532273,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.401073515423292e-06,
+      "loss": 0.7606,
+      "step": 107380
+    },
+    {
+      "epoch": 7.296167957602935,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.398950264981656e-06,
+      "loss": 0.8716,
+      "step": 107385
+    },
+    {
+      "epoch": 7.296507677673597,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.396827014540019e-06,
+      "loss": 0.8507,
+      "step": 107390
+    },
+    {
+      "epoch": 7.2968473977442585,
+      "grad_norm": 8.5,
+      "learning_rate": 4.394703764098383e-06,
+      "loss": 0.7691,
+      "step": 107395
+    },
+    {
+      "epoch": 7.297187117814921,
+      "grad_norm": 7.21875,
+      "learning_rate": 4.392580513656747e-06,
+      "loss": 0.8625,
+      "step": 107400
+    },
+    {
+      "epoch": 7.297526837885583,
+      "grad_norm": 8.0,
+      "learning_rate": 4.390457263215111e-06,
+      "loss": 0.7956,
+      "step": 107405
+    },
+    {
+      "epoch": 7.297866557956244,
+      "grad_norm": 8.0,
+      "learning_rate": 4.388334012773475e-06,
+      "loss": 0.8761,
+      "step": 107410
+    },
+    {
+      "epoch": 7.298206278026906,
+      "grad_norm": 8.375,
+      "learning_rate": 4.386210762331839e-06,
+      "loss": 0.8862,
+      "step": 107415
+    },
+    {
+      "epoch": 7.298545998097568,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.384087511890202e-06,
+      "loss": 0.7748,
+      "step": 107420
+    },
+    {
+      "epoch": 7.298885718168229,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.381964261448566e-06,
+      "loss": 0.7221,
+      "step": 107425
+    },
+    {
+      "epoch": 7.299225438238891,
+      "grad_norm": 8.375,
+      "learning_rate": 4.37984101100693e-06,
+      "loss": 1.1921,
+      "step": 107430
+    },
+    {
+      "epoch": 7.299565158309553,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.377717760565295e-06,
+      "loss": 0.8336,
+      "step": 107435
+    },
+    {
+      "epoch": 7.2999048783802145,
+      "grad_norm": 8.125,
+      "learning_rate": 4.375594510123658e-06,
+      "loss": 0.814,
+      "step": 107440
+    },
+    {
+      "epoch": 7.300244598450877,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.373471259682022e-06,
+      "loss": 0.7932,
+      "step": 107445
+    },
+    {
+      "epoch": 7.300584318521539,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.371348009240386e-06,
+      "loss": 0.9053,
+      "step": 107450
+    },
+    {
+      "epoch": 7.3009240385922,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.369224758798749e-06,
+      "loss": 0.8333,
+      "step": 107455
+    },
+    {
+      "epoch": 7.301263758662862,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.367101508357114e-06,
+      "loss": 0.7717,
+      "step": 107460
+    },
+    {
+      "epoch": 7.301603478733524,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.364978257915478e-06,
+      "loss": 0.884,
+      "step": 107465
+    },
+    {
+      "epoch": 7.301943198804185,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.362855007473842e-06,
+      "loss": 0.8764,
+      "step": 107470
+    },
+    {
+      "epoch": 7.302282918874847,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.360731757032205e-06,
+      "loss": 0.7931,
+      "step": 107475
+    },
+    {
+      "epoch": 7.302622638945509,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.358608506590569e-06,
+      "loss": 0.8854,
+      "step": 107480
+    },
+    {
+      "epoch": 7.3029623590161705,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.356485256148933e-06,
+      "loss": 0.895,
+      "step": 107485
+    },
+    {
+      "epoch": 7.303302079086833,
+      "grad_norm": 7.28125,
+      "learning_rate": 4.354362005707297e-06,
+      "loss": 0.8755,
+      "step": 107490
+    },
+    {
+      "epoch": 7.303641799157495,
+      "grad_norm": 7.625,
+      "learning_rate": 4.352238755265661e-06,
+      "loss": 0.7943,
+      "step": 107495
+    },
+    {
+      "epoch": 7.303981519228156,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.350115504824025e-06,
+      "loss": 0.8619,
+      "step": 107500
+    },
+    {
+      "epoch": 7.304321239298818,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.3479922543823886e-06,
+      "loss": 0.9553,
+      "step": 107505
+    },
+    {
+      "epoch": 7.30466095936948,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.3458690039407526e-06,
+      "loss": 0.8276,
+      "step": 107510
+    },
+    {
+      "epoch": 7.305000679440141,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.343745753499117e-06,
+      "loss": 0.7992,
+      "step": 107515
+    },
+    {
+      "epoch": 7.305340399510803,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.341622503057481e-06,
+      "loss": 0.9567,
+      "step": 107520
+    },
+    {
+      "epoch": 7.3056801195814645,
+      "grad_norm": 7.5,
+      "learning_rate": 4.3394992526158446e-06,
+      "loss": 0.7903,
+      "step": 107525
+    },
+    {
+      "epoch": 7.3060198396521265,
+      "grad_norm": 8.5,
+      "learning_rate": 4.337376002174209e-06,
+      "loss": 0.8249,
+      "step": 107530
+    },
+    {
+      "epoch": 7.306359559722789,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.335252751732573e-06,
+      "loss": 0.8336,
+      "step": 107535
+    },
+    {
+      "epoch": 7.30669927979345,
+      "grad_norm": 8.375,
+      "learning_rate": 4.333129501290937e-06,
+      "loss": 0.7611,
+      "step": 107540
+    },
+    {
+      "epoch": 7.307038999864112,
+      "grad_norm": 8.125,
+      "learning_rate": 4.331006250849301e-06,
+      "loss": 0.9009,
+      "step": 107545
+    },
+    {
+      "epoch": 7.307378719934774,
+      "grad_norm": 7.5,
+      "learning_rate": 4.328883000407665e-06,
+      "loss": 0.8216,
+      "step": 107550
+    },
+    {
+      "epoch": 7.307718440005435,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.326759749966029e-06,
+      "loss": 0.9305,
+      "step": 107555
+    },
+    {
+      "epoch": 7.308058160076097,
+      "grad_norm": 7.625,
+      "learning_rate": 4.324636499524392e-06,
+      "loss": 0.9148,
+      "step": 107560
+    },
+    {
+      "epoch": 7.308397880146759,
+      "grad_norm": 8.375,
+      "learning_rate": 4.322513249082756e-06,
+      "loss": 0.8379,
+      "step": 107565
+    },
+    {
+      "epoch": 7.3087376002174205,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.32038999864112e-06,
+      "loss": 0.918,
+      "step": 107570
+    },
+    {
+      "epoch": 7.3090773202880825,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.318266748199484e-06,
+      "loss": 0.8569,
+      "step": 107575
+    },
+    {
+      "epoch": 7.309417040358745,
+      "grad_norm": 8.0,
+      "learning_rate": 4.316143497757848e-06,
+      "loss": 0.8026,
+      "step": 107580
+    },
+    {
+      "epoch": 7.309756760429406,
+      "grad_norm": 8.125,
+      "learning_rate": 4.314020247316212e-06,
+      "loss": 0.8559,
+      "step": 107585
+    },
+    {
+      "epoch": 7.310096480500068,
+      "grad_norm": 8.75,
+      "learning_rate": 4.311896996874575e-06,
+      "loss": 0.867,
+      "step": 107590
+    },
+    {
+      "epoch": 7.31043620057073,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.309773746432939e-06,
+      "loss": 0.8644,
+      "step": 107595
+    },
+    {
+      "epoch": 7.310775920641391,
+      "grad_norm": 8.5,
+      "learning_rate": 4.307650495991304e-06,
+      "loss": 0.877,
+      "step": 107600
+    },
+    {
+      "epoch": 7.311115640712053,
+      "grad_norm": 8.875,
+      "learning_rate": 4.305527245549668e-06,
+      "loss": 0.8473,
+      "step": 107605
+    },
+    {
+      "epoch": 7.311455360782715,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.303403995108031e-06,
+      "loss": 0.8942,
+      "step": 107610
+    },
+    {
+      "epoch": 7.3117950808533765,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.301280744666395e-06,
+      "loss": 0.9095,
+      "step": 107615
+    },
+    {
+      "epoch": 7.312134800924039,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.299157494224759e-06,
+      "loss": 0.9276,
+      "step": 107620
+    },
+    {
+      "epoch": 7.312474520994701,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.297034243783123e-06,
+      "loss": 0.8407,
+      "step": 107625
+    },
+    {
+      "epoch": 7.312814241065362,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.294910993341487e-06,
+      "loss": 0.9612,
+      "step": 107630
+    },
+    {
+      "epoch": 7.313153961136024,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.292787742899851e-06,
+      "loss": 0.7667,
+      "step": 107635
+    },
+    {
+      "epoch": 7.313493681206686,
+      "grad_norm": 8.75,
+      "learning_rate": 4.290664492458215e-06,
+      "loss": 0.8656,
+      "step": 107640
+    },
+    {
+      "epoch": 7.313833401277347,
+      "grad_norm": 8.875,
+      "learning_rate": 4.288541242016578e-06,
+      "loss": 0.8938,
+      "step": 107645
+    },
+    {
+      "epoch": 7.314173121348009,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.286417991574942e-06,
+      "loss": 0.8056,
+      "step": 107650
+    },
+    {
+      "epoch": 7.314512841418671,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.284294741133307e-06,
+      "loss": 0.808,
+      "step": 107655
+    },
+    {
+      "epoch": 7.3148525614893325,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.28217149069167e-06,
+      "loss": 0.8146,
+      "step": 107660
+    },
+    {
+      "epoch": 7.315192281559995,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.280048240250034e-06,
+      "loss": 0.8068,
+      "step": 107665
+    },
+    {
+      "epoch": 7.315532001630657,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.277924989808398e-06,
+      "loss": 0.813,
+      "step": 107670
+    },
+    {
+      "epoch": 7.315871721701318,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.275801739366761e-06,
+      "loss": 0.9397,
+      "step": 107675
+    },
+    {
+      "epoch": 7.31621144177198,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.273678488925126e-06,
+      "loss": 0.8842,
+      "step": 107680
+    },
+    {
+      "epoch": 7.316551161842642,
+      "grad_norm": 7.75,
+      "learning_rate": 4.27155523848349e-06,
+      "loss": 0.8801,
+      "step": 107685
+    },
+    {
+      "epoch": 7.316890881913303,
+      "grad_norm": 8.625,
+      "learning_rate": 4.269431988041854e-06,
+      "loss": 0.8496,
+      "step": 107690
+    },
+    {
+      "epoch": 7.317230601983965,
+      "grad_norm": 7.125,
+      "learning_rate": 4.267308737600217e-06,
+      "loss": 0.7168,
+      "step": 107695
+    },
+    {
+      "epoch": 7.317570322054627,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.265185487158581e-06,
+      "loss": 0.8015,
+      "step": 107700
+    },
+    {
+      "epoch": 7.3179100421252885,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.263062236716945e-06,
+      "loss": 0.8144,
+      "step": 107705
+    },
+    {
+      "epoch": 7.318249762195951,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.260938986275309e-06,
+      "loss": 0.7748,
+      "step": 107710
+    },
+    {
+      "epoch": 7.318589482266613,
+      "grad_norm": 8.125,
+      "learning_rate": 4.258815735833673e-06,
+      "loss": 0.805,
+      "step": 107715
+    },
+    {
+      "epoch": 7.318929202337274,
+      "grad_norm": 9.0625,
+      "learning_rate": 4.256692485392037e-06,
+      "loss": 0.8997,
+      "step": 107720
+    },
+    {
+      "epoch": 7.319268922407936,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.254569234950401e-06,
+      "loss": 0.8148,
+      "step": 107725
+    },
+    {
+      "epoch": 7.319608642478598,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.2524459845087645e-06,
+      "loss": 0.7799,
+      "step": 107730
+    },
+    {
+      "epoch": 7.319948362549259,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.250322734067129e-06,
+      "loss": 0.871,
+      "step": 107735
+    },
+    {
+      "epoch": 7.320288082619921,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.248199483625493e-06,
+      "loss": 0.8244,
+      "step": 107740
+    },
+    {
+      "epoch": 7.320627802690583,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.2460762331838565e-06,
+      "loss": 0.8933,
+      "step": 107745
+    },
+    {
+      "epoch": 7.3209675227612445,
+      "grad_norm": 9.125,
+      "learning_rate": 4.2439529827422205e-06,
+      "loss": 0.8678,
+      "step": 107750
+    },
+    {
+      "epoch": 7.321307242831907,
+      "grad_norm": 6.90625,
+      "learning_rate": 4.2418297323005845e-06,
+      "loss": 0.7698,
+      "step": 107755
+    },
+    {
+      "epoch": 7.321646962902569,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.2397064818589485e-06,
+      "loss": 0.844,
+      "step": 107760
+    },
+    {
+      "epoch": 7.32198668297323,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.2375832314173125e-06,
+      "loss": 0.8523,
+      "step": 107765
+    },
+    {
+      "epoch": 7.322326403043892,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.2354599809756765e-06,
+      "loss": 0.8966,
+      "step": 107770
+    },
+    {
+      "epoch": 7.322666123114554,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.2333367305340405e-06,
+      "loss": 0.7781,
+      "step": 107775
+    },
+    {
+      "epoch": 7.323005843185215,
+      "grad_norm": 7.125,
+      "learning_rate": 4.231213480092404e-06,
+      "loss": 0.802,
+      "step": 107780
+    },
+    {
+      "epoch": 7.323345563255877,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.229090229650768e-06,
+      "loss": 0.9348,
+      "step": 107785
+    },
+    {
+      "epoch": 7.323685283326539,
+      "grad_norm": 8.125,
+      "learning_rate": 4.226966979209132e-06,
+      "loss": 0.8211,
+      "step": 107790
+    },
+    {
+      "epoch": 7.3240250033972005,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.224843728767496e-06,
+      "loss": 0.8942,
+      "step": 107795
+    },
+    {
+      "epoch": 7.324364723467863,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.22272047832586e-06,
+      "loss": 0.7993,
+      "step": 107800
+    },
+    {
+      "epoch": 7.324704443538524,
+      "grad_norm": 8.0,
+      "learning_rate": 4.220597227884224e-06,
+      "loss": 0.7554,
+      "step": 107805
+    },
+    {
+      "epoch": 7.325044163609186,
+      "grad_norm": 7.625,
+      "learning_rate": 4.218473977442588e-06,
+      "loss": 0.8548,
+      "step": 107810
+    },
+    {
+      "epoch": 7.325383883679848,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.216350727000951e-06,
+      "loss": 0.8272,
+      "step": 107815
+    },
+    {
+      "epoch": 7.325723603750509,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.214227476559316e-06,
+      "loss": 0.8039,
+      "step": 107820
+    },
+    {
+      "epoch": 7.326063323821171,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.21210422611768e-06,
+      "loss": 0.8063,
+      "step": 107825
+    },
+    {
+      "epoch": 7.326403043891833,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.209980975676043e-06,
+      "loss": 0.8972,
+      "step": 107830
+    },
+    {
+      "epoch": 7.3267427639624945,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.207857725234407e-06,
+      "loss": 0.7499,
+      "step": 107835
+    },
+    {
+      "epoch": 7.3270824840331565,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.205734474792771e-06,
+      "loss": 0.8342,
+      "step": 107840
+    },
+    {
+      "epoch": 7.327422204103819,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.203611224351135e-06,
+      "loss": 0.8417,
+      "step": 107845
+    },
+    {
+      "epoch": 7.32776192417448,
+      "grad_norm": 7.125,
+      "learning_rate": 4.201487973909499e-06,
+      "loss": 0.7809,
+      "step": 107850
+    },
+    {
+      "epoch": 7.328101644245142,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.199364723467863e-06,
+      "loss": 0.8974,
+      "step": 107855
+    },
+    {
+      "epoch": 7.328441364315804,
+      "grad_norm": 7.46875,
+      "learning_rate": 4.197241473026227e-06,
+      "loss": 0.9101,
+      "step": 107860
+    },
+    {
+      "epoch": 7.328781084386465,
+      "grad_norm": 8.125,
+      "learning_rate": 4.19511822258459e-06,
+      "loss": 0.8068,
+      "step": 107865
+    },
+    {
+      "epoch": 7.329120804457127,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.192994972142954e-06,
+      "loss": 0.8538,
+      "step": 107870
+    },
+    {
+      "epoch": 7.329460524527789,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.190871721701319e-06,
+      "loss": 0.852,
+      "step": 107875
+    },
+    {
+      "epoch": 7.3298002445984505,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.188748471259682e-06,
+      "loss": 0.8503,
+      "step": 107880
+    },
+    {
+      "epoch": 7.3301399646691126,
+      "grad_norm": 7.875,
+      "learning_rate": 4.186625220818046e-06,
+      "loss": 0.804,
+      "step": 107885
+    },
+    {
+      "epoch": 7.330479684739775,
+      "grad_norm": 8.375,
+      "learning_rate": 4.18450197037641e-06,
+      "loss": 0.897,
+      "step": 107890
+    },
+    {
+      "epoch": 7.330819404810436,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.182378719934774e-06,
+      "loss": 0.8288,
+      "step": 107895
+    },
+    {
+      "epoch": 7.331159124881098,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.180255469493138e-06,
+      "loss": 0.847,
+      "step": 107900
+    },
+    {
+      "epoch": 7.33149884495176,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.178132219051502e-06,
+      "loss": 0.8252,
+      "step": 107905
+    },
+    {
+      "epoch": 7.331838565022421,
+      "grad_norm": 7.625,
+      "learning_rate": 4.176008968609866e-06,
+      "loss": 0.85,
+      "step": 107910
+    },
+    {
+      "epoch": 7.332178285093083,
+      "grad_norm": 7.59375,
+      "learning_rate": 4.173885718168229e-06,
+      "loss": 0.8662,
+      "step": 107915
+    },
+    {
+      "epoch": 7.332518005163745,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.171762467726593e-06,
+      "loss": 0.8142,
+      "step": 107920
+    },
+    {
+      "epoch": 7.3328577252344065,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.169639217284957e-06,
+      "loss": 0.8403,
+      "step": 107925
+    },
+    {
+      "epoch": 7.333197445305069,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.167515966843321e-06,
+      "loss": 0.8407,
+      "step": 107930
+    },
+    {
+      "epoch": 7.333537165375731,
+      "grad_norm": 8.5,
+      "learning_rate": 4.165392716401685e-06,
+      "loss": 0.8952,
+      "step": 107935
+    },
+    {
+      "epoch": 7.333876885446392,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.163269465960049e-06,
+      "loss": 0.8504,
+      "step": 107940
+    },
+    {
+      "epoch": 7.334216605517054,
+      "grad_norm": 8.625,
+      "learning_rate": 4.161146215518413e-06,
+      "loss": 0.7941,
+      "step": 107945
+    },
+    {
+      "epoch": 7.334556325587716,
+      "grad_norm": 8.5,
+      "learning_rate": 4.1590229650767764e-06,
+      "loss": 0.8505,
+      "step": 107950
+    },
+    {
+      "epoch": 7.334896045658377,
+      "grad_norm": 8.875,
+      "learning_rate": 4.1568997146351404e-06,
+      "loss": 0.8071,
+      "step": 107955
+    },
+    {
+      "epoch": 7.335235765729039,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.154776464193505e-06,
+      "loss": 0.9015,
+      "step": 107960
+    },
+    {
+      "epoch": 7.335575485799701,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.1526532137518684e-06,
+      "loss": 0.7035,
+      "step": 107965
+    },
+    {
+      "epoch": 7.3359152058703625,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.1505299633102324e-06,
+      "loss": 0.8211,
+      "step": 107970
+    },
+    {
+      "epoch": 7.336254925941025,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.1484067128685965e-06,
+      "loss": 0.7933,
+      "step": 107975
+    },
+    {
+      "epoch": 7.336594646011687,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.1462834624269605e-06,
+      "loss": 0.8838,
+      "step": 107980
+    },
+    {
+      "epoch": 7.336934366082348,
+      "grad_norm": 7.0,
+      "learning_rate": 4.1441602119853245e-06,
+      "loss": 0.933,
+      "step": 107985
+    },
+    {
+      "epoch": 7.33727408615301,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.1420369615436885e-06,
+      "loss": 0.8365,
+      "step": 107990
+    },
+    {
+      "epoch": 7.337613806223672,
+      "grad_norm": 7.625,
+      "learning_rate": 4.1399137111020525e-06,
+      "loss": 0.8803,
+      "step": 107995
+    },
+    {
+      "epoch": 7.337953526294333,
+      "grad_norm": 8.0,
+      "learning_rate": 4.137790460660416e-06,
+      "loss": 0.8794,
+      "step": 108000
+    },
+    {
+      "epoch": 7.338293246364995,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.13566721021878e-06,
+      "loss": 0.856,
+      "step": 108005
+    },
+    {
+      "epoch": 7.338632966435657,
+      "grad_norm": 8.75,
+      "learning_rate": 4.133543959777144e-06,
+      "loss": 0.762,
+      "step": 108010
+    },
+    {
+      "epoch": 7.3389726865063185,
+      "grad_norm": 7.5,
+      "learning_rate": 4.131420709335508e-06,
+      "loss": 0.8505,
+      "step": 108015
+    },
+    {
+      "epoch": 7.339312406576981,
+      "grad_norm": 8.9375,
+      "learning_rate": 4.129297458893872e-06,
+      "loss": 0.9096,
+      "step": 108020
+    },
+    {
+      "epoch": 7.339652126647643,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.127174208452236e-06,
+      "loss": 0.9247,
+      "step": 108025
+    },
+    {
+      "epoch": 7.339991846718304,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.1250509580106e-06,
+      "loss": 0.8619,
+      "step": 108030
+    },
+    {
+      "epoch": 7.340331566788966,
+      "grad_norm": 8.5,
+      "learning_rate": 4.122927707568963e-06,
+      "loss": 0.8218,
+      "step": 108035
+    },
+    {
+      "epoch": 7.340671286859628,
+      "grad_norm": 7.4375,
+      "learning_rate": 4.120804457127328e-06,
+      "loss": 0.9483,
+      "step": 108040
+    },
+    {
+      "epoch": 7.341011006930289,
+      "grad_norm": 8.875,
+      "learning_rate": 4.118681206685692e-06,
+      "loss": 0.8563,
+      "step": 108045
+    },
+    {
+      "epoch": 7.341350727000951,
+      "grad_norm": 7.65625,
+      "learning_rate": 4.116557956244055e-06,
+      "loss": 0.9147,
+      "step": 108050
+    },
+    {
+      "epoch": 7.341690447071613,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.114434705802419e-06,
+      "loss": 0.8345,
+      "step": 108055
+    },
+    {
+      "epoch": 7.3420301671422745,
+      "grad_norm": 7.03125,
+      "learning_rate": 4.112311455360783e-06,
+      "loss": 0.8394,
+      "step": 108060
+    },
+    {
+      "epoch": 7.342369887212937,
+      "grad_norm": 8.5625,
+      "learning_rate": 4.110188204919147e-06,
+      "loss": 0.8234,
+      "step": 108065
+    },
+    {
+      "epoch": 7.342709607283599,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.108064954477511e-06,
+      "loss": 0.8818,
+      "step": 108070
+    },
+    {
+      "epoch": 7.34304932735426,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.105941704035875e-06,
+      "loss": 0.7431,
+      "step": 108075
+    },
+    {
+      "epoch": 7.343389047424922,
+      "grad_norm": 7.8125,
+      "learning_rate": 4.103818453594239e-06,
+      "loss": 0.8343,
+      "step": 108080
+    },
+    {
+      "epoch": 7.343728767495584,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.101695203152602e-06,
+      "loss": 0.8934,
+      "step": 108085
+    },
+    {
+      "epoch": 7.344068487566245,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.099571952710966e-06,
+      "loss": 0.8819,
+      "step": 108090
+    },
+    {
+      "epoch": 7.344408207636907,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.097448702269331e-06,
+      "loss": 0.768,
+      "step": 108095
+    },
+    {
+      "epoch": 7.344747927707569,
+      "grad_norm": 8.0,
+      "learning_rate": 4.095325451827694e-06,
+      "loss": 0.8232,
+      "step": 108100
+    },
+    {
+      "epoch": 7.3450876477782305,
+      "grad_norm": 7.28125,
+      "learning_rate": 4.093202201386058e-06,
+      "loss": 0.852,
+      "step": 108105
+    },
+    {
+      "epoch": 7.345427367848893,
+      "grad_norm": 6.78125,
+      "learning_rate": 4.091078950944422e-06,
+      "loss": 0.8547,
+      "step": 108110
+    },
+    {
+      "epoch": 7.345767087919555,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.088955700502786e-06,
+      "loss": 0.7822,
+      "step": 108115
+    },
+    {
+      "epoch": 7.346106807990216,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.08683245006115e-06,
+      "loss": 0.7494,
+      "step": 108120
+    },
+    {
+      "epoch": 7.346446528060878,
+      "grad_norm": 8.75,
+      "learning_rate": 4.084709199619514e-06,
+      "loss": 0.8535,
+      "step": 108125
+    },
+    {
+      "epoch": 7.34678624813154,
+      "grad_norm": 8.5,
+      "learning_rate": 4.082585949177878e-06,
+      "loss": 0.7814,
+      "step": 108130
+    },
+    {
+      "epoch": 7.347125968202201,
+      "grad_norm": 7.875,
+      "learning_rate": 4.080462698736241e-06,
+      "loss": 0.8275,
+      "step": 108135
+    },
+    {
+      "epoch": 7.347465688272863,
+      "grad_norm": 7.53125,
+      "learning_rate": 4.078339448294605e-06,
+      "loss": 0.8388,
+      "step": 108140
+    },
+    {
+      "epoch": 7.347805408343525,
+      "grad_norm": 8.0,
+      "learning_rate": 4.076216197852969e-06,
+      "loss": 0.876,
+      "step": 108145
+    },
+    {
+      "epoch": 7.3481451284141865,
+      "grad_norm": 8.25,
+      "learning_rate": 4.074092947411333e-06,
+      "loss": 0.8414,
+      "step": 108150
+    },
+    {
+      "epoch": 7.348484848484849,
+      "grad_norm": 6.84375,
+      "learning_rate": 4.071969696969697e-06,
+      "loss": 0.7971,
+      "step": 108155
+    },
+    {
+      "epoch": 7.348824568555511,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.069846446528061e-06,
+      "loss": 0.8449,
+      "step": 108160
+    },
+    {
+      "epoch": 7.349164288626172,
+      "grad_norm": 8.875,
+      "learning_rate": 4.067723196086425e-06,
+      "loss": 0.8488,
+      "step": 108165
+    },
+    {
+      "epoch": 7.349504008696834,
+      "grad_norm": 7.875,
+      "learning_rate": 4.065599945644788e-06,
+      "loss": 0.8066,
+      "step": 108170
+    },
+    {
+      "epoch": 7.349843728767496,
+      "grad_norm": 8.125,
+      "learning_rate": 4.063476695203152e-06,
+      "loss": 0.9399,
+      "step": 108175
+    },
+    {
+      "epoch": 7.350183448838157,
+      "grad_norm": 7.25,
+      "learning_rate": 4.061353444761517e-06,
+      "loss": 0.8599,
+      "step": 108180
+    },
+    {
+      "epoch": 7.350523168908819,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.05923019431988e-06,
+      "loss": 0.738,
+      "step": 108185
+    },
+    {
+      "epoch": 7.350862888979481,
+      "grad_norm": 8.25,
+      "learning_rate": 4.057106943878244e-06,
+      "loss": 0.8682,
+      "step": 108190
+    },
+    {
+      "epoch": 7.3512026090501426,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.054983693436608e-06,
+      "loss": 0.9236,
+      "step": 108195
+    },
+    {
+      "epoch": 7.351542329120805,
+      "grad_norm": 8.125,
+      "learning_rate": 4.052860442994972e-06,
+      "loss": 0.8177,
+      "step": 108200
+    },
+    {
+      "epoch": 7.351882049191466,
+      "grad_norm": 7.71875,
+      "learning_rate": 4.050737192553336e-06,
+      "loss": 0.7206,
+      "step": 108205
+    },
+    {
+      "epoch": 7.352221769262128,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.0486139421117e-06,
+      "loss": 0.8249,
+      "step": 108210
+    },
+    {
+      "epoch": 7.35256148933279,
+      "grad_norm": 9.1875,
+      "learning_rate": 4.046490691670064e-06,
+      "loss": 0.8912,
+      "step": 108215
+    },
+    {
+      "epoch": 7.352901209403451,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.0443674412284276e-06,
+      "loss": 0.8881,
+      "step": 108220
+    },
+    {
+      "epoch": 7.353240929474113,
+      "grad_norm": 7.5,
+      "learning_rate": 4.0422441907867916e-06,
+      "loss": 0.9328,
+      "step": 108225
+    },
+    {
+      "epoch": 7.353580649544775,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.0401209403451556e-06,
+      "loss": 0.9201,
+      "step": 108230
+    },
+    {
+      "epoch": 7.3539203696154365,
+      "grad_norm": 7.625,
+      "learning_rate": 4.03799768990352e-06,
+      "loss": 0.8041,
+      "step": 108235
+    },
+    {
+      "epoch": 7.354260089686099,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.0358744394618836e-06,
+      "loss": 0.8457,
+      "step": 108240
+    },
+    {
+      "epoch": 7.354599809756761,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.033751189020248e-06,
+      "loss": 0.8343,
+      "step": 108245
+    },
+    {
+      "epoch": 7.354939529827422,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.031627938578612e-06,
+      "loss": 0.8905,
+      "step": 108250
+    },
+    {
+      "epoch": 7.355279249898084,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.029504688136975e-06,
+      "loss": 0.8175,
+      "step": 108255
+    },
+    {
+      "epoch": 7.355618969968746,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.02738143769534e-06,
+      "loss": 0.8584,
+      "step": 108260
+    },
+    {
+      "epoch": 7.355958690039407,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.025258187253704e-06,
+      "loss": 0.8411,
+      "step": 108265
+    },
+    {
+      "epoch": 7.356298410110069,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.023134936812067e-06,
+      "loss": 0.8442,
+      "step": 108270
+    },
+    {
+      "epoch": 7.356638130180731,
+      "grad_norm": 8.25,
+      "learning_rate": 4.021011686370431e-06,
+      "loss": 0.8759,
+      "step": 108275
+    },
+    {
+      "epoch": 7.3569778502513925,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.018888435928795e-06,
+      "loss": 0.7956,
+      "step": 108280
+    },
+    {
+      "epoch": 7.357317570322055,
+      "grad_norm": 7.625,
+      "learning_rate": 4.016765185487159e-06,
+      "loss": 0.8949,
+      "step": 108285
+    },
+    {
+      "epoch": 7.357657290392717,
+      "grad_norm": 7.875,
+      "learning_rate": 4.014641935045523e-06,
+      "loss": 0.9011,
+      "step": 108290
+    },
+    {
+      "epoch": 7.357997010463378,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.012518684603887e-06,
+      "loss": 0.8185,
+      "step": 108295
+    },
+    {
+      "epoch": 7.35833673053404,
+      "grad_norm": 7.5625,
+      "learning_rate": 4.010395434162251e-06,
+      "loss": 0.8697,
+      "step": 108300
+    },
+    {
+      "epoch": 7.358676450604702,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.008272183720614e-06,
+      "loss": 0.8709,
+      "step": 108305
+    },
+    {
+      "epoch": 7.359016170675363,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.006148933278978e-06,
+      "loss": 0.7768,
+      "step": 108310
+    },
+    {
+      "epoch": 7.359355890746025,
+      "grad_norm": 7.875,
+      "learning_rate": 4.004025682837343e-06,
+      "loss": 0.8725,
+      "step": 108315
+    },
+    {
+      "epoch": 7.359695610816687,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.001902432395707e-06,
+      "loss": 0.8646,
+      "step": 108320
+    },
+    {
+      "epoch": 7.3600353308873485,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.99977918195407e-06,
+      "loss": 0.8719,
+      "step": 108325
+    },
+    {
+      "epoch": 7.360375050958011,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.997655931512434e-06,
+      "loss": 0.9041,
+      "step": 108330
+    },
+    {
+      "epoch": 7.360714771028673,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.995532681070798e-06,
+      "loss": 0.8578,
+      "step": 108335
+    },
+    {
+      "epoch": 7.361054491099334,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.993409430629162e-06,
+      "loss": 0.7945,
+      "step": 108340
+    },
+    {
+      "epoch": 7.361394211169996,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.991286180187526e-06,
+      "loss": 0.7344,
+      "step": 108345
+    },
+    {
+      "epoch": 7.361733931240658,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.98916292974589e-06,
+      "loss": 0.7381,
+      "step": 108350
+    },
+    {
+      "epoch": 7.362073651311319,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.987039679304253e-06,
+      "loss": 0.7926,
+      "step": 108355
+    },
+    {
+      "epoch": 7.362413371381981,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.984916428862617e-06,
+      "loss": 0.8628,
+      "step": 108360
+    },
+    {
+      "epoch": 7.362753091452643,
+      "grad_norm": 7.125,
+      "learning_rate": 3.982793178420981e-06,
+      "loss": 0.9348,
+      "step": 108365
+    },
+    {
+      "epoch": 7.3630928115233045,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.980669927979345e-06,
+      "loss": 0.8555,
+      "step": 108370
+    },
+    {
+      "epoch": 7.363432531593967,
+      "grad_norm": 6.84375,
+      "learning_rate": 3.978546677537709e-06,
+      "loss": 0.8636,
+      "step": 108375
+    },
+    {
+      "epoch": 7.363772251664629,
+      "grad_norm": 8.125,
+      "learning_rate": 3.976423427096073e-06,
+      "loss": 0.8769,
+      "step": 108380
+    },
+    {
+      "epoch": 7.36411197173529,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.974300176654437e-06,
+      "loss": 0.9235,
+      "step": 108385
+    },
+    {
+      "epoch": 7.364451691805952,
+      "grad_norm": 9.1875,
+      "learning_rate": 3.9721769262128e-06,
+      "loss": 0.7424,
+      "step": 108390
+    },
+    {
+      "epoch": 7.364791411876614,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.970053675771164e-06,
+      "loss": 0.8234,
+      "step": 108395
+    },
+    {
+      "epoch": 7.365131131947275,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.967930425329529e-06,
+      "loss": 0.8208,
+      "step": 108400
+    },
+    {
+      "epoch": 7.365470852017937,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.965807174887893e-06,
+      "loss": 0.8146,
+      "step": 108405
+    },
+    {
+      "epoch": 7.365810572088599,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.963683924446256e-06,
+      "loss": 0.812,
+      "step": 108410
+    },
+    {
+      "epoch": 7.3661502921592605,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.96156067400462e-06,
+      "loss": 0.8247,
+      "step": 108415
+    },
+    {
+      "epoch": 7.366490012229923,
+      "grad_norm": 9.8125,
+      "learning_rate": 3.959437423562984e-06,
+      "loss": 0.8252,
+      "step": 108420
+    },
+    {
+      "epoch": 7.366829732300585,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.957314173121348e-06,
+      "loss": 0.9235,
+      "step": 108425
+    },
+    {
+      "epoch": 7.367169452371246,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.955190922679712e-06,
+      "loss": 0.8218,
+      "step": 108430
+    },
+    {
+      "epoch": 7.367509172441908,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.953067672238076e-06,
+      "loss": 0.796,
+      "step": 108435
+    },
+    {
+      "epoch": 7.36784889251257,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.9509444217964395e-06,
+      "loss": 0.8247,
+      "step": 108440
+    },
+    {
+      "epoch": 7.368188612583231,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.9488211713548035e-06,
+      "loss": 0.8736,
+      "step": 108445
+    },
+    {
+      "epoch": 7.368528332653893,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.9466979209131675e-06,
+      "loss": 0.8254,
+      "step": 108450
+    },
+    {
+      "epoch": 7.368868052724555,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.944574670471532e-06,
+      "loss": 0.8585,
+      "step": 108455
+    },
+    {
+      "epoch": 7.3692077727952165,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.9424514200298955e-06,
+      "loss": 0.8697,
+      "step": 108460
+    },
+    {
+      "epoch": 7.369547492865879,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.9403281695882595e-06,
+      "loss": 0.8661,
+      "step": 108465
+    },
+    {
+      "epoch": 7.369887212936541,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.9382049191466235e-06,
+      "loss": 0.8437,
+      "step": 108470
+    },
+    {
+      "epoch": 7.370226933007202,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.936081668704987e-06,
+      "loss": 0.8801,
+      "step": 108475
+    },
+    {
+      "epoch": 7.370566653077864,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.9339584182633515e-06,
+      "loss": 0.929,
+      "step": 108480
+    },
+    {
+      "epoch": 7.370906373148525,
+      "grad_norm": 8.25,
+      "learning_rate": 3.9318351678217155e-06,
+      "loss": 0.9492,
+      "step": 108485
+    },
+    {
+      "epoch": 7.371246093219187,
+      "grad_norm": 8.375,
+      "learning_rate": 3.9297119173800795e-06,
+      "loss": 0.8009,
+      "step": 108490
+    },
+    {
+      "epoch": 7.371585813289849,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.927588666938443e-06,
+      "loss": 0.8952,
+      "step": 108495
+    },
+    {
+      "epoch": 7.3719255333605105,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.925465416496807e-06,
+      "loss": 0.8985,
+      "step": 108500
+    },
+    {
+      "epoch": 7.372265253431173,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.923342166055171e-06,
+      "loss": 0.8542,
+      "step": 108505
+    },
+    {
+      "epoch": 7.372604973501835,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.921218915613535e-06,
+      "loss": 0.8373,
+      "step": 108510
+    },
+    {
+      "epoch": 7.372944693572496,
+      "grad_norm": 8.125,
+      "learning_rate": 3.919095665171899e-06,
+      "loss": 0.882,
+      "step": 108515
+    },
+    {
+      "epoch": 7.373284413643158,
+      "grad_norm": 7.375,
+      "learning_rate": 3.916972414730263e-06,
+      "loss": 0.8153,
+      "step": 108520
+    },
+    {
+      "epoch": 7.37362413371382,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.914849164288626e-06,
+      "loss": 0.7991,
+      "step": 108525
+    },
+    {
+      "epoch": 7.373963853784481,
+      "grad_norm": 5.59375,
+      "learning_rate": 3.91272591384699e-06,
+      "loss": 0.6999,
+      "step": 108530
+    },
+    {
+      "epoch": 7.374303573855143,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.910602663405354e-06,
+      "loss": 0.8428,
+      "step": 108535
+    },
+    {
+      "epoch": 7.374643293925805,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.908479412963719e-06,
+      "loss": 0.8924,
+      "step": 108540
+    },
+    {
+      "epoch": 7.3749830139964665,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.906356162522082e-06,
+      "loss": 0.8148,
+      "step": 108545
+    },
+    {
+      "epoch": 7.375322734067129,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.904232912080446e-06,
+      "loss": 0.8002,
+      "step": 108550
+    },
+    {
+      "epoch": 7.375662454137791,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.90210966163881e-06,
+      "loss": 0.786,
+      "step": 108555
+    },
+    {
+      "epoch": 7.376002174208452,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.899986411197173e-06,
+      "loss": 0.8656,
+      "step": 108560
+    },
+    {
+      "epoch": 7.376341894279114,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.897863160755538e-06,
+      "loss": 0.8433,
+      "step": 108565
+    },
+    {
+      "epoch": 7.376681614349776,
+      "grad_norm": 8.625,
+      "learning_rate": 3.895739910313902e-06,
+      "loss": 0.9742,
+      "step": 108570
+    },
+    {
+      "epoch": 7.377021334420437,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.893616659872265e-06,
+      "loss": 0.875,
+      "step": 108575
+    },
+    {
+      "epoch": 7.377361054491099,
+      "grad_norm": 7.75,
+      "learning_rate": 3.891493409430629e-06,
+      "loss": 0.7862,
+      "step": 108580
+    },
+    {
+      "epoch": 7.377700774561761,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.889370158988993e-06,
+      "loss": 0.8187,
+      "step": 108585
+    },
+    {
+      "epoch": 7.3780404946324225,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.887246908547357e-06,
+      "loss": 0.8752,
+      "step": 108590
+    },
+    {
+      "epoch": 7.378380214703085,
+      "grad_norm": 8.125,
+      "learning_rate": 3.885123658105721e-06,
+      "loss": 0.9975,
+      "step": 108595
+    },
+    {
+      "epoch": 7.378719934773747,
+      "grad_norm": 9.0,
+      "learning_rate": 3.883000407664085e-06,
+      "loss": 0.8255,
+      "step": 108600
+    },
+    {
+      "epoch": 7.379059654844408,
+      "grad_norm": 8.375,
+      "learning_rate": 3.880877157222449e-06,
+      "loss": 0.8063,
+      "step": 108605
+    },
+    {
+      "epoch": 7.37939937491507,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.878753906780812e-06,
+      "loss": 0.862,
+      "step": 108610
+    },
+    {
+      "epoch": 7.379739094985732,
+      "grad_norm": 8.75,
+      "learning_rate": 3.876630656339176e-06,
+      "loss": 0.9319,
+      "step": 108615
+    },
+    {
+      "epoch": 7.380078815056393,
+      "grad_norm": 8.25,
+      "learning_rate": 3.874507405897541e-06,
+      "loss": 0.7996,
+      "step": 108620
+    },
+    {
+      "epoch": 7.380418535127055,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.872384155455905e-06,
+      "loss": 0.9053,
+      "step": 108625
+    },
+    {
+      "epoch": 7.380758255197717,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.870260905014268e-06,
+      "loss": 0.7823,
+      "step": 108630
+    },
+    {
+      "epoch": 7.3810979752683785,
+      "grad_norm": 8.0,
+      "learning_rate": 3.868137654572632e-06,
+      "loss": 0.8555,
+      "step": 108635
+    },
+    {
+      "epoch": 7.381437695339041,
+      "grad_norm": 7.875,
+      "learning_rate": 3.866014404130996e-06,
+      "loss": 0.8224,
+      "step": 108640
+    },
+    {
+      "epoch": 7.381777415409703,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.86389115368936e-06,
+      "loss": 0.7975,
+      "step": 108645
+    },
+    {
+      "epoch": 7.382117135480364,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.861767903247724e-06,
+      "loss": 0.8996,
+      "step": 108650
+    },
+    {
+      "epoch": 7.382456855551026,
+      "grad_norm": 8.375,
+      "learning_rate": 3.859644652806088e-06,
+      "loss": 0.8965,
+      "step": 108655
+    },
+    {
+      "epoch": 7.382796575621688,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.8575214023644514e-06,
+      "loss": 0.8549,
+      "step": 108660
+    },
+    {
+      "epoch": 7.383136295692349,
+      "grad_norm": 8.25,
+      "learning_rate": 3.8553981519228154e-06,
+      "loss": 0.7552,
+      "step": 108665
+    },
+    {
+      "epoch": 7.383476015763011,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.8532749014811794e-06,
+      "loss": 0.8957,
+      "step": 108670
+    },
+    {
+      "epoch": 7.383815735833673,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.851151651039544e-06,
+      "loss": 0.848,
+      "step": 108675
+    },
+    {
+      "epoch": 7.3841554559043345,
+      "grad_norm": 8.25,
+      "learning_rate": 3.8490284005979074e-06,
+      "loss": 0.8423,
+      "step": 108680
+    },
+    {
+      "epoch": 7.384495175974997,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.8469051501562714e-06,
+      "loss": 0.8856,
+      "step": 108685
+    },
+    {
+      "epoch": 7.384834896045659,
+      "grad_norm": 8.0,
+      "learning_rate": 3.8447818997146355e-06,
+      "loss": 0.8627,
+      "step": 108690
+    },
+    {
+      "epoch": 7.38517461611632,
+      "grad_norm": 8.375,
+      "learning_rate": 3.842658649272999e-06,
+      "loss": 0.8353,
+      "step": 108695
+    },
+    {
+      "epoch": 7.385514336186982,
+      "grad_norm": 8.125,
+      "learning_rate": 3.8405353988313635e-06,
+      "loss": 0.8868,
+      "step": 108700
+    },
+    {
+      "epoch": 7.385854056257644,
+      "grad_norm": 8.25,
+      "learning_rate": 3.8384121483897275e-06,
+      "loss": 0.8173,
+      "step": 108705
+    },
+    {
+      "epoch": 7.386193776328305,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.8362888979480915e-06,
+      "loss": 0.883,
+      "step": 108710
+    },
+    {
+      "epoch": 7.386533496398967,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.834165647506455e-06,
+      "loss": 0.8111,
+      "step": 108715
+    },
+    {
+      "epoch": 7.386873216469629,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.832042397064819e-06,
+      "loss": 0.8679,
+      "step": 108720
+    },
+    {
+      "epoch": 7.3872129365402905,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.829919146623183e-06,
+      "loss": 0.7782,
+      "step": 108725
+    },
+    {
+      "epoch": 7.387552656610953,
+      "grad_norm": 7.75,
+      "learning_rate": 3.827795896181547e-06,
+      "loss": 0.8322,
+      "step": 108730
+    },
+    {
+      "epoch": 7.387892376681615,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.825672645739911e-06,
+      "loss": 0.8348,
+      "step": 108735
+    },
+    {
+      "epoch": 7.388232096752276,
+      "grad_norm": 9.0,
+      "learning_rate": 3.823549395298275e-06,
+      "loss": 0.7707,
+      "step": 108740
+    },
+    {
+      "epoch": 7.388571816822938,
+      "grad_norm": 8.25,
+      "learning_rate": 3.821426144856638e-06,
+      "loss": 0.9163,
+      "step": 108745
+    },
+    {
+      "epoch": 7.3889115368936,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.819302894415002e-06,
+      "loss": 0.7742,
+      "step": 108750
+    },
+    {
+      "epoch": 7.389251256964261,
+      "grad_norm": 8.9375,
+      "learning_rate": 3.817179643973366e-06,
+      "loss": 0.8994,
+      "step": 108755
+    },
+    {
+      "epoch": 7.389590977034923,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.815056393531731e-06,
+      "loss": 0.8208,
+      "step": 108760
+    },
+    {
+      "epoch": 7.389930697105585,
+      "grad_norm": 8.125,
+      "learning_rate": 3.812933143090094e-06,
+      "loss": 0.83,
+      "step": 108765
+    },
+    {
+      "epoch": 7.3902704171762466,
+      "grad_norm": 9.125,
+      "learning_rate": 3.810809892648458e-06,
+      "loss": 0.8966,
+      "step": 108770
+    },
+    {
+      "epoch": 7.390610137246909,
+      "grad_norm": 7.25,
+      "learning_rate": 3.808686642206822e-06,
+      "loss": 0.8867,
+      "step": 108775
+    },
+    {
+      "epoch": 7.390949857317571,
+      "grad_norm": 7.75,
+      "learning_rate": 3.8065633917651854e-06,
+      "loss": 0.8469,
+      "step": 108780
+    },
+    {
+      "epoch": 7.391289577388232,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.8044401413235494e-06,
+      "loss": 0.9434,
+      "step": 108785
+    },
+    {
+      "epoch": 7.391629297458894,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.8023168908819134e-06,
+      "loss": 0.8643,
+      "step": 108790
+    },
+    {
+      "epoch": 7.391969017529556,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.800193640440278e-06,
+      "loss": 0.8974,
+      "step": 108795
+    },
+    {
+      "epoch": 7.392308737600217,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.798070389998641e-06,
+      "loss": 0.7217,
+      "step": 108800
+    },
+    {
+      "epoch": 7.392648457670879,
+      "grad_norm": 8.625,
+      "learning_rate": 3.7959471395570054e-06,
+      "loss": 0.8713,
+      "step": 108805
+    },
+    {
+      "epoch": 7.392988177741541,
+      "grad_norm": 7.625,
+      "learning_rate": 3.7938238891153694e-06,
+      "loss": 0.8447,
+      "step": 108810
+    },
+    {
+      "epoch": 7.393327897812203,
+      "grad_norm": 8.375,
+      "learning_rate": 3.7917006386737326e-06,
+      "loss": 0.9218,
+      "step": 108815
+    },
+    {
+      "epoch": 7.393667617882865,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.789577388232097e-06,
+      "loss": 0.8547,
+      "step": 108820
+    },
+    {
+      "epoch": 7.394007337953527,
+      "grad_norm": 7.625,
+      "learning_rate": 3.787454137790461e-06,
+      "loss": 0.8228,
+      "step": 108825
+    },
+    {
+      "epoch": 7.394347058024188,
+      "grad_norm": 8.375,
+      "learning_rate": 3.7853308873488246e-06,
+      "loss": 0.9404,
+      "step": 108830
+    },
+    {
+      "epoch": 7.39468677809485,
+      "grad_norm": 8.0,
+      "learning_rate": 3.7832076369071886e-06,
+      "loss": 0.8708,
+      "step": 108835
+    },
+    {
+      "epoch": 7.395026498165512,
+      "grad_norm": 8.875,
+      "learning_rate": 3.7810843864655526e-06,
+      "loss": 0.7194,
+      "step": 108840
+    },
+    {
+      "epoch": 7.395366218236173,
+      "grad_norm": 8.125,
+      "learning_rate": 3.7789611360239166e-06,
+      "loss": 0.7644,
+      "step": 108845
+    },
+    {
+      "epoch": 7.395705938306835,
+      "grad_norm": 7.625,
+      "learning_rate": 3.77683788558228e-06,
+      "loss": 0.8988,
+      "step": 108850
+    },
+    {
+      "epoch": 7.396045658377497,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.774714635140644e-06,
+      "loss": 0.8298,
+      "step": 108855
+    },
+    {
+      "epoch": 7.396385378448159,
+      "grad_norm": 8.125,
+      "learning_rate": 3.772591384699008e-06,
+      "loss": 0.8496,
+      "step": 108860
+    },
+    {
+      "epoch": 7.396725098518821,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.7704681342573718e-06,
+      "loss": 0.8157,
+      "step": 108865
+    },
+    {
+      "epoch": 7.397064818589483,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.7683448838157358e-06,
+      "loss": 0.7987,
+      "step": 108870
+    },
+    {
+      "epoch": 7.397404538660144,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.7662216333741e-06,
+      "loss": 0.9311,
+      "step": 108875
+    },
+    {
+      "epoch": 7.397744258730806,
+      "grad_norm": 7.75,
+      "learning_rate": 3.764098382932464e-06,
+      "loss": 0.8361,
+      "step": 108880
+    },
+    {
+      "epoch": 7.398083978801467,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.7619751324908274e-06,
+      "loss": 0.8582,
+      "step": 108885
+    },
+    {
+      "epoch": 7.398423698872129,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.759851882049192e-06,
+      "loss": 0.9036,
+      "step": 108890
+    },
+    {
+      "epoch": 7.398763418942791,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.757728631607556e-06,
+      "loss": 0.7126,
+      "step": 108895
+    },
+    {
+      "epoch": 7.3991031390134525,
+      "grad_norm": 8.375,
+      "learning_rate": 3.7556053811659194e-06,
+      "loss": 0.7847,
+      "step": 108900
+    },
+    {
+      "epoch": 7.399442859084115,
+      "grad_norm": 8.875,
+      "learning_rate": 3.7534821307242834e-06,
+      "loss": 0.9117,
+      "step": 108905
+    },
+    {
+      "epoch": 7.399782579154777,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.7513588802826474e-06,
+      "loss": 0.8629,
+      "step": 108910
+    },
+    {
+      "epoch": 7.400122299225438,
+      "grad_norm": 7.03125,
+      "learning_rate": 3.749235629841011e-06,
+      "loss": 0.7565,
+      "step": 108915
+    },
+    {
+      "epoch": 7.4004620192961,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.747112379399375e-06,
+      "loss": 0.798,
+      "step": 108920
+    },
+    {
+      "epoch": 7.400801739366762,
+      "grad_norm": 8.75,
+      "learning_rate": 3.744989128957739e-06,
+      "loss": 0.8307,
+      "step": 108925
+    },
+    {
+      "epoch": 7.401141459437423,
+      "grad_norm": 8.625,
+      "learning_rate": 3.742865878516103e-06,
+      "loss": 0.9402,
+      "step": 108930
+    },
+    {
+      "epoch": 7.401481179508085,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.7407426280744666e-06,
+      "loss": 0.8247,
+      "step": 108935
+    },
+    {
+      "epoch": 7.401820899578747,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.7386193776328306e-06,
+      "loss": 0.808,
+      "step": 108940
+    },
+    {
+      "epoch": 7.4021606196494085,
+      "grad_norm": 7.75,
+      "learning_rate": 3.736496127191195e-06,
+      "loss": 0.8353,
+      "step": 108945
+    },
+    {
+      "epoch": 7.402500339720071,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.734372876749558e-06,
+      "loss": 0.8958,
+      "step": 108950
+    },
+    {
+      "epoch": 7.402840059790733,
+      "grad_norm": 8.875,
+      "learning_rate": 3.732249626307922e-06,
+      "loss": 0.8576,
+      "step": 108955
+    },
+    {
+      "epoch": 7.403179779861394,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.7301263758662866e-06,
+      "loss": 0.8811,
+      "step": 108960
+    },
+    {
+      "epoch": 7.403519499932056,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.7280031254246506e-06,
+      "loss": 0.8175,
+      "step": 108965
+    },
+    {
+      "epoch": 7.403859220002718,
+      "grad_norm": 8.25,
+      "learning_rate": 3.725879874983014e-06,
+      "loss": 0.8316,
+      "step": 108970
+    },
+    {
+      "epoch": 7.404198940073379,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.723756624541378e-06,
+      "loss": 0.888,
+      "step": 108975
+    },
+    {
+      "epoch": 7.404538660144041,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.721633374099742e-06,
+      "loss": 0.8448,
+      "step": 108980
+    },
+    {
+      "epoch": 7.404878380214703,
+      "grad_norm": 8.0,
+      "learning_rate": 3.7195101236581058e-06,
+      "loss": 0.8019,
+      "step": 108985
+    },
+    {
+      "epoch": 7.4052181002853645,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.7173868732164698e-06,
+      "loss": 0.7837,
+      "step": 108990
+    },
+    {
+      "epoch": 7.405557820356027,
+      "grad_norm": 7.625,
+      "learning_rate": 3.7152636227748338e-06,
+      "loss": 0.8499,
+      "step": 108995
+    },
+    {
+      "epoch": 7.405897540426689,
+      "grad_norm": 7.1875,
+      "learning_rate": 3.7131403723331973e-06,
+      "loss": 0.8071,
+      "step": 109000
+    },
+    {
+      "epoch": 7.40623726049735,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.7110171218915613e-06,
+      "loss": 0.8168,
+      "step": 109005
+    },
+    {
+      "epoch": 7.406576980568012,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.7088938714499253e-06,
+      "loss": 0.9576,
+      "step": 109010
+    },
+    {
+      "epoch": 7.406916700638674,
+      "grad_norm": 7.75,
+      "learning_rate": 3.7067706210082898e-06,
+      "loss": 0.7608,
+      "step": 109015
+    },
+    {
+      "epoch": 7.407256420709335,
+      "grad_norm": 8.375,
+      "learning_rate": 3.704647370566653e-06,
+      "loss": 0.764,
+      "step": 109020
+    },
+    {
+      "epoch": 7.407596140779997,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.7025241201250174e-06,
+      "loss": 0.8873,
+      "step": 109025
+    },
+    {
+      "epoch": 7.407935860850659,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.7004008696833814e-06,
+      "loss": 0.8358,
+      "step": 109030
+    },
+    {
+      "epoch": 7.4082755809213205,
+      "grad_norm": 8.125,
+      "learning_rate": 3.6982776192417445e-06,
+      "loss": 0.9144,
+      "step": 109035
+    },
+    {
+      "epoch": 7.408615300991983,
+      "grad_norm": 8.0,
+      "learning_rate": 3.696154368800109e-06,
+      "loss": 0.8305,
+      "step": 109040
+    },
+    {
+      "epoch": 7.408955021062645,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.694031118358473e-06,
+      "loss": 0.9503,
+      "step": 109045
+    },
+    {
+      "epoch": 7.409294741133306,
+      "grad_norm": 7.0,
+      "learning_rate": 3.691907867916837e-06,
+      "loss": 0.8289,
+      "step": 109050
+    },
+    {
+      "epoch": 7.409634461203968,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.6897846174752005e-06,
+      "loss": 0.8409,
+      "step": 109055
+    },
+    {
+      "epoch": 7.40997418127463,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.6876613670335645e-06,
+      "loss": 0.8465,
+      "step": 109060
+    },
+    {
+      "epoch": 7.410313901345291,
+      "grad_norm": 8.0,
+      "learning_rate": 3.6855381165919285e-06,
+      "loss": 0.8976,
+      "step": 109065
+    },
+    {
+      "epoch": 7.410653621415953,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.683414866150292e-06,
+      "loss": 0.8079,
+      "step": 109070
+    },
+    {
+      "epoch": 7.410993341486615,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.681291615708656e-06,
+      "loss": 0.6723,
+      "step": 109075
+    },
+    {
+      "epoch": 7.411333061557277,
+      "grad_norm": 9.0,
+      "learning_rate": 3.67916836526702e-06,
+      "loss": 0.9111,
+      "step": 109080
+    },
+    {
+      "epoch": 7.411672781627939,
+      "grad_norm": 8.375,
+      "learning_rate": 3.6770451148253837e-06,
+      "loss": 0.7803,
+      "step": 109085
+    },
+    {
+      "epoch": 7.412012501698601,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.6749218643837477e-06,
+      "loss": 0.7948,
+      "step": 109090
+    },
+    {
+      "epoch": 7.412352221769262,
+      "grad_norm": 8.25,
+      "learning_rate": 3.672798613942112e-06,
+      "loss": 0.8302,
+      "step": 109095
+    },
+    {
+      "epoch": 7.412691941839924,
+      "grad_norm": 8.0,
+      "learning_rate": 3.670675363500476e-06,
+      "loss": 0.758,
+      "step": 109100
+    },
+    {
+      "epoch": 7.413031661910586,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.6685521130588393e-06,
+      "loss": 0.9169,
+      "step": 109105
+    },
+    {
+      "epoch": 7.413371381981247,
+      "grad_norm": 8.5,
+      "learning_rate": 3.6664288626172037e-06,
+      "loss": 0.7957,
+      "step": 109110
+    },
+    {
+      "epoch": 7.413711102051909,
+      "grad_norm": 7.875,
+      "learning_rate": 3.6643056121755677e-06,
+      "loss": 0.8084,
+      "step": 109115
+    },
+    {
+      "epoch": 7.414050822122571,
+      "grad_norm": 8.125,
+      "learning_rate": 3.6621823617339313e-06,
+      "loss": 0.7901,
+      "step": 109120
+    },
+    {
+      "epoch": 7.414390542193233,
+      "grad_norm": 8.125,
+      "learning_rate": 3.6600591112922953e-06,
+      "loss": 0.8165,
+      "step": 109125
+    },
+    {
+      "epoch": 7.414730262263895,
+      "grad_norm": 8.375,
+      "learning_rate": 3.6579358608506593e-06,
+      "loss": 0.8345,
+      "step": 109130
+    },
+    {
+      "epoch": 7.415069982334557,
+      "grad_norm": 7.5,
+      "learning_rate": 3.6558126104090233e-06,
+      "loss": 0.8395,
+      "step": 109135
+    },
+    {
+      "epoch": 7.415409702405218,
+      "grad_norm": 8.375,
+      "learning_rate": 3.653689359967387e-06,
+      "loss": 0.8262,
+      "step": 109140
+    },
+    {
+      "epoch": 7.41574942247588,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.651566109525751e-06,
+      "loss": 0.794,
+      "step": 109145
+    },
+    {
+      "epoch": 7.416089142546542,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.649442859084115e-06,
+      "loss": 0.7284,
+      "step": 109150
+    },
+    {
+      "epoch": 7.416428862617203,
+      "grad_norm": 7.875,
+      "learning_rate": 3.6473196086424785e-06,
+      "loss": 0.8327,
+      "step": 109155
+    },
+    {
+      "epoch": 7.416768582687865,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.6451963582008425e-06,
+      "loss": 0.801,
+      "step": 109160
+    },
+    {
+      "epoch": 7.4171083027585265,
+      "grad_norm": 8.25,
+      "learning_rate": 3.643073107759207e-06,
+      "loss": 0.7792,
+      "step": 109165
+    },
+    {
+      "epoch": 7.417448022829189,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.64094985731757e-06,
+      "loss": 0.8508,
+      "step": 109170
+    },
+    {
+      "epoch": 7.417787742899851,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.638826606875934e-06,
+      "loss": 0.8592,
+      "step": 109175
+    },
+    {
+      "epoch": 7.418127462970512,
+      "grad_norm": 8.25,
+      "learning_rate": 3.6367033564342985e-06,
+      "loss": 0.7981,
+      "step": 109180
+    },
+    {
+      "epoch": 7.418467183041174,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.6345801059926625e-06,
+      "loss": 0.8793,
+      "step": 109185
+    },
+    {
+      "epoch": 7.418806903111836,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.632456855551026e-06,
+      "loss": 0.8386,
+      "step": 109190
+    },
+    {
+      "epoch": 7.419146623182497,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.63033360510939e-06,
+      "loss": 0.7975,
+      "step": 109195
+    },
+    {
+      "epoch": 7.419486343253159,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.628210354667754e-06,
+      "loss": 0.8353,
+      "step": 109200
+    },
+    {
+      "epoch": 7.419826063323821,
+      "grad_norm": 8.375,
+      "learning_rate": 3.6260871042261177e-06,
+      "loss": 0.7971,
+      "step": 109205
+    },
+    {
+      "epoch": 7.4201657833944825,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.6239638537844817e-06,
+      "loss": 0.8516,
+      "step": 109210
+    },
+    {
+      "epoch": 7.420505503465145,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.6218406033428457e-06,
+      "loss": 0.9217,
+      "step": 109215
+    },
+    {
+      "epoch": 7.420845223535807,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.6197173529012097e-06,
+      "loss": 0.8129,
+      "step": 109220
+    },
+    {
+      "epoch": 7.421184943606468,
+      "grad_norm": 7.5,
+      "learning_rate": 3.6175941024595733e-06,
+      "loss": 0.9452,
+      "step": 109225
+    },
+    {
+      "epoch": 7.42152466367713,
+      "grad_norm": 6.9375,
+      "learning_rate": 3.6154708520179373e-06,
+      "loss": 0.8004,
+      "step": 109230
+    },
+    {
+      "epoch": 7.421864383747792,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.6133476015763017e-06,
+      "loss": 0.8397,
+      "step": 109235
+    },
+    {
+      "epoch": 7.422204103818453,
+      "grad_norm": 8.0,
+      "learning_rate": 3.611224351134665e-06,
+      "loss": 0.8692,
+      "step": 109240
+    },
+    {
+      "epoch": 7.422543823889115,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.609101100693029e-06,
+      "loss": 0.9892,
+      "step": 109245
+    },
+    {
+      "epoch": 7.422883543959777,
+      "grad_norm": 7.25,
+      "learning_rate": 3.6069778502513933e-06,
+      "loss": 0.775,
+      "step": 109250
+    },
+    {
+      "epoch": 7.4232232640304385,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.6048545998097565e-06,
+      "loss": 0.8395,
+      "step": 109255
+    },
+    {
+      "epoch": 7.423562984101101,
+      "grad_norm": 8.875,
+      "learning_rate": 3.602731349368121e-06,
+      "loss": 0.8318,
+      "step": 109260
+    },
+    {
+      "epoch": 7.423902704171763,
+      "grad_norm": 8.875,
+      "learning_rate": 3.600608098926485e-06,
+      "loss": 0.9187,
+      "step": 109265
+    },
+    {
+      "epoch": 7.424242424242424,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.598484848484849e-06,
+      "loss": 0.8687,
+      "step": 109270
+    },
+    {
+      "epoch": 7.424582144313086,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.5963615980432125e-06,
+      "loss": 0.903,
+      "step": 109275
+    },
+    {
+      "epoch": 7.424921864383748,
+      "grad_norm": 6.5,
+      "learning_rate": 3.5942383476015765e-06,
+      "loss": 0.7945,
+      "step": 109280
+    },
+    {
+      "epoch": 7.425261584454409,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.5921150971599405e-06,
+      "loss": 0.903,
+      "step": 109285
+    },
+    {
+      "epoch": 7.425601304525071,
+      "grad_norm": 7.875,
+      "learning_rate": 3.589991846718304e-06,
+      "loss": 0.9315,
+      "step": 109290
+    },
+    {
+      "epoch": 7.425941024595733,
+      "grad_norm": 7.125,
+      "learning_rate": 3.587868596276668e-06,
+      "loss": 0.7064,
+      "step": 109295
+    },
+    {
+      "epoch": 7.4262807446663945,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.585745345835032e-06,
+      "loss": 0.8898,
+      "step": 109300
+    },
+    {
+      "epoch": 7.426620464737057,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.5836220953933965e-06,
+      "loss": 0.8073,
+      "step": 109305
+    },
+    {
+      "epoch": 7.426960184807719,
+      "grad_norm": 7.625,
+      "learning_rate": 3.5814988449517596e-06,
+      "loss": 0.7942,
+      "step": 109310
+    },
+    {
+      "epoch": 7.42729990487838,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.5793755945101237e-06,
+      "loss": 0.8113,
+      "step": 109315
+    },
+    {
+      "epoch": 7.427639624949042,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.577252344068488e-06,
+      "loss": 0.8742,
+      "step": 109320
+    },
+    {
+      "epoch": 7.427979345019704,
+      "grad_norm": 9.4375,
+      "learning_rate": 3.5751290936268512e-06,
+      "loss": 0.9392,
+      "step": 109325
+    },
+    {
+      "epoch": 7.428319065090365,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.5730058431852157e-06,
+      "loss": 0.9195,
+      "step": 109330
+    },
+    {
+      "epoch": 7.428658785161027,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.5708825927435797e-06,
+      "loss": 0.7775,
+      "step": 109335
+    },
+    {
+      "epoch": 7.428998505231689,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.568759342301943e-06,
+      "loss": 0.856,
+      "step": 109340
+    },
+    {
+      "epoch": 7.4293382253023506,
+      "grad_norm": 8.125,
+      "learning_rate": 3.5666360918603073e-06,
+      "loss": 0.8129,
+      "step": 109345
+    },
+    {
+      "epoch": 7.429677945373013,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.5645128414186713e-06,
+      "loss": 0.8877,
+      "step": 109350
+    },
+    {
+      "epoch": 7.430017665443675,
+      "grad_norm": 8.75,
+      "learning_rate": 3.5623895909770353e-06,
+      "loss": 0.8088,
+      "step": 109355
+    },
+    {
+      "epoch": 7.430357385514336,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.560266340535399e-06,
+      "loss": 0.7499,
+      "step": 109360
+    },
+    {
+      "epoch": 7.430697105584998,
+      "grad_norm": 9.125,
+      "learning_rate": 3.558143090093763e-06,
+      "loss": 0.8141,
+      "step": 109365
+    },
+    {
+      "epoch": 7.43103682565566,
+      "grad_norm": 8.0,
+      "learning_rate": 3.556019839652127e-06,
+      "loss": 0.8148,
+      "step": 109370
+    },
+    {
+      "epoch": 7.431376545726321,
+      "grad_norm": 8.0,
+      "learning_rate": 3.5538965892104904e-06,
+      "loss": 0.9174,
+      "step": 109375
+    },
+    {
+      "epoch": 7.431716265796983,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.5517733387688544e-06,
+      "loss": 0.8995,
+      "step": 109380
+    },
+    {
+      "epoch": 7.432055985867645,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.549650088327219e-06,
+      "loss": 0.8294,
+      "step": 109385
+    },
+    {
+      "epoch": 7.432395705938307,
+      "grad_norm": 8.125,
+      "learning_rate": 3.547526837885583e-06,
+      "loss": 0.8769,
+      "step": 109390
+    },
+    {
+      "epoch": 7.432735426008969,
+      "grad_norm": 8.375,
+      "learning_rate": 3.545403587443946e-06,
+      "loss": 0.8994,
+      "step": 109395
+    },
+    {
+      "epoch": 7.433075146079631,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.5432803370023104e-06,
+      "loss": 0.8477,
+      "step": 109400
+    },
+    {
+      "epoch": 7.433414866150292,
+      "grad_norm": 7.75,
+      "learning_rate": 3.5411570865606745e-06,
+      "loss": 0.9213,
+      "step": 109405
+    },
+    {
+      "epoch": 7.433754586220954,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.539033836119038e-06,
+      "loss": 0.9109,
+      "step": 109410
+    },
+    {
+      "epoch": 7.434094306291616,
+      "grad_norm": 6.59375,
+      "learning_rate": 3.536910585677402e-06,
+      "loss": 0.7842,
+      "step": 109415
+    },
+    {
+      "epoch": 7.434434026362277,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.534787335235766e-06,
+      "loss": 0.864,
+      "step": 109420
+    },
+    {
+      "epoch": 7.434773746432939,
+      "grad_norm": 8.875,
+      "learning_rate": 3.5326640847941296e-06,
+      "loss": 0.8465,
+      "step": 109425
+    },
+    {
+      "epoch": 7.435113466503601,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.5305408343524936e-06,
+      "loss": 0.7521,
+      "step": 109430
+    },
+    {
+      "epoch": 7.435453186574263,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.5284175839108576e-06,
+      "loss": 0.6906,
+      "step": 109435
+    },
+    {
+      "epoch": 7.435792906644925,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.5262943334692216e-06,
+      "loss": 0.8069,
+      "step": 109440
+    },
+    {
+      "epoch": 7.436132626715587,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.524171083027585e-06,
+      "loss": 0.7719,
+      "step": 109445
+    },
+    {
+      "epoch": 7.436472346786248,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.5220478325859492e-06,
+      "loss": 0.9609,
+      "step": 109450
+    },
+    {
+      "epoch": 7.43681206685691,
+      "grad_norm": 8.0,
+      "learning_rate": 3.5199245821443136e-06,
+      "loss": 0.88,
+      "step": 109455
+    },
+    {
+      "epoch": 7.437151786927572,
+      "grad_norm": 8.25,
+      "learning_rate": 3.517801331702677e-06,
+      "loss": 0.8498,
+      "step": 109460
+    },
+    {
+      "epoch": 7.437491506998233,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.515678081261041e-06,
+      "loss": 0.7351,
+      "step": 109465
+    },
+    {
+      "epoch": 7.437831227068895,
+      "grad_norm": 8.75,
+      "learning_rate": 3.5135548308194052e-06,
+      "loss": 0.9011,
+      "step": 109470
+    },
+    {
+      "epoch": 7.438170947139557,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.5114315803777692e-06,
+      "loss": 0.9092,
+      "step": 109475
+    },
+    {
+      "epoch": 7.438510667210219,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.509308329936133e-06,
+      "loss": 0.9521,
+      "step": 109480
+    },
+    {
+      "epoch": 7.438850387280881,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.507185079494497e-06,
+      "loss": 0.8778,
+      "step": 109485
+    },
+    {
+      "epoch": 7.439190107351543,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.505061829052861e-06,
+      "loss": 0.89,
+      "step": 109490
+    },
+    {
+      "epoch": 7.439529827422204,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.5029385786112244e-06,
+      "loss": 0.8349,
+      "step": 109495
+    },
+    {
+      "epoch": 7.439869547492866,
+      "grad_norm": 7.25,
+      "learning_rate": 3.5008153281695884e-06,
+      "loss": 0.7926,
+      "step": 109500
+    },
+    {
+      "epoch": 7.440209267563528,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.4986920777279524e-06,
+      "loss": 0.7784,
+      "step": 109505
+    },
+    {
+      "epoch": 7.440548987634189,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.496568827286316e-06,
+      "loss": 0.7951,
+      "step": 109510
+    },
+    {
+      "epoch": 7.440888707704851,
+      "grad_norm": 8.25,
+      "learning_rate": 3.49444557684468e-06,
+      "loss": 0.8477,
+      "step": 109515
+    },
+    {
+      "epoch": 7.441228427775513,
+      "grad_norm": 6.34375,
+      "learning_rate": 3.492322326403044e-06,
+      "loss": 0.9465,
+      "step": 109520
+    },
+    {
+      "epoch": 7.441568147846175,
+      "grad_norm": 8.25,
+      "learning_rate": 3.4901990759614084e-06,
+      "loss": 0.804,
+      "step": 109525
+    },
+    {
+      "epoch": 7.441907867916837,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.4880758255197716e-06,
+      "loss": 0.7566,
+      "step": 109530
+    },
+    {
+      "epoch": 7.442247587987499,
+      "grad_norm": 8.0,
+      "learning_rate": 3.4859525750781356e-06,
+      "loss": 0.832,
+      "step": 109535
+    },
+    {
+      "epoch": 7.44258730805816,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.4838293246365e-06,
+      "loss": 0.8701,
+      "step": 109540
+    },
+    {
+      "epoch": 7.442927028128822,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.481706074194863e-06,
+      "loss": 0.8353,
+      "step": 109545
+    },
+    {
+      "epoch": 7.443266748199484,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.4795828237532276e-06,
+      "loss": 0.823,
+      "step": 109550
+    },
+    {
+      "epoch": 7.443606468270145,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.4774595733115916e-06,
+      "loss": 0.8816,
+      "step": 109555
+    },
+    {
+      "epoch": 7.443946188340807,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.4753363228699556e-06,
+      "loss": 0.7863,
+      "step": 109560
+    },
+    {
+      "epoch": 7.444285908411469,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.473213072428319e-06,
+      "loss": 0.8444,
+      "step": 109565
+    },
+    {
+      "epoch": 7.444625628482131,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.471089821986683e-06,
+      "loss": 0.9025,
+      "step": 109570
+    },
+    {
+      "epoch": 7.444965348552793,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.468966571545047e-06,
+      "loss": 0.8895,
+      "step": 109575
+    },
+    {
+      "epoch": 7.445305068623454,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.4668433211034108e-06,
+      "loss": 0.8578,
+      "step": 109580
+    },
+    {
+      "epoch": 7.445644788694116,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.4647200706617748e-06,
+      "loss": 0.8779,
+      "step": 109585
+    },
+    {
+      "epoch": 7.445984508764778,
+      "grad_norm": 8.0,
+      "learning_rate": 3.4625968202201388e-06,
+      "loss": 0.791,
+      "step": 109590
+    },
+    {
+      "epoch": 7.446324228835439,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.4604735697785024e-06,
+      "loss": 0.8661,
+      "step": 109595
+    },
+    {
+      "epoch": 7.446663948906101,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.4583503193368664e-06,
+      "loss": 0.862,
+      "step": 109600
+    },
+    {
+      "epoch": 7.447003668976763,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.4562270688952304e-06,
+      "loss": 0.8329,
+      "step": 109605
+    },
+    {
+      "epoch": 7.4473433890474245,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.454103818453595e-06,
+      "loss": 0.6899,
+      "step": 109610
+    },
+    {
+      "epoch": 7.447683109118087,
+      "grad_norm": 7.5,
+      "learning_rate": 3.451980568011958e-06,
+      "loss": 0.875,
+      "step": 109615
+    },
+    {
+      "epoch": 7.448022829188749,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.4498573175703224e-06,
+      "loss": 0.8555,
+      "step": 109620
+    },
+    {
+      "epoch": 7.44836254925941,
+      "grad_norm": 7.5,
+      "learning_rate": 3.4477340671286864e-06,
+      "loss": 0.7675,
+      "step": 109625
+    },
+    {
+      "epoch": 7.448702269330072,
+      "grad_norm": 8.125,
+      "learning_rate": 3.4456108166870495e-06,
+      "loss": 0.8381,
+      "step": 109630
+    },
+    {
+      "epoch": 7.449041989400734,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.443487566245414e-06,
+      "loss": 0.7958,
+      "step": 109635
+    },
+    {
+      "epoch": 7.449381709471395,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.441364315803778e-06,
+      "loss": 0.8675,
+      "step": 109640
+    },
+    {
+      "epoch": 7.449721429542057,
+      "grad_norm": 7.75,
+      "learning_rate": 3.439241065362142e-06,
+      "loss": 0.8326,
+      "step": 109645
+    },
+    {
+      "epoch": 7.450061149612719,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.4371178149205056e-06,
+      "loss": 0.8029,
+      "step": 109650
+    },
+    {
+      "epoch": 7.4504008696833806,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.4349945644788696e-06,
+      "loss": 0.821,
+      "step": 109655
+    },
+    {
+      "epoch": 7.450740589754043,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.4328713140372336e-06,
+      "loss": 0.8712,
+      "step": 109660
+    },
+    {
+      "epoch": 7.451080309824705,
+      "grad_norm": 8.625,
+      "learning_rate": 3.430748063595597e-06,
+      "loss": 0.7996,
+      "step": 109665
+    },
+    {
+      "epoch": 7.451420029895366,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.428624813153961e-06,
+      "loss": 0.9565,
+      "step": 109670
+    },
+    {
+      "epoch": 7.451759749966028,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.426501562712325e-06,
+      "loss": 0.859,
+      "step": 109675
+    },
+    {
+      "epoch": 7.45209947003669,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.4243783122706887e-06,
+      "loss": 0.7625,
+      "step": 109680
+    },
+    {
+      "epoch": 7.452439190107351,
+      "grad_norm": 8.5,
+      "learning_rate": 3.4222550618290527e-06,
+      "loss": 0.7813,
+      "step": 109685
+    },
+    {
+      "epoch": 7.452778910178013,
+      "grad_norm": 7.875,
+      "learning_rate": 3.420131811387417e-06,
+      "loss": 0.8695,
+      "step": 109690
+    },
+    {
+      "epoch": 7.453118630248675,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.418008560945781e-06,
+      "loss": 0.8283,
+      "step": 109695
+    },
+    {
+      "epoch": 7.453458350319337,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.4158853105041443e-06,
+      "loss": 0.8342,
+      "step": 109700
+    },
+    {
+      "epoch": 7.453798070389999,
+      "grad_norm": 6.65625,
+      "learning_rate": 3.4137620600625088e-06,
+      "loss": 0.8248,
+      "step": 109705
+    },
+    {
+      "epoch": 7.454137790460661,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.4116388096208728e-06,
+      "loss": 0.7989,
+      "step": 109710
+    },
+    {
+      "epoch": 7.454477510531322,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.4095155591792363e-06,
+      "loss": 0.821,
+      "step": 109715
+    },
+    {
+      "epoch": 7.454817230601984,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.4073923087376003e-06,
+      "loss": 0.8839,
+      "step": 109720
+    },
+    {
+      "epoch": 7.455156950672646,
+      "grad_norm": 7.75,
+      "learning_rate": 3.4052690582959643e-06,
+      "loss": 0.9078,
+      "step": 109725
+    },
+    {
+      "epoch": 7.455496670743307,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.4031458078543283e-06,
+      "loss": 0.7572,
+      "step": 109730
+    },
+    {
+      "epoch": 7.455836390813969,
+      "grad_norm": 8.125,
+      "learning_rate": 3.401022557412692e-06,
+      "loss": 0.8071,
+      "step": 109735
+    },
+    {
+      "epoch": 7.456176110884631,
+      "grad_norm": 8.625,
+      "learning_rate": 3.398899306971056e-06,
+      "loss": 0.8657,
+      "step": 109740
+    },
+    {
+      "epoch": 7.456515830955293,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.3967760565294204e-06,
+      "loss": 0.7612,
+      "step": 109745
+    },
+    {
+      "epoch": 7.456855551025955,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.3946528060877835e-06,
+      "loss": 0.8568,
+      "step": 109750
+    },
+    {
+      "epoch": 7.457195271096617,
+      "grad_norm": 8.375,
+      "learning_rate": 3.3925295556461475e-06,
+      "loss": 0.9019,
+      "step": 109755
+    },
+    {
+      "epoch": 7.457534991167278,
+      "grad_norm": 8.125,
+      "learning_rate": 3.390406305204512e-06,
+      "loss": 0.8048,
+      "step": 109760
+    },
+    {
+      "epoch": 7.45787471123794,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.388283054762875e-06,
+      "loss": 0.8514,
+      "step": 109765
+    },
+    {
+      "epoch": 7.458214431308602,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.3861598043212395e-06,
+      "loss": 0.7491,
+      "step": 109770
+    },
+    {
+      "epoch": 7.458554151379263,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.3840365538796035e-06,
+      "loss": 0.7469,
+      "step": 109775
+    },
+    {
+      "epoch": 7.458893871449925,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.3819133034379675e-06,
+      "loss": 0.8515,
+      "step": 109780
+    },
+    {
+      "epoch": 7.459233591520587,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.379790052996331e-06,
+      "loss": 0.8173,
+      "step": 109785
+    },
+    {
+      "epoch": 7.459573311591249,
+      "grad_norm": 8.25,
+      "learning_rate": 3.377666802554695e-06,
+      "loss": 0.7299,
+      "step": 109790
+    },
+    {
+      "epoch": 7.459913031661911,
+      "grad_norm": 9.4375,
+      "learning_rate": 3.375543552113059e-06,
+      "loss": 0.8444,
+      "step": 109795
+    },
+    {
+      "epoch": 7.460252751732573,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.3734203016714227e-06,
+      "loss": 0.8265,
+      "step": 109800
+    },
+    {
+      "epoch": 7.460592471803234,
+      "grad_norm": 7.875,
+      "learning_rate": 3.3712970512297867e-06,
+      "loss": 0.8568,
+      "step": 109805
+    },
+    {
+      "epoch": 7.460932191873896,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.3691738007881507e-06,
+      "loss": 0.9613,
+      "step": 109810
+    },
+    {
+      "epoch": 7.461271911944558,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.367050550346515e-06,
+      "loss": 0.9211,
+      "step": 109815
+    },
+    {
+      "epoch": 7.461611632015219,
+      "grad_norm": 8.25,
+      "learning_rate": 3.3649272999048783e-06,
+      "loss": 0.8306,
+      "step": 109820
+    },
+    {
+      "epoch": 7.461951352085881,
+      "grad_norm": 7.25,
+      "learning_rate": 3.3628040494632423e-06,
+      "loss": 0.7806,
+      "step": 109825
+    },
+    {
+      "epoch": 7.462291072156543,
+      "grad_norm": 8.375,
+      "learning_rate": 3.3606807990216067e-06,
+      "loss": 0.763,
+      "step": 109830
+    },
+    {
+      "epoch": 7.462630792227205,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.35855754857997e-06,
+      "loss": 0.8222,
+      "step": 109835
+    },
+    {
+      "epoch": 7.462970512297867,
+      "grad_norm": 8.0,
+      "learning_rate": 3.3564342981383343e-06,
+      "loss": 0.8593,
+      "step": 109840
+    },
+    {
+      "epoch": 7.463310232368528,
+      "grad_norm": 8.125,
+      "learning_rate": 3.3543110476966983e-06,
+      "loss": 0.8841,
+      "step": 109845
+    },
+    {
+      "epoch": 7.46364995243919,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.3521877972550615e-06,
+      "loss": 0.7331,
+      "step": 109850
+    },
+    {
+      "epoch": 7.463989672509852,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.350064546813426e-06,
+      "loss": 0.7677,
+      "step": 109855
+    },
+    {
+      "epoch": 7.464329392580513,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.34794129637179e-06,
+      "loss": 0.826,
+      "step": 109860
+    },
+    {
+      "epoch": 7.464669112651175,
+      "grad_norm": 7.15625,
+      "learning_rate": 3.345818045930154e-06,
+      "loss": 0.8243,
+      "step": 109865
+    },
+    {
+      "epoch": 7.465008832721837,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.3436947954885175e-06,
+      "loss": 0.8052,
+      "step": 109870
+    },
+    {
+      "epoch": 7.4653485527924985,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.3415715450468815e-06,
+      "loss": 0.8657,
+      "step": 109875
+    },
+    {
+      "epoch": 7.465688272863161,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.3394482946052455e-06,
+      "loss": 0.856,
+      "step": 109880
+    },
+    {
+      "epoch": 7.466027992933823,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.337325044163609e-06,
+      "loss": 0.846,
+      "step": 109885
+    },
+    {
+      "epoch": 7.466367713004484,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.335201793721973e-06,
+      "loss": 0.7935,
+      "step": 109890
+    },
+    {
+      "epoch": 7.466707433075146,
+      "grad_norm": 8.25,
+      "learning_rate": 3.333078543280337e-06,
+      "loss": 0.8185,
+      "step": 109895
+    },
+    {
+      "epoch": 7.467047153145808,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.3309552928387015e-06,
+      "loss": 0.8977,
+      "step": 109900
+    },
+    {
+      "epoch": 7.467386873216469,
+      "grad_norm": 8.625,
+      "learning_rate": 3.3288320423970647e-06,
+      "loss": 0.7728,
+      "step": 109905
+    },
+    {
+      "epoch": 7.467726593287131,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.326708791955429e-06,
+      "loss": 0.8508,
+      "step": 109910
+    },
+    {
+      "epoch": 7.468066313357793,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.324585541513793e-06,
+      "loss": 0.8926,
+      "step": 109915
+    },
+    {
+      "epoch": 7.4684060334284545,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.3224622910721563e-06,
+      "loss": 0.8904,
+      "step": 109920
+    },
+    {
+      "epoch": 7.468745753499117,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.3203390406305207e-06,
+      "loss": 0.834,
+      "step": 109925
+    },
+    {
+      "epoch": 7.469085473569779,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.3182157901888847e-06,
+      "loss": 0.7776,
+      "step": 109930
+    },
+    {
+      "epoch": 7.46942519364044,
+      "grad_norm": 8.125,
+      "learning_rate": 3.3160925397472483e-06,
+      "loss": 0.7514,
+      "step": 109935
+    },
+    {
+      "epoch": 7.469764913711102,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.3139692893056123e-06,
+      "loss": 0.8567,
+      "step": 109940
+    },
+    {
+      "epoch": 7.470104633781764,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.3118460388639763e-06,
+      "loss": 0.8151,
+      "step": 109945
+    },
+    {
+      "epoch": 7.470444353852425,
+      "grad_norm": 8.625,
+      "learning_rate": 3.3097227884223403e-06,
+      "loss": 0.7915,
+      "step": 109950
+    },
+    {
+      "epoch": 7.470784073923087,
+      "grad_norm": 8.125,
+      "learning_rate": 3.307599537980704e-06,
+      "loss": 0.8406,
+      "step": 109955
+    },
+    {
+      "epoch": 7.471123793993749,
+      "grad_norm": 8.5,
+      "learning_rate": 3.305476287539068e-06,
+      "loss": 0.8754,
+      "step": 109960
+    },
+    {
+      "epoch": 7.471463514064411,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.303353037097432e-06,
+      "loss": 0.8955,
+      "step": 109965
+    },
+    {
+      "epoch": 7.471803234135073,
+      "grad_norm": 8.875,
+      "learning_rate": 3.3012297866557955e-06,
+      "loss": 0.7608,
+      "step": 109970
+    },
+    {
+      "epoch": 7.472142954205735,
+      "grad_norm": 7.75,
+      "learning_rate": 3.2991065362141595e-06,
+      "loss": 0.7548,
+      "step": 109975
+    },
+    {
+      "epoch": 7.472482674276396,
+      "grad_norm": 7.5,
+      "learning_rate": 3.296983285772524e-06,
+      "loss": 0.9,
+      "step": 109980
+    },
+    {
+      "epoch": 7.472822394347058,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.294860035330888e-06,
+      "loss": 0.9584,
+      "step": 109985
+    },
+    {
+      "epoch": 7.47316211441772,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.292736784889251e-06,
+      "loss": 0.868,
+      "step": 109990
+    },
+    {
+      "epoch": 7.473501834488381,
+      "grad_norm": 8.125,
+      "learning_rate": 3.2906135344476155e-06,
+      "loss": 0.9578,
+      "step": 109995
+    },
+    {
+      "epoch": 7.473841554559043,
+      "grad_norm": 8.0,
+      "learning_rate": 3.2884902840059795e-06,
+      "loss": 0.7685,
+      "step": 110000
+    },
+    {
+      "epoch": 7.474181274629705,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.286367033564343e-06,
+      "loss": 0.7779,
+      "step": 110005
+    },
+    {
+      "epoch": 7.474520994700367,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.284243783122707e-06,
+      "loss": 0.8578,
+      "step": 110010
+    },
+    {
+      "epoch": 7.474860714771029,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.282120532681071e-06,
+      "loss": 0.7643,
+      "step": 110015
+    },
+    {
+      "epoch": 7.475200434841691,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.2799972822394346e-06,
+      "loss": 0.8548,
+      "step": 110020
+    },
+    {
+      "epoch": 7.475540154912352,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.2778740317977986e-06,
+      "loss": 0.9063,
+      "step": 110025
+    },
+    {
+      "epoch": 7.475879874983014,
+      "grad_norm": 7.09375,
+      "learning_rate": 3.2757507813561627e-06,
+      "loss": 0.8828,
+      "step": 110030
+    },
+    {
+      "epoch": 7.476219595053676,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.273627530914527e-06,
+      "loss": 0.76,
+      "step": 110035
+    },
+    {
+      "epoch": 7.476559315124337,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.2715042804728902e-06,
+      "loss": 0.808,
+      "step": 110040
+    },
+    {
+      "epoch": 7.476899035194999,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.2693810300312542e-06,
+      "loss": 0.8199,
+      "step": 110045
+    },
+    {
+      "epoch": 7.477238755265661,
+      "grad_norm": 8.0,
+      "learning_rate": 3.2672577795896187e-06,
+      "loss": 0.8033,
+      "step": 110050
+    },
+    {
+      "epoch": 7.477578475336323,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.265134529147982e-06,
+      "loss": 0.8371,
+      "step": 110055
+    },
+    {
+      "epoch": 7.477918195406985,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.2630112787063463e-06,
+      "loss": 0.854,
+      "step": 110060
+    },
+    {
+      "epoch": 7.478257915477647,
+      "grad_norm": 8.125,
+      "learning_rate": 3.2608880282647103e-06,
+      "loss": 0.9551,
+      "step": 110065
+    },
+    {
+      "epoch": 7.478597635548308,
+      "grad_norm": 8.25,
+      "learning_rate": 3.2587647778230743e-06,
+      "loss": 0.7525,
+      "step": 110070
+    },
+    {
+      "epoch": 7.47893735561897,
+      "grad_norm": 8.375,
+      "learning_rate": 3.256641527381438e-06,
+      "loss": 0.9613,
+      "step": 110075
+    },
+    {
+      "epoch": 7.479277075689632,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.254518276939802e-06,
+      "loss": 0.9072,
+      "step": 110080
+    },
+    {
+      "epoch": 7.479616795760293,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.252395026498166e-06,
+      "loss": 0.8738,
+      "step": 110085
+    },
+    {
+      "epoch": 7.479956515830955,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.2502717760565294e-06,
+      "loss": 0.7558,
+      "step": 110090
+    },
+    {
+      "epoch": 7.480296235901617,
+      "grad_norm": 8.625,
+      "learning_rate": 3.2481485256148934e-06,
+      "loss": 0.8798,
+      "step": 110095
+    },
+    {
+      "epoch": 7.480635955972279,
+      "grad_norm": 7.0,
+      "learning_rate": 3.2460252751732574e-06,
+      "loss": 0.7454,
+      "step": 110100
+    },
+    {
+      "epoch": 7.480975676042941,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.243902024731621e-06,
+      "loss": 0.8612,
+      "step": 110105
+    },
+    {
+      "epoch": 7.481315396113603,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.241778774289985e-06,
+      "loss": 0.8213,
+      "step": 110110
+    },
+    {
+      "epoch": 7.481655116184264,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.239655523848349e-06,
+      "loss": 0.8152,
+      "step": 110115
+    },
+    {
+      "epoch": 7.481994836254926,
+      "grad_norm": 6.46875,
+      "learning_rate": 3.2375322734067135e-06,
+      "loss": 0.8919,
+      "step": 110120
+    },
+    {
+      "epoch": 7.482334556325588,
+      "grad_norm": 8.9375,
+      "learning_rate": 3.2354090229650766e-06,
+      "loss": 0.8479,
+      "step": 110125
+    },
+    {
+      "epoch": 7.482674276396249,
+      "grad_norm": 7.75,
+      "learning_rate": 3.233285772523441e-06,
+      "loss": 0.8334,
+      "step": 110130
+    },
+    {
+      "epoch": 7.483013996466911,
+      "grad_norm": 9.5625,
+      "learning_rate": 3.231162522081805e-06,
+      "loss": 0.8754,
+      "step": 110135
+    },
+    {
+      "epoch": 7.483353716537573,
+      "grad_norm": 7.75,
+      "learning_rate": 3.229039271640168e-06,
+      "loss": 0.7921,
+      "step": 110140
+    },
+    {
+      "epoch": 7.483693436608235,
+      "grad_norm": 8.875,
+      "learning_rate": 3.2269160211985326e-06,
+      "loss": 0.8524,
+      "step": 110145
+    },
+    {
+      "epoch": 7.484033156678897,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.2247927707568966e-06,
+      "loss": 0.8215,
+      "step": 110150
+    },
+    {
+      "epoch": 7.484372876749559,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.2226695203152606e-06,
+      "loss": 0.939,
+      "step": 110155
+    },
+    {
+      "epoch": 7.48471259682022,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.220546269873624e-06,
+      "loss": 0.796,
+      "step": 110160
+    },
+    {
+      "epoch": 7.485052316890882,
+      "grad_norm": 7.875,
+      "learning_rate": 3.2184230194319882e-06,
+      "loss": 0.8718,
+      "step": 110165
+    },
+    {
+      "epoch": 7.485392036961544,
+      "grad_norm": 8.5,
+      "learning_rate": 3.2162997689903522e-06,
+      "loss": 0.8658,
+      "step": 110170
+    },
+    {
+      "epoch": 7.485731757032205,
+      "grad_norm": 6.71875,
+      "learning_rate": 3.214176518548716e-06,
+      "loss": 0.8327,
+      "step": 110175
+    },
+    {
+      "epoch": 7.486071477102867,
+      "grad_norm": 8.375,
+      "learning_rate": 3.21205326810708e-06,
+      "loss": 0.7832,
+      "step": 110180
+    },
+    {
+      "epoch": 7.486411197173529,
+      "grad_norm": 6.5625,
+      "learning_rate": 3.209930017665444e-06,
+      "loss": 0.7543,
+      "step": 110185
+    },
+    {
+      "epoch": 7.486750917244191,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.2078067672238074e-06,
+      "loss": 0.851,
+      "step": 110190
+    },
+    {
+      "epoch": 7.487090637314853,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.2056835167821714e-06,
+      "loss": 0.8671,
+      "step": 110195
+    },
+    {
+      "epoch": 7.487430357385515,
+      "grad_norm": 8.25,
+      "learning_rate": 3.203560266340536e-06,
+      "loss": 0.7889,
+      "step": 110200
+    },
+    {
+      "epoch": 7.487770077456176,
+      "grad_norm": 8.75,
+      "learning_rate": 3.2014370158989e-06,
+      "loss": 0.9718,
+      "step": 110205
+    },
+    {
+      "epoch": 7.488109797526838,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.199313765457263e-06,
+      "loss": 0.8735,
+      "step": 110210
+    },
+    {
+      "epoch": 7.4884495175975,
+      "grad_norm": 8.375,
+      "learning_rate": 3.1971905150156274e-06,
+      "loss": 0.9847,
+      "step": 110215
+    },
+    {
+      "epoch": 7.488789237668161,
+      "grad_norm": 7.6875,
+      "learning_rate": 3.1950672645739914e-06,
+      "loss": 0.777,
+      "step": 110220
+    },
+    {
+      "epoch": 7.489128957738823,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.192944014132355e-06,
+      "loss": 0.793,
+      "step": 110225
+    },
+    {
+      "epoch": 7.489468677809485,
+      "grad_norm": 6.90625,
+      "learning_rate": 3.190820763690719e-06,
+      "loss": 0.8623,
+      "step": 110230
+    },
+    {
+      "epoch": 7.489808397880147,
+      "grad_norm": 7.5,
+      "learning_rate": 3.188697513249083e-06,
+      "loss": 0.8879,
+      "step": 110235
+    },
+    {
+      "epoch": 7.490148117950809,
+      "grad_norm": 7.875,
+      "learning_rate": 3.186574262807447e-06,
+      "loss": 0.7959,
+      "step": 110240
+    },
+    {
+      "epoch": 7.490487838021471,
+      "grad_norm": 8.75,
+      "learning_rate": 3.1844510123658106e-06,
+      "loss": 0.8462,
+      "step": 110245
+    },
+    {
+      "epoch": 7.490827558092132,
+      "grad_norm": 8.125,
+      "learning_rate": 3.1823277619241746e-06,
+      "loss": 0.8284,
+      "step": 110250
+    },
+    {
+      "epoch": 7.491167278162794,
+      "grad_norm": 8.25,
+      "learning_rate": 3.1802045114825386e-06,
+      "loss": 0.8093,
+      "step": 110255
+    },
+    {
+      "epoch": 7.491506998233455,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.178081261040902e-06,
+      "loss": 0.809,
+      "step": 110260
+    },
+    {
+      "epoch": 7.491846718304117,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.175958010599266e-06,
+      "loss": 0.8213,
+      "step": 110265
+    },
+    {
+      "epoch": 7.492186438374779,
+      "grad_norm": 8.0,
+      "learning_rate": 3.1738347601576306e-06,
+      "loss": 0.8557,
+      "step": 110270
+    },
+    {
+      "epoch": 7.492526158445441,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.1717115097159938e-06,
+      "loss": 0.8532,
+      "step": 110275
+    },
+    {
+      "epoch": 7.492865878516103,
+      "grad_norm": 7.875,
+      "learning_rate": 3.1695882592743578e-06,
+      "loss": 0.8,
+      "step": 110280
+    },
+    {
+      "epoch": 7.493205598586765,
+      "grad_norm": 9.5625,
+      "learning_rate": 3.167465008832722e-06,
+      "loss": 0.8319,
+      "step": 110285
+    },
+    {
+      "epoch": 7.493545318657426,
+      "grad_norm": 8.375,
+      "learning_rate": 3.165341758391086e-06,
+      "loss": 0.9003,
+      "step": 110290
+    },
+    {
+      "epoch": 7.493885038728088,
+      "grad_norm": 8.25,
+      "learning_rate": 3.1632185079494498e-06,
+      "loss": 0.8287,
+      "step": 110295
+    },
+    {
+      "epoch": 7.49422475879875,
+      "grad_norm": 7.28125,
+      "learning_rate": 3.1610952575078138e-06,
+      "loss": 0.7907,
+      "step": 110300
+    },
+    {
+      "epoch": 7.494564478869411,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.1589720070661778e-06,
+      "loss": 0.8947,
+      "step": 110305
+    },
+    {
+      "epoch": 7.494904198940073,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.1568487566245414e-06,
+      "loss": 0.7663,
+      "step": 110310
+    },
+    {
+      "epoch": 7.495243919010735,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.1547255061829054e-06,
+      "loss": 0.802,
+      "step": 110315
+    },
+    {
+      "epoch": 7.495583639081397,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.1526022557412694e-06,
+      "loss": 0.7217,
+      "step": 110320
+    },
+    {
+      "epoch": 7.495923359152059,
+      "grad_norm": 9.125,
+      "learning_rate": 3.1504790052996334e-06,
+      "loss": 0.9032,
+      "step": 110325
+    },
+    {
+      "epoch": 7.496263079222721,
+      "grad_norm": 7.5,
+      "learning_rate": 3.148355754857997e-06,
+      "loss": 0.78,
+      "step": 110330
+    },
+    {
+      "epoch": 7.496602799293382,
+      "grad_norm": 8.25,
+      "learning_rate": 3.146232504416361e-06,
+      "loss": 0.8981,
+      "step": 110335
+    },
+    {
+      "epoch": 7.496942519364044,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.1441092539747254e-06,
+      "loss": 0.8866,
+      "step": 110340
+    },
+    {
+      "epoch": 7.497282239434706,
+      "grad_norm": 8.0,
+      "learning_rate": 3.1419860035330885e-06,
+      "loss": 0.7919,
+      "step": 110345
+    },
+    {
+      "epoch": 7.497621959505367,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.1398627530914525e-06,
+      "loss": 0.8304,
+      "step": 110350
+    },
+    {
+      "epoch": 7.497961679576029,
+      "grad_norm": 8.625,
+      "learning_rate": 3.137739502649817e-06,
+      "loss": 0.8323,
+      "step": 110355
+    },
+    {
+      "epoch": 7.498301399646691,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.13561625220818e-06,
+      "loss": 0.9323,
+      "step": 110360
+    },
+    {
+      "epoch": 7.498641119717353,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.1334930017665446e-06,
+      "loss": 0.827,
+      "step": 110365
+    },
+    {
+      "epoch": 7.498980839788015,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.1313697513249086e-06,
+      "loss": 0.8273,
+      "step": 110370
+    },
+    {
+      "epoch": 7.499320559858677,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.1292465008832726e-06,
+      "loss": 0.7921,
+      "step": 110375
+    },
+    {
+      "epoch": 7.499660279929338,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.127123250441636e-06,
+      "loss": 0.8077,
+      "step": 110380
+    },
+    {
+      "epoch": 7.5,
+      "grad_norm": 9.25,
+      "learning_rate": 3.125e-06,
+      "loss": 0.8013,
+      "step": 110385
+    },
+    {
+      "epoch": 7.500339720070662,
+      "grad_norm": 8.125,
+      "learning_rate": 3.1228767495583637e-06,
+      "loss": 0.8784,
+      "step": 110390
+    },
+    {
+      "epoch": 7.500679440141323,
+      "grad_norm": 9.125,
+      "learning_rate": 3.120753499116728e-06,
+      "loss": 0.8489,
+      "step": 110395
+    },
+    {
+      "epoch": 7.501019160211985,
+      "grad_norm": 8.375,
+      "learning_rate": 3.1186302486750917e-06,
+      "loss": 0.783,
+      "step": 110400
+    },
+    {
+      "epoch": 7.501358880282647,
+      "grad_norm": 9.125,
+      "learning_rate": 3.1165069982334557e-06,
+      "loss": 0.8112,
+      "step": 110405
+    },
+    {
+      "epoch": 7.501698600353309,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.1143837477918197e-06,
+      "loss": 0.7669,
+      "step": 110410
+    },
+    {
+      "epoch": 7.502038320423971,
+      "grad_norm": 7.875,
+      "learning_rate": 3.1122604973501833e-06,
+      "loss": 0.8191,
+      "step": 110415
+    },
+    {
+      "epoch": 7.502378040494633,
+      "grad_norm": 7.46875,
+      "learning_rate": 3.1101372469085478e-06,
+      "loss": 0.8961,
+      "step": 110420
+    },
+    {
+      "epoch": 7.502717760565294,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.1080139964669113e-06,
+      "loss": 0.9325,
+      "step": 110425
+    },
+    {
+      "epoch": 7.503057480635956,
+      "grad_norm": 8.0,
+      "learning_rate": 3.1058907460252753e-06,
+      "loss": 0.8744,
+      "step": 110430
+    },
+    {
+      "epoch": 7.503397200706618,
+      "grad_norm": 8.875,
+      "learning_rate": 3.1037674955836393e-06,
+      "loss": 0.772,
+      "step": 110435
+    },
+    {
+      "epoch": 7.503736920777279,
+      "grad_norm": 6.96875,
+      "learning_rate": 3.1016442451420033e-06,
+      "loss": 0.8937,
+      "step": 110440
+    },
+    {
+      "epoch": 7.504076640847941,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.099520994700367e-06,
+      "loss": 0.8674,
+      "step": 110445
+    },
+    {
+      "epoch": 7.504416360918603,
+      "grad_norm": 8.375,
+      "learning_rate": 3.097397744258731e-06,
+      "loss": 0.9111,
+      "step": 110450
+    },
+    {
+      "epoch": 7.504756080989265,
+      "grad_norm": 8.625,
+      "learning_rate": 3.095274493817095e-06,
+      "loss": 0.8858,
+      "step": 110455
+    },
+    {
+      "epoch": 7.505095801059927,
+      "grad_norm": 7.71875,
+      "learning_rate": 3.0931512433754585e-06,
+      "loss": 0.7623,
+      "step": 110460
+    },
+    {
+      "epoch": 7.505435521130589,
+      "grad_norm": 7.34375,
+      "learning_rate": 3.091027992933823e-06,
+      "loss": 0.7769,
+      "step": 110465
+    },
+    {
+      "epoch": 7.50577524120125,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0889047424921865e-06,
+      "loss": 0.8453,
+      "step": 110470
+    },
+    {
+      "epoch": 7.506114961271912,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0867814920505505e-06,
+      "loss": 0.8777,
+      "step": 110475
+    },
+    {
+      "epoch": 7.506454681342574,
+      "grad_norm": 8.25,
+      "learning_rate": 3.0846582416089145e-06,
+      "loss": 0.9055,
+      "step": 110480
+    },
+    {
+      "epoch": 7.506794401413235,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.082534991167278e-06,
+      "loss": 0.8746,
+      "step": 110485
+    },
+    {
+      "epoch": 7.507134121483897,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.0804117407256425e-06,
+      "loss": 0.8594,
+      "step": 110490
+    },
+    {
+      "epoch": 7.5074738415545585,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.078288490284006e-06,
+      "loss": 0.8728,
+      "step": 110495
+    },
+    {
+      "epoch": 7.507813561625221,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.0761652398423697e-06,
+      "loss": 0.8515,
+      "step": 110500
+    },
+    {
+      "epoch": 7.508153281695883,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.074041989400734e-06,
+      "loss": 0.9387,
+      "step": 110505
+    },
+    {
+      "epoch": 7.508493001766544,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0719187389590977e-06,
+      "loss": 0.7676,
+      "step": 110510
+    },
+    {
+      "epoch": 7.508832721837206,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.0697954885174617e-06,
+      "loss": 0.8969,
+      "step": 110515
+    },
+    {
+      "epoch": 7.509172441907868,
+      "grad_norm": 6.625,
+      "learning_rate": 3.0676722380758257e-06,
+      "loss": 0.794,
+      "step": 110520
+    },
+    {
+      "epoch": 7.509512161978529,
+      "grad_norm": 8.125,
+      "learning_rate": 3.0655489876341897e-06,
+      "loss": 0.7963,
+      "step": 110525
+    },
+    {
+      "epoch": 7.509851882049191,
+      "grad_norm": 9.125,
+      "learning_rate": 3.0634257371925533e-06,
+      "loss": 0.8971,
+      "step": 110530
+    },
+    {
+      "epoch": 7.510191602119853,
+      "grad_norm": 7.96875,
+      "learning_rate": 3.0613024867509173e-06,
+      "loss": 0.9044,
+      "step": 110535
+    },
+    {
+      "epoch": 7.510531322190515,
+      "grad_norm": 8.25,
+      "learning_rate": 3.0591792363092813e-06,
+      "loss": 0.8781,
+      "step": 110540
+    },
+    {
+      "epoch": 7.510871042261177,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0570559858676453e-06,
+      "loss": 0.8568,
+      "step": 110545
+    },
+    {
+      "epoch": 7.511210762331839,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.0549327354260093e-06,
+      "loss": 0.8603,
+      "step": 110550
+    },
+    {
+      "epoch": 7.5115504824025,
+      "grad_norm": 8.25,
+      "learning_rate": 3.052809484984373e-06,
+      "loss": 0.8437,
+      "step": 110555
+    },
+    {
+      "epoch": 7.511890202473162,
+      "grad_norm": 8.125,
+      "learning_rate": 3.050686234542737e-06,
+      "loss": 0.8285,
+      "step": 110560
+    },
+    {
+      "epoch": 7.512229922543824,
+      "grad_norm": 8.375,
+      "learning_rate": 3.048562984101101e-06,
+      "loss": 0.784,
+      "step": 110565
+    },
+    {
+      "epoch": 7.512569642614485,
+      "grad_norm": 8.0,
+      "learning_rate": 3.0464397336594645e-06,
+      "loss": 0.7844,
+      "step": 110570
+    },
+    {
+      "epoch": 7.512909362685147,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.044316483217829e-06,
+      "loss": 0.8151,
+      "step": 110575
+    },
+    {
+      "epoch": 7.513249082755809,
+      "grad_norm": 8.0,
+      "learning_rate": 3.0421932327761925e-06,
+      "loss": 1.0034,
+      "step": 110580
+    },
+    {
+      "epoch": 7.513588802826471,
+      "grad_norm": 7.21875,
+      "learning_rate": 3.0400699823345565e-06,
+      "loss": 0.761,
+      "step": 110585
+    },
+    {
+      "epoch": 7.513928522897133,
+      "grad_norm": 9.375,
+      "learning_rate": 3.0379467318929205e-06,
+      "loss": 0.8151,
+      "step": 110590
+    },
+    {
+      "epoch": 7.514268242967795,
+      "grad_norm": 7.875,
+      "learning_rate": 3.035823481451284e-06,
+      "loss": 0.803,
+      "step": 110595
+    },
+    {
+      "epoch": 7.514607963038456,
+      "grad_norm": 7.84375,
+      "learning_rate": 3.0337002310096485e-06,
+      "loss": 0.8892,
+      "step": 110600
+    },
+    {
+      "epoch": 7.514947683109118,
+      "grad_norm": 7.40625,
+      "learning_rate": 3.031576980568012e-06,
+      "loss": 0.8587,
+      "step": 110605
+    },
+    {
+      "epoch": 7.51528740317978,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.029453730126376e-06,
+      "loss": 0.8845,
+      "step": 110610
+    },
+    {
+      "epoch": 7.515627123250441,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.02733047968474e-06,
+      "loss": 0.8297,
+      "step": 110615
+    },
+    {
+      "epoch": 7.515966843321103,
+      "grad_norm": 8.5,
+      "learning_rate": 3.0252072292431037e-06,
+      "loss": 0.8316,
+      "step": 110620
+    },
+    {
+      "epoch": 7.516306563391765,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.0230839788014677e-06,
+      "loss": 0.8525,
+      "step": 110625
+    },
+    {
+      "epoch": 7.516646283462427,
+      "grad_norm": 8.375,
+      "learning_rate": 3.0209607283598317e-06,
+      "loss": 0.9396,
+      "step": 110630
+    },
+    {
+      "epoch": 7.516986003533089,
+      "grad_norm": 8.125,
+      "learning_rate": 3.0188374779181957e-06,
+      "loss": 0.7814,
+      "step": 110635
+    },
+    {
+      "epoch": 7.517325723603751,
+      "grad_norm": 7.4375,
+      "learning_rate": 3.0167142274765593e-06,
+      "loss": 0.7707,
+      "step": 110640
+    },
+    {
+      "epoch": 7.517665443674412,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.0145909770349233e-06,
+      "loss": 0.8562,
+      "step": 110645
+    },
+    {
+      "epoch": 7.518005163745074,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.0124677265932873e-06,
+      "loss": 0.859,
+      "step": 110650
+    },
+    {
+      "epoch": 7.518344883815736,
+      "grad_norm": 8.5625,
+      "learning_rate": 3.0103444761516513e-06,
+      "loss": 0.6722,
+      "step": 110655
+    },
+    {
+      "epoch": 7.518684603886397,
+      "grad_norm": 8.375,
+      "learning_rate": 3.0082212257100153e-06,
+      "loss": 0.7824,
+      "step": 110660
+    },
+    {
+      "epoch": 7.519024323957059,
+      "grad_norm": 7.625,
+      "learning_rate": 3.006097975268379e-06,
+      "loss": 0.8709,
+      "step": 110665
+    },
+    {
+      "epoch": 7.519364044027721,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.003974724826743e-06,
+      "loss": 0.9298,
+      "step": 110670
+    },
+    {
+      "epoch": 7.519703764098383,
+      "grad_norm": 8.1875,
+      "learning_rate": 3.001851474385107e-06,
+      "loss": 0.8403,
+      "step": 110675
+    },
+    {
+      "epoch": 7.520043484169045,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.9997282239434704e-06,
+      "loss": 0.9526,
+      "step": 110680
+    },
+    {
+      "epoch": 7.520383204239707,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.997604973501835e-06,
+      "loss": 0.8784,
+      "step": 110685
+    },
+    {
+      "epoch": 7.520722924310368,
+      "grad_norm": 8.625,
+      "learning_rate": 2.9954817230601985e-06,
+      "loss": 0.7722,
+      "step": 110690
+    },
+    {
+      "epoch": 7.52106264438103,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9933584726185625e-06,
+      "loss": 0.8415,
+      "step": 110695
+    },
+    {
+      "epoch": 7.521402364451692,
+      "grad_norm": 8.0,
+      "learning_rate": 2.9912352221769265e-06,
+      "loss": 0.7179,
+      "step": 110700
+    },
+    {
+      "epoch": 7.521742084522353,
+      "grad_norm": 8.375,
+      "learning_rate": 2.98911197173529e-06,
+      "loss": 0.865,
+      "step": 110705
+    },
+    {
+      "epoch": 7.522081804593015,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.986988721293654e-06,
+      "loss": 0.8849,
+      "step": 110710
+    },
+    {
+      "epoch": 7.522421524663677,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.984865470852018e-06,
+      "loss": 0.8243,
+      "step": 110715
+    },
+    {
+      "epoch": 7.522761244734339,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.982742220410382e-06,
+      "loss": 0.7764,
+      "step": 110720
+    },
+    {
+      "epoch": 7.523100964805001,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.980618969968746e-06,
+      "loss": 0.8199,
+      "step": 110725
+    },
+    {
+      "epoch": 7.523440684875663,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.9784957195271096e-06,
+      "loss": 0.8482,
+      "step": 110730
+    },
+    {
+      "epoch": 7.523780404946324,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.9763724690854736e-06,
+      "loss": 0.8944,
+      "step": 110735
+    },
+    {
+      "epoch": 7.524120125016986,
+      "grad_norm": 6.375,
+      "learning_rate": 2.9742492186438376e-06,
+      "loss": 0.8517,
+      "step": 110740
+    },
+    {
+      "epoch": 7.524459845087648,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.9721259682022017e-06,
+      "loss": 0.9263,
+      "step": 110745
+    },
+    {
+      "epoch": 7.524799565158309,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.9700027177605652e-06,
+      "loss": 0.8404,
+      "step": 110750
+    },
+    {
+      "epoch": 7.525139285228971,
+      "grad_norm": 8.75,
+      "learning_rate": 2.9678794673189292e-06,
+      "loss": 0.7673,
+      "step": 110755
+    },
+    {
+      "epoch": 7.525479005299633,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9657562168772932e-06,
+      "loss": 0.8325,
+      "step": 110760
+    },
+    {
+      "epoch": 7.525818725370295,
+      "grad_norm": 8.25,
+      "learning_rate": 2.9636329664356572e-06,
+      "loss": 0.8649,
+      "step": 110765
+    },
+    {
+      "epoch": 7.526158445440957,
+      "grad_norm": 8.0,
+      "learning_rate": 2.9615097159940212e-06,
+      "loss": 0.7803,
+      "step": 110770
+    },
+    {
+      "epoch": 7.526498165511619,
+      "grad_norm": 8.25,
+      "learning_rate": 2.959386465552385e-06,
+      "loss": 0.893,
+      "step": 110775
+    },
+    {
+      "epoch": 7.52683788558228,
+      "grad_norm": 7.375,
+      "learning_rate": 2.9572632151107493e-06,
+      "loss": 0.8136,
+      "step": 110780
+    },
+    {
+      "epoch": 7.527177605652942,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.955139964669113e-06,
+      "loss": 0.8408,
+      "step": 110785
+    },
+    {
+      "epoch": 7.527517325723604,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.9530167142274764e-06,
+      "loss": 0.7619,
+      "step": 110790
+    },
+    {
+      "epoch": 7.527857045794265,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.950893463785841e-06,
+      "loss": 0.8527,
+      "step": 110795
+    },
+    {
+      "epoch": 7.528196765864927,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9487702133442044e-06,
+      "loss": 0.9224,
+      "step": 110800
+    },
+    {
+      "epoch": 7.528536485935589,
+      "grad_norm": 7.5,
+      "learning_rate": 2.9466469629025684e-06,
+      "loss": 0.8507,
+      "step": 110805
+    },
+    {
+      "epoch": 7.528876206006251,
+      "grad_norm": 7.75,
+      "learning_rate": 2.9445237124609324e-06,
+      "loss": 0.8255,
+      "step": 110810
+    },
+    {
+      "epoch": 7.529215926076913,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.942400462019296e-06,
+      "loss": 0.8682,
+      "step": 110815
+    },
+    {
+      "epoch": 7.529555646147575,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.94027721157766e-06,
+      "loss": 0.7955,
+      "step": 110820
+    },
+    {
+      "epoch": 7.529895366218236,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.938153961136024e-06,
+      "loss": 0.7903,
+      "step": 110825
+    },
+    {
+      "epoch": 7.530235086288898,
+      "grad_norm": 9.0,
+      "learning_rate": 2.936030710694388e-06,
+      "loss": 0.8892,
+      "step": 110830
+    },
+    {
+      "epoch": 7.53057480635956,
+      "grad_norm": 8.375,
+      "learning_rate": 2.933907460252752e-06,
+      "loss": 0.8832,
+      "step": 110835
+    },
+    {
+      "epoch": 7.530914526430221,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.9317842098111156e-06,
+      "loss": 0.8224,
+      "step": 110840
+    },
+    {
+      "epoch": 7.531254246500883,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.9296609593694796e-06,
+      "loss": 0.864,
+      "step": 110845
+    },
+    {
+      "epoch": 7.5315939665715455,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.9275377089278436e-06,
+      "loss": 0.8332,
+      "step": 110850
+    },
+    {
+      "epoch": 7.531933686642207,
+      "grad_norm": 8.125,
+      "learning_rate": 2.9254144584862076e-06,
+      "loss": 0.7943,
+      "step": 110855
+    },
+    {
+      "epoch": 7.532273406712869,
+      "grad_norm": 7.875,
+      "learning_rate": 2.923291208044571e-06,
+      "loss": 0.8387,
+      "step": 110860
+    },
+    {
+      "epoch": 7.532613126783531,
+      "grad_norm": 8.0,
+      "learning_rate": 2.9211679576029356e-06,
+      "loss": 0.868,
+      "step": 110865
+    },
+    {
+      "epoch": 7.532952846854192,
+      "grad_norm": 8.25,
+      "learning_rate": 2.919044707161299e-06,
+      "loss": 0.8275,
+      "step": 110870
+    },
+    {
+      "epoch": 7.533292566924854,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.916921456719663e-06,
+      "loss": 0.8877,
+      "step": 110875
+    },
+    {
+      "epoch": 7.533632286995516,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.9147982062780272e-06,
+      "loss": 0.9005,
+      "step": 110880
+    },
+    {
+      "epoch": 7.533972007066177,
+      "grad_norm": 7.75,
+      "learning_rate": 2.912674955836391e-06,
+      "loss": 0.7762,
+      "step": 110885
+    },
+    {
+      "epoch": 7.534311727136839,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.910551705394755e-06,
+      "loss": 0.819,
+      "step": 110890
+    },
+    {
+      "epoch": 7.5346514472075015,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.908428454953119e-06,
+      "loss": 0.9313,
+      "step": 110895
+    },
+    {
+      "epoch": 7.534991167278163,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.9063052045114824e-06,
+      "loss": 0.7891,
+      "step": 110900
+    },
+    {
+      "epoch": 7.535330887348825,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.904181954069847e-06,
+      "loss": 0.8886,
+      "step": 110905
+    },
+    {
+      "epoch": 7.535670607419487,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.9020587036282104e-06,
+      "loss": 0.8781,
+      "step": 110910
+    },
+    {
+      "epoch": 7.536010327490148,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8999354531865744e-06,
+      "loss": 0.9214,
+      "step": 110915
+    },
+    {
+      "epoch": 7.53635004756081,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.8978122027449384e-06,
+      "loss": 0.7907,
+      "step": 110920
+    },
+    {
+      "epoch": 7.536689767631472,
+      "grad_norm": 9.125,
+      "learning_rate": 2.895688952303302e-06,
+      "loss": 0.7946,
+      "step": 110925
+    },
+    {
+      "epoch": 7.537029487702133,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.893565701861666e-06,
+      "loss": 0.9097,
+      "step": 110930
+    },
+    {
+      "epoch": 7.537369207772795,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.89144245142003e-06,
+      "loss": 0.9872,
+      "step": 110935
+    },
+    {
+      "epoch": 7.5377089278434575,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.889319200978394e-06,
+      "loss": 0.8255,
+      "step": 110940
+    },
+    {
+      "epoch": 7.538048647914119,
+      "grad_norm": 8.125,
+      "learning_rate": 2.887195950536758e-06,
+      "loss": 0.8553,
+      "step": 110945
+    },
+    {
+      "epoch": 7.538388367984781,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.885072700095122e-06,
+      "loss": 0.7805,
+      "step": 110950
+    },
+    {
+      "epoch": 7.538728088055443,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.8829494496534856e-06,
+      "loss": 0.8399,
+      "step": 110955
+    },
+    {
+      "epoch": 7.539067808126104,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8808261992118496e-06,
+      "loss": 0.9263,
+      "step": 110960
+    },
+    {
+      "epoch": 7.539407528196766,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8787029487702136e-06,
+      "loss": 0.7749,
+      "step": 110965
+    },
+    {
+      "epoch": 7.539747248267427,
+      "grad_norm": 8.0,
+      "learning_rate": 2.876579698328577e-06,
+      "loss": 0.8599,
+      "step": 110970
+    },
+    {
+      "epoch": 7.540086968338089,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8744564478869416e-06,
+      "loss": 0.9519,
+      "step": 110975
+    },
+    {
+      "epoch": 7.540426688408751,
+      "grad_norm": 8.625,
+      "learning_rate": 2.872333197445305e-06,
+      "loss": 0.7993,
+      "step": 110980
+    },
+    {
+      "epoch": 7.540766408479413,
+      "grad_norm": 8.375,
+      "learning_rate": 2.870209947003669e-06,
+      "loss": 0.9105,
+      "step": 110985
+    },
+    {
+      "epoch": 7.541106128550075,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.868086696562033e-06,
+      "loss": 0.8103,
+      "step": 110990
+    },
+    {
+      "epoch": 7.541445848620737,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.8659634461203968e-06,
+      "loss": 0.8295,
+      "step": 110995
+    },
+    {
+      "epoch": 7.541785568691398,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.8638401956787608e-06,
+      "loss": 0.9379,
+      "step": 111000
+    },
+    {
+      "epoch": 7.54212528876206,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.8617169452371248e-06,
+      "loss": 0.913,
+      "step": 111005
+    },
+    {
+      "epoch": 7.542465008832722,
+      "grad_norm": 8.375,
+      "learning_rate": 2.8595936947954884e-06,
+      "loss": 0.8864,
+      "step": 111010
+    },
+    {
+      "epoch": 7.542804728903383,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.8574704443538528e-06,
+      "loss": 0.8812,
+      "step": 111015
+    },
+    {
+      "epoch": 7.543144448974045,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.8553471939122164e-06,
+      "loss": 0.8894,
+      "step": 111020
+    },
+    {
+      "epoch": 7.543484169044707,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.8532239434705804e-06,
+      "loss": 0.8404,
+      "step": 111025
+    },
+    {
+      "epoch": 7.543823889115369,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.8511006930289444e-06,
+      "loss": 0.8086,
+      "step": 111030
+    },
+    {
+      "epoch": 7.544163609186031,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.8489774425873084e-06,
+      "loss": 0.9226,
+      "step": 111035
+    },
+    {
+      "epoch": 7.544503329256693,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.846854192145672e-06,
+      "loss": 0.7979,
+      "step": 111040
+    },
+    {
+      "epoch": 7.544843049327354,
+      "grad_norm": 7.125,
+      "learning_rate": 2.844730941704036e-06,
+      "loss": 0.8449,
+      "step": 111045
+    },
+    {
+      "epoch": 7.545182769398016,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8426076912624e-06,
+      "loss": 0.8351,
+      "step": 111050
+    },
+    {
+      "epoch": 7.545522489468678,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.840484440820764e-06,
+      "loss": 0.8831,
+      "step": 111055
+    },
+    {
+      "epoch": 7.545862209539339,
+      "grad_norm": 8.0,
+      "learning_rate": 2.838361190379128e-06,
+      "loss": 0.7869,
+      "step": 111060
+    },
+    {
+      "epoch": 7.546201929610001,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.8362379399374915e-06,
+      "loss": 0.9331,
+      "step": 111065
+    },
+    {
+      "epoch": 7.546541649680663,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8341146894958555e-06,
+      "loss": 0.8814,
+      "step": 111070
+    },
+    {
+      "epoch": 7.546881369751325,
+      "grad_norm": 7.375,
+      "learning_rate": 2.8319914390542196e-06,
+      "loss": 0.8796,
+      "step": 111075
+    },
+    {
+      "epoch": 7.547221089821987,
+      "grad_norm": 9.3125,
+      "learning_rate": 2.829868188612583e-06,
+      "loss": 0.9221,
+      "step": 111080
+    },
+    {
+      "epoch": 7.547560809892649,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.8277449381709476e-06,
+      "loss": 0.7531,
+      "step": 111085
+    },
+    {
+      "epoch": 7.54790052996331,
+      "grad_norm": 7.125,
+      "learning_rate": 2.825621687729311e-06,
+      "loss": 0.8511,
+      "step": 111090
+    },
+    {
+      "epoch": 7.548240250033972,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.823498437287675e-06,
+      "loss": 0.8287,
+      "step": 111095
+    },
+    {
+      "epoch": 7.548579970104634,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.821375186846039e-06,
+      "loss": 0.8457,
+      "step": 111100
+    },
+    {
+      "epoch": 7.548919690175295,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.8192519364044027e-06,
+      "loss": 0.8291,
+      "step": 111105
+    },
+    {
+      "epoch": 7.549259410245957,
+      "grad_norm": 6.28125,
+      "learning_rate": 2.8171286859627667e-06,
+      "loss": 0.7827,
+      "step": 111110
+    },
+    {
+      "epoch": 7.5495991303166194,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.8150054355211307e-06,
+      "loss": 0.7983,
+      "step": 111115
+    },
+    {
+      "epoch": 7.549938850387281,
+      "grad_norm": 8.375,
+      "learning_rate": 2.8128821850794947e-06,
+      "loss": 0.9327,
+      "step": 111120
+    },
+    {
+      "epoch": 7.550278570457943,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8107589346378587e-06,
+      "loss": 0.8736,
+      "step": 111125
+    },
+    {
+      "epoch": 7.550618290528605,
+      "grad_norm": 8.0,
+      "learning_rate": 2.8086356841962223e-06,
+      "loss": 0.8441,
+      "step": 111130
+    },
+    {
+      "epoch": 7.550958010599266,
+      "grad_norm": 8.125,
+      "learning_rate": 2.8065124337545863e-06,
+      "loss": 0.7607,
+      "step": 111135
+    },
+    {
+      "epoch": 7.551297730669928,
+      "grad_norm": 8.25,
+      "learning_rate": 2.8043891833129503e-06,
+      "loss": 0.8531,
+      "step": 111140
+    },
+    {
+      "epoch": 7.55163745074059,
+      "grad_norm": 7.625,
+      "learning_rate": 2.8022659328713143e-06,
+      "loss": 0.8604,
+      "step": 111145
+    },
+    {
+      "epoch": 7.551977170811251,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.800142682429678e-06,
+      "loss": 0.885,
+      "step": 111150
+    },
+    {
+      "epoch": 7.552316890881913,
+      "grad_norm": 7.125,
+      "learning_rate": 2.798019431988042e-06,
+      "loss": 0.8206,
+      "step": 111155
+    },
+    {
+      "epoch": 7.5526566109525755,
+      "grad_norm": 7.25,
+      "learning_rate": 2.795896181546406e-06,
+      "loss": 0.8575,
+      "step": 111160
+    },
+    {
+      "epoch": 7.552996331023237,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.79377293110477e-06,
+      "loss": 0.9012,
+      "step": 111165
+    },
+    {
+      "epoch": 7.553336051093899,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.791649680663134e-06,
+      "loss": 0.8409,
+      "step": 111170
+    },
+    {
+      "epoch": 7.55367577116456,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.7895264302214975e-06,
+      "loss": 0.7694,
+      "step": 111175
+    },
+    {
+      "epoch": 7.554015491235222,
+      "grad_norm": 8.5,
+      "learning_rate": 2.7874031797798615e-06,
+      "loss": 0.8545,
+      "step": 111180
+    },
+    {
+      "epoch": 7.554355211305884,
+      "grad_norm": 7.625,
+      "learning_rate": 2.7852799293382255e-06,
+      "loss": 0.9103,
+      "step": 111185
+    },
+    {
+      "epoch": 7.554694931376545,
+      "grad_norm": 9.0,
+      "learning_rate": 2.783156678896589e-06,
+      "loss": 0.8348,
+      "step": 111190
+    },
+    {
+      "epoch": 7.555034651447207,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.7810334284549535e-06,
+      "loss": 0.7493,
+      "step": 111195
+    },
+    {
+      "epoch": 7.555374371517869,
+      "grad_norm": 8.5,
+      "learning_rate": 2.778910178013317e-06,
+      "loss": 0.8568,
+      "step": 111200
+    },
+    {
+      "epoch": 7.555714091588531,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.776786927571681e-06,
+      "loss": 0.8998,
+      "step": 111205
+    },
+    {
+      "epoch": 7.556053811659193,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.774663677130045e-06,
+      "loss": 0.778,
+      "step": 111210
+    },
+    {
+      "epoch": 7.556393531729855,
+      "grad_norm": 8.5,
+      "learning_rate": 2.7725404266884087e-06,
+      "loss": 0.7809,
+      "step": 111215
+    },
+    {
+      "epoch": 7.556733251800516,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.7704171762467727e-06,
+      "loss": 0.777,
+      "step": 111220
+    },
+    {
+      "epoch": 7.557072971871178,
+      "grad_norm": 8.0,
+      "learning_rate": 2.7682939258051367e-06,
+      "loss": 0.7987,
+      "step": 111225
+    },
+    {
+      "epoch": 7.55741269194184,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7661706753635007e-06,
+      "loss": 0.7917,
+      "step": 111230
+    },
+    {
+      "epoch": 7.557752412012501,
+      "grad_norm": 9.25,
+      "learning_rate": 2.7640474249218647e-06,
+      "loss": 0.792,
+      "step": 111235
+    },
+    {
+      "epoch": 7.558092132083163,
+      "grad_norm": 7.875,
+      "learning_rate": 2.7619241744802283e-06,
+      "loss": 0.7835,
+      "step": 111240
+    },
+    {
+      "epoch": 7.558431852153825,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.7598009240385923e-06,
+      "loss": 0.8258,
+      "step": 111245
+    },
+    {
+      "epoch": 7.558771572224487,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.7576776735969563e-06,
+      "loss": 0.9064,
+      "step": 111250
+    },
+    {
+      "epoch": 7.559111292295149,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.7555544231553203e-06,
+      "loss": 0.8334,
+      "step": 111255
+    },
+    {
+      "epoch": 7.559451012365811,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.753431172713684e-06,
+      "loss": 0.8801,
+      "step": 111260
+    },
+    {
+      "epoch": 7.559790732436472,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.751307922272048e-06,
+      "loss": 0.8636,
+      "step": 111265
+    },
+    {
+      "epoch": 7.560130452507134,
+      "grad_norm": 8.875,
+      "learning_rate": 2.749184671830412e-06,
+      "loss": 0.8711,
+      "step": 111270
+    },
+    {
+      "epoch": 7.560470172577796,
+      "grad_norm": 7.875,
+      "learning_rate": 2.747061421388776e-06,
+      "loss": 0.8998,
+      "step": 111275
+    },
+    {
+      "epoch": 7.560809892648457,
+      "grad_norm": 8.25,
+      "learning_rate": 2.74493817094714e-06,
+      "loss": 0.91,
+      "step": 111280
+    },
+    {
+      "epoch": 7.561149612719119,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7428149205055035e-06,
+      "loss": 0.7789,
+      "step": 111285
+    },
+    {
+      "epoch": 7.561489332789781,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7406916700638675e-06,
+      "loss": 0.802,
+      "step": 111290
+    },
+    {
+      "epoch": 7.561829052860443,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.7385684196222315e-06,
+      "loss": 0.7993,
+      "step": 111295
+    },
+    {
+      "epoch": 7.562168772931105,
+      "grad_norm": 9.375,
+      "learning_rate": 2.736445169180595e-06,
+      "loss": 0.9126,
+      "step": 111300
+    },
+    {
+      "epoch": 7.562508493001767,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.7343219187389595e-06,
+      "loss": 0.8661,
+      "step": 111305
+    },
+    {
+      "epoch": 7.562848213072428,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.732198668297323e-06,
+      "loss": 0.819,
+      "step": 111310
+    },
+    {
+      "epoch": 7.56318793314309,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.730075417855687e-06,
+      "loss": 0.8796,
+      "step": 111315
+    },
+    {
+      "epoch": 7.563527653213752,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.727952167414051e-06,
+      "loss": 0.7985,
+      "step": 111320
+    },
+    {
+      "epoch": 7.563867373284413,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.7258289169724147e-06,
+      "loss": 0.8235,
+      "step": 111325
+    },
+    {
+      "epoch": 7.564207093355075,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.7237056665307787e-06,
+      "loss": 0.7837,
+      "step": 111330
+    },
+    {
+      "epoch": 7.564546813425737,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.7215824160891427e-06,
+      "loss": 0.8721,
+      "step": 111335
+    },
+    {
+      "epoch": 7.564886533496399,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.7194591656475067e-06,
+      "loss": 0.8079,
+      "step": 111340
+    },
+    {
+      "epoch": 7.565226253567061,
+      "grad_norm": 8.25,
+      "learning_rate": 2.7173359152058707e-06,
+      "loss": 0.8178,
+      "step": 111345
+    },
+    {
+      "epoch": 7.565565973637723,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.7152126647642343e-06,
+      "loss": 0.8767,
+      "step": 111350
+    },
+    {
+      "epoch": 7.565905693708384,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.7130894143225983e-06,
+      "loss": 0.7216,
+      "step": 111355
+    },
+    {
+      "epoch": 7.566245413779046,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.7109661638809623e-06,
+      "loss": 0.8134,
+      "step": 111360
+    },
+    {
+      "epoch": 7.566585133849708,
+      "grad_norm": 7.25,
+      "learning_rate": 2.7088429134393263e-06,
+      "loss": 0.8887,
+      "step": 111365
+    },
+    {
+      "epoch": 7.566924853920369,
+      "grad_norm": 8.125,
+      "learning_rate": 2.70671966299769e-06,
+      "loss": 0.8804,
+      "step": 111370
+    },
+    {
+      "epoch": 7.567264573991031,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.704596412556054e-06,
+      "loss": 0.856,
+      "step": 111375
+    },
+    {
+      "epoch": 7.567604294061693,
+      "grad_norm": 7.625,
+      "learning_rate": 2.702473162114418e-06,
+      "loss": 0.8613,
+      "step": 111380
+    },
+    {
+      "epoch": 7.567944014132355,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.7003499116727814e-06,
+      "loss": 0.8185,
+      "step": 111385
+    },
+    {
+      "epoch": 7.568283734203017,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.698226661231146e-06,
+      "loss": 0.8654,
+      "step": 111390
+    },
+    {
+      "epoch": 7.568623454273679,
+      "grad_norm": 8.625,
+      "learning_rate": 2.6961034107895094e-06,
+      "loss": 0.8366,
+      "step": 111395
+    },
+    {
+      "epoch": 7.56896317434434,
+      "grad_norm": 8.25,
+      "learning_rate": 2.6939801603478735e-06,
+      "loss": 0.8566,
+      "step": 111400
+    },
+    {
+      "epoch": 7.569302894415002,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.6918569099062375e-06,
+      "loss": 0.8499,
+      "step": 111405
+    },
+    {
+      "epoch": 7.569642614485664,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.689733659464601e-06,
+      "loss": 0.8824,
+      "step": 111410
+    },
+    {
+      "epoch": 7.569982334556325,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.6876104090229655e-06,
+      "loss": 0.8094,
+      "step": 111415
+    },
+    {
+      "epoch": 7.570322054626987,
+      "grad_norm": 7.0,
+      "learning_rate": 2.685487158581329e-06,
+      "loss": 0.8126,
+      "step": 111420
+    },
+    {
+      "epoch": 7.5706617746976494,
+      "grad_norm": 8.5,
+      "learning_rate": 2.683363908139693e-06,
+      "loss": 0.9577,
+      "step": 111425
+    },
+    {
+      "epoch": 7.571001494768311,
+      "grad_norm": 9.125,
+      "learning_rate": 2.681240657698057e-06,
+      "loss": 0.8444,
+      "step": 111430
+    },
+    {
+      "epoch": 7.571341214838973,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.6791174072564206e-06,
+      "loss": 0.7526,
+      "step": 111435
+    },
+    {
+      "epoch": 7.571680934909635,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.6769941568147846e-06,
+      "loss": 0.8528,
+      "step": 111440
+    },
+    {
+      "epoch": 7.572020654980296,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6748709063731486e-06,
+      "loss": 0.8245,
+      "step": 111445
+    },
+    {
+      "epoch": 7.572360375050958,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.6727476559315126e-06,
+      "loss": 0.7842,
+      "step": 111450
+    },
+    {
+      "epoch": 7.57270009512162,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.6706244054898766e-06,
+      "loss": 0.946,
+      "step": 111455
+    },
+    {
+      "epoch": 7.573039815192281,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6685011550482402e-06,
+      "loss": 0.7582,
+      "step": 111460
+    },
+    {
+      "epoch": 7.573379535262943,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.6663779046066042e-06,
+      "loss": 0.7856,
+      "step": 111465
+    },
+    {
+      "epoch": 7.5737192553336055,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.6642546541649682e-06,
+      "loss": 0.8177,
+      "step": 111470
+    },
+    {
+      "epoch": 7.574058975404267,
+      "grad_norm": 7.625,
+      "learning_rate": 2.6621314037233322e-06,
+      "loss": 0.8486,
+      "step": 111475
+    },
+    {
+      "epoch": 7.574398695474929,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.660008153281696e-06,
+      "loss": 0.9525,
+      "step": 111480
+    },
+    {
+      "epoch": 7.574738415545591,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.6578849028400602e-06,
+      "loss": 0.8399,
+      "step": 111485
+    },
+    {
+      "epoch": 7.575078135616252,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.655761652398424e-06,
+      "loss": 0.8788,
+      "step": 111490
+    },
+    {
+      "epoch": 7.575417855686914,
+      "grad_norm": 8.75,
+      "learning_rate": 2.6536384019567874e-06,
+      "loss": 0.8518,
+      "step": 111495
+    },
+    {
+      "epoch": 7.575757575757576,
+      "grad_norm": 8.125,
+      "learning_rate": 2.651515151515152e-06,
+      "loss": 0.8135,
+      "step": 111500
+    },
+    {
+      "epoch": 7.576097295828237,
+      "grad_norm": 8.125,
+      "learning_rate": 2.6493919010735154e-06,
+      "loss": 0.8165,
+      "step": 111505
+    },
+    {
+      "epoch": 7.576437015898899,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.6472686506318794e-06,
+      "loss": 0.7815,
+      "step": 111510
+    },
+    {
+      "epoch": 7.5767767359695615,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.6451454001902434e-06,
+      "loss": 0.7714,
+      "step": 111515
+    },
+    {
+      "epoch": 7.577116456040223,
+      "grad_norm": 8.125,
+      "learning_rate": 2.643022149748607e-06,
+      "loss": 0.9205,
+      "step": 111520
+    },
+    {
+      "epoch": 7.577456176110885,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6408988993069714e-06,
+      "loss": 0.7703,
+      "step": 111525
+    },
+    {
+      "epoch": 7.577795896181547,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.638775648865335e-06,
+      "loss": 0.8156,
+      "step": 111530
+    },
+    {
+      "epoch": 7.578135616252208,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.636652398423699e-06,
+      "loss": 0.8598,
+      "step": 111535
+    },
+    {
+      "epoch": 7.57847533632287,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.634529147982063e-06,
+      "loss": 0.7959,
+      "step": 111540
+    },
+    {
+      "epoch": 7.578815056393532,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.6324058975404266e-06,
+      "loss": 0.7385,
+      "step": 111545
+    },
+    {
+      "epoch": 7.579154776464193,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.6302826470987906e-06,
+      "loss": 0.8414,
+      "step": 111550
+    },
+    {
+      "epoch": 7.579494496534855,
+      "grad_norm": 8.375,
+      "learning_rate": 2.6281593966571546e-06,
+      "loss": 0.8013,
+      "step": 111555
+    },
+    {
+      "epoch": 7.5798342166055175,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.6260361462155186e-06,
+      "loss": 0.8617,
+      "step": 111560
+    },
+    {
+      "epoch": 7.580173936676179,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.623912895773882e-06,
+      "loss": 0.7949,
+      "step": 111565
+    },
+    {
+      "epoch": 7.580513656746841,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.6217896453322466e-06,
+      "loss": 0.7075,
+      "step": 111570
+    },
+    {
+      "epoch": 7.580853376817503,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.61966639489061e-06,
+      "loss": 0.8646,
+      "step": 111575
+    },
+    {
+      "epoch": 7.581193096888164,
+      "grad_norm": 8.0,
+      "learning_rate": 2.617543144448974e-06,
+      "loss": 0.8463,
+      "step": 111580
+    },
+    {
+      "epoch": 7.581532816958826,
+      "grad_norm": 8.125,
+      "learning_rate": 2.615419894007338e-06,
+      "loss": 0.8118,
+      "step": 111585
+    },
+    {
+      "epoch": 7.581872537029488,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.6132966435657018e-06,
+      "loss": 0.9067,
+      "step": 111590
+    },
+    {
+      "epoch": 7.582212257100149,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.6111733931240662e-06,
+      "loss": 0.867,
+      "step": 111595
+    },
+    {
+      "epoch": 7.582551977170811,
+      "grad_norm": 8.375,
+      "learning_rate": 2.60905014268243e-06,
+      "loss": 0.8405,
+      "step": 111600
+    },
+    {
+      "epoch": 7.5828916972414735,
+      "grad_norm": 7.875,
+      "learning_rate": 2.6069268922407934e-06,
+      "loss": 0.8888,
+      "step": 111605
+    },
+    {
+      "epoch": 7.583231417312135,
+      "grad_norm": 8.375,
+      "learning_rate": 2.604803641799158e-06,
+      "loss": 0.7834,
+      "step": 111610
+    },
+    {
+      "epoch": 7.583571137382797,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.6026803913575214e-06,
+      "loss": 0.8359,
+      "step": 111615
+    },
+    {
+      "epoch": 7.583910857453459,
+      "grad_norm": 8.0,
+      "learning_rate": 2.6005571409158854e-06,
+      "loss": 0.8671,
+      "step": 111620
+    },
+    {
+      "epoch": 7.58425057752412,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.5984338904742494e-06,
+      "loss": 0.8482,
+      "step": 111625
+    },
+    {
+      "epoch": 7.584590297594782,
+      "grad_norm": 8.625,
+      "learning_rate": 2.596310640032613e-06,
+      "loss": 0.8113,
+      "step": 111630
+    },
+    {
+      "epoch": 7.584930017665444,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.5941873895909774e-06,
+      "loss": 0.8804,
+      "step": 111635
+    },
+    {
+      "epoch": 7.585269737736105,
+      "grad_norm": 8.625,
+      "learning_rate": 2.592064139149341e-06,
+      "loss": 0.7532,
+      "step": 111640
+    },
+    {
+      "epoch": 7.585609457806767,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.589940888707705e-06,
+      "loss": 0.8051,
+      "step": 111645
+    },
+    {
+      "epoch": 7.585949177877429,
+      "grad_norm": 8.875,
+      "learning_rate": 2.587817638266069e-06,
+      "loss": 0.7444,
+      "step": 111650
+    },
+    {
+      "epoch": 7.586288897948091,
+      "grad_norm": 8.25,
+      "learning_rate": 2.585694387824433e-06,
+      "loss": 0.86,
+      "step": 111655
+    },
+    {
+      "epoch": 7.586628618018753,
+      "grad_norm": 7.0,
+      "learning_rate": 2.5835711373827966e-06,
+      "loss": 0.8616,
+      "step": 111660
+    },
+    {
+      "epoch": 7.586968338089414,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.5814478869411606e-06,
+      "loss": 0.8117,
+      "step": 111665
+    },
+    {
+      "epoch": 7.587308058160076,
+      "grad_norm": 8.875,
+      "learning_rate": 2.5793246364995246e-06,
+      "loss": 0.8689,
+      "step": 111670
+    },
+    {
+      "epoch": 7.587647778230738,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.577201386057888e-06,
+      "loss": 0.8551,
+      "step": 111675
+    },
+    {
+      "epoch": 7.587987498301399,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.5750781356162526e-06,
+      "loss": 0.7415,
+      "step": 111680
+    },
+    {
+      "epoch": 7.588327218372061,
+      "grad_norm": 8.125,
+      "learning_rate": 2.572954885174616e-06,
+      "loss": 0.8585,
+      "step": 111685
+    },
+    {
+      "epoch": 7.588666938442723,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.57083163473298e-06,
+      "loss": 0.9458,
+      "step": 111690
+    },
+    {
+      "epoch": 7.589006658513385,
+      "grad_norm": 6.53125,
+      "learning_rate": 2.568708384291344e-06,
+      "loss": 0.831,
+      "step": 111695
+    },
+    {
+      "epoch": 7.589346378584047,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.5665851338497078e-06,
+      "loss": 0.765,
+      "step": 111700
+    },
+    {
+      "epoch": 7.589686098654709,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.564461883408072e-06,
+      "loss": 0.7726,
+      "step": 111705
+    },
+    {
+      "epoch": 7.59002581872537,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.5623386329664358e-06,
+      "loss": 0.9264,
+      "step": 111710
+    },
+    {
+      "epoch": 7.590365538796032,
+      "grad_norm": 6.21875,
+      "learning_rate": 2.5602153825247993e-06,
+      "loss": 0.8548,
+      "step": 111715
+    },
+    {
+      "epoch": 7.590705258866694,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.5580921320831638e-06,
+      "loss": 0.891,
+      "step": 111720
+    },
+    {
+      "epoch": 7.591044978937355,
+      "grad_norm": 9.0,
+      "learning_rate": 2.5559688816415273e-06,
+      "loss": 0.8833,
+      "step": 111725
+    },
+    {
+      "epoch": 7.591384699008017,
+      "grad_norm": 7.75,
+      "learning_rate": 2.5538456311998914e-06,
+      "loss": 0.7782,
+      "step": 111730
+    },
+    {
+      "epoch": 7.5917244190786795,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.5517223807582554e-06,
+      "loss": 0.8636,
+      "step": 111735
+    },
+    {
+      "epoch": 7.592064139149341,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.5495991303166194e-06,
+      "loss": 0.9171,
+      "step": 111740
+    },
+    {
+      "epoch": 7.592403859220003,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.547475879874983e-06,
+      "loss": 0.8138,
+      "step": 111745
+    },
+    {
+      "epoch": 7.592743579290665,
+      "grad_norm": 8.0,
+      "learning_rate": 2.545352629433347e-06,
+      "loss": 0.7657,
+      "step": 111750
+    },
+    {
+      "epoch": 7.593083299361326,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.543229378991711e-06,
+      "loss": 0.8743,
+      "step": 111755
+    },
+    {
+      "epoch": 7.593423019431988,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.541106128550075e-06,
+      "loss": 0.8825,
+      "step": 111760
+    },
+    {
+      "epoch": 7.59376273950265,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.538982878108439e-06,
+      "loss": 0.8719,
+      "step": 111765
+    },
+    {
+      "epoch": 7.594102459573311,
+      "grad_norm": 8.5,
+      "learning_rate": 2.5368596276668025e-06,
+      "loss": 0.881,
+      "step": 111770
+    },
+    {
+      "epoch": 7.594442179643973,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.5347363772251665e-06,
+      "loss": 0.9084,
+      "step": 111775
+    },
+    {
+      "epoch": 7.5947818997146355,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5326131267835305e-06,
+      "loss": 0.7824,
+      "step": 111780
+    },
+    {
+      "epoch": 7.595121619785297,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.530489876341894e-06,
+      "loss": 0.819,
+      "step": 111785
+    },
+    {
+      "epoch": 7.595461339855959,
+      "grad_norm": 7.25,
+      "learning_rate": 2.5283666259002586e-06,
+      "loss": 0.7912,
+      "step": 111790
+    },
+    {
+      "epoch": 7.595801059926621,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.526243375458622e-06,
+      "loss": 0.7686,
+      "step": 111795
+    },
+    {
+      "epoch": 7.596140779997282,
+      "grad_norm": 8.5,
+      "learning_rate": 2.524120125016986e-06,
+      "loss": 0.7954,
+      "step": 111800
+    },
+    {
+      "epoch": 7.596480500067944,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.52199687457535e-06,
+      "loss": 0.9035,
+      "step": 111805
+    },
+    {
+      "epoch": 7.596820220138606,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.5198736241337137e-06,
+      "loss": 0.8739,
+      "step": 111810
+    },
+    {
+      "epoch": 7.597159940209267,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.517750373692078e-06,
+      "loss": 0.8344,
+      "step": 111815
+    },
+    {
+      "epoch": 7.597499660279929,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.5156271232504417e-06,
+      "loss": 0.9245,
+      "step": 111820
+    },
+    {
+      "epoch": 7.5978393803505915,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.5135038728088057e-06,
+      "loss": 0.8343,
+      "step": 111825
+    },
+    {
+      "epoch": 7.598179100421253,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.5113806223671697e-06,
+      "loss": 0.7372,
+      "step": 111830
+    },
+    {
+      "epoch": 7.598518820491915,
+      "grad_norm": 8.0,
+      "learning_rate": 2.5092573719255333e-06,
+      "loss": 0.8043,
+      "step": 111835
+    },
+    {
+      "epoch": 7.598858540562577,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.5071341214838973e-06,
+      "loss": 0.8597,
+      "step": 111840
+    },
+    {
+      "epoch": 7.599198260633238,
+      "grad_norm": 8.5,
+      "learning_rate": 2.5050108710422613e-06,
+      "loss": 0.8888,
+      "step": 111845
+    },
+    {
+      "epoch": 7.5995379807039,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.5028876206006253e-06,
+      "loss": 0.7893,
+      "step": 111850
+    },
+    {
+      "epoch": 7.599877700774561,
+      "grad_norm": 7.625,
+      "learning_rate": 2.500764370158989e-06,
+      "loss": 0.8845,
+      "step": 111855
+    },
+    {
+      "epoch": 7.600217420845223,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.498641119717353e-06,
+      "loss": 0.8701,
+      "step": 111860
+    },
+    {
+      "epoch": 7.600557140915885,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.496517869275717e-06,
+      "loss": 0.7832,
+      "step": 111865
+    },
+    {
+      "epoch": 7.600896860986547,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.494394618834081e-06,
+      "loss": 0.8356,
+      "step": 111870
+    },
+    {
+      "epoch": 7.601236581057209,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.492271368392445e-06,
+      "loss": 0.7961,
+      "step": 111875
+    },
+    {
+      "epoch": 7.601576301127871,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.4901481179508085e-06,
+      "loss": 0.9076,
+      "step": 111880
+    },
+    {
+      "epoch": 7.601916021198532,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4880248675091725e-06,
+      "loss": 0.829,
+      "step": 111885
+    },
+    {
+      "epoch": 7.602255741269194,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.4859016170675365e-06,
+      "loss": 0.8594,
+      "step": 111890
+    },
+    {
+      "epoch": 7.602595461339856,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.4837783666259e-06,
+      "loss": 0.8393,
+      "step": 111895
+    },
+    {
+      "epoch": 7.602935181410517,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.4816551161842645e-06,
+      "loss": 0.8369,
+      "step": 111900
+    },
+    {
+      "epoch": 7.603274901481179,
+      "grad_norm": 7.28125,
+      "learning_rate": 2.479531865742628e-06,
+      "loss": 0.8359,
+      "step": 111905
+    },
+    {
+      "epoch": 7.603614621551841,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.477408615300992e-06,
+      "loss": 0.907,
+      "step": 111910
+    },
+    {
+      "epoch": 7.603954341622503,
+      "grad_norm": 8.625,
+      "learning_rate": 2.475285364859356e-06,
+      "loss": 0.7858,
+      "step": 111915
+    },
+    {
+      "epoch": 7.604294061693165,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.4731621144177197e-06,
+      "loss": 0.783,
+      "step": 111920
+    },
+    {
+      "epoch": 7.604633781763827,
+      "grad_norm": 8.75,
+      "learning_rate": 2.4710388639760837e-06,
+      "loss": 0.8482,
+      "step": 111925
+    },
+    {
+      "epoch": 7.604973501834488,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.4689156135344477e-06,
+      "loss": 0.7871,
+      "step": 111930
+    },
+    {
+      "epoch": 7.60531322190515,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4667923630928117e-06,
+      "loss": 0.8309,
+      "step": 111935
+    },
+    {
+      "epoch": 7.605652941975812,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.4646691126511757e-06,
+      "loss": 0.871,
+      "step": 111940
+    },
+    {
+      "epoch": 7.605992662046473,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.4625458622095393e-06,
+      "loss": 0.8816,
+      "step": 111945
+    },
+    {
+      "epoch": 7.606332382117135,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.4604226117679033e-06,
+      "loss": 0.8379,
+      "step": 111950
+    },
+    {
+      "epoch": 7.606672102187797,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4582993613262673e-06,
+      "loss": 0.8127,
+      "step": 111955
+    },
+    {
+      "epoch": 7.607011822258459,
+      "grad_norm": 7.0,
+      "learning_rate": 2.4561761108846313e-06,
+      "loss": 0.8572,
+      "step": 111960
+    },
+    {
+      "epoch": 7.607351542329121,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.454052860442995e-06,
+      "loss": 0.89,
+      "step": 111965
+    },
+    {
+      "epoch": 7.607691262399783,
+      "grad_norm": 9.3125,
+      "learning_rate": 2.451929610001359e-06,
+      "loss": 0.9284,
+      "step": 111970
+    },
+    {
+      "epoch": 7.608030982470444,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.449806359559723e-06,
+      "loss": 0.8081,
+      "step": 111975
+    },
+    {
+      "epoch": 7.608370702541106,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.447683109118087e-06,
+      "loss": 0.8124,
+      "step": 111980
+    },
+    {
+      "epoch": 7.608710422611768,
+      "grad_norm": 8.5,
+      "learning_rate": 2.445559858676451e-06,
+      "loss": 0.7819,
+      "step": 111985
+    },
+    {
+      "epoch": 7.609050142682429,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4434366082348145e-06,
+      "loss": 0.8916,
+      "step": 111990
+    },
+    {
+      "epoch": 7.609389862753091,
+      "grad_norm": 8.25,
+      "learning_rate": 2.441313357793179e-06,
+      "loss": 0.9731,
+      "step": 111995
+    },
+    {
+      "epoch": 7.6097295828237534,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.4391901073515425e-06,
+      "loss": 0.8198,
+      "step": 112000
+    },
+    {
+      "epoch": 7.610069302894415,
+      "grad_norm": 8.125,
+      "learning_rate": 2.437066856909906e-06,
+      "loss": 0.8563,
+      "step": 112005
+    },
+    {
+      "epoch": 7.610409022965077,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4349436064682705e-06,
+      "loss": 0.8628,
+      "step": 112010
+    },
+    {
+      "epoch": 7.610748743035739,
+      "grad_norm": 7.0625,
+      "learning_rate": 2.432820356026634e-06,
+      "loss": 0.8365,
+      "step": 112015
+    },
+    {
+      "epoch": 7.6110884631064,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.430697105584998e-06,
+      "loss": 0.8751,
+      "step": 112020
+    },
+    {
+      "epoch": 7.611428183177062,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.428573855143362e-06,
+      "loss": 0.8832,
+      "step": 112025
+    },
+    {
+      "epoch": 7.611767903247724,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4264506047017257e-06,
+      "loss": 0.9179,
+      "step": 112030
+    },
+    {
+      "epoch": 7.612107623318385,
+      "grad_norm": 8.5,
+      "learning_rate": 2.4243273542600897e-06,
+      "loss": 0.8107,
+      "step": 112035
+    },
+    {
+      "epoch": 7.612447343389047,
+      "grad_norm": 7.40625,
+      "learning_rate": 2.4222041038184537e-06,
+      "loss": 0.8547,
+      "step": 112040
+    },
+    {
+      "epoch": 7.6127870634597095,
+      "grad_norm": 8.5,
+      "learning_rate": 2.4200808533768177e-06,
+      "loss": 0.7582,
+      "step": 112045
+    },
+    {
+      "epoch": 7.613126783530371,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.4179576029351817e-06,
+      "loss": 0.7734,
+      "step": 112050
+    },
+    {
+      "epoch": 7.613466503601033,
+      "grad_norm": 8.625,
+      "learning_rate": 2.4158343524935453e-06,
+      "loss": 0.8816,
+      "step": 112055
+    },
+    {
+      "epoch": 7.613806223671695,
+      "grad_norm": 9.375,
+      "learning_rate": 2.4137111020519093e-06,
+      "loss": 0.784,
+      "step": 112060
+    },
+    {
+      "epoch": 7.614145943742356,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.4115878516102733e-06,
+      "loss": 0.7643,
+      "step": 112065
+    },
+    {
+      "epoch": 7.614485663813018,
+      "grad_norm": 9.4375,
+      "learning_rate": 2.4094646011686373e-06,
+      "loss": 0.7961,
+      "step": 112070
+    },
+    {
+      "epoch": 7.61482538388368,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.407341350727001e-06,
+      "loss": 0.8251,
+      "step": 112075
+    },
+    {
+      "epoch": 7.615165103954341,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4052181002853653e-06,
+      "loss": 0.8531,
+      "step": 112080
+    },
+    {
+      "epoch": 7.615504824025003,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.403094849843729e-06,
+      "loss": 0.8854,
+      "step": 112085
+    },
+    {
+      "epoch": 7.6158445440956655,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.400971599402093e-06,
+      "loss": 0.8735,
+      "step": 112090
+    },
+    {
+      "epoch": 7.616184264166327,
+      "grad_norm": 8.875,
+      "learning_rate": 2.398848348960457e-06,
+      "loss": 0.8104,
+      "step": 112095
+    },
+    {
+      "epoch": 7.616523984236989,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3967250985188204e-06,
+      "loss": 0.7348,
+      "step": 112100
+    },
+    {
+      "epoch": 7.616863704307651,
+      "grad_norm": 8.625,
+      "learning_rate": 2.394601848077185e-06,
+      "loss": 0.9343,
+      "step": 112105
+    },
+    {
+      "epoch": 7.617203424378312,
+      "grad_norm": 7.46875,
+      "learning_rate": 2.3924785976355484e-06,
+      "loss": 0.8451,
+      "step": 112110
+    },
+    {
+      "epoch": 7.617543144448974,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.390355347193912e-06,
+      "loss": 0.8689,
+      "step": 112115
+    },
+    {
+      "epoch": 7.617882864519636,
+      "grad_norm": 8.75,
+      "learning_rate": 2.3882320967522765e-06,
+      "loss": 0.913,
+      "step": 112120
+    },
+    {
+      "epoch": 7.618222584590297,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.38610884631064e-06,
+      "loss": 0.8843,
+      "step": 112125
+    },
+    {
+      "epoch": 7.618562304660959,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.383985595869004e-06,
+      "loss": 0.8593,
+      "step": 112130
+    },
+    {
+      "epoch": 7.6189020247316215,
+      "grad_norm": 9.125,
+      "learning_rate": 2.381862345427368e-06,
+      "loss": 0.9073,
+      "step": 112135
+    },
+    {
+      "epoch": 7.619241744802283,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3797390949857316e-06,
+      "loss": 0.8415,
+      "step": 112140
+    },
+    {
+      "epoch": 7.619581464872945,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.3776158445440956e-06,
+      "loss": 0.7576,
+      "step": 112145
+    },
+    {
+      "epoch": 7.619921184943607,
+      "grad_norm": 6.8125,
+      "learning_rate": 2.3754925941024596e-06,
+      "loss": 0.7562,
+      "step": 112150
+    },
+    {
+      "epoch": 7.620260905014268,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.3733693436608236e-06,
+      "loss": 0.8992,
+      "step": 112155
+    },
+    {
+      "epoch": 7.62060062508493,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.3712460932191876e-06,
+      "loss": 0.9197,
+      "step": 112160
+    },
+    {
+      "epoch": 7.620940345155592,
+      "grad_norm": 7.625,
+      "learning_rate": 2.3691228427775516e-06,
+      "loss": 0.7885,
+      "step": 112165
+    },
+    {
+      "epoch": 7.621280065226253,
+      "grad_norm": 7.5,
+      "learning_rate": 2.3669995923359152e-06,
+      "loss": 0.8456,
+      "step": 112170
+    },
+    {
+      "epoch": 7.621619785296915,
+      "grad_norm": 8.125,
+      "learning_rate": 2.3648763418942792e-06,
+      "loss": 0.8673,
+      "step": 112175
+    },
+    {
+      "epoch": 7.6219595053675775,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.3627530914526432e-06,
+      "loss": 0.744,
+      "step": 112180
+    },
+    {
+      "epoch": 7.622299225438239,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.360629841011007e-06,
+      "loss": 0.8889,
+      "step": 112185
+    },
+    {
+      "epoch": 7.622638945508901,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3585065905693712e-06,
+      "loss": 0.8425,
+      "step": 112190
+    },
+    {
+      "epoch": 7.622978665579563,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.356383340127735e-06,
+      "loss": 0.7664,
+      "step": 112195
+    },
+    {
+      "epoch": 7.623318385650224,
+      "grad_norm": 7.75,
+      "learning_rate": 2.354260089686099e-06,
+      "loss": 0.9326,
+      "step": 112200
+    },
+    {
+      "epoch": 7.623658105720886,
+      "grad_norm": 8.5,
+      "learning_rate": 2.352136839244463e-06,
+      "loss": 0.8283,
+      "step": 112205
+    },
+    {
+      "epoch": 7.623997825791548,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3500135888028264e-06,
+      "loss": 0.8384,
+      "step": 112210
+    },
+    {
+      "epoch": 7.624337545862209,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.3478903383611904e-06,
+      "loss": 0.8543,
+      "step": 112215
+    },
+    {
+      "epoch": 7.624677265932871,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.3457670879195544e-06,
+      "loss": 0.8076,
+      "step": 112220
+    },
+    {
+      "epoch": 7.6250169860035335,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.343643837477918e-06,
+      "loss": 0.8947,
+      "step": 112225
+    },
+    {
+      "epoch": 7.625356706074195,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.3415205870362824e-06,
+      "loss": 0.8087,
+      "step": 112230
+    },
+    {
+      "epoch": 7.625696426144857,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.339397336594646e-06,
+      "loss": 0.8481,
+      "step": 112235
+    },
+    {
+      "epoch": 7.626036146215519,
+      "grad_norm": 8.625,
+      "learning_rate": 2.33727408615301e-06,
+      "loss": 0.8692,
+      "step": 112240
+    },
+    {
+      "epoch": 7.62637586628618,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.335150835711374e-06,
+      "loss": 0.7451,
+      "step": 112245
+    },
+    {
+      "epoch": 7.626715586356842,
+      "grad_norm": 8.625,
+      "learning_rate": 2.333027585269738e-06,
+      "loss": 0.8448,
+      "step": 112250
+    },
+    {
+      "epoch": 7.627055306427504,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.3309043348281016e-06,
+      "loss": 0.8612,
+      "step": 112255
+    },
+    {
+      "epoch": 7.627395026498165,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3287810843864656e-06,
+      "loss": 0.8362,
+      "step": 112260
+    },
+    {
+      "epoch": 7.627734746568827,
+      "grad_norm": 8.5,
+      "learning_rate": 2.3266578339448296e-06,
+      "loss": 0.8646,
+      "step": 112265
+    },
+    {
+      "epoch": 7.6280744666394895,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3245345835031936e-06,
+      "loss": 0.8321,
+      "step": 112270
+    },
+    {
+      "epoch": 7.628414186710151,
+      "grad_norm": 7.875,
+      "learning_rate": 2.3224113330615576e-06,
+      "loss": 0.8211,
+      "step": 112275
+    },
+    {
+      "epoch": 7.628753906780813,
+      "grad_norm": 7.375,
+      "learning_rate": 2.320288082619921e-06,
+      "loss": 0.8314,
+      "step": 112280
+    },
+    {
+      "epoch": 7.629093626851475,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.318164832178285e-06,
+      "loss": 0.9859,
+      "step": 112285
+    },
+    {
+      "epoch": 7.629433346922136,
+      "grad_norm": 9.6875,
+      "learning_rate": 2.316041581736649e-06,
+      "loss": 0.7992,
+      "step": 112290
+    },
+    {
+      "epoch": 7.629773066992798,
+      "grad_norm": 8.25,
+      "learning_rate": 2.3139183312950128e-06,
+      "loss": 0.8511,
+      "step": 112295
+    },
+    {
+      "epoch": 7.63011278706346,
+      "grad_norm": 8.5,
+      "learning_rate": 2.311795080853377e-06,
+      "loss": 0.8561,
+      "step": 112300
+    },
+    {
+      "epoch": 7.630452507134121,
+      "grad_norm": 6.90625,
+      "learning_rate": 2.3096718304117408e-06,
+      "loss": 0.8447,
+      "step": 112305
+    },
+    {
+      "epoch": 7.6307922272047835,
+      "grad_norm": 8.5,
+      "learning_rate": 2.307548579970105e-06,
+      "loss": 0.8284,
+      "step": 112310
+    },
+    {
+      "epoch": 7.6311319472754455,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.305425329528469e-06,
+      "loss": 0.8467,
+      "step": 112315
+    },
+    {
+      "epoch": 7.631471667346107,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.3033020790868324e-06,
+      "loss": 0.7865,
+      "step": 112320
+    },
+    {
+      "epoch": 7.631811387416769,
+      "grad_norm": 8.625,
+      "learning_rate": 2.3011788286451964e-06,
+      "loss": 0.7997,
+      "step": 112325
+    },
+    {
+      "epoch": 7.63215110748743,
+      "grad_norm": 7.375,
+      "learning_rate": 2.2990555782035604e-06,
+      "loss": 0.912,
+      "step": 112330
+    },
+    {
+      "epoch": 7.632490827558092,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2969323277619244e-06,
+      "loss": 0.8032,
+      "step": 112335
+    },
+    {
+      "epoch": 7.632830547628754,
+      "grad_norm": 7.75,
+      "learning_rate": 2.2948090773202884e-06,
+      "loss": 0.8421,
+      "step": 112340
+    },
+    {
+      "epoch": 7.633170267699415,
+      "grad_norm": 8.8125,
+      "learning_rate": 2.292685826878652e-06,
+      "loss": 0.9036,
+      "step": 112345
+    },
+    {
+      "epoch": 7.633509987770077,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.290562576437016e-06,
+      "loss": 0.8311,
+      "step": 112350
+    },
+    {
+      "epoch": 7.6338497078407395,
+      "grad_norm": 8.625,
+      "learning_rate": 2.28843932599538e-06,
+      "loss": 0.8204,
+      "step": 112355
+    },
+    {
+      "epoch": 7.634189427911401,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.286316075553744e-06,
+      "loss": 0.8496,
+      "step": 112360
+    },
+    {
+      "epoch": 7.634529147982063,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2841928251121076e-06,
+      "loss": 0.8868,
+      "step": 112365
+    },
+    {
+      "epoch": 7.634868868052725,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2820695746704716e-06,
+      "loss": 0.8183,
+      "step": 112370
+    },
+    {
+      "epoch": 7.635208588123386,
+      "grad_norm": 7.625,
+      "learning_rate": 2.2799463242288356e-06,
+      "loss": 0.7155,
+      "step": 112375
+    },
+    {
+      "epoch": 7.635548308194048,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2778230737871996e-06,
+      "loss": 0.9262,
+      "step": 112380
+    },
+    {
+      "epoch": 7.63588802826471,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2756998233455636e-06,
+      "loss": 0.7772,
+      "step": 112385
+    },
+    {
+      "epoch": 7.636227748335371,
+      "grad_norm": 8.5,
+      "learning_rate": 2.273576572903927e-06,
+      "loss": 0.7922,
+      "step": 112390
+    },
+    {
+      "epoch": 7.636567468406033,
+      "grad_norm": 8.625,
+      "learning_rate": 2.271453322462291e-06,
+      "loss": 0.8555,
+      "step": 112395
+    },
+    {
+      "epoch": 7.6369071884766955,
+      "grad_norm": 8.375,
+      "learning_rate": 2.269330072020655e-06,
+      "loss": 0.8449,
+      "step": 112400
+    },
+    {
+      "epoch": 7.637246908547357,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2672068215790187e-06,
+      "loss": 0.8013,
+      "step": 112405
+    },
+    {
+      "epoch": 7.637586628618019,
+      "grad_norm": 7.3125,
+      "learning_rate": 2.265083571137383e-06,
+      "loss": 0.8125,
+      "step": 112410
+    },
+    {
+      "epoch": 7.637926348688681,
+      "grad_norm": 7.5,
+      "learning_rate": 2.2629603206957468e-06,
+      "loss": 0.7882,
+      "step": 112415
+    },
+    {
+      "epoch": 7.638266068759342,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.2608370702541108e-06,
+      "loss": 0.8395,
+      "step": 112420
+    },
+    {
+      "epoch": 7.638605788830004,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.2587138198124748e-06,
+      "loss": 0.8602,
+      "step": 112425
+    },
+    {
+      "epoch": 7.638945508900666,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2565905693708383e-06,
+      "loss": 0.8288,
+      "step": 112430
+    },
+    {
+      "epoch": 7.639285228971327,
+      "grad_norm": 8.375,
+      "learning_rate": 2.2544673189292023e-06,
+      "loss": 0.845,
+      "step": 112435
+    },
+    {
+      "epoch": 7.639624949041989,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2523440684875663e-06,
+      "loss": 0.874,
+      "step": 112440
+    },
+    {
+      "epoch": 7.6399646691126515,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.2502208180459304e-06,
+      "loss": 0.8006,
+      "step": 112445
+    },
+    {
+      "epoch": 7.640304389183313,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.2480975676042944e-06,
+      "loss": 0.8754,
+      "step": 112450
+    },
+    {
+      "epoch": 7.640644109253975,
+      "grad_norm": 8.25,
+      "learning_rate": 2.245974317162658e-06,
+      "loss": 0.8273,
+      "step": 112455
+    },
+    {
+      "epoch": 7.640983829324637,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.243851066721022e-06,
+      "loss": 0.8356,
+      "step": 112460
+    },
+    {
+      "epoch": 7.641323549395298,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.241727816279386e-06,
+      "loss": 0.7736,
+      "step": 112465
+    },
+    {
+      "epoch": 7.64166326946596,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.23960456583775e-06,
+      "loss": 0.7728,
+      "step": 112470
+    },
+    {
+      "epoch": 7.642002989536622,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2374813153961135e-06,
+      "loss": 0.8644,
+      "step": 112475
+    },
+    {
+      "epoch": 7.642342709607283,
+      "grad_norm": 9.625,
+      "learning_rate": 2.2353580649544775e-06,
+      "loss": 0.8607,
+      "step": 112480
+    },
+    {
+      "epoch": 7.642682429677945,
+      "grad_norm": 8.5,
+      "learning_rate": 2.2332348145128415e-06,
+      "loss": 0.9092,
+      "step": 112485
+    },
+    {
+      "epoch": 7.6430221497486075,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.2311115640712055e-06,
+      "loss": 0.7559,
+      "step": 112490
+    },
+    {
+      "epoch": 7.643361869819269,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2289883136295695e-06,
+      "loss": 0.7302,
+      "step": 112495
+    },
+    {
+      "epoch": 7.643701589889931,
+      "grad_norm": 8.125,
+      "learning_rate": 2.226865063187933e-06,
+      "loss": 0.8779,
+      "step": 112500
+    },
+    {
+      "epoch": 7.644041309960593,
+      "grad_norm": 7.78125,
+      "learning_rate": 2.224741812746297e-06,
+      "loss": 0.7661,
+      "step": 112505
+    },
+    {
+      "epoch": 7.644381030031254,
+      "grad_norm": 7.75,
+      "learning_rate": 2.222618562304661e-06,
+      "loss": 0.8424,
+      "step": 112510
+    },
+    {
+      "epoch": 7.644720750101916,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.2204953118630247e-06,
+      "loss": 0.9438,
+      "step": 112515
+    },
+    {
+      "epoch": 7.645060470172578,
+      "grad_norm": 8.25,
+      "learning_rate": 2.218372061421389e-06,
+      "loss": 0.766,
+      "step": 112520
+    },
+    {
+      "epoch": 7.645400190243239,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.2162488109797527e-06,
+      "loss": 0.8903,
+      "step": 112525
+    },
+    {
+      "epoch": 7.645739910313901,
+      "grad_norm": 8.75,
+      "learning_rate": 2.2141255605381167e-06,
+      "loss": 1.1514,
+      "step": 112530
+    },
+    {
+      "epoch": 7.6460796303845635,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2120023100964807e-06,
+      "loss": 0.8852,
+      "step": 112535
+    },
+    {
+      "epoch": 7.646419350455225,
+      "grad_norm": 7.25,
+      "learning_rate": 2.2098790596548443e-06,
+      "loss": 0.8302,
+      "step": 112540
+    },
+    {
+      "epoch": 7.646759070525887,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.2077558092132083e-06,
+      "loss": 0.9076,
+      "step": 112545
+    },
+    {
+      "epoch": 7.647098790596548,
+      "grad_norm": 8.125,
+      "learning_rate": 2.2056325587715723e-06,
+      "loss": 0.9145,
+      "step": 112550
+    },
+    {
+      "epoch": 7.64743851066721,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.2035093083299363e-06,
+      "loss": 0.8112,
+      "step": 112555
+    },
+    {
+      "epoch": 7.647778230737872,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.2013860578883003e-06,
+      "loss": 0.8545,
+      "step": 112560
+    },
+    {
+      "epoch": 7.648117950808533,
+      "grad_norm": 8.25,
+      "learning_rate": 2.199262807446664e-06,
+      "loss": 0.7136,
+      "step": 112565
+    },
+    {
+      "epoch": 7.648457670879195,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.197139557005028e-06,
+      "loss": 0.836,
+      "step": 112570
+    },
+    {
+      "epoch": 7.6487973909498574,
+      "grad_norm": 7.75,
+      "learning_rate": 2.195016306563392e-06,
+      "loss": 0.7808,
+      "step": 112575
+    },
+    {
+      "epoch": 7.649137111020519,
+      "grad_norm": 8.125,
+      "learning_rate": 2.192893056121756e-06,
+      "loss": 0.864,
+      "step": 112580
+    },
+    {
+      "epoch": 7.649476831091181,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1907698056801195e-06,
+      "loss": 0.7237,
+      "step": 112585
+    },
+    {
+      "epoch": 7.649816551161843,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.188646555238484e-06,
+      "loss": 0.8265,
+      "step": 112590
+    },
+    {
+      "epoch": 7.650156271232504,
+      "grad_norm": 8.625,
+      "learning_rate": 2.1865233047968475e-06,
+      "loss": 0.856,
+      "step": 112595
+    },
+    {
+      "epoch": 7.650495991303166,
+      "grad_norm": 8.375,
+      "learning_rate": 2.184400054355211e-06,
+      "loss": 0.8273,
+      "step": 112600
+    },
+    {
+      "epoch": 7.650835711373828,
+      "grad_norm": 8.125,
+      "learning_rate": 2.1822768039135755e-06,
+      "loss": 0.8698,
+      "step": 112605
+    },
+    {
+      "epoch": 7.651175431444489,
+      "grad_norm": 7.25,
+      "learning_rate": 2.180153553471939e-06,
+      "loss": 0.8202,
+      "step": 112610
+    },
+    {
+      "epoch": 7.651515151515151,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.178030303030303e-06,
+      "loss": 0.7664,
+      "step": 112615
+    },
+    {
+      "epoch": 7.6518548715858135,
+      "grad_norm": 7.21875,
+      "learning_rate": 2.175907052588667e-06,
+      "loss": 0.6629,
+      "step": 112620
+    },
+    {
+      "epoch": 7.652194591656475,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.1737838021470307e-06,
+      "loss": 0.8115,
+      "step": 112625
+    },
+    {
+      "epoch": 7.652534311727137,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.171660551705395e-06,
+      "loss": 0.8504,
+      "step": 112630
+    },
+    {
+      "epoch": 7.652874031797799,
+      "grad_norm": 7.4375,
+      "learning_rate": 2.1695373012637587e-06,
+      "loss": 0.8285,
+      "step": 112635
+    },
+    {
+      "epoch": 7.65321375186846,
+      "grad_norm": 6.9375,
+      "learning_rate": 2.1674140508221227e-06,
+      "loss": 0.9451,
+      "step": 112640
+    },
+    {
+      "epoch": 7.653553471939122,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1652908003804867e-06,
+      "loss": 0.7942,
+      "step": 112645
+    },
+    {
+      "epoch": 7.653893192009784,
+      "grad_norm": 9.375,
+      "learning_rate": 2.1631675499388503e-06,
+      "loss": 0.9075,
+      "step": 112650
+    },
+    {
+      "epoch": 7.654232912080445,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.1610442994972143e-06,
+      "loss": 0.8882,
+      "step": 112655
+    },
+    {
+      "epoch": 7.654572632151107,
+      "grad_norm": 6.4375,
+      "learning_rate": 2.1589210490555783e-06,
+      "loss": 0.7913,
+      "step": 112660
+    },
+    {
+      "epoch": 7.6549123522217695,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.1567977986139423e-06,
+      "loss": 0.83,
+      "step": 112665
+    },
+    {
+      "epoch": 7.655252072292431,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.1546745481723063e-06,
+      "loss": 0.8133,
+      "step": 112670
+    },
+    {
+      "epoch": 7.655591792363093,
+      "grad_norm": 8.25,
+      "learning_rate": 2.1525512977306703e-06,
+      "loss": 0.918,
+      "step": 112675
+    },
+    {
+      "epoch": 7.655931512433755,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.150428047289034e-06,
+      "loss": 0.8733,
+      "step": 112680
+    },
+    {
+      "epoch": 7.656271232504416,
+      "grad_norm": 8.875,
+      "learning_rate": 2.148304796847398e-06,
+      "loss": 0.8892,
+      "step": 112685
+    },
+    {
+      "epoch": 7.656610952575078,
+      "grad_norm": 8.875,
+      "learning_rate": 2.146181546405762e-06,
+      "loss": 0.8507,
+      "step": 112690
+    },
+    {
+      "epoch": 7.65695067264574,
+      "grad_norm": 7.25,
+      "learning_rate": 2.1440582959641255e-06,
+      "loss": 0.822,
+      "step": 112695
+    },
+    {
+      "epoch": 7.657290392716401,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.14193504552249e-06,
+      "loss": 0.7799,
+      "step": 112700
+    },
+    {
+      "epoch": 7.657630112787063,
+      "grad_norm": 7.15625,
+      "learning_rate": 2.1398117950808535e-06,
+      "loss": 0.8597,
+      "step": 112705
+    },
+    {
+      "epoch": 7.6579698328577255,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.137688544639217e-06,
+      "loss": 0.8035,
+      "step": 112710
+    },
+    {
+      "epoch": 7.658309552928387,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.1355652941975815e-06,
+      "loss": 0.8291,
+      "step": 112715
+    },
+    {
+      "epoch": 7.658649272999049,
+      "grad_norm": 7.25,
+      "learning_rate": 2.133442043755945e-06,
+      "loss": 0.7796,
+      "step": 112720
+    },
+    {
+      "epoch": 7.658988993069711,
+      "grad_norm": 9.5,
+      "learning_rate": 2.131318793314309e-06,
+      "loss": 0.8764,
+      "step": 112725
+    },
+    {
+      "epoch": 7.659328713140372,
+      "grad_norm": 7.875,
+      "learning_rate": 2.129195542872673e-06,
+      "loss": 0.862,
+      "step": 112730
+    },
+    {
+      "epoch": 7.659668433211034,
+      "grad_norm": 7.5,
+      "learning_rate": 2.1270722924310366e-06,
+      "loss": 0.7867,
+      "step": 112735
+    },
+    {
+      "epoch": 7.660008153281696,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.124949041989401e-06,
+      "loss": 0.8883,
+      "step": 112740
+    },
+    {
+      "epoch": 7.660347873352357,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1228257915477647e-06,
+      "loss": 0.9315,
+      "step": 112745
+    },
+    {
+      "epoch": 7.660687593423019,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.1207025411061287e-06,
+      "loss": 0.8366,
+      "step": 112750
+    },
+    {
+      "epoch": 7.6610273134936815,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.1185792906644927e-06,
+      "loss": 0.9639,
+      "step": 112755
+    },
+    {
+      "epoch": 7.661367033564343,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.1164560402228567e-06,
+      "loss": 0.8194,
+      "step": 112760
+    },
+    {
+      "epoch": 7.661706753635005,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1143327897812202e-06,
+      "loss": 0.8338,
+      "step": 112765
+    },
+    {
+      "epoch": 7.662046473705667,
+      "grad_norm": 8.5,
+      "learning_rate": 2.1122095393395843e-06,
+      "loss": 0.7907,
+      "step": 112770
+    },
+    {
+      "epoch": 7.662386193776328,
+      "grad_norm": 8.0,
+      "learning_rate": 2.1100862888979483e-06,
+      "loss": 0.7786,
+      "step": 112775
+    },
+    {
+      "epoch": 7.66272591384699,
+      "grad_norm": 8.125,
+      "learning_rate": 2.107963038456312e-06,
+      "loss": 0.7823,
+      "step": 112780
+    },
+    {
+      "epoch": 7.663065633917652,
+      "grad_norm": 8.375,
+      "learning_rate": 2.1058397880146763e-06,
+      "loss": 0.8476,
+      "step": 112785
+    },
+    {
+      "epoch": 7.663405353988313,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.10371653757304e-06,
+      "loss": 0.8056,
+      "step": 112790
+    },
+    {
+      "epoch": 7.663745074058975,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.101593287131404e-06,
+      "loss": 0.8516,
+      "step": 112795
+    },
+    {
+      "epoch": 7.6640847941296375,
+      "grad_norm": 8.0,
+      "learning_rate": 2.099470036689768e-06,
+      "loss": 0.9031,
+      "step": 112800
+    },
+    {
+      "epoch": 7.664424514200299,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.0973467862481314e-06,
+      "loss": 0.8294,
+      "step": 112805
+    },
+    {
+      "epoch": 7.664764234270961,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.095223535806496e-06,
+      "loss": 0.8166,
+      "step": 112810
+    },
+    {
+      "epoch": 7.665103954341623,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0931002853648594e-06,
+      "loss": 0.868,
+      "step": 112815
+    },
+    {
+      "epoch": 7.665443674412284,
+      "grad_norm": 7.375,
+      "learning_rate": 2.090977034923223e-06,
+      "loss": 0.7874,
+      "step": 112820
+    },
+    {
+      "epoch": 7.665783394482946,
+      "grad_norm": 7.125,
+      "learning_rate": 2.0888537844815874e-06,
+      "loss": 0.8171,
+      "step": 112825
+    },
+    {
+      "epoch": 7.666123114553608,
+      "grad_norm": 8.5,
+      "learning_rate": 2.086730534039951e-06,
+      "loss": 0.8451,
+      "step": 112830
+    },
+    {
+      "epoch": 7.666462834624269,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.084607283598315e-06,
+      "loss": 0.8892,
+      "step": 112835
+    },
+    {
+      "epoch": 7.666802554694931,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.082484033156679e-06,
+      "loss": 0.7649,
+      "step": 112840
+    },
+    {
+      "epoch": 7.6671422747655935,
+      "grad_norm": 8.625,
+      "learning_rate": 2.080360782715043e-06,
+      "loss": 0.8029,
+      "step": 112845
+    },
+    {
+      "epoch": 7.667481994836255,
+      "grad_norm": 9.25,
+      "learning_rate": 2.078237532273407e-06,
+      "loss": 0.8457,
+      "step": 112850
+    },
+    {
+      "epoch": 7.667821714906917,
+      "grad_norm": 9.25,
+      "learning_rate": 2.0761142818317706e-06,
+      "loss": 0.8399,
+      "step": 112855
+    },
+    {
+      "epoch": 7.668161434977579,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0739910313901346e-06,
+      "loss": 0.8489,
+      "step": 112860
+    },
+    {
+      "epoch": 7.66850115504824,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0718677809484986e-06,
+      "loss": 0.8101,
+      "step": 112865
+    },
+    {
+      "epoch": 7.668840875118902,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0697445305068626e-06,
+      "loss": 0.7504,
+      "step": 112870
+    },
+    {
+      "epoch": 7.669180595189564,
+      "grad_norm": 8.5,
+      "learning_rate": 2.0676212800652262e-06,
+      "loss": 0.8992,
+      "step": 112875
+    },
+    {
+      "epoch": 7.669520315260225,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.0654980296235902e-06,
+      "loss": 0.8108,
+      "step": 112880
+    },
+    {
+      "epoch": 7.6698600353308874,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0633747791819542e-06,
+      "loss": 0.8186,
+      "step": 112885
+    },
+    {
+      "epoch": 7.6701997554015495,
+      "grad_norm": 8.625,
+      "learning_rate": 2.061251528740318e-06,
+      "loss": 0.8823,
+      "step": 112890
+    },
+    {
+      "epoch": 7.670539475472211,
+      "grad_norm": 8.25,
+      "learning_rate": 2.0591282782986822e-06,
+      "loss": 0.8618,
+      "step": 112895
+    },
+    {
+      "epoch": 7.670879195542873,
+      "grad_norm": 8.75,
+      "learning_rate": 2.057005027857046e-06,
+      "loss": 0.7927,
+      "step": 112900
+    },
+    {
+      "epoch": 7.671218915613535,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.05488177741541e-06,
+      "loss": 0.8479,
+      "step": 112905
+    },
+    {
+      "epoch": 7.671558635684196,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.052758526973774e-06,
+      "loss": 0.7772,
+      "step": 112910
+    },
+    {
+      "epoch": 7.671898355754858,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0506352765321374e-06,
+      "loss": 0.7902,
+      "step": 112915
+    },
+    {
+      "epoch": 7.67223807582552,
+      "grad_norm": 8.25,
+      "learning_rate": 2.048512026090502e-06,
+      "loss": 0.789,
+      "step": 112920
+    },
+    {
+      "epoch": 7.672577795896181,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.0463887756488654e-06,
+      "loss": 0.9075,
+      "step": 112925
+    },
+    {
+      "epoch": 7.6729175159668435,
+      "grad_norm": 7.1875,
+      "learning_rate": 2.0442655252072294e-06,
+      "loss": 0.7879,
+      "step": 112930
+    },
+    {
+      "epoch": 7.6732572360375055,
+      "grad_norm": 6.96875,
+      "learning_rate": 2.0421422747655934e-06,
+      "loss": 0.8432,
+      "step": 112935
+    },
+    {
+      "epoch": 7.673596956108167,
+      "grad_norm": 9.1875,
+      "learning_rate": 2.040019024323957e-06,
+      "loss": 0.8347,
+      "step": 112940
+    },
+    {
+      "epoch": 7.673936676178829,
+      "grad_norm": 7.84375,
+      "learning_rate": 2.037895773882321e-06,
+      "loss": 0.8316,
+      "step": 112945
+    },
+    {
+      "epoch": 7.674276396249491,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.035772523440685e-06,
+      "loss": 0.915,
+      "step": 112950
+    },
+    {
+      "epoch": 7.674616116320152,
+      "grad_norm": 8.125,
+      "learning_rate": 2.033649272999049e-06,
+      "loss": 0.825,
+      "step": 112955
+    },
+    {
+      "epoch": 7.674955836390814,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.0315260225574126e-06,
+      "loss": 0.7869,
+      "step": 112960
+    },
+    {
+      "epoch": 7.675295556461476,
+      "grad_norm": 7.875,
+      "learning_rate": 2.0294027721157766e-06,
+      "loss": 0.8264,
+      "step": 112965
+    },
+    {
+      "epoch": 7.675635276532137,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0272795216741406e-06,
+      "loss": 0.8991,
+      "step": 112970
+    },
+    {
+      "epoch": 7.6759749966027995,
+      "grad_norm": 8.125,
+      "learning_rate": 2.0251562712325046e-06,
+      "loss": 0.812,
+      "step": 112975
+    },
+    {
+      "epoch": 7.6763147166734615,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.0230330207908686e-06,
+      "loss": 0.8484,
+      "step": 112980
+    },
+    {
+      "epoch": 7.676654436744123,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.020909770349232e-06,
+      "loss": 0.8414,
+      "step": 112985
+    },
+    {
+      "epoch": 7.676994156814785,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.018786519907596e-06,
+      "loss": 0.8684,
+      "step": 112990
+    },
+    {
+      "epoch": 7.677333876885447,
+      "grad_norm": 8.0,
+      "learning_rate": 2.01666326946596e-06,
+      "loss": 0.8186,
+      "step": 112995
+    },
+    {
+      "epoch": 7.677673596956108,
+      "grad_norm": 8.0,
+      "learning_rate": 2.0145400190243238e-06,
+      "loss": 0.8518,
+      "step": 113000
+    },
+    {
+      "epoch": 7.67801331702677,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.012416768582688e-06,
+      "loss": 0.8108,
+      "step": 113005
+    },
+    {
+      "epoch": 7.678353037097431,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.0102935181410518e-06,
+      "loss": 0.8489,
+      "step": 113010
+    },
+    {
+      "epoch": 7.678692757168093,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.0081702676994158e-06,
+      "loss": 0.8578,
+      "step": 113015
+    },
+    {
+      "epoch": 7.6790324772387555,
+      "grad_norm": 7.59375,
+      "learning_rate": 2.0060470172577798e-06,
+      "loss": 0.868,
+      "step": 113020
+    },
+    {
+      "epoch": 7.679372197309417,
+      "grad_norm": 6.78125,
+      "learning_rate": 2.0039237668161434e-06,
+      "loss": 0.8912,
+      "step": 113025
+    },
+    {
+      "epoch": 7.679711917380079,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.001800516374508e-06,
+      "loss": 0.7942,
+      "step": 113030
+    },
+    {
+      "epoch": 7.680051637450741,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.9996772659328714e-06,
+      "loss": 0.9044,
+      "step": 113035
+    },
+    {
+      "epoch": 7.680391357521402,
+      "grad_norm": 8.0,
+      "learning_rate": 1.9975540154912354e-06,
+      "loss": 0.8308,
+      "step": 113040
+    },
+    {
+      "epoch": 7.680731077592064,
+      "grad_norm": 7.5,
+      "learning_rate": 1.9954307650495994e-06,
+      "loss": 0.7844,
+      "step": 113045
+    },
+    {
+      "epoch": 7.681070797662726,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.993307514607963e-06,
+      "loss": 0.8449,
+      "step": 113050
+    },
+    {
+      "epoch": 7.681410517733387,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.991184264166327e-06,
+      "loss": 0.8031,
+      "step": 113055
+    },
+    {
+      "epoch": 7.681750237804049,
+      "grad_norm": 7.125,
+      "learning_rate": 1.989061013724691e-06,
+      "loss": 0.8244,
+      "step": 113060
+    },
+    {
+      "epoch": 7.6820899578747115,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.986937763283055e-06,
+      "loss": 0.7187,
+      "step": 113065
+    },
+    {
+      "epoch": 7.682429677945373,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9848145128414186e-06,
+      "loss": 0.837,
+      "step": 113070
+    },
+    {
+      "epoch": 7.682769398016035,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.9826912623997826e-06,
+      "loss": 0.8143,
+      "step": 113075
+    },
+    {
+      "epoch": 7.683109118086697,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9805680119581466e-06,
+      "loss": 0.8098,
+      "step": 113080
+    },
+    {
+      "epoch": 7.683448838157358,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.9784447615165106e-06,
+      "loss": 0.8401,
+      "step": 113085
+    },
+    {
+      "epoch": 7.68378855822802,
+      "grad_norm": 7.875,
+      "learning_rate": 1.9763215110748746e-06,
+      "loss": 0.9212,
+      "step": 113090
+    },
+    {
+      "epoch": 7.684128278298682,
+      "grad_norm": 8.5,
+      "learning_rate": 1.974198260633238e-06,
+      "loss": 0.885,
+      "step": 113095
+    },
+    {
+      "epoch": 7.684467998369343,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.9720750101916026e-06,
+      "loss": 0.7922,
+      "step": 113100
+    },
+    {
+      "epoch": 7.684807718440005,
+      "grad_norm": 8.25,
+      "learning_rate": 1.969951759749966e-06,
+      "loss": 0.8175,
+      "step": 113105
+    },
+    {
+      "epoch": 7.6851474385106675,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9678285093083297e-06,
+      "loss": 0.923,
+      "step": 113110
+    },
+    {
+      "epoch": 7.685487158581329,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.965705258866694e-06,
+      "loss": 0.8175,
+      "step": 113115
+    },
+    {
+      "epoch": 7.685826878651991,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9635820084250577e-06,
+      "loss": 0.8088,
+      "step": 113120
+    },
+    {
+      "epoch": 7.686166598722653,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9614587579834217e-06,
+      "loss": 0.7862,
+      "step": 113125
+    },
+    {
+      "epoch": 7.686506318793314,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.9593355075417858e-06,
+      "loss": 0.8028,
+      "step": 113130
+    },
+    {
+      "epoch": 7.686846038863976,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9572122571001493e-06,
+      "loss": 0.8092,
+      "step": 113135
+    },
+    {
+      "epoch": 7.687185758934638,
+      "grad_norm": 7.375,
+      "learning_rate": 1.9550890066585138e-06,
+      "loss": 0.8371,
+      "step": 113140
+    },
+    {
+      "epoch": 7.687525479005299,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.9529657562168773e-06,
+      "loss": 0.7935,
+      "step": 113145
+    },
+    {
+      "epoch": 7.687865199075961,
+      "grad_norm": 8.25,
+      "learning_rate": 1.9508425057752413e-06,
+      "loss": 0.8091,
+      "step": 113150
+    },
+    {
+      "epoch": 7.6882049191466235,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9487192553336053e-06,
+      "loss": 0.8633,
+      "step": 113155
+    },
+    {
+      "epoch": 7.688544639217285,
+      "grad_norm": 7.375,
+      "learning_rate": 1.946596004891969e-06,
+      "loss": 0.8064,
+      "step": 113160
+    },
+    {
+      "epoch": 7.688884359287947,
+      "grad_norm": 7.125,
+      "learning_rate": 1.944472754450333e-06,
+      "loss": 0.8399,
+      "step": 113165
+    },
+    {
+      "epoch": 7.689224079358609,
+      "grad_norm": 8.0,
+      "learning_rate": 1.942349504008697e-06,
+      "loss": 0.8959,
+      "step": 113170
+    },
+    {
+      "epoch": 7.68956379942927,
+      "grad_norm": 7.625,
+      "learning_rate": 1.940226253567061e-06,
+      "loss": 0.8859,
+      "step": 113175
+    },
+    {
+      "epoch": 7.689903519499932,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9381030031254245e-06,
+      "loss": 0.9252,
+      "step": 113180
+    },
+    {
+      "epoch": 7.690243239570594,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9359797526837885e-06,
+      "loss": 0.8249,
+      "step": 113185
+    },
+    {
+      "epoch": 7.690582959641255,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.9338565022421525e-06,
+      "loss": 0.8564,
+      "step": 113190
+    },
+    {
+      "epoch": 7.6909226797119175,
+      "grad_norm": 8.5,
+      "learning_rate": 1.9317332518005165e-06,
+      "loss": 0.7951,
+      "step": 113195
+    },
+    {
+      "epoch": 7.6912623997825795,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.9296100013588805e-06,
+      "loss": 0.8252,
+      "step": 113200
+    },
+    {
+      "epoch": 7.691602119853241,
+      "grad_norm": 8.0,
+      "learning_rate": 1.927486750917244e-06,
+      "loss": 0.7393,
+      "step": 113205
+    },
+    {
+      "epoch": 7.691941839923903,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.9253635004756085e-06,
+      "loss": 0.7946,
+      "step": 113210
+    },
+    {
+      "epoch": 7.692281559994565,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.923240250033972e-06,
+      "loss": 0.9079,
+      "step": 113215
+    },
+    {
+      "epoch": 7.692621280065226,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.9211169995923357e-06,
+      "loss": 0.8776,
+      "step": 113220
+    },
+    {
+      "epoch": 7.692961000135888,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9189937491507e-06,
+      "loss": 0.881,
+      "step": 113225
+    },
+    {
+      "epoch": 7.693300720206549,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.9168704987090637e-06,
+      "loss": 0.8309,
+      "step": 113230
+    },
+    {
+      "epoch": 7.693640440277211,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.9147472482674277e-06,
+      "loss": 0.7578,
+      "step": 113235
+    },
+    {
+      "epoch": 7.6939801603478735,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.9126239978257917e-06,
+      "loss": 0.9125,
+      "step": 113240
+    },
+    {
+      "epoch": 7.694319880418535,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.9105007473841553e-06,
+      "loss": 0.9247,
+      "step": 113245
+    },
+    {
+      "epoch": 7.694659600489197,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.9083774969425193e-06,
+      "loss": 0.7515,
+      "step": 113250
+    },
+    {
+      "epoch": 7.694999320559859,
+      "grad_norm": 6.875,
+      "learning_rate": 1.9062542465008833e-06,
+      "loss": 0.8602,
+      "step": 113255
+    },
+    {
+      "epoch": 7.69533904063052,
+      "grad_norm": 7.75,
+      "learning_rate": 1.9041309960592473e-06,
+      "loss": 0.8038,
+      "step": 113260
+    },
+    {
+      "epoch": 7.695678760701182,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.902007745617611e-06,
+      "loss": 0.8006,
+      "step": 113265
+    },
+    {
+      "epoch": 7.696018480771844,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.899884495175975e-06,
+      "loss": 0.8741,
+      "step": 113270
+    },
+    {
+      "epoch": 7.696358200842505,
+      "grad_norm": 6.625,
+      "learning_rate": 1.8977612447343391e-06,
+      "loss": 0.8937,
+      "step": 113275
+    },
+    {
+      "epoch": 7.696697920913167,
+      "grad_norm": 8.375,
+      "learning_rate": 1.8956379942927027e-06,
+      "loss": 0.8944,
+      "step": 113280
+    },
+    {
+      "epoch": 7.6970376409838295,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.893514743851067e-06,
+      "loss": 0.8386,
+      "step": 113285
+    },
+    {
+      "epoch": 7.697377361054491,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.8913914934094307e-06,
+      "loss": 0.8823,
+      "step": 113290
+    },
+    {
+      "epoch": 7.697717081125153,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.8892682429677947e-06,
+      "loss": 0.782,
+      "step": 113295
+    },
+    {
+      "epoch": 7.698056801195815,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8871449925261585e-06,
+      "loss": 0.8597,
+      "step": 113300
+    },
+    {
+      "epoch": 7.698396521266476,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.8850217420845223e-06,
+      "loss": 0.8738,
+      "step": 113305
+    },
+    {
+      "epoch": 7.698736241337138,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.8828984916428865e-06,
+      "loss": 0.8244,
+      "step": 113310
+    },
+    {
+      "epoch": 7.6990759614078,
+      "grad_norm": 8.5,
+      "learning_rate": 1.8807752412012503e-06,
+      "loss": 0.8572,
+      "step": 113315
+    },
+    {
+      "epoch": 7.699415681478461,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.8786519907596143e-06,
+      "loss": 0.8151,
+      "step": 113320
+    },
+    {
+      "epoch": 7.699755401549123,
+      "grad_norm": 8.25,
+      "learning_rate": 1.876528740317978e-06,
+      "loss": 0.779,
+      "step": 113325
+    },
+    {
+      "epoch": 7.7000951216197855,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8744054898763419e-06,
+      "loss": 0.76,
+      "step": 113330
+    },
+    {
+      "epoch": 7.700434841690447,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.8722822394347059e-06,
+      "loss": 0.8855,
+      "step": 113335
+    },
+    {
+      "epoch": 7.700774561761109,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.8701589889930697e-06,
+      "loss": 0.7614,
+      "step": 113340
+    },
+    {
+      "epoch": 7.701114281831771,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.868035738551434e-06,
+      "loss": 0.7899,
+      "step": 113345
+    },
+    {
+      "epoch": 7.701454001902432,
+      "grad_norm": 8.5,
+      "learning_rate": 1.8659124881097977e-06,
+      "loss": 0.9134,
+      "step": 113350
+    },
+    {
+      "epoch": 7.701793721973094,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8637892376681613e-06,
+      "loss": 0.8703,
+      "step": 113355
+    },
+    {
+      "epoch": 7.702133442043756,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.8616659872265255e-06,
+      "loss": 0.801,
+      "step": 113360
+    },
+    {
+      "epoch": 7.702473162114417,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8595427367848893e-06,
+      "loss": 0.7973,
+      "step": 113365
+    },
+    {
+      "epoch": 7.702812882185079,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.8574194863432533e-06,
+      "loss": 0.9193,
+      "step": 113370
+    },
+    {
+      "epoch": 7.7031526022557415,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.855296235901617e-06,
+      "loss": 0.7901,
+      "step": 113375
+    },
+    {
+      "epoch": 7.703492322326403,
+      "grad_norm": 8.625,
+      "learning_rate": 1.8531729854599813e-06,
+      "loss": 0.843,
+      "step": 113380
+    },
+    {
+      "epoch": 7.703832042397065,
+      "grad_norm": 8.5,
+      "learning_rate": 1.851049735018345e-06,
+      "loss": 0.8417,
+      "step": 113385
+    },
+    {
+      "epoch": 7.704171762467727,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.8489264845767087e-06,
+      "loss": 0.9144,
+      "step": 113390
+    },
+    {
+      "epoch": 7.704511482538388,
+      "grad_norm": 7.5,
+      "learning_rate": 1.8468032341350729e-06,
+      "loss": 0.8476,
+      "step": 113395
+    },
+    {
+      "epoch": 7.70485120260905,
+      "grad_norm": 8.0,
+      "learning_rate": 1.8446799836934367e-06,
+      "loss": 0.8589,
+      "step": 113400
+    },
+    {
+      "epoch": 7.705190922679712,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8425567332518007e-06,
+      "loss": 0.8397,
+      "step": 113405
+    },
+    {
+      "epoch": 7.705530642750373,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.8404334828101645e-06,
+      "loss": 0.8228,
+      "step": 113410
+    },
+    {
+      "epoch": 7.705870362821035,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8383102323685283e-06,
+      "loss": 0.8707,
+      "step": 113415
+    },
+    {
+      "epoch": 7.7062100828916975,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.8361869819268925e-06,
+      "loss": 0.8985,
+      "step": 113420
+    },
+    {
+      "epoch": 7.706549802962359,
+      "grad_norm": 9.0,
+      "learning_rate": 1.834063731485256e-06,
+      "loss": 0.7723,
+      "step": 113425
+    },
+    {
+      "epoch": 7.706889523033021,
+      "grad_norm": 7.625,
+      "learning_rate": 1.8319404810436203e-06,
+      "loss": 0.8184,
+      "step": 113430
+    },
+    {
+      "epoch": 7.707229243103683,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.829817230601984e-06,
+      "loss": 0.8405,
+      "step": 113435
+    },
+    {
+      "epoch": 7.707568963174344,
+      "grad_norm": 7.375,
+      "learning_rate": 1.8276939801603479e-06,
+      "loss": 0.7477,
+      "step": 113440
+    },
+    {
+      "epoch": 7.707908683245006,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.8255707297187119e-06,
+      "loss": 0.8201,
+      "step": 113445
+    },
+    {
+      "epoch": 7.708248403315668,
+      "grad_norm": 8.25,
+      "learning_rate": 1.8234474792770756e-06,
+      "loss": 0.784,
+      "step": 113450
+    },
+    {
+      "epoch": 7.708588123386329,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.8213242288354399e-06,
+      "loss": 0.8753,
+      "step": 113455
+    },
+    {
+      "epoch": 7.7089278434569914,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.8192009783938034e-06,
+      "loss": 0.8281,
+      "step": 113460
+    },
+    {
+      "epoch": 7.7092675635276535,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.8170777279521677e-06,
+      "loss": 0.7711,
+      "step": 113465
+    },
+    {
+      "epoch": 7.709607283598315,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.8149544775105315e-06,
+      "loss": 0.8324,
+      "step": 113470
+    },
+    {
+      "epoch": 7.709947003668977,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8128312270688952e-06,
+      "loss": 0.8499,
+      "step": 113475
+    },
+    {
+      "epoch": 7.710286723739639,
+      "grad_norm": 9.125,
+      "learning_rate": 1.8107079766272592e-06,
+      "loss": 0.8808,
+      "step": 113480
+    },
+    {
+      "epoch": 7.7106264438103,
+      "grad_norm": 8.5,
+      "learning_rate": 1.808584726185623e-06,
+      "loss": 0.8487,
+      "step": 113485
+    },
+    {
+      "epoch": 7.710966163880962,
+      "grad_norm": 8.125,
+      "learning_rate": 1.8064614757439873e-06,
+      "loss": 0.8391,
+      "step": 113490
+    },
+    {
+      "epoch": 7.711305883951624,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.804338225302351e-06,
+      "loss": 0.8618,
+      "step": 113495
+    },
+    {
+      "epoch": 7.711645604022285,
+      "grad_norm": 8.75,
+      "learning_rate": 1.8022149748607146e-06,
+      "loss": 0.842,
+      "step": 113500
+    },
+    {
+      "epoch": 7.7119853240929475,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.8000917244190788e-06,
+      "loss": 0.7885,
+      "step": 113505
+    },
+    {
+      "epoch": 7.7123250441636095,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7979684739774426e-06,
+      "loss": 0.8923,
+      "step": 113510
+    },
+    {
+      "epoch": 7.712664764234271,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.7958452235358066e-06,
+      "loss": 0.8969,
+      "step": 113515
+    },
+    {
+      "epoch": 7.713004484304933,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7937219730941704e-06,
+      "loss": 0.8213,
+      "step": 113520
+    },
+    {
+      "epoch": 7.713344204375595,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7915987226525342e-06,
+      "loss": 0.8881,
+      "step": 113525
+    },
+    {
+      "epoch": 7.713683924446256,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.7894754722108984e-06,
+      "loss": 0.9171,
+      "step": 113530
+    },
+    {
+      "epoch": 7.714023644516918,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.787352221769262e-06,
+      "loss": 0.8116,
+      "step": 113535
+    },
+    {
+      "epoch": 7.71436336458758,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7852289713276262e-06,
+      "loss": 0.825,
+      "step": 113540
+    },
+    {
+      "epoch": 7.714703084658241,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.78310572088599e-06,
+      "loss": 0.7561,
+      "step": 113545
+    },
+    {
+      "epoch": 7.7150428047289035,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.780982470444354e-06,
+      "loss": 0.8478,
+      "step": 113550
+    },
+    {
+      "epoch": 7.7153825247995655,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7788592200027178e-06,
+      "loss": 0.781,
+      "step": 113555
+    },
+    {
+      "epoch": 7.715722244870227,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7767359695610816e-06,
+      "loss": 0.8735,
+      "step": 113560
+    },
+    {
+      "epoch": 7.716061964940889,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7746127191194458e-06,
+      "loss": 0.8016,
+      "step": 113565
+    },
+    {
+      "epoch": 7.716401685011551,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7724894686778094e-06,
+      "loss": 0.7773,
+      "step": 113570
+    },
+    {
+      "epoch": 7.716741405082212,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.7703662182361736e-06,
+      "loss": 0.7643,
+      "step": 113575
+    },
+    {
+      "epoch": 7.717081125152874,
+      "grad_norm": 5.75,
+      "learning_rate": 1.7682429677945374e-06,
+      "loss": 0.94,
+      "step": 113580
+    },
+    {
+      "epoch": 7.717420845223536,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7661197173529012e-06,
+      "loss": 0.79,
+      "step": 113585
+    },
+    {
+      "epoch": 7.717760565294197,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.7639964669112652e-06,
+      "loss": 0.8266,
+      "step": 113590
+    },
+    {
+      "epoch": 7.7181002853648595,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.761873216469629e-06,
+      "loss": 0.7574,
+      "step": 113595
+    },
+    {
+      "epoch": 7.7184400054355216,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7597499660279932e-06,
+      "loss": 0.7928,
+      "step": 113600
+    },
+    {
+      "epoch": 7.718779725506183,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.7576267155863568e-06,
+      "loss": 0.8924,
+      "step": 113605
+    },
+    {
+      "epoch": 7.719119445576845,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7555034651447206e-06,
+      "loss": 0.8502,
+      "step": 113610
+    },
+    {
+      "epoch": 7.719459165647507,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.7533802147030848e-06,
+      "loss": 0.7707,
+      "step": 113615
+    },
+    {
+      "epoch": 7.719798885718168,
+      "grad_norm": 8.0,
+      "learning_rate": 1.7512569642614486e-06,
+      "loss": 0.8781,
+      "step": 113620
+    },
+    {
+      "epoch": 7.72013860578883,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7491337138198126e-06,
+      "loss": 0.8202,
+      "step": 113625
+    },
+    {
+      "epoch": 7.720478325859492,
+      "grad_norm": 8.75,
+      "learning_rate": 1.7470104633781764e-06,
+      "loss": 0.8447,
+      "step": 113630
+    },
+    {
+      "epoch": 7.720818045930153,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7448872129365406e-06,
+      "loss": 0.795,
+      "step": 113635
+    },
+    {
+      "epoch": 7.7211577660008155,
+      "grad_norm": 7.5,
+      "learning_rate": 1.7427639624949042e-06,
+      "loss": 0.8037,
+      "step": 113640
+    },
+    {
+      "epoch": 7.721497486071478,
+      "grad_norm": 8.5,
+      "learning_rate": 1.740640712053268e-06,
+      "loss": 0.8671,
+      "step": 113645
+    },
+    {
+      "epoch": 7.721837206142139,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.7385174616116322e-06,
+      "loss": 0.962,
+      "step": 113650
+    },
+    {
+      "epoch": 7.722176926212801,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.736394211169996e-06,
+      "loss": 0.8665,
+      "step": 113655
+    },
+    {
+      "epoch": 7.722516646283463,
+      "grad_norm": 8.0,
+      "learning_rate": 1.73427096072836e-06,
+      "loss": 0.802,
+      "step": 113660
+    },
+    {
+      "epoch": 7.722856366354124,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.7321477102867238e-06,
+      "loss": 0.8509,
+      "step": 113665
+    },
+    {
+      "epoch": 7.723196086424786,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.7300244598450876e-06,
+      "loss": 0.8199,
+      "step": 113670
+    },
+    {
+      "epoch": 7.723535806495448,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7279012094034518e-06,
+      "loss": 0.8635,
+      "step": 113675
+    },
+    {
+      "epoch": 7.723875526566109,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.7257779589618154e-06,
+      "loss": 0.8823,
+      "step": 113680
+    },
+    {
+      "epoch": 7.7242152466367715,
+      "grad_norm": 8.375,
+      "learning_rate": 1.7236547085201796e-06,
+      "loss": 0.8549,
+      "step": 113685
+    },
+    {
+      "epoch": 7.724554966707433,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.7215314580785434e-06,
+      "loss": 0.9174,
+      "step": 113690
+    },
+    {
+      "epoch": 7.724894686778095,
+      "grad_norm": 7.625,
+      "learning_rate": 1.7194082076369072e-06,
+      "loss": 0.8365,
+      "step": 113695
+    },
+    {
+      "epoch": 7.725234406848757,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7172849571952712e-06,
+      "loss": 0.8476,
+      "step": 113700
+    },
+    {
+      "epoch": 7.725574126919418,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.715161706753635e-06,
+      "loss": 0.9084,
+      "step": 113705
+    },
+    {
+      "epoch": 7.72591384699008,
+      "grad_norm": 8.25,
+      "learning_rate": 1.7130384563119992e-06,
+      "loss": 0.7951,
+      "step": 113710
+    },
+    {
+      "epoch": 7.726253567060742,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.7109152058703628e-06,
+      "loss": 0.8259,
+      "step": 113715
+    },
+    {
+      "epoch": 7.726593287131403,
+      "grad_norm": 8.0,
+      "learning_rate": 1.708791955428727e-06,
+      "loss": 0.8901,
+      "step": 113720
+    },
+    {
+      "epoch": 7.726933007202065,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.7066687049870908e-06,
+      "loss": 0.8486,
+      "step": 113725
+    },
+    {
+      "epoch": 7.7272727272727275,
+      "grad_norm": 7.375,
+      "learning_rate": 1.7045454545454546e-06,
+      "loss": 0.847,
+      "step": 113730
+    },
+    {
+      "epoch": 7.727612447343389,
+      "grad_norm": 9.125,
+      "learning_rate": 1.7024222041038186e-06,
+      "loss": 0.951,
+      "step": 113735
+    },
+    {
+      "epoch": 7.727952167414051,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.7002989536621824e-06,
+      "loss": 0.8535,
+      "step": 113740
+    },
+    {
+      "epoch": 7.728291887484713,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6981757032205466e-06,
+      "loss": 0.862,
+      "step": 113745
+    },
+    {
+      "epoch": 7.728631607555374,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6960524527789102e-06,
+      "loss": 0.818,
+      "step": 113750
+    },
+    {
+      "epoch": 7.728971327626036,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.693929202337274e-06,
+      "loss": 0.822,
+      "step": 113755
+    },
+    {
+      "epoch": 7.729311047696698,
+      "grad_norm": 7.375,
+      "learning_rate": 1.6918059518956382e-06,
+      "loss": 0.9959,
+      "step": 113760
+    },
+    {
+      "epoch": 7.729650767767359,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.689682701454002e-06,
+      "loss": 0.9226,
+      "step": 113765
+    },
+    {
+      "epoch": 7.7299904878380215,
+      "grad_norm": 9.5,
+      "learning_rate": 1.687559451012366e-06,
+      "loss": 0.8106,
+      "step": 113770
+    },
+    {
+      "epoch": 7.7303302079086835,
+      "grad_norm": 9.0,
+      "learning_rate": 1.6854362005707298e-06,
+      "loss": 0.9006,
+      "step": 113775
+    },
+    {
+      "epoch": 7.730669927979345,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.6833129501290935e-06,
+      "loss": 0.8235,
+      "step": 113780
+    },
+    {
+      "epoch": 7.731009648050007,
+      "grad_norm": 8.125,
+      "learning_rate": 1.6811896996874576e-06,
+      "loss": 0.781,
+      "step": 113785
+    },
+    {
+      "epoch": 7.731349368120669,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6790664492458213e-06,
+      "loss": 0.8134,
+      "step": 113790
+    },
+    {
+      "epoch": 7.73168908819133,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.6769431988041856e-06,
+      "loss": 0.8246,
+      "step": 113795
+    },
+    {
+      "epoch": 7.732028808261992,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6748199483625494e-06,
+      "loss": 0.8428,
+      "step": 113800
+    },
+    {
+      "epoch": 7.732368528332654,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6726966979209134e-06,
+      "loss": 0.8607,
+      "step": 113805
+    },
+    {
+      "epoch": 7.732708248403315,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.6705734474792771e-06,
+      "loss": 0.8696,
+      "step": 113810
+    },
+    {
+      "epoch": 7.7330479684739775,
+      "grad_norm": 7.1875,
+      "learning_rate": 1.668450197037641e-06,
+      "loss": 0.8419,
+      "step": 113815
+    },
+    {
+      "epoch": 7.7333876885446395,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6663269465960052e-06,
+      "loss": 0.8105,
+      "step": 113820
+    },
+    {
+      "epoch": 7.733727408615301,
+      "grad_norm": 7.875,
+      "learning_rate": 1.6642036961543687e-06,
+      "loss": 0.8217,
+      "step": 113825
+    },
+    {
+      "epoch": 7.734067128685963,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.662080445712733e-06,
+      "loss": 0.8216,
+      "step": 113830
+    },
+    {
+      "epoch": 7.734406848756625,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6599571952710967e-06,
+      "loss": 0.8686,
+      "step": 113835
+    },
+    {
+      "epoch": 7.734746568827286,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6578339448294605e-06,
+      "loss": 0.8239,
+      "step": 113840
+    },
+    {
+      "epoch": 7.735086288897948,
+      "grad_norm": 8.25,
+      "learning_rate": 1.6557106943878245e-06,
+      "loss": 0.7929,
+      "step": 113845
+    },
+    {
+      "epoch": 7.73542600896861,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.6535874439461883e-06,
+      "loss": 0.8514,
+      "step": 113850
+    },
+    {
+      "epoch": 7.735765729039271,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.6514641935045525e-06,
+      "loss": 0.8865,
+      "step": 113855
+    },
+    {
+      "epoch": 7.7361054491099335,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.6493409430629161e-06,
+      "loss": 0.9244,
+      "step": 113860
+    },
+    {
+      "epoch": 7.7364451691805955,
+      "grad_norm": 9.6875,
+      "learning_rate": 1.64721769262128e-06,
+      "loss": 0.8751,
+      "step": 113865
+    },
+    {
+      "epoch": 7.736784889251257,
+      "grad_norm": 8.375,
+      "learning_rate": 1.6450944421796441e-06,
+      "loss": 0.9307,
+      "step": 113870
+    },
+    {
+      "epoch": 7.737124609321919,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.642971191738008e-06,
+      "loss": 0.8233,
+      "step": 113875
+    },
+    {
+      "epoch": 7.737464329392581,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.640847941296372e-06,
+      "loss": 0.9531,
+      "step": 113880
+    },
+    {
+      "epoch": 7.737804049463242,
+      "grad_norm": 6.4375,
+      "learning_rate": 1.6387246908547357e-06,
+      "loss": 0.7616,
+      "step": 113885
+    },
+    {
+      "epoch": 7.738143769533904,
+      "grad_norm": 8.5,
+      "learning_rate": 1.6366014404131e-06,
+      "loss": 0.8075,
+      "step": 113890
+    },
+    {
+      "epoch": 7.738483489604566,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.6344781899714635e-06,
+      "loss": 0.8511,
+      "step": 113895
+    },
+    {
+      "epoch": 7.738823209675227,
+      "grad_norm": 8.0,
+      "learning_rate": 1.6323549395298273e-06,
+      "loss": 0.9091,
+      "step": 113900
+    },
+    {
+      "epoch": 7.7391629297458895,
+      "grad_norm": 6.8125,
+      "learning_rate": 1.6302316890881915e-06,
+      "loss": 0.8276,
+      "step": 113905
+    },
+    {
+      "epoch": 7.739502649816551,
+      "grad_norm": 6.59375,
+      "learning_rate": 1.6281084386465553e-06,
+      "loss": 0.8942,
+      "step": 113910
+    },
+    {
+      "epoch": 7.739842369887213,
+      "grad_norm": 8.625,
+      "learning_rate": 1.6259851882049193e-06,
+      "loss": 0.8433,
+      "step": 113915
+    },
+    {
+      "epoch": 7.740182089957875,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.6238619377632831e-06,
+      "loss": 0.8632,
+      "step": 113920
+    },
+    {
+      "epoch": 7.740521810028536,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.621738687321647e-06,
+      "loss": 0.8978,
+      "step": 113925
+    },
+    {
+      "epoch": 7.740861530099198,
+      "grad_norm": 8.75,
+      "learning_rate": 1.619615436880011e-06,
+      "loss": 0.9275,
+      "step": 113930
+    },
+    {
+      "epoch": 7.74120125016986,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.6174921864383747e-06,
+      "loss": 0.799,
+      "step": 113935
+    },
+    {
+      "epoch": 7.741540970240521,
+      "grad_norm": 8.0,
+      "learning_rate": 1.615368935996739e-06,
+      "loss": 0.8302,
+      "step": 113940
+    },
+    {
+      "epoch": 7.741880690311183,
+      "grad_norm": 8.75,
+      "learning_rate": 1.6132456855551027e-06,
+      "loss": 0.7586,
+      "step": 113945
+    },
+    {
+      "epoch": 7.7422204103818455,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.6111224351134665e-06,
+      "loss": 0.7593,
+      "step": 113950
+    },
+    {
+      "epoch": 7.742560130452507,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.6089991846718305e-06,
+      "loss": 0.7934,
+      "step": 113955
+    },
+    {
+      "epoch": 7.742899850523169,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.6068759342301943e-06,
+      "loss": 0.8677,
+      "step": 113960
+    },
+    {
+      "epoch": 7.743239570593831,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.6047526837885583e-06,
+      "loss": 0.8098,
+      "step": 113965
+    },
+    {
+      "epoch": 7.743579290664492,
+      "grad_norm": 8.75,
+      "learning_rate": 1.602629433346922e-06,
+      "loss": 0.8547,
+      "step": 113970
+    },
+    {
+      "epoch": 7.743919010735154,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.6005061829052863e-06,
+      "loss": 0.8742,
+      "step": 113975
+    },
+    {
+      "epoch": 7.744258730805816,
+      "grad_norm": 8.25,
+      "learning_rate": 1.59838293246365e-06,
+      "loss": 0.8476,
+      "step": 113980
+    },
+    {
+      "epoch": 7.744598450876477,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.596259682022014e-06,
+      "loss": 0.7923,
+      "step": 113985
+    },
+    {
+      "epoch": 7.744938170947139,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.594136431580378e-06,
+      "loss": 0.7817,
+      "step": 113990
+    },
+    {
+      "epoch": 7.7452778910178015,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5920131811387417e-06,
+      "loss": 0.7737,
+      "step": 113995
+    },
+    {
+      "epoch": 7.745617611088463,
+      "grad_norm": 9.4375,
+      "learning_rate": 1.589889930697106e-06,
+      "loss": 0.8814,
+      "step": 114000
+    },
+    {
+      "epoch": 7.745957331159125,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5877666802554695e-06,
+      "loss": 0.8689,
+      "step": 114005
+    },
+    {
+      "epoch": 7.746297051229787,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5856434298138333e-06,
+      "loss": 0.7676,
+      "step": 114010
+    },
+    {
+      "epoch": 7.746636771300448,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.5835201793721975e-06,
+      "loss": 0.814,
+      "step": 114015
+    },
+    {
+      "epoch": 7.74697649137111,
+      "grad_norm": 8.0,
+      "learning_rate": 1.5813969289305613e-06,
+      "loss": 0.7729,
+      "step": 114020
+    },
+    {
+      "epoch": 7.747316211441772,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.5792736784889253e-06,
+      "loss": 0.827,
+      "step": 114025
+    },
+    {
+      "epoch": 7.747655931512433,
+      "grad_norm": 8.125,
+      "learning_rate": 1.577150428047289e-06,
+      "loss": 0.836,
+      "step": 114030
+    },
+    {
+      "epoch": 7.7479956515830954,
+      "grad_norm": 8.125,
+      "learning_rate": 1.5750271776056529e-06,
+      "loss": 0.8451,
+      "step": 114035
+    },
+    {
+      "epoch": 7.7483353716537575,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.5729039271640169e-06,
+      "loss": 0.8786,
+      "step": 114040
+    },
+    {
+      "epoch": 7.748675091724419,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.5707806767223807e-06,
+      "loss": 0.8396,
+      "step": 114045
+    },
+    {
+      "epoch": 7.749014811795081,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5686574262807449e-06,
+      "loss": 0.8313,
+      "step": 114050
+    },
+    {
+      "epoch": 7.749354531865743,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.5665341758391087e-06,
+      "loss": 0.8898,
+      "step": 114055
+    },
+    {
+      "epoch": 7.749694251936404,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5644109253974727e-06,
+      "loss": 0.8209,
+      "step": 114060
+    },
+    {
+      "epoch": 7.750033972007066,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.5622876749558365e-06,
+      "loss": 0.7432,
+      "step": 114065
+    },
+    {
+      "epoch": 7.750373692077728,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5601644245142005e-06,
+      "loss": 0.8599,
+      "step": 114070
+    },
+    {
+      "epoch": 7.750713412148389,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.5580411740725643e-06,
+      "loss": 0.8902,
+      "step": 114075
+    },
+    {
+      "epoch": 7.7510531322190515,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.555917923630928e-06,
+      "loss": 0.7917,
+      "step": 114080
+    },
+    {
+      "epoch": 7.7513928522897135,
+      "grad_norm": 8.875,
+      "learning_rate": 1.553794673189292e-06,
+      "loss": 0.8015,
+      "step": 114085
+    },
+    {
+      "epoch": 7.751732572360375,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.551671422747656e-06,
+      "loss": 0.9268,
+      "step": 114090
+    },
+    {
+      "epoch": 7.752072292431037,
+      "grad_norm": 8.75,
+      "learning_rate": 1.5495481723060199e-06,
+      "loss": 0.8489,
+      "step": 114095
+    },
+    {
+      "epoch": 7.752412012501699,
+      "grad_norm": 6.75,
+      "learning_rate": 1.5474249218643839e-06,
+      "loss": 0.783,
+      "step": 114100
+    },
+    {
+      "epoch": 7.75275173257236,
+      "grad_norm": 7.625,
+      "learning_rate": 1.5453016714227477e-06,
+      "loss": 0.7672,
+      "step": 114105
+    },
+    {
+      "epoch": 7.753091452643022,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.5431784209811117e-06,
+      "loss": 0.8545,
+      "step": 114110
+    },
+    {
+      "epoch": 7.753431172713684,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.5410551705394755e-06,
+      "loss": 0.8682,
+      "step": 114115
+    },
+    {
+      "epoch": 7.753770892784345,
+      "grad_norm": 7.875,
+      "learning_rate": 1.5389319200978395e-06,
+      "loss": 0.9094,
+      "step": 114120
+    },
+    {
+      "epoch": 7.7541106128550075,
+      "grad_norm": 8.25,
+      "learning_rate": 1.5368086696562035e-06,
+      "loss": 0.7759,
+      "step": 114125
+    },
+    {
+      "epoch": 7.7544503329256695,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.5346854192145673e-06,
+      "loss": 0.8007,
+      "step": 114130
+    },
+    {
+      "epoch": 7.754790052996331,
+      "grad_norm": 9.0,
+      "learning_rate": 1.532562168772931e-06,
+      "loss": 0.8008,
+      "step": 114135
+    },
+    {
+      "epoch": 7.755129773066993,
+      "grad_norm": 9.25,
+      "learning_rate": 1.530438918331295e-06,
+      "loss": 0.8152,
+      "step": 114140
+    },
+    {
+      "epoch": 7.755469493137655,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.528315667889659e-06,
+      "loss": 0.8508,
+      "step": 114145
+    },
+    {
+      "epoch": 7.755809213208316,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5261924174480228e-06,
+      "loss": 0.8604,
+      "step": 114150
+    },
+    {
+      "epoch": 7.756148933278978,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.5240691670063869e-06,
+      "loss": 0.794,
+      "step": 114155
+    },
+    {
+      "epoch": 7.75648865334964,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.5219459165647509e-06,
+      "loss": 0.85,
+      "step": 114160
+    },
+    {
+      "epoch": 7.756828373420301,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5198226661231146e-06,
+      "loss": 0.8404,
+      "step": 114165
+    },
+    {
+      "epoch": 7.7571680934909635,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.5176994156814784e-06,
+      "loss": 0.8526,
+      "step": 114170
+    },
+    {
+      "epoch": 7.7575078135616256,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.5155761652398424e-06,
+      "loss": 0.9327,
+      "step": 114175
+    },
+    {
+      "epoch": 7.757847533632287,
+      "grad_norm": 7.625,
+      "learning_rate": 1.5134529147982064e-06,
+      "loss": 0.8022,
+      "step": 114180
+    },
+    {
+      "epoch": 7.758187253702949,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5113296643565702e-06,
+      "loss": 0.867,
+      "step": 114185
+    },
+    {
+      "epoch": 7.758526973773611,
+      "grad_norm": 8.5,
+      "learning_rate": 1.509206413914934e-06,
+      "loss": 0.8286,
+      "step": 114190
+    },
+    {
+      "epoch": 7.758866693844272,
+      "grad_norm": 8.5,
+      "learning_rate": 1.507083163473298e-06,
+      "loss": 0.798,
+      "step": 114195
+    },
+    {
+      "epoch": 7.759206413914934,
+      "grad_norm": 5.875,
+      "learning_rate": 1.504959913031662e-06,
+      "loss": 0.8174,
+      "step": 114200
+    },
+    {
+      "epoch": 7.759546133985596,
+      "grad_norm": 8.5,
+      "learning_rate": 1.5028366625900258e-06,
+      "loss": 0.7932,
+      "step": 114205
+    },
+    {
+      "epoch": 7.759885854056257,
+      "grad_norm": 7.75,
+      "learning_rate": 1.5007134121483898e-06,
+      "loss": 0.8561,
+      "step": 114210
+    },
+    {
+      "epoch": 7.7602255741269195,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4985901617067538e-06,
+      "loss": 0.8771,
+      "step": 114215
+    },
+    {
+      "epoch": 7.760565294197582,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.4964669112651176e-06,
+      "loss": 0.8945,
+      "step": 114220
+    },
+    {
+      "epoch": 7.760905014268243,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.4943436608234814e-06,
+      "loss": 0.8584,
+      "step": 114225
+    },
+    {
+      "epoch": 7.761244734338905,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.4922204103818454e-06,
+      "loss": 0.8641,
+      "step": 114230
+    },
+    {
+      "epoch": 7.761584454409567,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4900971599402094e-06,
+      "loss": 0.831,
+      "step": 114235
+    },
+    {
+      "epoch": 7.761924174480228,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4879739094985732e-06,
+      "loss": 0.7771,
+      "step": 114240
+    },
+    {
+      "epoch": 7.76226389455089,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4858506590569372e-06,
+      "loss": 0.8196,
+      "step": 114245
+    },
+    {
+      "epoch": 7.762603614621552,
+      "grad_norm": 8.125,
+      "learning_rate": 1.483727408615301e-06,
+      "loss": 0.7889,
+      "step": 114250
+    },
+    {
+      "epoch": 7.762943334692213,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.481604158173665e-06,
+      "loss": 0.8439,
+      "step": 114255
+    },
+    {
+      "epoch": 7.7632830547628755,
+      "grad_norm": 7.15625,
+      "learning_rate": 1.4794809077320288e-06,
+      "loss": 0.8582,
+      "step": 114260
+    },
+    {
+      "epoch": 7.763622774833538,
+      "grad_norm": 8.5,
+      "learning_rate": 1.4773576572903928e-06,
+      "loss": 0.8698,
+      "step": 114265
+    },
+    {
+      "epoch": 7.763962494904199,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4752344068487568e-06,
+      "loss": 0.8325,
+      "step": 114270
+    },
+    {
+      "epoch": 7.764302214974861,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.4731111564071206e-06,
+      "loss": 0.8301,
+      "step": 114275
+    },
+    {
+      "epoch": 7.764641935045523,
+      "grad_norm": 6.09375,
+      "learning_rate": 1.4709879059654844e-06,
+      "loss": 0.8216,
+      "step": 114280
+    },
+    {
+      "epoch": 7.764981655116184,
+      "grad_norm": 7.875,
+      "learning_rate": 1.4688646555238484e-06,
+      "loss": 1.0684,
+      "step": 114285
+    },
+    {
+      "epoch": 7.765321375186846,
+      "grad_norm": 8.125,
+      "learning_rate": 1.4667414050822124e-06,
+      "loss": 0.8947,
+      "step": 114290
+    },
+    {
+      "epoch": 7.765661095257508,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.4646181546405762e-06,
+      "loss": 0.8735,
+      "step": 114295
+    },
+    {
+      "epoch": 7.766000815328169,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.4624949041989402e-06,
+      "loss": 0.6942,
+      "step": 114300
+    },
+    {
+      "epoch": 7.7663405353988315,
+      "grad_norm": 8.625,
+      "learning_rate": 1.460371653757304e-06,
+      "loss": 0.9199,
+      "step": 114305
+    },
+    {
+      "epoch": 7.766680255469494,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.458248403315668e-06,
+      "loss": 0.8138,
+      "step": 114310
+    },
+    {
+      "epoch": 7.767019975540155,
+      "grad_norm": 8.25,
+      "learning_rate": 1.4561251528740318e-06,
+      "loss": 0.846,
+      "step": 114315
+    },
+    {
+      "epoch": 7.767359695610817,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4540019024323958e-06,
+      "loss": 0.9422,
+      "step": 114320
+    },
+    {
+      "epoch": 7.767699415681479,
+      "grad_norm": 7.5,
+      "learning_rate": 1.4518786519907598e-06,
+      "loss": 0.7598,
+      "step": 114325
+    },
+    {
+      "epoch": 7.76803913575214,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.4497554015491236e-06,
+      "loss": 0.8135,
+      "step": 114330
+    },
+    {
+      "epoch": 7.768378855822802,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.4476321511074874e-06,
+      "loss": 0.9318,
+      "step": 114335
+    },
+    {
+      "epoch": 7.768718575893464,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4455089006658514e-06,
+      "loss": 0.8266,
+      "step": 114340
+    },
+    {
+      "epoch": 7.7690582959641254,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4433856502242154e-06,
+      "loss": 0.7875,
+      "step": 114345
+    },
+    {
+      "epoch": 7.7693980160347875,
+      "grad_norm": 7.46875,
+      "learning_rate": 1.4412623997825792e-06,
+      "loss": 0.836,
+      "step": 114350
+    },
+    {
+      "epoch": 7.76973773610545,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.4391391493409432e-06,
+      "loss": 0.7475,
+      "step": 114355
+    },
+    {
+      "epoch": 7.770077456176111,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.437015898899307e-06,
+      "loss": 0.7911,
+      "step": 114360
+    },
+    {
+      "epoch": 7.770417176246773,
+      "grad_norm": 8.125,
+      "learning_rate": 1.434892648457671e-06,
+      "loss": 0.7948,
+      "step": 114365
+    },
+    {
+      "epoch": 7.770756896317434,
+      "grad_norm": 8.875,
+      "learning_rate": 1.4327693980160348e-06,
+      "loss": 0.8809,
+      "step": 114370
+    },
+    {
+      "epoch": 7.771096616388096,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.4306461475743988e-06,
+      "loss": 1.1377,
+      "step": 114375
+    },
+    {
+      "epoch": 7.771436336458758,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.4285228971327628e-06,
+      "loss": 0.8221,
+      "step": 114380
+    },
+    {
+      "epoch": 7.771776056529419,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.4263996466911266e-06,
+      "loss": 0.8303,
+      "step": 114385
+    },
+    {
+      "epoch": 7.7721157766000815,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.4242763962494904e-06,
+      "loss": 0.8066,
+      "step": 114390
+    },
+    {
+      "epoch": 7.7724554966707435,
+      "grad_norm": 7.75,
+      "learning_rate": 1.4221531458078544e-06,
+      "loss": 0.754,
+      "step": 114395
+    },
+    {
+      "epoch": 7.772795216741405,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.4200298953662184e-06,
+      "loss": 0.7931,
+      "step": 114400
+    },
+    {
+      "epoch": 7.773134936812067,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4179066449245822e-06,
+      "loss": 0.861,
+      "step": 114405
+    },
+    {
+      "epoch": 7.773474656882729,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4157833944829462e-06,
+      "loss": 0.8199,
+      "step": 114410
+    },
+    {
+      "epoch": 7.77381437695339,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.4136601440413102e-06,
+      "loss": 0.8698,
+      "step": 114415
+    },
+    {
+      "epoch": 7.774154097024052,
+      "grad_norm": 8.75,
+      "learning_rate": 1.411536893599674e-06,
+      "loss": 0.9259,
+      "step": 114420
+    },
+    {
+      "epoch": 7.774493817094714,
+      "grad_norm": 6.875,
+      "learning_rate": 1.4094136431580378e-06,
+      "loss": 0.8409,
+      "step": 114425
+    },
+    {
+      "epoch": 7.774833537165375,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.4072903927164018e-06,
+      "loss": 0.7886,
+      "step": 114430
+    },
+    {
+      "epoch": 7.7751732572360375,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.4051671422747658e-06,
+      "loss": 0.8402,
+      "step": 114435
+    },
+    {
+      "epoch": 7.7755129773066995,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.4030438918331296e-06,
+      "loss": 0.9199,
+      "step": 114440
+    },
+    {
+      "epoch": 7.775852697377361,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.4009206413914934e-06,
+      "loss": 0.7253,
+      "step": 114445
+    },
+    {
+      "epoch": 7.776192417448023,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3987973909498574e-06,
+      "loss": 0.8711,
+      "step": 114450
+    },
+    {
+      "epoch": 7.776532137518685,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3966741405082214e-06,
+      "loss": 0.7679,
+      "step": 114455
+    },
+    {
+      "epoch": 7.776871857589346,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3945508900665852e-06,
+      "loss": 0.8135,
+      "step": 114460
+    },
+    {
+      "epoch": 7.777211577660008,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.3924276396249492e-06,
+      "loss": 0.8696,
+      "step": 114465
+    },
+    {
+      "epoch": 7.77755129773067,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.3903043891833132e-06,
+      "loss": 0.8385,
+      "step": 114470
+    },
+    {
+      "epoch": 7.777891017801331,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.388181138741677e-06,
+      "loss": 0.7567,
+      "step": 114475
+    },
+    {
+      "epoch": 7.7782307378719935,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3860578883000407e-06,
+      "loss": 0.8153,
+      "step": 114480
+    },
+    {
+      "epoch": 7.778570457942656,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3839346378584048e-06,
+      "loss": 0.9251,
+      "step": 114485
+    },
+    {
+      "epoch": 7.778910178013317,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3818113874167688e-06,
+      "loss": 0.8817,
+      "step": 114490
+    },
+    {
+      "epoch": 7.779249898083979,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3796881369751325e-06,
+      "loss": 0.9542,
+      "step": 114495
+    },
+    {
+      "epoch": 7.779589618154641,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3775648865334966e-06,
+      "loss": 0.8278,
+      "step": 114500
+    },
+    {
+      "epoch": 7.779929338225302,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3754416360918603e-06,
+      "loss": 0.8015,
+      "step": 114505
+    },
+    {
+      "epoch": 7.780269058295964,
+      "grad_norm": 8.0,
+      "learning_rate": 1.3733183856502243e-06,
+      "loss": 0.8957,
+      "step": 114510
+    },
+    {
+      "epoch": 7.780608778366626,
+      "grad_norm": 6.78125,
+      "learning_rate": 1.3711951352085881e-06,
+      "loss": 0.8757,
+      "step": 114515
+    },
+    {
+      "epoch": 7.780948498437287,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.3690718847669521e-06,
+      "loss": 0.7343,
+      "step": 114520
+    },
+    {
+      "epoch": 7.7812882185079495,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.3669486343253161e-06,
+      "loss": 0.8972,
+      "step": 114525
+    },
+    {
+      "epoch": 7.781627938578612,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.36482538388368e-06,
+      "loss": 0.9117,
+      "step": 114530
+    },
+    {
+      "epoch": 7.781967658649273,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.3627021334420437e-06,
+      "loss": 0.8203,
+      "step": 114535
+    },
+    {
+      "epoch": 7.782307378719935,
+      "grad_norm": 8.375,
+      "learning_rate": 1.3605788830004077e-06,
+      "loss": 0.8209,
+      "step": 114540
+    },
+    {
+      "epoch": 7.782647098790597,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3584556325587717e-06,
+      "loss": 0.9206,
+      "step": 114545
+    },
+    {
+      "epoch": 7.782986818861258,
+      "grad_norm": 7.0,
+      "learning_rate": 1.3563323821171355e-06,
+      "loss": 0.8209,
+      "step": 114550
+    },
+    {
+      "epoch": 7.78332653893192,
+      "grad_norm": 7.25,
+      "learning_rate": 1.3542091316754995e-06,
+      "loss": 0.8045,
+      "step": 114555
+    },
+    {
+      "epoch": 7.783666259002582,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.3520858812338633e-06,
+      "loss": 0.7948,
+      "step": 114560
+    },
+    {
+      "epoch": 7.784005979073243,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3499626307922273e-06,
+      "loss": 0.9493,
+      "step": 114565
+    },
+    {
+      "epoch": 7.7843456991439055,
+      "grad_norm": 8.75,
+      "learning_rate": 1.3478393803505911e-06,
+      "loss": 0.8009,
+      "step": 114570
+    },
+    {
+      "epoch": 7.784685419214568,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.3457161299089551e-06,
+      "loss": 0.8347,
+      "step": 114575
+    },
+    {
+      "epoch": 7.785025139285229,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.3435928794673191e-06,
+      "loss": 0.8323,
+      "step": 114580
+    },
+    {
+      "epoch": 7.785364859355891,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.3414696290256827e-06,
+      "loss": 0.8162,
+      "step": 114585
+    },
+    {
+      "epoch": 7.785704579426552,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.3393463785840467e-06,
+      "loss": 0.882,
+      "step": 114590
+    },
+    {
+      "epoch": 7.786044299497214,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.3372231281424107e-06,
+      "loss": 0.8769,
+      "step": 114595
+    },
+    {
+      "epoch": 7.786384019567876,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3350998777007747e-06,
+      "loss": 0.9771,
+      "step": 114600
+    },
+    {
+      "epoch": 7.786723739638537,
+      "grad_norm": 8.875,
+      "learning_rate": 1.3329766272591385e-06,
+      "loss": 0.9427,
+      "step": 114605
+    },
+    {
+      "epoch": 7.787063459709199,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3308533768175025e-06,
+      "loss": 0.8383,
+      "step": 114610
+    },
+    {
+      "epoch": 7.7874031797798615,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.3287301263758663e-06,
+      "loss": 0.8235,
+      "step": 114615
+    },
+    {
+      "epoch": 7.787742899850523,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.3266068759342303e-06,
+      "loss": 0.852,
+      "step": 114620
+    },
+    {
+      "epoch": 7.788082619921185,
+      "grad_norm": 8.75,
+      "learning_rate": 1.3244836254925941e-06,
+      "loss": 0.8409,
+      "step": 114625
+    },
+    {
+      "epoch": 7.788422339991847,
+      "grad_norm": 6.71875,
+      "learning_rate": 1.3223603750509581e-06,
+      "loss": 0.8276,
+      "step": 114630
+    },
+    {
+      "epoch": 7.788762060062508,
+      "grad_norm": 8.125,
+      "learning_rate": 1.3202371246093221e-06,
+      "loss": 0.819,
+      "step": 114635
+    },
+    {
+      "epoch": 7.78910178013317,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.318113874167686e-06,
+      "loss": 0.8125,
+      "step": 114640
+    },
+    {
+      "epoch": 7.789441500203832,
+      "grad_norm": 7.75,
+      "learning_rate": 1.3159906237260497e-06,
+      "loss": 0.8831,
+      "step": 114645
+    },
+    {
+      "epoch": 7.789781220274493,
+      "grad_norm": 7.875,
+      "learning_rate": 1.3138673732844137e-06,
+      "loss": 0.7872,
+      "step": 114650
+    },
+    {
+      "epoch": 7.7901209403451555,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3117441228427777e-06,
+      "loss": 0.8668,
+      "step": 114655
+    },
+    {
+      "epoch": 7.7904606604158175,
+      "grad_norm": 8.5,
+      "learning_rate": 1.3096208724011415e-06,
+      "loss": 0.8165,
+      "step": 114660
+    },
+    {
+      "epoch": 7.790800380486479,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.3074976219595055e-06,
+      "loss": 0.8108,
+      "step": 114665
+    },
+    {
+      "epoch": 7.791140100557141,
+      "grad_norm": 9.25,
+      "learning_rate": 1.3053743715178693e-06,
+      "loss": 0.9151,
+      "step": 114670
+    },
+    {
+      "epoch": 7.791479820627803,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.303251121076233e-06,
+      "loss": 0.8038,
+      "step": 114675
+    },
+    {
+      "epoch": 7.791819540698464,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.301127870634597e-06,
+      "loss": 0.7283,
+      "step": 114680
+    },
+    {
+      "epoch": 7.792159260769126,
+      "grad_norm": 8.625,
+      "learning_rate": 1.299004620192961e-06,
+      "loss": 0.8383,
+      "step": 114685
+    },
+    {
+      "epoch": 7.792498980839788,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.296881369751325e-06,
+      "loss": 0.8238,
+      "step": 114690
+    },
+    {
+      "epoch": 7.792838700910449,
+      "grad_norm": 7.125,
+      "learning_rate": 1.2947581193096889e-06,
+      "loss": 0.787,
+      "step": 114695
+    },
+    {
+      "epoch": 7.7931784209811115,
+      "grad_norm": 8.25,
+      "learning_rate": 1.2926348688680527e-06,
+      "loss": 0.7842,
+      "step": 114700
+    },
+    {
+      "epoch": 7.7935181410517735,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.2905116184264167e-06,
+      "loss": 0.8086,
+      "step": 114705
+    },
+    {
+      "epoch": 7.793857861122435,
+      "grad_norm": 8.625,
+      "learning_rate": 1.2883883679847807e-06,
+      "loss": 0.7617,
+      "step": 114710
+    },
+    {
+      "epoch": 7.794197581193097,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2862651175431445e-06,
+      "loss": 0.7252,
+      "step": 114715
+    },
+    {
+      "epoch": 7.794537301263759,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2841418671015085e-06,
+      "loss": 0.8344,
+      "step": 114720
+    },
+    {
+      "epoch": 7.79487702133442,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.2820186166598725e-06,
+      "loss": 0.8792,
+      "step": 114725
+    },
+    {
+      "epoch": 7.795216741405082,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.279895366218236e-06,
+      "loss": 0.7976,
+      "step": 114730
+    },
+    {
+      "epoch": 7.795556461475744,
+      "grad_norm": 6.75,
+      "learning_rate": 1.2777721157766e-06,
+      "loss": 0.847,
+      "step": 114735
+    },
+    {
+      "epoch": 7.795896181546405,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.275648865334964e-06,
+      "loss": 0.7845,
+      "step": 114740
+    },
+    {
+      "epoch": 7.7962359016170675,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.273525614893328e-06,
+      "loss": 0.8448,
+      "step": 114745
+    },
+    {
+      "epoch": 7.7965756216877296,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.2714023644516919e-06,
+      "loss": 0.8958,
+      "step": 114750
+    },
+    {
+      "epoch": 7.796915341758391,
+      "grad_norm": 9.125,
+      "learning_rate": 1.2692791140100557e-06,
+      "loss": 0.7732,
+      "step": 114755
+    },
+    {
+      "epoch": 7.797255061829053,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2671558635684197e-06,
+      "loss": 0.7911,
+      "step": 114760
+    },
+    {
+      "epoch": 7.797594781899715,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2650326131267835e-06,
+      "loss": 0.8458,
+      "step": 114765
+    },
+    {
+      "epoch": 7.797934501970376,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.2629093626851475e-06,
+      "loss": 0.826,
+      "step": 114770
+    },
+    {
+      "epoch": 7.798274222041038,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.2607861122435115e-06,
+      "loss": 0.7434,
+      "step": 114775
+    },
+    {
+      "epoch": 7.7986139421117,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2586628618018755e-06,
+      "loss": 0.8107,
+      "step": 114780
+    },
+    {
+      "epoch": 7.798953662182361,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.256539611360239e-06,
+      "loss": 0.8658,
+      "step": 114785
+    },
+    {
+      "epoch": 7.7992933822530235,
+      "grad_norm": 7.875,
+      "learning_rate": 1.254416360918603e-06,
+      "loss": 0.8454,
+      "step": 114790
+    },
+    {
+      "epoch": 7.799633102323686,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.252293110476967e-06,
+      "loss": 0.7982,
+      "step": 114795
+    },
+    {
+      "epoch": 7.799972822394347,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.250169860035331e-06,
+      "loss": 0.8275,
+      "step": 114800
+    },
+    {
+      "epoch": 7.800312542465009,
+      "grad_norm": 7.03125,
+      "learning_rate": 1.2480466095936949e-06,
+      "loss": 0.8094,
+      "step": 114805
+    },
+    {
+      "epoch": 7.800652262535671,
+      "grad_norm": 8.125,
+      "learning_rate": 1.2459233591520589e-06,
+      "loss": 0.8532,
+      "step": 114810
+    },
+    {
+      "epoch": 7.800991982606332,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2438001087104227e-06,
+      "loss": 0.8276,
+      "step": 114815
+    },
+    {
+      "epoch": 7.801331702676994,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.2416768582687864e-06,
+      "loss": 0.8247,
+      "step": 114820
+    },
+    {
+      "epoch": 7.801671422747656,
+      "grad_norm": 8.5,
+      "learning_rate": 1.2395536078271505e-06,
+      "loss": 0.7831,
+      "step": 114825
+    },
+    {
+      "epoch": 7.802011142818317,
+      "grad_norm": 8.875,
+      "learning_rate": 1.2374303573855145e-06,
+      "loss": 0.7832,
+      "step": 114830
+    },
+    {
+      "epoch": 7.8023508628889795,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2353071069438785e-06,
+      "loss": 0.8968,
+      "step": 114835
+    },
+    {
+      "epoch": 7.802690582959642,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.233183856502242e-06,
+      "loss": 0.7682,
+      "step": 114840
+    },
+    {
+      "epoch": 7.803030303030303,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.231060606060606e-06,
+      "loss": 0.8344,
+      "step": 114845
+    },
+    {
+      "epoch": 7.803370023100965,
+      "grad_norm": 8.75,
+      "learning_rate": 1.22893735561897e-06,
+      "loss": 0.8086,
+      "step": 114850
+    },
+    {
+      "epoch": 7.803709743171627,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.226814105177334e-06,
+      "loss": 0.8632,
+      "step": 114855
+    },
+    {
+      "epoch": 7.804049463242288,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.2246908547356978e-06,
+      "loss": 0.8426,
+      "step": 114860
+    },
+    {
+      "epoch": 7.80438918331295,
+      "grad_norm": 9.25,
+      "learning_rate": 1.2225676042940618e-06,
+      "loss": 0.765,
+      "step": 114865
+    },
+    {
+      "epoch": 7.804728903383612,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.2204443538524256e-06,
+      "loss": 0.8223,
+      "step": 114870
+    },
+    {
+      "epoch": 7.805068623454273,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2183211034107894e-06,
+      "loss": 0.7578,
+      "step": 114875
+    },
+    {
+      "epoch": 7.8054083435249355,
+      "grad_norm": 9.0,
+      "learning_rate": 1.2161978529691534e-06,
+      "loss": 0.7779,
+      "step": 114880
+    },
+    {
+      "epoch": 7.805748063595598,
+      "grad_norm": 7.5,
+      "learning_rate": 1.2140746025275174e-06,
+      "loss": 0.8369,
+      "step": 114885
+    },
+    {
+      "epoch": 7.806087783666259,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.2119513520858814e-06,
+      "loss": 0.7565,
+      "step": 114890
+    },
+    {
+      "epoch": 7.806427503736921,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2098281016442452e-06,
+      "loss": 0.9902,
+      "step": 114895
+    },
+    {
+      "epoch": 7.806767223807583,
+      "grad_norm": 6.5625,
+      "learning_rate": 1.207704851202609e-06,
+      "loss": 0.8007,
+      "step": 114900
+    },
+    {
+      "epoch": 7.807106943878244,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.205581600760973e-06,
+      "loss": 0.742,
+      "step": 114905
+    },
+    {
+      "epoch": 7.807446663948906,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.2034583503193368e-06,
+      "loss": 0.8176,
+      "step": 114910
+    },
+    {
+      "epoch": 7.807786384019568,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.2013350998777008e-06,
+      "loss": 0.8477,
+      "step": 114915
+    },
+    {
+      "epoch": 7.8081261040902294,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1992118494360648e-06,
+      "loss": 0.7995,
+      "step": 114920
+    },
+    {
+      "epoch": 7.8084658241608915,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1970885989944286e-06,
+      "loss": 0.8109,
+      "step": 114925
+    },
+    {
+      "epoch": 7.808805544231554,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1949653485527924e-06,
+      "loss": 0.827,
+      "step": 114930
+    },
+    {
+      "epoch": 7.809145264302215,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.1928420981111564e-06,
+      "loss": 0.7791,
+      "step": 114935
+    },
+    {
+      "epoch": 7.809484984372877,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1907188476695204e-06,
+      "loss": 0.8612,
+      "step": 114940
+    },
+    {
+      "epoch": 7.809824704443539,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.1885955972278844e-06,
+      "loss": 0.8451,
+      "step": 114945
+    },
+    {
+      "epoch": 7.8101644245142,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.1864723467862482e-06,
+      "loss": 0.8584,
+      "step": 114950
+    },
+    {
+      "epoch": 7.810504144584862,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.184349096344612e-06,
+      "loss": 0.8776,
+      "step": 114955
+    },
+    {
+      "epoch": 7.810843864655524,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.182225845902976e-06,
+      "loss": 0.8444,
+      "step": 114960
+    },
+    {
+      "epoch": 7.8111835847261855,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.1801025954613398e-06,
+      "loss": 0.7622,
+      "step": 114965
+    },
+    {
+      "epoch": 7.8115233047968475,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.1779793450197038e-06,
+      "loss": 0.8376,
+      "step": 114970
+    },
+    {
+      "epoch": 7.81186302486751,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1758560945780678e-06,
+      "loss": 0.8508,
+      "step": 114975
+    },
+    {
+      "epoch": 7.812202744938171,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.1737328441364318e-06,
+      "loss": 0.6987,
+      "step": 114980
+    },
+    {
+      "epoch": 7.812542465008833,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.1716095936947954e-06,
+      "loss": 0.8698,
+      "step": 114985
+    },
+    {
+      "epoch": 7.812882185079495,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1694863432531594e-06,
+      "loss": 0.8819,
+      "step": 114990
+    },
+    {
+      "epoch": 7.813221905150156,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.1673630928115234e-06,
+      "loss": 0.8328,
+      "step": 114995
+    },
+    {
+      "epoch": 7.813561625220818,
+      "grad_norm": 8.375,
+      "learning_rate": 1.1652398423698872e-06,
+      "loss": 0.9245,
+      "step": 115000
+    },
+    {
+      "epoch": 7.81390134529148,
+      "grad_norm": 9.375,
+      "learning_rate": 1.1631165919282512e-06,
+      "loss": 0.9144,
+      "step": 115005
+    },
+    {
+      "epoch": 7.8142410653621415,
+      "grad_norm": 8.125,
+      "learning_rate": 1.160993341486615e-06,
+      "loss": 0.8491,
+      "step": 115010
+    },
+    {
+      "epoch": 7.8145807854328035,
+      "grad_norm": 7.0,
+      "learning_rate": 1.158870091044979e-06,
+      "loss": 0.7341,
+      "step": 115015
+    },
+    {
+      "epoch": 7.814920505503466,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.1567468406033428e-06,
+      "loss": 0.8231,
+      "step": 115020
+    },
+    {
+      "epoch": 7.815260225574127,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.1546235901617068e-06,
+      "loss": 0.772,
+      "step": 115025
+    },
+    {
+      "epoch": 7.815599945644789,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.1525003397200708e-06,
+      "loss": 0.8135,
+      "step": 115030
+    },
+    {
+      "epoch": 7.815939665715451,
+      "grad_norm": 8.0,
+      "learning_rate": 1.1503770892784348e-06,
+      "loss": 0.8325,
+      "step": 115035
+    },
+    {
+      "epoch": 7.816279385786112,
+      "grad_norm": 7.125,
+      "learning_rate": 1.1482538388367984e-06,
+      "loss": 0.8971,
+      "step": 115040
+    },
+    {
+      "epoch": 7.816619105856774,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1461305883951624e-06,
+      "loss": 0.8931,
+      "step": 115045
+    },
+    {
+      "epoch": 7.816958825927435,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1440073379535264e-06,
+      "loss": 0.7763,
+      "step": 115050
+    },
+    {
+      "epoch": 7.8172985459980975,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1418840875118902e-06,
+      "loss": 0.8449,
+      "step": 115055
+    },
+    {
+      "epoch": 7.81763826606876,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.1397608370702542e-06,
+      "loss": 0.8506,
+      "step": 115060
+    },
+    {
+      "epoch": 7.817977986139421,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1376375866286182e-06,
+      "loss": 0.9348,
+      "step": 115065
+    },
+    {
+      "epoch": 7.818317706210083,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.135514336186982e-06,
+      "loss": 0.9342,
+      "step": 115070
+    },
+    {
+      "epoch": 7.818657426280745,
+      "grad_norm": 9.0,
+      "learning_rate": 1.1333910857453458e-06,
+      "loss": 0.8247,
+      "step": 115075
+    },
+    {
+      "epoch": 7.818997146351406,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1312678353037098e-06,
+      "loss": 0.8374,
+      "step": 115080
+    },
+    {
+      "epoch": 7.819336866422068,
+      "grad_norm": 6.5,
+      "learning_rate": 1.1291445848620738e-06,
+      "loss": 0.8138,
+      "step": 115085
+    },
+    {
+      "epoch": 7.81967658649273,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1270213344204376e-06,
+      "loss": 0.8829,
+      "step": 115090
+    },
+    {
+      "epoch": 7.820016306563391,
+      "grad_norm": 8.25,
+      "learning_rate": 1.1248980839788014e-06,
+      "loss": 0.833,
+      "step": 115095
+    },
+    {
+      "epoch": 7.8203560266340535,
+      "grad_norm": 8.625,
+      "learning_rate": 1.1227748335371654e-06,
+      "loss": 0.7859,
+      "step": 115100
+    },
+    {
+      "epoch": 7.820695746704716,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.1206515830955294e-06,
+      "loss": 0.9171,
+      "step": 115105
+    },
+    {
+      "epoch": 7.821035466775377,
+      "grad_norm": 7.53125,
+      "learning_rate": 1.1185283326538932e-06,
+      "loss": 0.8939,
+      "step": 115110
+    },
+    {
+      "epoch": 7.821375186846039,
+      "grad_norm": 8.125,
+      "learning_rate": 1.1164050822122572e-06,
+      "loss": 0.8664,
+      "step": 115115
+    },
+    {
+      "epoch": 7.821714906916701,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.1142818317706212e-06,
+      "loss": 0.8668,
+      "step": 115120
+    },
+    {
+      "epoch": 7.822054626987362,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.112158581328985e-06,
+      "loss": 0.7421,
+      "step": 115125
+    },
+    {
+      "epoch": 7.822394347058024,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.1100353308873488e-06,
+      "loss": 0.8029,
+      "step": 115130
+    },
+    {
+      "epoch": 7.822734067128686,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.1079120804457128e-06,
+      "loss": 0.8434,
+      "step": 115135
+    },
+    {
+      "epoch": 7.823073787199347,
+      "grad_norm": 7.625,
+      "learning_rate": 1.1057888300040768e-06,
+      "loss": 0.8001,
+      "step": 115140
+    },
+    {
+      "epoch": 7.8234135072700095,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.1036655795624406e-06,
+      "loss": 0.7893,
+      "step": 115145
+    },
+    {
+      "epoch": 7.823753227340672,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.1015423291208046e-06,
+      "loss": 0.8465,
+      "step": 115150
+    },
+    {
+      "epoch": 7.824092947411333,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0994190786791684e-06,
+      "loss": 0.7413,
+      "step": 115155
+    },
+    {
+      "epoch": 7.824432667481995,
+      "grad_norm": 9.1875,
+      "learning_rate": 1.0972958282375324e-06,
+      "loss": 0.9112,
+      "step": 115160
+    },
+    {
+      "epoch": 7.824772387552657,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.0951725777958961e-06,
+      "loss": 0.95,
+      "step": 115165
+    },
+    {
+      "epoch": 7.825112107623318,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0930493273542602e-06,
+      "loss": 0.8475,
+      "step": 115170
+    },
+    {
+      "epoch": 7.82545182769398,
+      "grad_norm": 7.875,
+      "learning_rate": 1.0909260769126242e-06,
+      "loss": 0.8362,
+      "step": 115175
+    },
+    {
+      "epoch": 7.825791547764642,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.088802826470988e-06,
+      "loss": 0.8164,
+      "step": 115180
+    },
+    {
+      "epoch": 7.826131267835303,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.0866795760293517e-06,
+      "loss": 0.8996,
+      "step": 115185
+    },
+    {
+      "epoch": 7.8264709879059655,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0845563255877157e-06,
+      "loss": 0.7754,
+      "step": 115190
+    },
+    {
+      "epoch": 7.826810707976628,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0824330751460797e-06,
+      "loss": 0.8241,
+      "step": 115195
+    },
+    {
+      "epoch": 7.827150428047289,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0803098247044435e-06,
+      "loss": 0.8271,
+      "step": 115200
+    },
+    {
+      "epoch": 7.827490148117951,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0781865742628075e-06,
+      "loss": 0.9497,
+      "step": 115205
+    },
+    {
+      "epoch": 7.827829868188613,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0760633238211713e-06,
+      "loss": 0.8095,
+      "step": 115210
+    },
+    {
+      "epoch": 7.828169588259274,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0739400733795353e-06,
+      "loss": 0.8078,
+      "step": 115215
+    },
+    {
+      "epoch": 7.828509308329936,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0718168229378991e-06,
+      "loss": 0.7773,
+      "step": 115220
+    },
+    {
+      "epoch": 7.828849028400598,
+      "grad_norm": 6.6875,
+      "learning_rate": 1.0696935724962631e-06,
+      "loss": 0.8683,
+      "step": 115225
+    },
+    {
+      "epoch": 7.8291887484712595,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.0675703220546271e-06,
+      "loss": 0.8256,
+      "step": 115230
+    },
+    {
+      "epoch": 7.8295284685419215,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.065447071612991e-06,
+      "loss": 0.8599,
+      "step": 115235
+    },
+    {
+      "epoch": 7.829868188612584,
+      "grad_norm": 8.9375,
+      "learning_rate": 1.0633238211713547e-06,
+      "loss": 0.802,
+      "step": 115240
+    },
+    {
+      "epoch": 7.830207908683245,
+      "grad_norm": 7.71875,
+      "learning_rate": 1.0612005707297187e-06,
+      "loss": 0.8282,
+      "step": 115245
+    },
+    {
+      "epoch": 7.830547628753907,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0590773202880827e-06,
+      "loss": 0.8453,
+      "step": 115250
+    },
+    {
+      "epoch": 7.830887348824569,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0569540698464465e-06,
+      "loss": 0.784,
+      "step": 115255
+    },
+    {
+      "epoch": 7.83122706889523,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.0548308194048105e-06,
+      "loss": 0.8074,
+      "step": 115260
+    },
+    {
+      "epoch": 7.831566788965892,
+      "grad_norm": 7.6875,
+      "learning_rate": 1.0527075689631743e-06,
+      "loss": 0.8745,
+      "step": 115265
+    },
+    {
+      "epoch": 7.831906509036553,
+      "grad_norm": 8.75,
+      "learning_rate": 1.0505843185215383e-06,
+      "loss": 0.8618,
+      "step": 115270
+    },
+    {
+      "epoch": 7.8322462291072155,
+      "grad_norm": 8.375,
+      "learning_rate": 1.0484610680799021e-06,
+      "loss": 0.9182,
+      "step": 115275
+    },
+    {
+      "epoch": 7.8325859491778775,
+      "grad_norm": 7.0625,
+      "learning_rate": 1.0463378176382661e-06,
+      "loss": 0.7767,
+      "step": 115280
+    },
+    {
+      "epoch": 7.832925669248539,
+      "grad_norm": 8.25,
+      "learning_rate": 1.0442145671966301e-06,
+      "loss": 0.8385,
+      "step": 115285
+    },
+    {
+      "epoch": 7.833265389319201,
+      "grad_norm": 7.78125,
+      "learning_rate": 1.042091316754994e-06,
+      "loss": 0.8269,
+      "step": 115290
+    },
+    {
+      "epoch": 7.833605109389863,
+      "grad_norm": 7.75,
+      "learning_rate": 1.0399680663133577e-06,
+      "loss": 0.8224,
+      "step": 115295
+    },
+    {
+      "epoch": 7.833944829460524,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0378448158717217e-06,
+      "loss": 0.7042,
+      "step": 115300
+    },
+    {
+      "epoch": 7.834284549531186,
+      "grad_norm": 7.625,
+      "learning_rate": 1.0357215654300857e-06,
+      "loss": 0.8612,
+      "step": 115305
+    },
+    {
+      "epoch": 7.834624269601848,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.0335983149884495e-06,
+      "loss": 0.9712,
+      "step": 115310
+    },
+    {
+      "epoch": 7.834963989672509,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.0314750645468135e-06,
+      "loss": 0.8295,
+      "step": 115315
+    },
+    {
+      "epoch": 7.8353037097431715,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.0293518141051775e-06,
+      "loss": 0.8675,
+      "step": 115320
+    },
+    {
+      "epoch": 7.8356434298138335,
+      "grad_norm": 7.25,
+      "learning_rate": 1.0272285636635413e-06,
+      "loss": 0.9113,
+      "step": 115325
+    },
+    {
+      "epoch": 7.835983149884495,
+      "grad_norm": 7.65625,
+      "learning_rate": 1.025105313221905e-06,
+      "loss": 0.8486,
+      "step": 115330
+    },
+    {
+      "epoch": 7.836322869955157,
+      "grad_norm": 8.375,
+      "learning_rate": 1.022982062780269e-06,
+      "loss": 0.9506,
+      "step": 115335
+    },
+    {
+      "epoch": 7.836662590025819,
+      "grad_norm": 7.4375,
+      "learning_rate": 1.0208588123386331e-06,
+      "loss": 0.7835,
+      "step": 115340
+    },
+    {
+      "epoch": 7.83700231009648,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.018735561896997e-06,
+      "loss": 0.7935,
+      "step": 115345
+    },
+    {
+      "epoch": 7.837342030167142,
+      "grad_norm": 8.125,
+      "learning_rate": 1.0166123114553607e-06,
+      "loss": 0.7851,
+      "step": 115350
+    },
+    {
+      "epoch": 7.837681750237804,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.0144890610137247e-06,
+      "loss": 0.8348,
+      "step": 115355
+    },
+    {
+      "epoch": 7.838021470308465,
+      "grad_norm": 6.5,
+      "learning_rate": 1.0123658105720887e-06,
+      "loss": 0.7516,
+      "step": 115360
+    },
+    {
+      "epoch": 7.8383611903791275,
+      "grad_norm": 9.0625,
+      "learning_rate": 1.0102425601304525e-06,
+      "loss": 0.89,
+      "step": 115365
+    },
+    {
+      "epoch": 7.83870091044979,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0081193096888165e-06,
+      "loss": 0.9247,
+      "step": 115370
+    },
+    {
+      "epoch": 7.839040630520451,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.0059960592471805e-06,
+      "loss": 0.7764,
+      "step": 115375
+    },
+    {
+      "epoch": 7.839380350591113,
+      "grad_norm": 8.5,
+      "learning_rate": 1.0038728088055443e-06,
+      "loss": 0.843,
+      "step": 115380
+    },
+    {
+      "epoch": 7.839720070661775,
+      "grad_norm": 7.75,
+      "learning_rate": 1.001749558363908e-06,
+      "loss": 0.8571,
+      "step": 115385
+    },
+    {
+      "epoch": 7.840059790732436,
+      "grad_norm": 7.84375,
+      "learning_rate": 9.99626307922272e-07,
+      "loss": 0.8795,
+      "step": 115390
+    },
+    {
+      "epoch": 7.840399510803098,
+      "grad_norm": 7.03125,
+      "learning_rate": 9.97503057480636e-07,
+      "loss": 0.8271,
+      "step": 115395
+    },
+    {
+      "epoch": 7.84073923087376,
+      "grad_norm": 8.5,
+      "learning_rate": 9.953798070389999e-07,
+      "loss": 0.8512,
+      "step": 115400
+    },
+    {
+      "epoch": 7.841078950944421,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.932565565973639e-07,
+      "loss": 0.8169,
+      "step": 115405
+    },
+    {
+      "epoch": 7.8414186710150835,
+      "grad_norm": 8.75,
+      "learning_rate": 9.911333061557277e-07,
+      "loss": 0.7802,
+      "step": 115410
+    },
+    {
+      "epoch": 7.841758391085746,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.890100557140917e-07,
+      "loss": 0.7862,
+      "step": 115415
+    },
+    {
+      "epoch": 7.842098111156407,
+      "grad_norm": 7.75,
+      "learning_rate": 9.868868052724555e-07,
+      "loss": 0.7924,
+      "step": 115420
+    },
+    {
+      "epoch": 7.842437831227069,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.847635548308195e-07,
+      "loss": 0.8505,
+      "step": 115425
+    },
+    {
+      "epoch": 7.842777551297731,
+      "grad_norm": 7.375,
+      "learning_rate": 9.826403043891835e-07,
+      "loss": 0.8769,
+      "step": 115430
+    },
+    {
+      "epoch": 7.843117271368392,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.805170539475473e-07,
+      "loss": 0.9069,
+      "step": 115435
+    },
+    {
+      "epoch": 7.843456991439054,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.78393803505911e-07,
+      "loss": 0.9174,
+      "step": 115440
+    },
+    {
+      "epoch": 7.843796711509716,
+      "grad_norm": 8.75,
+      "learning_rate": 9.76270553064275e-07,
+      "loss": 0.82,
+      "step": 115445
+    },
+    {
+      "epoch": 7.844136431580377,
+      "grad_norm": 7.4375,
+      "learning_rate": 9.74147302622639e-07,
+      "loss": 0.8741,
+      "step": 115450
+    },
+    {
+      "epoch": 7.8444761516510395,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.720240521810029e-07,
+      "loss": 0.798,
+      "step": 115455
+    },
+    {
+      "epoch": 7.844815871721702,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.699008017393669e-07,
+      "loss": 0.7543,
+      "step": 115460
+    },
+    {
+      "epoch": 7.845155591792363,
+      "grad_norm": 8.375,
+      "learning_rate": 9.677775512977307e-07,
+      "loss": 0.8515,
+      "step": 115465
+    },
+    {
+      "epoch": 7.845495311863025,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.656543008560947e-07,
+      "loss": 0.835,
+      "step": 115470
+    },
+    {
+      "epoch": 7.845835031933687,
+      "grad_norm": 8.0625,
+      "learning_rate": 9.635310504144585e-07,
+      "loss": 0.8095,
+      "step": 115475
+    },
+    {
+      "epoch": 7.846174752004348,
+      "grad_norm": 7.90625,
+      "learning_rate": 9.614077999728225e-07,
+      "loss": 0.8764,
+      "step": 115480
+    },
+    {
+      "epoch": 7.84651447207501,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.592845495311865e-07,
+      "loss": 0.84,
+      "step": 115485
+    },
+    {
+      "epoch": 7.846854192145672,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.571612990895503e-07,
+      "loss": 0.7425,
+      "step": 115490
+    },
+    {
+      "epoch": 7.8471939122163334,
+      "grad_norm": 8.125,
+      "learning_rate": 9.55038048647914e-07,
+      "loss": 0.8487,
+      "step": 115495
+    },
+    {
+      "epoch": 7.8475336322869955,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.529147982062781e-07,
+      "loss": 0.9631,
+      "step": 115500
+    },
+    {
+      "epoch": 7.847873352357658,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.50791547764642e-07,
+      "loss": 0.8475,
+      "step": 115505
+    },
+    {
+      "epoch": 7.848213072428319,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.48668297323006e-07,
+      "loss": 0.8956,
+      "step": 115510
+    },
+    {
+      "epoch": 7.848552792498981,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.465450468813699e-07,
+      "loss": 0.8321,
+      "step": 115515
+    },
+    {
+      "epoch": 7.848892512569643,
+      "grad_norm": 9.0,
+      "learning_rate": 9.444217964397336e-07,
+      "loss": 0.9118,
+      "step": 115520
+    },
+    {
+      "epoch": 7.849232232640304,
+      "grad_norm": 7.96875,
+      "learning_rate": 9.422985459980975e-07,
+      "loss": 0.8788,
+      "step": 115525
+    },
+    {
+      "epoch": 7.849571952710966,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.401752955564615e-07,
+      "loss": 0.7607,
+      "step": 115530
+    },
+    {
+      "epoch": 7.849911672781628,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.380520451148254e-07,
+      "loss": 0.9492,
+      "step": 115535
+    },
+    {
+      "epoch": 7.8502513928522895,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.359287946731893e-07,
+      "loss": 0.8486,
+      "step": 115540
+    },
+    {
+      "epoch": 7.8505911129229515,
+      "grad_norm": 8.5,
+      "learning_rate": 9.338055442315533e-07,
+      "loss": 0.8203,
+      "step": 115545
+    },
+    {
+      "epoch": 7.850930832993614,
+      "grad_norm": 9.1875,
+      "learning_rate": 9.316822937899171e-07,
+      "loss": 0.9895,
+      "step": 115550
+    },
+    {
+      "epoch": 7.851270553064275,
+      "grad_norm": 8.8125,
+      "learning_rate": 9.29559043348281e-07,
+      "loss": 0.8102,
+      "step": 115555
+    },
+    {
+      "epoch": 7.851610273134937,
+      "grad_norm": 8.1875,
+      "learning_rate": 9.274357929066449e-07,
+      "loss": 0.8118,
+      "step": 115560
+    },
+    {
+      "epoch": 7.851949993205599,
+      "grad_norm": 8.875,
+      "learning_rate": 9.253125424650089e-07,
+      "loss": 0.7984,
+      "step": 115565
+    },
+    {
+      "epoch": 7.85228971327626,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.231892920233728e-07,
+      "loss": 0.8584,
+      "step": 115570
+    },
+    {
+      "epoch": 7.852629433346922,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.210660415817366e-07,
+      "loss": 0.877,
+      "step": 115575
+    },
+    {
+      "epoch": 7.852969153417584,
+      "grad_norm": 8.4375,
+      "learning_rate": 9.189427911401005e-07,
+      "loss": 0.9237,
+      "step": 115580
+    },
+    {
+      "epoch": 7.8533088734882455,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.168195406984645e-07,
+      "loss": 0.8515,
+      "step": 115585
+    },
+    {
+      "epoch": 7.8536485935589075,
+      "grad_norm": 8.125,
+      "learning_rate": 9.146962902568284e-07,
+      "loss": 0.8264,
+      "step": 115590
+    },
+    {
+      "epoch": 7.85398831362957,
+      "grad_norm": 8.3125,
+      "learning_rate": 9.125730398151923e-07,
+      "loss": 0.8791,
+      "step": 115595
+    },
+    {
+      "epoch": 7.854328033700231,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.104497893735563e-07,
+      "loss": 0.8608,
+      "step": 115600
+    },
+    {
+      "epoch": 7.854667753770893,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.0832653893192e-07,
+      "loss": 0.9032,
+      "step": 115605
+    },
+    {
+      "epoch": 7.855007473841555,
+      "grad_norm": 7.59375,
+      "learning_rate": 9.06203288490284e-07,
+      "loss": 0.7494,
+      "step": 115610
+    },
+    {
+      "epoch": 7.855347193912216,
+      "grad_norm": 8.6875,
+      "learning_rate": 9.040800380486479e-07,
+      "loss": 0.8552,
+      "step": 115615
+    },
+    {
+      "epoch": 7.855686913982878,
+      "grad_norm": 7.5625,
+      "learning_rate": 9.019567876070119e-07,
+      "loss": 0.8195,
+      "step": 115620
+    },
+    {
+      "epoch": 7.85602663405354,
+      "grad_norm": 7.125,
+      "learning_rate": 8.998335371653758e-07,
+      "loss": 0.8006,
+      "step": 115625
+    },
+    {
+      "epoch": 7.8563663541242015,
+      "grad_norm": 7.25,
+      "learning_rate": 8.977102867237397e-07,
+      "loss": 0.7773,
+      "step": 115630
+    },
+    {
+      "epoch": 7.8567060741948636,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.955870362821035e-07,
+      "loss": 0.7982,
+      "step": 115635
+    },
+    {
+      "epoch": 7.857045794265526,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.934637858404675e-07,
+      "loss": 0.8619,
+      "step": 115640
+    },
+    {
+      "epoch": 7.857385514336187,
+      "grad_norm": 8.6875,
+      "learning_rate": 8.913405353988314e-07,
+      "loss": 0.827,
+      "step": 115645
+    },
+    {
+      "epoch": 7.857725234406849,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.892172849571953e-07,
+      "loss": 0.818,
+      "step": 115650
+    },
+    {
+      "epoch": 7.858064954477511,
+      "grad_norm": 6.78125,
+      "learning_rate": 8.870940345155593e-07,
+      "loss": 0.7678,
+      "step": 115655
+    },
+    {
+      "epoch": 7.858404674548172,
+      "grad_norm": 7.625,
+      "learning_rate": 8.84970784073923e-07,
+      "loss": 0.8113,
+      "step": 115660
+    },
+    {
+      "epoch": 7.858744394618834,
+      "grad_norm": 7.96875,
+      "learning_rate": 8.82847533632287e-07,
+      "loss": 0.9149,
+      "step": 115665
+    },
+    {
+      "epoch": 7.859084114689496,
+      "grad_norm": 9.0625,
+      "learning_rate": 8.807242831906509e-07,
+      "loss": 0.8614,
+      "step": 115670
+    },
+    {
+      "epoch": 7.8594238347601575,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.786010327490149e-07,
+      "loss": 0.8484,
+      "step": 115675
+    },
+    {
+      "epoch": 7.85976355483082,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.764777823073788e-07,
+      "loss": 0.8082,
+      "step": 115680
+    },
+    {
+      "epoch": 7.860103274901482,
+      "grad_norm": 8.5625,
+      "learning_rate": 8.743545318657427e-07,
+      "loss": 0.8832,
+      "step": 115685
+    },
+    {
+      "epoch": 7.860442994972143,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.722312814241065e-07,
+      "loss": 0.8278,
+      "step": 115690
+    },
+    {
+      "epoch": 7.860782715042805,
+      "grad_norm": 7.9375,
+      "learning_rate": 8.701080309824704e-07,
+      "loss": 0.791,
+      "step": 115695
+    },
+    {
+      "epoch": 7.861122435113467,
+      "grad_norm": 7.78125,
+      "learning_rate": 8.679847805408344e-07,
+      "loss": 0.8445,
+      "step": 115700
+    },
+    {
+      "epoch": 7.861462155184128,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.658615300991983e-07,
+      "loss": 0.8188,
+      "step": 115705
+    },
+    {
+      "epoch": 7.86180187525479,
+      "grad_norm": 8.25,
+      "learning_rate": 8.637382796575623e-07,
+      "loss": 0.8802,
+      "step": 115710
+    },
+    {
+      "epoch": 7.862141595325452,
+      "grad_norm": 7.4375,
+      "learning_rate": 8.616150292159262e-07,
+      "loss": 0.9161,
+      "step": 115715
+    },
+    {
+      "epoch": 7.8624813153961135,
+      "grad_norm": 8.5,
+      "learning_rate": 8.5949177877429e-07,
+      "loss": 0.8897,
+      "step": 115720
+    },
+    {
+      "epoch": 7.862821035466776,
+      "grad_norm": 7.3125,
+      "learning_rate": 8.573685283326539e-07,
+      "loss": 0.7912,
+      "step": 115725
+    },
+    {
+      "epoch": 7.863160755537437,
+      "grad_norm": 7.8125,
+      "learning_rate": 8.552452778910179e-07,
+      "loss": 0.8154,
+      "step": 115730
+    },
+    {
+      "epoch": 7.863500475608099,
+      "grad_norm": 7.65625,
+      "learning_rate": 8.531220274493818e-07,
+      "loss": 0.8311,
+      "step": 115735
+    },
+    {
+      "epoch": 7.863840195678761,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.509987770077457e-07,
+      "loss": 0.757,
+      "step": 115740
+    },
+    {
+      "epoch": 7.864179915749422,
+      "grad_norm": 8.5,
+      "learning_rate": 8.488755265661095e-07,
+      "loss": 0.864,
+      "step": 115745
+    },
+    {
+      "epoch": 7.864519635820084,
+      "grad_norm": 6.625,
+      "learning_rate": 8.467522761244734e-07,
+      "loss": 0.8271,
+      "step": 115750
+    },
+    {
+      "epoch": 7.864859355890746,
+      "grad_norm": 7.5,
+      "learning_rate": 8.446290256828374e-07,
+      "loss": 0.8382,
+      "step": 115755
+    },
+    {
+      "epoch": 7.865199075961407,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.425057752412013e-07,
+      "loss": 0.9251,
+      "step": 115760
+    },
+    {
+      "epoch": 7.8655387960320695,
+      "grad_norm": 8.1875,
+      "learning_rate": 8.403825247995653e-07,
+      "loss": 0.9048,
+      "step": 115765
+    },
+    {
+      "epoch": 7.865878516102732,
+      "grad_norm": 8.375,
+      "learning_rate": 8.382592743579292e-07,
+      "loss": 0.8223,
+      "step": 115770
+    },
+    {
+      "epoch": 7.866218236173393,
+      "grad_norm": 8.125,
+      "learning_rate": 8.36136023916293e-07,
+      "loss": 0.8629,
+      "step": 115775
+    },
+    {
+      "epoch": 7.866557956244055,
+      "grad_norm": 8.4375,
+      "learning_rate": 8.340127734746569e-07,
+      "loss": 0.8205,
+      "step": 115780
+    },
+    {
+      "epoch": 7.866897676314717,
+      "grad_norm": 7.03125,
+      "learning_rate": 8.318895230330208e-07,
+      "loss": 0.8513,
+      "step": 115785
+    },
+    {
+      "epoch": 7.867237396385378,
+      "grad_norm": 8.8125,
+      "learning_rate": 8.297662725913848e-07,
+      "loss": 0.9224,
+      "step": 115790
+    },
+    {
+      "epoch": 7.86757711645604,
+      "grad_norm": 8.125,
+      "learning_rate": 8.276430221497487e-07,
+      "loss": 0.8473,
+      "step": 115795
+    },
+    {
+      "epoch": 7.867916836526702,
+      "grad_norm": 7.90625,
+      "learning_rate": 8.255197717081127e-07,
+      "loss": 0.8491,
+      "step": 115800
+    },
+    {
+      "epoch": 7.8682565565973634,
+      "grad_norm": 8.375,
+      "learning_rate": 8.233965212664764e-07,
+      "loss": 0.89,
+      "step": 115805
+    },
+    {
+      "epoch": 7.8685962766680255,
+      "grad_norm": 8.125,
+      "learning_rate": 8.212732708248404e-07,
+      "loss": 0.8663,
+      "step": 115810
+    },
+    {
+      "epoch": 7.868935996738688,
+      "grad_norm": 7.59375,
+      "learning_rate": 8.191500203832043e-07,
+      "loss": 0.906,
+      "step": 115815
+    },
+    {
+      "epoch": 7.869275716809349,
+      "grad_norm": 8.25,
+      "learning_rate": 8.170267699415683e-07,
+      "loss": 0.9465,
+      "step": 115820
+    },
+    {
+      "epoch": 7.869615436880011,
+      "grad_norm": 7.625,
+      "learning_rate": 8.149035194999322e-07,
+      "loss": 0.9001,
+      "step": 115825
+    },
+    {
+      "epoch": 7.869955156950673,
+      "grad_norm": 8.125,
+      "learning_rate": 8.12780269058296e-07,
+      "loss": 0.799,
+      "step": 115830
+    },
+    {
+      "epoch": 7.870294877021334,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.106570186166599e-07,
+      "loss": 0.8677,
+      "step": 115835
+    },
+    {
+      "epoch": 7.870634597091996,
+      "grad_norm": 8.625,
+      "learning_rate": 8.085337681750238e-07,
+      "loss": 0.8434,
+      "step": 115840
+    },
+    {
+      "epoch": 7.870974317162658,
+      "grad_norm": 8.5,
+      "learning_rate": 8.064105177333878e-07,
+      "loss": 0.8209,
+      "step": 115845
+    },
+    {
+      "epoch": 7.8713140372333195,
+      "grad_norm": 7.875,
+      "learning_rate": 8.042872672917517e-07,
+      "loss": 0.7719,
+      "step": 115850
+    },
+    {
+      "epoch": 7.8716537573039815,
+      "grad_norm": 8.375,
+      "learning_rate": 8.021640168501157e-07,
+      "loss": 0.7993,
+      "step": 115855
+    },
+    {
+      "epoch": 7.871993477374644,
+      "grad_norm": 8.25,
+      "learning_rate": 8.000407664084793e-07,
+      "loss": 0.9176,
+      "step": 115860
+    },
+    {
+      "epoch": 7.872333197445305,
+      "grad_norm": 9.25,
+      "learning_rate": 7.979175159668433e-07,
+      "loss": 0.9279,
+      "step": 115865
+    },
+    {
+      "epoch": 7.872672917515967,
+      "grad_norm": 8.3125,
+      "learning_rate": 7.957942655252072e-07,
+      "loss": 0.9202,
+      "step": 115870
+    },
+    {
+      "epoch": 7.873012637586629,
+      "grad_norm": 7.875,
+      "learning_rate": 7.936710150835711e-07,
+      "loss": 0.9005,
+      "step": 115875
+    },
+    {
+      "epoch": 7.87335235765729,
+      "grad_norm": 7.9375,
+      "learning_rate": 7.915477646419351e-07,
+      "loss": 0.8625,
+      "step": 115880
+    },
+    {
+      "epoch": 7.873692077727952,
+      "grad_norm": 8.5,
+      "learning_rate": 7.89424514200299e-07,
+      "loss": 0.8671,
+      "step": 115885
+    },
+    {
+      "epoch": 7.874031797798614,
+      "grad_norm": 8.125,
+      "learning_rate": 7.873012637586628e-07,
+      "loss": 0.8895,
+      "step": 115890
+    },
+    {
+      "epoch": 7.8743715178692755,
+      "grad_norm": 7.1875,
+      "learning_rate": 7.851780133170267e-07,
+      "loss": 0.8055,
+      "step": 115895
+    },
+    {
+      "epoch": 7.8747112379399375,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.830547628753907e-07,
+      "loss": 0.8319,
+      "step": 115900
+    },
+    {
+      "epoch": 7.8750509580106,
+      "grad_norm": 8.1875,
+      "learning_rate": 7.809315124337546e-07,
+      "loss": 0.7544,
+      "step": 115905
+    },
+    {
+      "epoch": 7.875390678081261,
+      "grad_norm": 9.875,
+      "learning_rate": 7.788082619921185e-07,
+      "loss": 0.8431,
+      "step": 115910
+    },
+    {
+      "epoch": 7.875730398151923,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.766850115504824e-07,
+      "loss": 0.8181,
+      "step": 115915
+    },
+    {
+      "epoch": 7.876070118222585,
+      "grad_norm": 7.90625,
+      "learning_rate": 7.745617611088463e-07,
+      "loss": 0.7794,
+      "step": 115920
+    },
+    {
+      "epoch": 7.876409838293246,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.724385106672102e-07,
+      "loss": 0.867,
+      "step": 115925
+    },
+    {
+      "epoch": 7.876749558363908,
+      "grad_norm": 8.5,
+      "learning_rate": 7.703152602255741e-07,
+      "loss": 0.8642,
+      "step": 115930
+    },
+    {
+      "epoch": 7.87708927843457,
+      "grad_norm": 9.375,
+      "learning_rate": 7.681920097839381e-07,
+      "loss": 0.8419,
+      "step": 115935
+    },
+    {
+      "epoch": 7.8774289985052315,
+      "grad_norm": 7.375,
+      "learning_rate": 7.660687593423019e-07,
+      "loss": 0.8299,
+      "step": 115940
+    },
+    {
+      "epoch": 7.877768718575894,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.639455089006659e-07,
+      "loss": 0.8588,
+      "step": 115945
+    },
+    {
+      "epoch": 7.878108438646555,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.618222584590297e-07,
+      "loss": 0.7581,
+      "step": 115950
+    },
+    {
+      "epoch": 7.878448158717217,
+      "grad_norm": 8.5,
+      "learning_rate": 7.596990080173937e-07,
+      "loss": 0.8389,
+      "step": 115955
+    },
+    {
+      "epoch": 7.878787878787879,
+      "grad_norm": 7.75,
+      "learning_rate": 7.575757575757576e-07,
+      "loss": 0.866,
+      "step": 115960
+    },
+    {
+      "epoch": 7.87912759885854,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.554525071341215e-07,
+      "loss": 0.766,
+      "step": 115965
+    },
+    {
+      "epoch": 7.879467318929202,
+      "grad_norm": 7.625,
+      "learning_rate": 7.533292566924854e-07,
+      "loss": 0.836,
+      "step": 115970
+    },
+    {
+      "epoch": 7.879807038999864,
+      "grad_norm": 9.0,
+      "learning_rate": 7.512060062508493e-07,
+      "loss": 0.7688,
+      "step": 115975
+    },
+    {
+      "epoch": 7.880146759070525,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.490827558092132e-07,
+      "loss": 0.8593,
+      "step": 115980
+    },
+    {
+      "epoch": 7.8804864791411875,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.469595053675771e-07,
+      "loss": 0.8382,
+      "step": 115985
+    },
+    {
+      "epoch": 7.88082619921185,
+      "grad_norm": 7.40625,
+      "learning_rate": 7.448362549259411e-07,
+      "loss": 0.8551,
+      "step": 115990
+    },
+    {
+      "epoch": 7.881165919282511,
+      "grad_norm": 8.0625,
+      "learning_rate": 7.427130044843049e-07,
+      "loss": 0.7546,
+      "step": 115995
+    },
+    {
+      "epoch": 7.881505639353173,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.405897540426689e-07,
+      "loss": 0.9128,
+      "step": 116000
+    },
+    {
+      "epoch": 7.881845359423835,
+      "grad_norm": 7.46875,
+      "learning_rate": 7.384665036010328e-07,
+      "loss": 0.8951,
+      "step": 116005
+    },
+    {
+      "epoch": 7.882185079494496,
+      "grad_norm": 9.5625,
+      "learning_rate": 7.363432531593967e-07,
+      "loss": 0.932,
+      "step": 116010
+    },
+    {
+      "epoch": 7.882524799565158,
+      "grad_norm": 8.0,
+      "learning_rate": 7.342200027177606e-07,
+      "loss": 0.866,
+      "step": 116015
+    },
+    {
+      "epoch": 7.88286451963582,
+      "grad_norm": 7.65625,
+      "learning_rate": 7.320967522761245e-07,
+      "loss": 0.8467,
+      "step": 116020
+    },
+    {
+      "epoch": 7.883204239706481,
+      "grad_norm": 8.6875,
+      "learning_rate": 7.299735018344884e-07,
+      "loss": 0.9207,
+      "step": 116025
+    },
+    {
+      "epoch": 7.8835439597771435,
+      "grad_norm": 9.1875,
+      "learning_rate": 7.278502513928523e-07,
+      "loss": 0.8857,
+      "step": 116030
+    },
+    {
+      "epoch": 7.883883679847806,
+      "grad_norm": 7.875,
+      "learning_rate": 7.257270009512162e-07,
+      "loss": 0.7453,
+      "step": 116035
+    },
+    {
+      "epoch": 7.884223399918467,
+      "grad_norm": 7.84375,
+      "learning_rate": 7.236037505095801e-07,
+      "loss": 0.9337,
+      "step": 116040
+    },
+    {
+      "epoch": 7.884563119989129,
+      "grad_norm": 8.4375,
+      "learning_rate": 7.214805000679441e-07,
+      "loss": 0.8416,
+      "step": 116045
+    },
+    {
+      "epoch": 7.884902840059791,
+      "grad_norm": 7.71875,
+      "learning_rate": 7.193572496263079e-07,
+      "loss": 0.8166,
+      "step": 116050
+    },
+    {
+      "epoch": 7.885242560130452,
+      "grad_norm": 8.375,
+      "learning_rate": 7.172339991846719e-07,
+      "loss": 0.9845,
+      "step": 116055
+    },
+    {
+      "epoch": 7.885582280201114,
+      "grad_norm": 8.25,
+      "learning_rate": 7.151107487430358e-07,
+      "loss": 0.8026,
+      "step": 116060
+    },
+    {
+      "epoch": 7.885922000271776,
+      "grad_norm": 7.375,
+      "learning_rate": 7.129874983013997e-07,
+      "loss": 0.8701,
+      "step": 116065
+    },
+    {
+      "epoch": 7.886261720342437,
+      "grad_norm": 7.78125,
+      "learning_rate": 7.108642478597636e-07,
+      "loss": 0.8712,
+      "step": 116070
+    },
+    {
+      "epoch": 7.8866014404130995,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.087409974181275e-07,
+      "loss": 0.7782,
+      "step": 116075
+    },
+    {
+      "epoch": 7.886941160483762,
+      "grad_norm": 8.125,
+      "learning_rate": 7.066177469764914e-07,
+      "loss": 0.7561,
+      "step": 116080
+    },
+    {
+      "epoch": 7.887280880554423,
+      "grad_norm": 8.5625,
+      "learning_rate": 7.044944965348553e-07,
+      "loss": 0.8072,
+      "step": 116085
+    },
+    {
+      "epoch": 7.887620600625085,
+      "grad_norm": 7.21875,
+      "learning_rate": 7.023712460932193e-07,
+      "loss": 0.7729,
+      "step": 116090
+    },
+    {
+      "epoch": 7.887960320695747,
+      "grad_norm": 7.8125,
+      "learning_rate": 7.002479956515831e-07,
+      "loss": 0.879,
+      "step": 116095
+    },
+    {
+      "epoch": 7.888300040766408,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.981247452099471e-07,
+      "loss": 0.9289,
+      "step": 116100
+    },
+    {
+      "epoch": 7.88863976083707,
+      "grad_norm": 7.28125,
+      "learning_rate": 6.96001494768311e-07,
+      "loss": 0.7811,
+      "step": 116105
+    },
+    {
+      "epoch": 7.888979480907732,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.938782443266749e-07,
+      "loss": 0.7977,
+      "step": 116110
+    },
+    {
+      "epoch": 7.8893192009783935,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.917549938850388e-07,
+      "loss": 0.8576,
+      "step": 116115
+    },
+    {
+      "epoch": 7.8896589210490555,
+      "grad_norm": 8.1875,
+      "learning_rate": 6.896317434434027e-07,
+      "loss": 0.8121,
+      "step": 116120
+    },
+    {
+      "epoch": 7.889998641119718,
+      "grad_norm": 8.0,
+      "learning_rate": 6.875084930017666e-07,
+      "loss": 0.8539,
+      "step": 116125
+    },
+    {
+      "epoch": 7.890338361190379,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.853852425601305e-07,
+      "loss": 0.8464,
+      "step": 116130
+    },
+    {
+      "epoch": 7.890678081261041,
+      "grad_norm": 6.84375,
+      "learning_rate": 6.832619921184944e-07,
+      "loss": 0.8573,
+      "step": 116135
+    },
+    {
+      "epoch": 7.891017801331703,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.811387416768583e-07,
+      "loss": 0.8174,
+      "step": 116140
+    },
+    {
+      "epoch": 7.891357521402364,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.790154912352223e-07,
+      "loss": 0.8866,
+      "step": 116145
+    },
+    {
+      "epoch": 7.891697241473026,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.768922407935861e-07,
+      "loss": 0.7613,
+      "step": 116150
+    },
+    {
+      "epoch": 7.892036961543688,
+      "grad_norm": 8.8125,
+      "learning_rate": 6.747689903519501e-07,
+      "loss": 0.8657,
+      "step": 116155
+    },
+    {
+      "epoch": 7.8923766816143495,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.72645739910314e-07,
+      "loss": 0.775,
+      "step": 116160
+    },
+    {
+      "epoch": 7.8927164016850115,
+      "grad_norm": 7.5,
+      "learning_rate": 6.705224894686779e-07,
+      "loss": 0.8002,
+      "step": 116165
+    },
+    {
+      "epoch": 7.893056121755674,
+      "grad_norm": 7.5625,
+      "learning_rate": 6.683992390270418e-07,
+      "loss": 0.9037,
+      "step": 116170
+    },
+    {
+      "epoch": 7.893395841826335,
+      "grad_norm": 7.15625,
+      "learning_rate": 6.662759885854057e-07,
+      "loss": 0.9149,
+      "step": 116175
+    },
+    {
+      "epoch": 7.893735561896997,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.641527381437696e-07,
+      "loss": 0.8008,
+      "step": 116180
+    },
+    {
+      "epoch": 7.894075281967659,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.620294877021335e-07,
+      "loss": 0.8861,
+      "step": 116185
+    },
+    {
+      "epoch": 7.89441500203832,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.599062372604974e-07,
+      "loss": 0.8332,
+      "step": 116190
+    },
+    {
+      "epoch": 7.894754722108982,
+      "grad_norm": 7.5,
+      "learning_rate": 6.577829868188613e-07,
+      "loss": 0.8175,
+      "step": 116195
+    },
+    {
+      "epoch": 7.895094442179644,
+      "grad_norm": 8.0,
+      "learning_rate": 6.556597363772253e-07,
+      "loss": 0.8256,
+      "step": 116200
+    },
+    {
+      "epoch": 7.8954341622503055,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.53536485935589e-07,
+      "loss": 0.874,
+      "step": 116205
+    },
+    {
+      "epoch": 7.8957738823209676,
+      "grad_norm": 8.5625,
+      "learning_rate": 6.51413235493953e-07,
+      "loss": 0.8362,
+      "step": 116210
+    },
+    {
+      "epoch": 7.89611360239163,
+      "grad_norm": 7.65625,
+      "learning_rate": 6.49289985052317e-07,
+      "loss": 0.9389,
+      "step": 116215
+    },
+    {
+      "epoch": 7.896453322462291,
+      "grad_norm": 8.6875,
+      "learning_rate": 6.471667346106808e-07,
+      "loss": 0.9436,
+      "step": 116220
+    },
+    {
+      "epoch": 7.896793042532953,
+      "grad_norm": 6.84375,
+      "learning_rate": 6.450434841690447e-07,
+      "loss": 0.8378,
+      "step": 116225
+    },
+    {
+      "epoch": 7.897132762603615,
+      "grad_norm": 8.125,
+      "learning_rate": 6.429202337274086e-07,
+      "loss": 0.8661,
+      "step": 116230
+    },
+    {
+      "epoch": 7.897472482674276,
+      "grad_norm": 8.3125,
+      "learning_rate": 6.407969832857725e-07,
+      "loss": 0.7969,
+      "step": 116235
+    },
+    {
+      "epoch": 7.897812202744938,
+      "grad_norm": 7.71875,
+      "learning_rate": 6.386737328441364e-07,
+      "loss": 0.8805,
+      "step": 116240
+    },
+    {
+      "epoch": 7.8981519228156,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.365504824025004e-07,
+      "loss": 0.8181,
+      "step": 116245
+    },
+    {
+      "epoch": 7.8984916428862615,
+      "grad_norm": 8.0,
+      "learning_rate": 6.344272319608642e-07,
+      "loss": 0.8585,
+      "step": 116250
+    },
+    {
+      "epoch": 7.898831362956924,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.323039815192282e-07,
+      "loss": 0.7452,
+      "step": 116255
+    },
+    {
+      "epoch": 7.899171083027586,
+      "grad_norm": 8.375,
+      "learning_rate": 6.301807310775921e-07,
+      "loss": 0.8108,
+      "step": 116260
+    },
+    {
+      "epoch": 7.899510803098247,
+      "grad_norm": 7.96875,
+      "learning_rate": 6.28057480635956e-07,
+      "loss": 0.9261,
+      "step": 116265
+    },
+    {
+      "epoch": 7.899850523168909,
+      "grad_norm": 9.125,
+      "learning_rate": 6.259342301943199e-07,
+      "loss": 0.8683,
+      "step": 116270
+    },
+    {
+      "epoch": 7.900190243239571,
+      "grad_norm": 6.625,
+      "learning_rate": 6.238109797526837e-07,
+      "loss": 0.8446,
+      "step": 116275
+    },
+    {
+      "epoch": 7.900529963310232,
+      "grad_norm": 8.25,
+      "learning_rate": 6.216877293110477e-07,
+      "loss": 0.8736,
+      "step": 116280
+    },
+    {
+      "epoch": 7.900869683380894,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.195644788694116e-07,
+      "loss": 0.8389,
+      "step": 116285
+    },
+    {
+      "epoch": 7.901209403451556,
+      "grad_norm": 8.4375,
+      "learning_rate": 6.174412284277755e-07,
+      "loss": 0.9308,
+      "step": 116290
+    },
+    {
+      "epoch": 7.9015491235222175,
+      "grad_norm": 7.375,
+      "learning_rate": 6.153179779861394e-07,
+      "loss": 0.8122,
+      "step": 116295
+    },
+    {
+      "epoch": 7.90188884359288,
+      "grad_norm": 7.4375,
+      "learning_rate": 6.131947275445034e-07,
+      "loss": 0.8383,
+      "step": 116300
+    },
+    {
+      "epoch": 7.902228563663542,
+      "grad_norm": 8.9375,
+      "learning_rate": 6.110714771028672e-07,
+      "loss": 0.821,
+      "step": 116305
+    },
+    {
+      "epoch": 7.902568283734203,
+      "grad_norm": 8.25,
+      "learning_rate": 6.089482266612312e-07,
+      "loss": 0.7711,
+      "step": 116310
+    },
+    {
+      "epoch": 7.902908003804865,
+      "grad_norm": 7.25,
+      "learning_rate": 6.068249762195951e-07,
+      "loss": 0.8108,
+      "step": 116315
+    },
+    {
+      "epoch": 7.903247723875527,
+      "grad_norm": 8.75,
+      "learning_rate": 6.04701725777959e-07,
+      "loss": 0.852,
+      "step": 116320
+    },
+    {
+      "epoch": 7.903587443946188,
+      "grad_norm": 8.375,
+      "learning_rate": 6.025784753363229e-07,
+      "loss": 0.9218,
+      "step": 116325
+    },
+    {
+      "epoch": 7.90392716401685,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.004552248946868e-07,
+      "loss": 0.8596,
+      "step": 116330
+    },
+    {
+      "epoch": 7.904266884087512,
+      "grad_norm": 7.875,
+      "learning_rate": 5.983319744530507e-07,
+      "loss": 0.896,
+      "step": 116335
+    },
+    {
+      "epoch": 7.9046066041581735,
+      "grad_norm": 7.0,
+      "learning_rate": 5.962087240114146e-07,
+      "loss": 0.9151,
+      "step": 116340
+    },
+    {
+      "epoch": 7.904946324228836,
+      "grad_norm": 8.125,
+      "learning_rate": 5.940854735697786e-07,
+      "loss": 0.8001,
+      "step": 116345
+    },
+    {
+      "epoch": 7.905286044299498,
+      "grad_norm": 8.875,
+      "learning_rate": 5.919622231281424e-07,
+      "loss": 0.8619,
+      "step": 116350
+    },
+    {
+      "epoch": 7.905625764370159,
+      "grad_norm": 8.75,
+      "learning_rate": 5.898389726865064e-07,
+      "loss": 0.8781,
+      "step": 116355
+    },
+    {
+      "epoch": 7.905965484440821,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.877157222448702e-07,
+      "loss": 0.8486,
+      "step": 116360
+    },
+    {
+      "epoch": 7.906305204511483,
+      "grad_norm": 9.3125,
+      "learning_rate": 5.855924718032342e-07,
+      "loss": 0.9877,
+      "step": 116365
+    },
+    {
+      "epoch": 7.906644924582144,
+      "grad_norm": 7.71875,
+      "learning_rate": 5.834692213615981e-07,
+      "loss": 0.8965,
+      "step": 116370
+    },
+    {
+      "epoch": 7.906984644652806,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.813459709199619e-07,
+      "loss": 0.7632,
+      "step": 116375
+    },
+    {
+      "epoch": 7.907324364723468,
+      "grad_norm": 6.71875,
+      "learning_rate": 5.792227204783259e-07,
+      "loss": 0.8058,
+      "step": 116380
+    },
+    {
+      "epoch": 7.9076640847941295,
+      "grad_norm": 7.375,
+      "learning_rate": 5.770994700366898e-07,
+      "loss": 0.845,
+      "step": 116385
+    },
+    {
+      "epoch": 7.908003804864792,
+      "grad_norm": 8.6875,
+      "learning_rate": 5.749762195950537e-07,
+      "loss": 0.8785,
+      "step": 116390
+    },
+    {
+      "epoch": 7.908343524935454,
+      "grad_norm": 7.375,
+      "learning_rate": 5.728529691534176e-07,
+      "loss": 0.7446,
+      "step": 116395
+    },
+    {
+      "epoch": 7.908683245006115,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.707297187117816e-07,
+      "loss": 0.8076,
+      "step": 116400
+    },
+    {
+      "epoch": 7.909022965076777,
+      "grad_norm": 8.75,
+      "learning_rate": 5.686064682701454e-07,
+      "loss": 0.8595,
+      "step": 116405
+    },
+    {
+      "epoch": 7.909362685147439,
+      "grad_norm": 7.875,
+      "learning_rate": 5.664832178285094e-07,
+      "loss": 0.806,
+      "step": 116410
+    },
+    {
+      "epoch": 7.9097024052181,
+      "grad_norm": 8.625,
+      "learning_rate": 5.643599673868733e-07,
+      "loss": 0.9006,
+      "step": 116415
+    },
+    {
+      "epoch": 7.910042125288762,
+      "grad_norm": 7.4375,
+      "learning_rate": 5.622367169452371e-07,
+      "loss": 0.8108,
+      "step": 116420
+    },
+    {
+      "epoch": 7.9103818453594235,
+      "grad_norm": 7.28125,
+      "learning_rate": 5.601134665036011e-07,
+      "loss": 0.8846,
+      "step": 116425
+    },
+    {
+      "epoch": 7.9107215654300855,
+      "grad_norm": 6.4375,
+      "learning_rate": 5.57990216061965e-07,
+      "loss": 0.8615,
+      "step": 116430
+    },
+    {
+      "epoch": 7.911061285500748,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.558669656203289e-07,
+      "loss": 0.7905,
+      "step": 116435
+    },
+    {
+      "epoch": 7.911401005571409,
+      "grad_norm": 7.625,
+      "learning_rate": 5.537437151786928e-07,
+      "loss": 0.9105,
+      "step": 116440
+    },
+    {
+      "epoch": 7.911740725642071,
+      "grad_norm": 8.25,
+      "learning_rate": 5.516204647370567e-07,
+      "loss": 0.8654,
+      "step": 116445
+    },
+    {
+      "epoch": 7.912080445712733,
+      "grad_norm": 8.375,
+      "learning_rate": 5.494972142954206e-07,
+      "loss": 0.8441,
+      "step": 116450
+    },
+    {
+      "epoch": 7.912420165783394,
+      "grad_norm": 8.75,
+      "learning_rate": 5.473739638537846e-07,
+      "loss": 0.8703,
+      "step": 116455
+    },
+    {
+      "epoch": 7.912759885854056,
+      "grad_norm": 8.625,
+      "learning_rate": 5.452507134121484e-07,
+      "loss": 0.801,
+      "step": 116460
+    },
+    {
+      "epoch": 7.913099605924718,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.431274629705123e-07,
+      "loss": 0.8801,
+      "step": 116465
+    },
+    {
+      "epoch": 7.9134393259953795,
+      "grad_norm": 7.59375,
+      "learning_rate": 5.410042125288763e-07,
+      "loss": 0.6865,
+      "step": 116470
+    },
+    {
+      "epoch": 7.9137790460660415,
+      "grad_norm": 8.25,
+      "learning_rate": 5.388809620872401e-07,
+      "loss": 0.9646,
+      "step": 116475
+    },
+    {
+      "epoch": 7.914118766136704,
+      "grad_norm": 6.59375,
+      "learning_rate": 5.367577116456041e-07,
+      "loss": 0.8829,
+      "step": 116480
+    },
+    {
+      "epoch": 7.914458486207365,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.34634461203968e-07,
+      "loss": 0.756,
+      "step": 116485
+    },
+    {
+      "epoch": 7.914798206278027,
+      "grad_norm": 8.25,
+      "learning_rate": 5.325112107623319e-07,
+      "loss": 0.7793,
+      "step": 116490
+    },
+    {
+      "epoch": 7.915137926348689,
+      "grad_norm": 8.1875,
+      "learning_rate": 5.303879603206958e-07,
+      "loss": 0.8222,
+      "step": 116495
+    },
+    {
+      "epoch": 7.91547764641935,
+      "grad_norm": 6.84375,
+      "learning_rate": 5.282647098790598e-07,
+      "loss": 0.8153,
+      "step": 116500
+    },
+    {
+      "epoch": 7.915817366490012,
+      "grad_norm": 7.5,
+      "learning_rate": 5.261414594374236e-07,
+      "loss": 0.8361,
+      "step": 116505
+    },
+    {
+      "epoch": 7.916157086560674,
+      "grad_norm": 8.125,
+      "learning_rate": 5.240182089957875e-07,
+      "loss": 0.8868,
+      "step": 116510
+    },
+    {
+      "epoch": 7.9164968066313355,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.218949585541515e-07,
+      "loss": 0.8781,
+      "step": 116515
+    },
+    {
+      "epoch": 7.916836526701998,
+      "grad_norm": 7.9375,
+      "learning_rate": 5.197717081125153e-07,
+      "loss": 0.7865,
+      "step": 116520
+    },
+    {
+      "epoch": 7.91717624677266,
+      "grad_norm": 8.25,
+      "learning_rate": 5.176484576708793e-07,
+      "loss": 0.8165,
+      "step": 116525
+    },
+    {
+      "epoch": 7.917515966843321,
+      "grad_norm": 8.125,
+      "learning_rate": 5.15525207229243e-07,
+      "loss": 0.9136,
+      "step": 116530
+    },
+    {
+      "epoch": 7.917855686913983,
+      "grad_norm": 7.53125,
+      "learning_rate": 5.134019567876071e-07,
+      "loss": 0.829,
+      "step": 116535
+    },
+    {
+      "epoch": 7.918195406984645,
+      "grad_norm": 7.78125,
+      "learning_rate": 5.11278706345971e-07,
+      "loss": 0.8498,
+      "step": 116540
+    },
+    {
+      "epoch": 7.918535127055306,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.091554559043349e-07,
+      "loss": 0.8733,
+      "step": 116545
+    },
+    {
+      "epoch": 7.918874847125968,
+      "grad_norm": 7.125,
+      "learning_rate": 5.070322054626987e-07,
+      "loss": 0.7608,
+      "step": 116550
+    },
+    {
+      "epoch": 7.91921456719663,
+      "grad_norm": 7.65625,
+      "learning_rate": 5.049089550210626e-07,
+      "loss": 0.8943,
+      "step": 116555
+    },
+    {
+      "epoch": 7.9195542872672915,
+      "grad_norm": 8.25,
+      "learning_rate": 5.027857045794265e-07,
+      "loss": 0.9571,
+      "step": 116560
+    },
+    {
+      "epoch": 7.919894007337954,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.006624541377904e-07,
+      "loss": 0.8667,
+      "step": 116565
+    },
+    {
+      "epoch": 7.920233727408616,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.985392036961544e-07,
+      "loss": 0.8034,
+      "step": 116570
+    },
+    {
+      "epoch": 7.920573447479277,
+      "grad_norm": 6.28125,
+      "learning_rate": 4.964159532545182e-07,
+      "loss": 0.7957,
+      "step": 116575
+    },
+    {
+      "epoch": 7.920913167549939,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.942927028128822e-07,
+      "loss": 0.8154,
+      "step": 116580
+    },
+    {
+      "epoch": 7.921252887620601,
+      "grad_norm": 7.40625,
+      "learning_rate": 4.921694523712461e-07,
+      "loss": 0.8259,
+      "step": 116585
+    },
+    {
+      "epoch": 7.921592607691262,
+      "grad_norm": 7.5,
+      "learning_rate": 4.9004620192961e-07,
+      "loss": 0.7903,
+      "step": 116590
+    },
+    {
+      "epoch": 7.921932327761924,
+      "grad_norm": 7.5,
+      "learning_rate": 4.879229514879739e-07,
+      "loss": 0.851,
+      "step": 116595
+    },
+    {
+      "epoch": 7.922272047832586,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.857997010463378e-07,
+      "loss": 0.8982,
+      "step": 116600
+    },
+    {
+      "epoch": 7.9226117679032475,
+      "grad_norm": 7.875,
+      "learning_rate": 4.836764506047017e-07,
+      "loss": 0.8484,
+      "step": 116605
+    },
+    {
+      "epoch": 7.92295148797391,
+      "grad_norm": 8.75,
+      "learning_rate": 4.815532001630656e-07,
+      "loss": 0.8056,
+      "step": 116610
+    },
+    {
+      "epoch": 7.923291208044572,
+      "grad_norm": 7.78125,
+      "learning_rate": 4.794299497214295e-07,
+      "loss": 0.8336,
+      "step": 116615
+    },
+    {
+      "epoch": 7.923630928115233,
+      "grad_norm": 7.84375,
+      "learning_rate": 4.773066992797934e-07,
+      "loss": 0.8232,
+      "step": 116620
+    },
+    {
+      "epoch": 7.923970648185895,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.7518344883815743e-07,
+      "loss": 0.9306,
+      "step": 116625
+    },
+    {
+      "epoch": 7.924310368256556,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.730601983965213e-07,
+      "loss": 0.9042,
+      "step": 116630
+    },
+    {
+      "epoch": 7.924650088327218,
+      "grad_norm": 9.4375,
+      "learning_rate": 4.7093694795488517e-07,
+      "loss": 0.8679,
+      "step": 116635
+    },
+    {
+      "epoch": 7.92498980839788,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.688136975132491e-07,
+      "loss": 0.8822,
+      "step": 116640
+    },
+    {
+      "epoch": 7.925329528468541,
+      "grad_norm": 7.90625,
+      "learning_rate": 4.6669044707161297e-07,
+      "loss": 0.8935,
+      "step": 116645
+    },
+    {
+      "epoch": 7.9256692485392035,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.645671966299769e-07,
+      "loss": 0.8735,
+      "step": 116650
+    },
+    {
+      "epoch": 7.926008968609866,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.6244394618834087e-07,
+      "loss": 0.8353,
+      "step": 116655
+    },
+    {
+      "epoch": 7.926348688680527,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.603206957467047e-07,
+      "loss": 0.8382,
+      "step": 116660
+    },
+    {
+      "epoch": 7.926688408751189,
+      "grad_norm": 8.0,
+      "learning_rate": 4.5819744530506867e-07,
+      "loss": 0.8799,
+      "step": 116665
+    },
+    {
+      "epoch": 7.927028128821851,
+      "grad_norm": 7.15625,
+      "learning_rate": 4.560741948634326e-07,
+      "loss": 0.8742,
+      "step": 116670
+    },
+    {
+      "epoch": 7.927367848892512,
+      "grad_norm": 8.125,
+      "learning_rate": 4.5395094442179646e-07,
+      "loss": 0.8319,
+      "step": 116675
+    },
+    {
+      "epoch": 7.927707568963174,
+      "grad_norm": 7.3125,
+      "learning_rate": 4.5182769398016036e-07,
+      "loss": 0.8084,
+      "step": 116680
+    },
+    {
+      "epoch": 7.928047289033836,
+      "grad_norm": 7.875,
+      "learning_rate": 4.4970444353852426e-07,
+      "loss": 0.8925,
+      "step": 116685
+    },
+    {
+      "epoch": 7.9283870091044975,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.4758119309688816e-07,
+      "loss": 0.8716,
+      "step": 116690
+    },
+    {
+      "epoch": 7.9287267291751595,
+      "grad_norm": 9.125,
+      "learning_rate": 4.454579426552521e-07,
+      "loss": 0.8179,
+      "step": 116695
+    },
+    {
+      "epoch": 7.929066449245822,
+      "grad_norm": 7.25,
+      "learning_rate": 4.4333469221361595e-07,
+      "loss": 0.8359,
+      "step": 116700
+    },
+    {
+      "epoch": 7.929406169316483,
+      "grad_norm": 7.625,
+      "learning_rate": 4.412114417719799e-07,
+      "loss": 0.7983,
+      "step": 116705
+    },
+    {
+      "epoch": 7.929745889387145,
+      "grad_norm": 8.875,
+      "learning_rate": 4.3908819133034385e-07,
+      "loss": 0.9317,
+      "step": 116710
+    },
+    {
+      "epoch": 7.930085609457807,
+      "grad_norm": 8.4375,
+      "learning_rate": 4.369649408887077e-07,
+      "loss": 0.826,
+      "step": 116715
+    },
+    {
+      "epoch": 7.930425329528468,
+      "grad_norm": 7.1875,
+      "learning_rate": 4.3484169044707165e-07,
+      "loss": 0.8783,
+      "step": 116720
+    },
+    {
+      "epoch": 7.93076504959913,
+      "grad_norm": 7.09375,
+      "learning_rate": 4.3271844000543555e-07,
+      "loss": 0.8839,
+      "step": 116725
+    },
+    {
+      "epoch": 7.931104769669792,
+      "grad_norm": 8.0625,
+      "learning_rate": 4.3059518956379945e-07,
+      "loss": 0.897,
+      "step": 116730
+    },
+    {
+      "epoch": 7.9314444897404535,
+      "grad_norm": 8.3125,
+      "learning_rate": 4.2847193912216334e-07,
+      "loss": 0.7936,
+      "step": 116735
+    },
+    {
+      "epoch": 7.9317842098111155,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.263486886805273e-07,
+      "loss": 0.9251,
+      "step": 116740
+    },
+    {
+      "epoch": 7.932123929881778,
+      "grad_norm": 7.0625,
+      "learning_rate": 4.2422543823889114e-07,
+      "loss": 0.8285,
+      "step": 116745
+    },
+    {
+      "epoch": 7.932463649952439,
+      "grad_norm": 8.6875,
+      "learning_rate": 4.221021877972551e-07,
+      "loss": 0.8363,
+      "step": 116750
+    },
+    {
+      "epoch": 7.932803370023101,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.1997893735561904e-07,
+      "loss": 0.7829,
+      "step": 116755
+    },
+    {
+      "epoch": 7.933143090093763,
+      "grad_norm": 7.96875,
+      "learning_rate": 4.178556869139829e-07,
+      "loss": 0.862,
+      "step": 116760
+    },
+    {
+      "epoch": 7.933482810164424,
+      "grad_norm": 8.5,
+      "learning_rate": 4.1573243647234684e-07,
+      "loss": 0.8492,
+      "step": 116765
+    },
+    {
+      "epoch": 7.933822530235086,
+      "grad_norm": 8.375,
+      "learning_rate": 4.136091860307107e-07,
+      "loss": 0.938,
+      "step": 116770
+    },
+    {
+      "epoch": 7.934162250305748,
+      "grad_norm": 7.6875,
+      "learning_rate": 4.1148593558907463e-07,
+      "loss": 0.7639,
+      "step": 116775
+    },
+    {
+      "epoch": 7.9345019703764095,
+      "grad_norm": 10.0,
+      "learning_rate": 4.0936268514743853e-07,
+      "loss": 0.7539,
+      "step": 116780
+    },
+    {
+      "epoch": 7.9348416904470715,
+      "grad_norm": 7.34375,
+      "learning_rate": 4.072394347058024e-07,
+      "loss": 0.9129,
+      "step": 116785
+    },
+    {
+      "epoch": 7.935181410517734,
+      "grad_norm": 8.875,
+      "learning_rate": 4.0511618426416633e-07,
+      "loss": 0.7721,
+      "step": 116790
+    },
+    {
+      "epoch": 7.935521130588395,
+      "grad_norm": 8.8125,
+      "learning_rate": 4.029929338225303e-07,
+      "loss": 0.7719,
+      "step": 116795
+    },
+    {
+      "epoch": 7.935860850659057,
+      "grad_norm": 8.0,
+      "learning_rate": 4.008696833808941e-07,
+      "loss": 0.7525,
+      "step": 116800
+    },
+    {
+      "epoch": 7.936200570729719,
+      "grad_norm": 7.25,
+      "learning_rate": 3.987464329392581e-07,
+      "loss": 0.8017,
+      "step": 116805
+    },
+    {
+      "epoch": 7.93654029080038,
+      "grad_norm": 8.5,
+      "learning_rate": 3.96623182497622e-07,
+      "loss": 0.7629,
+      "step": 116810
+    },
+    {
+      "epoch": 7.936880010871042,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.9449993205598587e-07,
+      "loss": 0.8286,
+      "step": 116815
+    },
+    {
+      "epoch": 7.937219730941704,
+      "grad_norm": 8.6875,
+      "learning_rate": 3.923766816143498e-07,
+      "loss": 0.8181,
+      "step": 116820
+    },
+    {
+      "epoch": 7.9375594510123655,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.902534311727137e-07,
+      "loss": 0.8268,
+      "step": 116825
+    },
+    {
+      "epoch": 7.937899171083028,
+      "grad_norm": 8.125,
+      "learning_rate": 3.8813018073107756e-07,
+      "loss": 0.8077,
+      "step": 116830
+    },
+    {
+      "epoch": 7.93823889115369,
+      "grad_norm": 9.0625,
+      "learning_rate": 3.860069302894415e-07,
+      "loss": 0.8853,
+      "step": 116835
+    },
+    {
+      "epoch": 7.938578611224351,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.838836798478054e-07,
+      "loss": 0.6806,
+      "step": 116840
+    },
+    {
+      "epoch": 7.938918331295013,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.817604294061693e-07,
+      "loss": 0.9401,
+      "step": 116845
+    },
+    {
+      "epoch": 7.939258051365675,
+      "grad_norm": 8.125,
+      "learning_rate": 3.7963717896453326e-07,
+      "loss": 0.7977,
+      "step": 116850
+    },
+    {
+      "epoch": 7.939597771436336,
+      "grad_norm": 8.75,
+      "learning_rate": 3.7751392852289716e-07,
+      "loss": 0.7942,
+      "step": 116855
+    },
+    {
+      "epoch": 7.939937491506998,
+      "grad_norm": 8.5,
+      "learning_rate": 3.7539067808126106e-07,
+      "loss": 0.9081,
+      "step": 116860
+    },
+    {
+      "epoch": 7.94027721157766,
+      "grad_norm": 8.375,
+      "learning_rate": 3.73267427639625e-07,
+      "loss": 0.8402,
+      "step": 116865
+    },
+    {
+      "epoch": 7.9406169316483215,
+      "grad_norm": 8.125,
+      "learning_rate": 3.711441771979889e-07,
+      "loss": 0.7603,
+      "step": 116870
+    },
+    {
+      "epoch": 7.940956651718984,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.6902092675635275e-07,
+      "loss": 0.8735,
+      "step": 116875
+    },
+    {
+      "epoch": 7.941296371789646,
+      "grad_norm": 7.5625,
+      "learning_rate": 3.6689767631471665e-07,
+      "loss": 0.8339,
+      "step": 116880
+    },
+    {
+      "epoch": 7.941636091860307,
+      "grad_norm": 8.8125,
+      "learning_rate": 3.647744258730806e-07,
+      "loss": 0.9044,
+      "step": 116885
+    },
+    {
+      "epoch": 7.941975811930969,
+      "grad_norm": 8.375,
+      "learning_rate": 3.626511754314445e-07,
+      "loss": 0.8336,
+      "step": 116890
+    },
+    {
+      "epoch": 7.942315532001631,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.605279249898084e-07,
+      "loss": 0.9174,
+      "step": 116895
+    },
+    {
+      "epoch": 7.942655252072292,
+      "grad_norm": 8.625,
+      "learning_rate": 3.5840467454817235e-07,
+      "loss": 0.8126,
+      "step": 116900
+    },
+    {
+      "epoch": 7.942994972142954,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.5628142410653625e-07,
+      "loss": 0.858,
+      "step": 116905
+    },
+    {
+      "epoch": 7.943334692213616,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.5415817366490014e-07,
+      "loss": 0.8972,
+      "step": 116910
+    },
+    {
+      "epoch": 7.9436744122842775,
+      "grad_norm": 7.25,
+      "learning_rate": 3.5203492322326404e-07,
+      "loss": 0.8601,
+      "step": 116915
+    },
+    {
+      "epoch": 7.94401413235494,
+      "grad_norm": 7.5,
+      "learning_rate": 3.49911672781628e-07,
+      "loss": 0.9055,
+      "step": 116920
+    },
+    {
+      "epoch": 7.944353852425602,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.4778842233999184e-07,
+      "loss": 0.8022,
+      "step": 116925
+    },
+    {
+      "epoch": 7.944693572496263,
+      "grad_norm": 7.0,
+      "learning_rate": 3.4566517189835574e-07,
+      "loss": 0.7975,
+      "step": 116930
+    },
+    {
+      "epoch": 7.945033292566925,
+      "grad_norm": 8.0,
+      "learning_rate": 3.435419214567197e-07,
+      "loss": 0.8973,
+      "step": 116935
+    },
+    {
+      "epoch": 7.945373012637587,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.414186710150836e-07,
+      "loss": 0.7892,
+      "step": 116940
+    },
+    {
+      "epoch": 7.945712732708248,
+      "grad_norm": 7.5,
+      "learning_rate": 3.392954205734475e-07,
+      "loss": 0.7997,
+      "step": 116945
+    },
+    {
+      "epoch": 7.94605245277891,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.371721701318114e-07,
+      "loss": 0.8122,
+      "step": 116950
+    },
+    {
+      "epoch": 7.946392172849572,
+      "grad_norm": 8.25,
+      "learning_rate": 3.3504891969017533e-07,
+      "loss": 0.8322,
+      "step": 116955
+    },
+    {
+      "epoch": 7.9467318929202335,
+      "grad_norm": 7.59375,
+      "learning_rate": 3.3292566924853923e-07,
+      "loss": 0.8623,
+      "step": 116960
+    },
+    {
+      "epoch": 7.947071612990896,
+      "grad_norm": 6.75,
+      "learning_rate": 3.3080241880690313e-07,
+      "loss": 0.7873,
+      "step": 116965
+    },
+    {
+      "epoch": 7.947411333061558,
+      "grad_norm": 7.3125,
+      "learning_rate": 3.28679168365267e-07,
+      "loss": 0.783,
+      "step": 116970
+    },
+    {
+      "epoch": 7.947751053132219,
+      "grad_norm": 7.75,
+      "learning_rate": 3.265559179236309e-07,
+      "loss": 0.773,
+      "step": 116975
+    },
+    {
+      "epoch": 7.948090773202881,
+      "grad_norm": 7.53125,
+      "learning_rate": 3.244326674819948e-07,
+      "loss": 0.8386,
+      "step": 116980
+    },
+    {
+      "epoch": 7.948430493273543,
+      "grad_norm": 6.78125,
+      "learning_rate": 3.2230941704035877e-07,
+      "loss": 0.756,
+      "step": 116985
+    },
+    {
+      "epoch": 7.948770213344204,
+      "grad_norm": 8.3125,
+      "learning_rate": 3.2018616659872267e-07,
+      "loss": 0.894,
+      "step": 116990
+    },
+    {
+      "epoch": 7.949109933414866,
+      "grad_norm": 8.75,
+      "learning_rate": 3.1806291615708657e-07,
+      "loss": 0.7791,
+      "step": 116995
+    },
+    {
+      "epoch": 7.949449653485528,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.1593966571545047e-07,
+      "loss": 0.9702,
+      "step": 117000
+    },
+    {
+      "epoch": 7.9497893735561895,
+      "grad_norm": 8.0625,
+      "learning_rate": 3.138164152738144e-07,
+      "loss": 0.8549,
+      "step": 117005
+    },
+    {
+      "epoch": 7.950129093626852,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.116931648321783e-07,
+      "loss": 0.8115,
+      "step": 117010
+    },
+    {
+      "epoch": 7.950468813697514,
+      "grad_norm": 7.25,
+      "learning_rate": 3.095699143905422e-07,
+      "loss": 0.8502,
+      "step": 117015
+    },
+    {
+      "epoch": 7.950808533768175,
+      "grad_norm": 7.8125,
+      "learning_rate": 3.074466639489061e-07,
+      "loss": 0.8148,
+      "step": 117020
+    },
+    {
+      "epoch": 7.951148253838837,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.0532341350727e-07,
+      "loss": 0.802,
+      "step": 117025
+    },
+    {
+      "epoch": 7.951487973909499,
+      "grad_norm": 8.4375,
+      "learning_rate": 3.032001630656339e-07,
+      "loss": 0.8362,
+      "step": 117030
+    },
+    {
+      "epoch": 7.95182769398016,
+      "grad_norm": 7.65625,
+      "learning_rate": 3.010769126239978e-07,
+      "loss": 0.8192,
+      "step": 117035
+    },
+    {
+      "epoch": 7.952167414050822,
+      "grad_norm": 6.34375,
+      "learning_rate": 2.9895366218236176e-07,
+      "loss": 0.9322,
+      "step": 117040
+    },
+    {
+      "epoch": 7.952507134121484,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.9683041174072565e-07,
+      "loss": 0.8485,
+      "step": 117045
+    },
+    {
+      "epoch": 7.9528468541921455,
+      "grad_norm": 9.3125,
+      "learning_rate": 2.9470716129908955e-07,
+      "loss": 0.8582,
+      "step": 117050
+    },
+    {
+      "epoch": 7.953186574262808,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.925839108574535e-07,
+      "loss": 0.8383,
+      "step": 117055
+    },
+    {
+      "epoch": 7.95352629433347,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.904606604158174e-07,
+      "loss": 0.8815,
+      "step": 117060
+    },
+    {
+      "epoch": 7.953866014404131,
+      "grad_norm": 7.03125,
+      "learning_rate": 2.883374099741813e-07,
+      "loss": 0.7385,
+      "step": 117065
+    },
+    {
+      "epoch": 7.954205734474793,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.862141595325452e-07,
+      "loss": 0.8562,
+      "step": 117070
+    },
+    {
+      "epoch": 7.954545454545455,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.840909090909091e-07,
+      "loss": 0.8385,
+      "step": 117075
+    },
+    {
+      "epoch": 7.954885174616116,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.81967658649273e-07,
+      "loss": 0.8179,
+      "step": 117080
+    },
+    {
+      "epoch": 7.955224894686778,
+      "grad_norm": 8.1875,
+      "learning_rate": 2.798444082076369e-07,
+      "loss": 0.8051,
+      "step": 117085
+    },
+    {
+      "epoch": 7.95556461475744,
+      "grad_norm": 6.3125,
+      "learning_rate": 2.7772115776600084e-07,
+      "loss": 0.7743,
+      "step": 117090
+    },
+    {
+      "epoch": 7.9559043348281016,
+      "grad_norm": 8.3125,
+      "learning_rate": 2.7559790732436474e-07,
+      "loss": 0.845,
+      "step": 117095
+    },
+    {
+      "epoch": 7.956244054898764,
+      "grad_norm": 8.375,
+      "learning_rate": 2.7347465688272864e-07,
+      "loss": 0.9106,
+      "step": 117100
+    },
+    {
+      "epoch": 7.956583774969425,
+      "grad_norm": 8.5,
+      "learning_rate": 2.713514064410926e-07,
+      "loss": 0.8893,
+      "step": 117105
+    },
+    {
+      "epoch": 7.956923495040087,
+      "grad_norm": 7.09375,
+      "learning_rate": 2.692281559994565e-07,
+      "loss": 0.7721,
+      "step": 117110
+    },
+    {
+      "epoch": 7.957263215110749,
+      "grad_norm": 6.65625,
+      "learning_rate": 2.671049055578204e-07,
+      "loss": 0.8493,
+      "step": 117115
+    },
+    {
+      "epoch": 7.95760293518141,
+      "grad_norm": 8.5,
+      "learning_rate": 2.649816551161843e-07,
+      "loss": 0.7679,
+      "step": 117120
+    },
+    {
+      "epoch": 7.957942655252072,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.628584046745482e-07,
+      "loss": 0.8066,
+      "step": 117125
+    },
+    {
+      "epoch": 7.958282375322734,
+      "grad_norm": 6.6875,
+      "learning_rate": 2.607351542329121e-07,
+      "loss": 0.7594,
+      "step": 117130
+    },
+    {
+      "epoch": 7.9586220953933955,
+      "grad_norm": 8.25,
+      "learning_rate": 2.58611903791276e-07,
+      "loss": 0.9042,
+      "step": 117135
+    },
+    {
+      "epoch": 7.958961815464058,
+      "grad_norm": 7.8125,
+      "learning_rate": 2.5648865334963993e-07,
+      "loss": 0.9203,
+      "step": 117140
+    },
+    {
+      "epoch": 7.95930153553472,
+      "grad_norm": 7.9375,
+      "learning_rate": 2.543654029080038e-07,
+      "loss": 0.9074,
+      "step": 117145
+    },
+    {
+      "epoch": 7.959641255605381,
+      "grad_norm": 7.34375,
+      "learning_rate": 2.522421524663677e-07,
+      "loss": 0.8031,
+      "step": 117150
+    },
+    {
+      "epoch": 7.959980975676043,
+      "grad_norm": 8.0,
+      "learning_rate": 2.501189020247317e-07,
+      "loss": 0.766,
+      "step": 117155
+    },
+    {
+      "epoch": 7.960320695746705,
+      "grad_norm": 8.0625,
+      "learning_rate": 2.4799565158309557e-07,
+      "loss": 0.8302,
+      "step": 117160
+    },
+    {
+      "epoch": 7.960660415817366,
+      "grad_norm": 7.375,
+      "learning_rate": 2.4587240114145947e-07,
+      "loss": 0.8091,
+      "step": 117165
+    },
+    {
+      "epoch": 7.961000135888028,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.437491506998233e-07,
+      "loss": 0.792,
+      "step": 117170
+    },
+    {
+      "epoch": 7.96133985595869,
+      "grad_norm": 8.375,
+      "learning_rate": 2.4162590025818727e-07,
+      "loss": 0.8363,
+      "step": 117175
+    },
+    {
+      "epoch": 7.9616795760293515,
+      "grad_norm": 8.4375,
+      "learning_rate": 2.3950264981655116e-07,
+      "loss": 0.706,
+      "step": 117180
+    },
+    {
+      "epoch": 7.962019296100014,
+      "grad_norm": 7.65625,
+      "learning_rate": 2.3737939937491506e-07,
+      "loss": 0.7962,
+      "step": 117185
+    },
+    {
+      "epoch": 7.962359016170676,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.35256148933279e-07,
+      "loss": 0.8226,
+      "step": 117190
+    },
+    {
+      "epoch": 7.962698736241337,
+      "grad_norm": 7.96875,
+      "learning_rate": 2.331328984916429e-07,
+      "loss": 0.8412,
+      "step": 117195
+    },
+    {
+      "epoch": 7.963038456311999,
+      "grad_norm": 8.625,
+      "learning_rate": 2.310096480500068e-07,
+      "loss": 0.9393,
+      "step": 117200
+    },
+    {
+      "epoch": 7.963378176382661,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.288863976083707e-07,
+      "loss": 0.81,
+      "step": 117205
+    },
+    {
+      "epoch": 7.963717896453322,
+      "grad_norm": 8.6875,
+      "learning_rate": 2.2676314716673463e-07,
+      "loss": 0.8933,
+      "step": 117210
+    },
+    {
+      "epoch": 7.964057616523984,
+      "grad_norm": 8.0,
+      "learning_rate": 2.2463989672509853e-07,
+      "loss": 0.8196,
+      "step": 117215
+    },
+    {
+      "epoch": 7.964397336594646,
+      "grad_norm": 7.71875,
+      "learning_rate": 2.2251664628346243e-07,
+      "loss": 0.8562,
+      "step": 117220
+    },
+    {
+      "epoch": 7.9647370566653075,
+      "grad_norm": 9.0625,
+      "learning_rate": 2.2039339584182635e-07,
+      "loss": 0.8463,
+      "step": 117225
+    },
+    {
+      "epoch": 7.96507677673597,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1827014540019025e-07,
+      "loss": 0.7678,
+      "step": 117230
+    },
+    {
+      "epoch": 7.965416496806632,
+      "grad_norm": 7.90625,
+      "learning_rate": 2.1614689495855415e-07,
+      "loss": 0.875,
+      "step": 117235
+    },
+    {
+      "epoch": 7.965756216877293,
+      "grad_norm": 7.6875,
+      "learning_rate": 2.1402364451691805e-07,
+      "loss": 0.8527,
+      "step": 117240
+    },
+    {
+      "epoch": 7.966095936947955,
+      "grad_norm": 6.59375,
+      "learning_rate": 2.11900394075282e-07,
+      "loss": 0.9197,
+      "step": 117245
+    },
+    {
+      "epoch": 7.966435657018617,
+      "grad_norm": 7.375,
+      "learning_rate": 2.097771436336459e-07,
+      "loss": 0.7491,
+      "step": 117250
+    },
+    {
+      "epoch": 7.966775377089278,
+      "grad_norm": 7.53125,
+      "learning_rate": 2.0765389319200977e-07,
+      "loss": 0.7724,
+      "step": 117255
+    },
+    {
+      "epoch": 7.96711509715994,
+      "grad_norm": 7.625,
+      "learning_rate": 2.0553064275037372e-07,
+      "loss": 0.7864,
+      "step": 117260
+    },
+    {
+      "epoch": 7.967454817230602,
+      "grad_norm": 8.375,
+      "learning_rate": 2.0340739230873762e-07,
+      "loss": 0.9071,
+      "step": 117265
+    },
+    {
+      "epoch": 7.9677945373012635,
+      "grad_norm": 8.5,
+      "learning_rate": 2.012841418671015e-07,
+      "loss": 0.8457,
+      "step": 117270
+    },
+    {
+      "epoch": 7.968134257371926,
+      "grad_norm": 7.21875,
+      "learning_rate": 1.9916089142546544e-07,
+      "loss": 0.8598,
+      "step": 117275
+    },
+    {
+      "epoch": 7.968473977442588,
+      "grad_norm": 7.09375,
+      "learning_rate": 1.9703764098382934e-07,
+      "loss": 0.8463,
+      "step": 117280
+    },
+    {
+      "epoch": 7.968813697513249,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.9491439054219323e-07,
+      "loss": 0.8129,
+      "step": 117285
+    },
+    {
+      "epoch": 7.969153417583911,
+      "grad_norm": 8.375,
+      "learning_rate": 1.9279114010055716e-07,
+      "loss": 0.8587,
+      "step": 117290
+    },
+    {
+      "epoch": 7.969493137654573,
+      "grad_norm": 8.125,
+      "learning_rate": 1.9066788965892106e-07,
+      "loss": 0.8313,
+      "step": 117295
+    },
+    {
+      "epoch": 7.969832857725234,
+      "grad_norm": 9.0,
+      "learning_rate": 1.8854463921728495e-07,
+      "loss": 0.8578,
+      "step": 117300
+    },
+    {
+      "epoch": 7.970172577795896,
+      "grad_norm": 7.8125,
+      "learning_rate": 1.8642138877564888e-07,
+      "loss": 0.8663,
+      "step": 117305
+    },
+    {
+      "epoch": 7.9705122978665575,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.8429813833401278e-07,
+      "loss": 0.8115,
+      "step": 117310
+    },
+    {
+      "epoch": 7.9708520179372195,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.821748878923767e-07,
+      "loss": 0.7754,
+      "step": 117315
+    },
+    {
+      "epoch": 7.971191738007882,
+      "grad_norm": 7.5625,
+      "learning_rate": 1.800516374507406e-07,
+      "loss": 0.8702,
+      "step": 117320
+    },
+    {
+      "epoch": 7.971531458078543,
+      "grad_norm": 7.59375,
+      "learning_rate": 1.779283870091045e-07,
+      "loss": 0.8506,
+      "step": 117325
+    },
+    {
+      "epoch": 7.971871178149205,
+      "grad_norm": 8.625,
+      "learning_rate": 1.7580513656746842e-07,
+      "loss": 0.8259,
+      "step": 117330
+    },
+    {
+      "epoch": 7.972210898219867,
+      "grad_norm": 8.5,
+      "learning_rate": 1.7368188612583232e-07,
+      "loss": 0.8749,
+      "step": 117335
+    },
+    {
+      "epoch": 7.972550618290528,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.7155863568419624e-07,
+      "loss": 0.8603,
+      "step": 117340
+    },
+    {
+      "epoch": 7.97289033836119,
+      "grad_norm": 7.5,
+      "learning_rate": 1.6943538524256014e-07,
+      "loss": 0.8609,
+      "step": 117345
+    },
+    {
+      "epoch": 7.973230058431852,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.6731213480092404e-07,
+      "loss": 0.7554,
+      "step": 117350
+    },
+    {
+      "epoch": 7.9735697785025135,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.6518888435928794e-07,
+      "loss": 0.8963,
+      "step": 117355
+    },
+    {
+      "epoch": 7.9739094985731755,
+      "grad_norm": 7.9375,
+      "learning_rate": 1.6306563391765186e-07,
+      "loss": 0.9495,
+      "step": 117360
+    },
+    {
+      "epoch": 7.974249218643838,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.6094238347601579e-07,
+      "loss": 0.7968,
+      "step": 117365
+    },
+    {
+      "epoch": 7.974588938714499,
+      "grad_norm": 8.0625,
+      "learning_rate": 1.5881913303437968e-07,
+      "loss": 0.763,
+      "step": 117370
+    },
+    {
+      "epoch": 7.974928658785161,
+      "grad_norm": 6.90625,
+      "learning_rate": 1.5669588259274358e-07,
+      "loss": 0.9657,
+      "step": 117375
+    },
+    {
+      "epoch": 7.975268378855823,
+      "grad_norm": 8.375,
+      "learning_rate": 1.5457263215110748e-07,
+      "loss": 0.8238,
+      "step": 117380
+    },
+    {
+      "epoch": 7.975608098926484,
+      "grad_norm": 7.5,
+      "learning_rate": 1.524493817094714e-07,
+      "loss": 0.8497,
+      "step": 117385
+    },
+    {
+      "epoch": 7.975947818997146,
+      "grad_norm": 9.125,
+      "learning_rate": 1.5032613126783533e-07,
+      "loss": 0.7956,
+      "step": 117390
+    },
+    {
+      "epoch": 7.976287539067808,
+      "grad_norm": 7.625,
+      "learning_rate": 1.4820288082619923e-07,
+      "loss": 0.8564,
+      "step": 117395
+    },
+    {
+      "epoch": 7.9766272591384695,
+      "grad_norm": 8.625,
+      "learning_rate": 1.4607963038456313e-07,
+      "loss": 0.8115,
+      "step": 117400
+    },
+    {
+      "epoch": 7.976966979209132,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.4395637994292702e-07,
+      "loss": 0.8629,
+      "step": 117405
+    },
+    {
+      "epoch": 7.977306699279794,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.4183312950129095e-07,
+      "loss": 0.8376,
+      "step": 117410
+    },
+    {
+      "epoch": 7.977646419350455,
+      "grad_norm": 7.90625,
+      "learning_rate": 1.3970987905965485e-07,
+      "loss": 0.8319,
+      "step": 117415
+    },
+    {
+      "epoch": 7.977986139421117,
+      "grad_norm": 8.25,
+      "learning_rate": 1.3758662861801877e-07,
+      "loss": 0.8112,
+      "step": 117420
+    },
+    {
+      "epoch": 7.978325859491779,
+      "grad_norm": 8.4375,
+      "learning_rate": 1.3546337817638267e-07,
+      "loss": 0.9347,
+      "step": 117425
+    },
+    {
+      "epoch": 7.97866557956244,
+      "grad_norm": 6.96875,
+      "learning_rate": 1.3334012773474657e-07,
+      "loss": 0.8238,
+      "step": 117430
+    },
+    {
+      "epoch": 7.979005299633102,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.312168772931105e-07,
+      "loss": 0.8652,
+      "step": 117435
+    },
+    {
+      "epoch": 7.979345019703764,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.290936268514744e-07,
+      "loss": 0.8744,
+      "step": 117440
+    },
+    {
+      "epoch": 7.9796847397744255,
+      "grad_norm": 8.1875,
+      "learning_rate": 1.2697037640983829e-07,
+      "loss": 1.0284,
+      "step": 117445
+    },
+    {
+      "epoch": 7.980024459845088,
+      "grad_norm": 7.84375,
+      "learning_rate": 1.248471259682022e-07,
+      "loss": 0.8303,
+      "step": 117450
+    },
+    {
+      "epoch": 7.98036417991575,
+      "grad_norm": 8.25,
+      "learning_rate": 1.227238755265661e-07,
+      "loss": 0.8484,
+      "step": 117455
+    },
+    {
+      "epoch": 7.980703899986411,
+      "grad_norm": 8.0,
+      "learning_rate": 1.2060062508493003e-07,
+      "loss": 0.8659,
+      "step": 117460
+    },
+    {
+      "epoch": 7.981043620057073,
+      "grad_norm": 7.34375,
+      "learning_rate": 1.1847737464329392e-07,
+      "loss": 0.8094,
+      "step": 117465
+    },
+    {
+      "epoch": 7.981383340127735,
+      "grad_norm": 8.5,
+      "learning_rate": 1.1635412420165784e-07,
+      "loss": 0.8348,
+      "step": 117470
+    },
+    {
+      "epoch": 7.981723060198396,
+      "grad_norm": 8.8125,
+      "learning_rate": 1.1423087376002174e-07,
+      "loss": 0.8709,
+      "step": 117475
+    },
+    {
+      "epoch": 7.982062780269058,
+      "grad_norm": 8.75,
+      "learning_rate": 1.1210762331838565e-07,
+      "loss": 0.839,
+      "step": 117480
+    },
+    {
+      "epoch": 7.98240250033972,
+      "grad_norm": 8.0,
+      "learning_rate": 1.0998437287674958e-07,
+      "loss": 0.8549,
+      "step": 117485
+    },
+    {
+      "epoch": 7.9827422204103815,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0786112243511346e-07,
+      "loss": 0.9169,
+      "step": 117490
+    },
+    {
+      "epoch": 7.983081940481044,
+      "grad_norm": 7.96875,
+      "learning_rate": 1.0573787199347739e-07,
+      "loss": 0.7412,
+      "step": 117495
+    },
+    {
+      "epoch": 7.983421660551706,
+      "grad_norm": 8.875,
+      "learning_rate": 1.0361462155184128e-07,
+      "loss": 0.8865,
+      "step": 117500
+    },
+    {
+      "epoch": 7.983761380622367,
+      "grad_norm": 6.84375,
+      "learning_rate": 1.014913711102052e-07,
+      "loss": 0.7543,
+      "step": 117505
+    },
+    {
+      "epoch": 7.984101100693029,
+      "grad_norm": 8.25,
+      "learning_rate": 9.936812066856912e-08,
+      "loss": 0.8702,
+      "step": 117510
+    },
+    {
+      "epoch": 7.984440820763691,
+      "grad_norm": 8.125,
+      "learning_rate": 9.7244870226933e-08,
+      "loss": 0.7872,
+      "step": 117515
+    },
+    {
+      "epoch": 7.984780540834352,
+      "grad_norm": 7.71875,
+      "learning_rate": 9.512161978529693e-08,
+      "loss": 0.8626,
+      "step": 117520
+    },
+    {
+      "epoch": 7.985120260905014,
+      "grad_norm": 8.125,
+      "learning_rate": 9.299836934366083e-08,
+      "loss": 0.9052,
+      "step": 117525
+    },
+    {
+      "epoch": 7.985459980975676,
+      "grad_norm": 7.40625,
+      "learning_rate": 9.087511890202474e-08,
+      "loss": 0.7172,
+      "step": 117530
+    },
+    {
+      "epoch": 7.9857997010463375,
+      "grad_norm": 8.0,
+      "learning_rate": 8.875186846038865e-08,
+      "loss": 0.9501,
+      "step": 117535
+    },
+    {
+      "epoch": 7.986139421117,
+      "grad_norm": 6.9375,
+      "learning_rate": 8.662861801875255e-08,
+      "loss": 0.7314,
+      "step": 117540
+    },
+    {
+      "epoch": 7.986479141187662,
+      "grad_norm": 8.0625,
+      "learning_rate": 8.450536757711646e-08,
+      "loss": 0.8426,
+      "step": 117545
+    },
+    {
+      "epoch": 7.986818861258323,
+      "grad_norm": 8.3125,
+      "learning_rate": 8.238211713548037e-08,
+      "loss": 0.8679,
+      "step": 117550
+    },
+    {
+      "epoch": 7.987158581328985,
+      "grad_norm": 7.84375,
+      "learning_rate": 8.025886669384428e-08,
+      "loss": 0.8585,
+      "step": 117555
+    },
+    {
+      "epoch": 7.987498301399647,
+      "grad_norm": 8.625,
+      "learning_rate": 7.813561625220819e-08,
+      "loss": 0.8793,
+      "step": 117560
+    },
+    {
+      "epoch": 7.987838021470308,
+      "grad_norm": 7.5625,
+      "learning_rate": 7.601236581057209e-08,
+      "loss": 0.9857,
+      "step": 117565
+    },
+    {
+      "epoch": 7.98817774154097,
+      "grad_norm": 7.96875,
+      "learning_rate": 7.3889115368936e-08,
+      "loss": 0.8464,
+      "step": 117570
+    },
+    {
+      "epoch": 7.988517461611632,
+      "grad_norm": 8.8125,
+      "learning_rate": 7.17658649272999e-08,
+      "loss": 0.8017,
+      "step": 117575
+    },
+    {
+      "epoch": 7.9888571816822935,
+      "grad_norm": 8.125,
+      "learning_rate": 6.964261448566382e-08,
+      "loss": 0.8998,
+      "step": 117580
+    },
+    {
+      "epoch": 7.989196901752956,
+      "grad_norm": 8.0625,
+      "learning_rate": 6.751936404402772e-08,
+      "loss": 0.73,
+      "step": 117585
+    },
+    {
+      "epoch": 7.989536621823618,
+      "grad_norm": 7.84375,
+      "learning_rate": 6.539611360239163e-08,
+      "loss": 0.8284,
+      "step": 117590
+    },
+    {
+      "epoch": 7.989876341894279,
+      "grad_norm": 7.40625,
+      "learning_rate": 6.327286316075554e-08,
+      "loss": 0.9402,
+      "step": 117595
+    },
+    {
+      "epoch": 7.990216061964941,
+      "grad_norm": 7.25,
+      "learning_rate": 6.114961271911944e-08,
+      "loss": 0.8238,
+      "step": 117600
+    },
+    {
+      "epoch": 7.990555782035603,
+      "grad_norm": 8.4375,
+      "learning_rate": 5.902636227748335e-08,
+      "loss": 0.9472,
+      "step": 117605
+    },
+    {
+      "epoch": 7.990895502106264,
+      "grad_norm": 8.375,
+      "learning_rate": 5.690311183584727e-08,
+      "loss": 0.7952,
+      "step": 117610
+    },
+    {
+      "epoch": 7.991235222176926,
+      "grad_norm": 6.90625,
+      "learning_rate": 5.4779861394211175e-08,
+      "loss": 0.8304,
+      "step": 117615
+    },
+    {
+      "epoch": 7.991574942247588,
+      "grad_norm": 8.25,
+      "learning_rate": 5.265661095257508e-08,
+      "loss": 0.8798,
+      "step": 117620
+    },
+    {
+      "epoch": 7.9919146623182495,
+      "grad_norm": 8.0625,
+      "learning_rate": 5.0533360510938985e-08,
+      "loss": 0.8511,
+      "step": 117625
+    },
+    {
+      "epoch": 7.992254382388912,
+      "grad_norm": 8.5,
+      "learning_rate": 4.8410110069302896e-08,
+      "loss": 0.879,
+      "step": 117630
+    },
+    {
+      "epoch": 7.992594102459574,
+      "grad_norm": 7.9375,
+      "learning_rate": 4.628685962766681e-08,
+      "loss": 0.813,
+      "step": 117635
+    },
+    {
+      "epoch": 7.992933822530235,
+      "grad_norm": 7.25,
+      "learning_rate": 4.416360918603071e-08,
+      "loss": 0.8164,
+      "step": 117640
+    },
+    {
+      "epoch": 7.993273542600897,
+      "grad_norm": 8.1875,
+      "learning_rate": 4.204035874439462e-08,
+      "loss": 0.7967,
+      "step": 117645
+    },
+    {
+      "epoch": 7.993613262671559,
+      "grad_norm": 7.78125,
+      "learning_rate": 3.991710830275853e-08,
+      "loss": 0.9422,
+      "step": 117650
+    },
+    {
+      "epoch": 7.99395298274222,
+      "grad_norm": 7.90625,
+      "learning_rate": 3.779385786112244e-08,
+      "loss": 0.8696,
+      "step": 117655
+    },
+    {
+      "epoch": 7.994292702812882,
+      "grad_norm": 8.25,
+      "learning_rate": 3.567060741948635e-08,
+      "loss": 0.9345,
+      "step": 117660
+    },
+    {
+      "epoch": 7.994632422883544,
+      "grad_norm": 7.9375,
+      "learning_rate": 3.3547356977850254e-08,
+      "loss": 0.8566,
+      "step": 117665
+    },
+    {
+      "epoch": 7.9949721429542056,
+      "grad_norm": 8.9375,
+      "learning_rate": 3.142410653621416e-08,
+      "loss": 0.7653,
+      "step": 117670
+    },
+    {
+      "epoch": 7.995311863024868,
+      "grad_norm": 8.9375,
+      "learning_rate": 2.930085609457807e-08,
+      "loss": 0.8379,
+      "step": 117675
+    },
+    {
+      "epoch": 7.99565158309553,
+      "grad_norm": 7.75,
+      "learning_rate": 2.7177605652941978e-08,
+      "loss": 0.7154,
+      "step": 117680
+    },
+    {
+      "epoch": 7.995991303166191,
+      "grad_norm": 8.5625,
+      "learning_rate": 2.5054355211305883e-08,
+      "loss": 0.9074,
+      "step": 117685
+    },
+    {
+      "epoch": 7.996331023236853,
+      "grad_norm": 7.75,
+      "learning_rate": 2.293110476966979e-08,
+      "loss": 0.7705,
+      "step": 117690
+    },
+    {
+      "epoch": 7.996670743307515,
+      "grad_norm": 7.5625,
+      "learning_rate": 2.0807854328033702e-08,
+      "loss": 0.7935,
+      "step": 117695
+    },
+    {
+      "epoch": 7.997010463378176,
+      "grad_norm": 7.3125,
+      "learning_rate": 1.868460388639761e-08,
+      "loss": 0.8,
+      "step": 117700
+    },
+    {
+      "epoch": 7.997350183448838,
+      "grad_norm": 7.28125,
+      "learning_rate": 1.6561353444761518e-08,
+      "loss": 0.9078,
+      "step": 117705
+    },
+    {
+      "epoch": 7.9976899035195,
+      "grad_norm": 7.40625,
+      "learning_rate": 1.4438103003125426e-08,
+      "loss": 0.8068,
+      "step": 117710
+    },
+    {
+      "epoch": 7.998029623590162,
+      "grad_norm": 8.3125,
+      "learning_rate": 1.2314852561489333e-08,
+      "loss": 0.891,
+      "step": 117715
+    },
+    {
+      "epoch": 7.998369343660824,
+      "grad_norm": 8.5625,
+      "learning_rate": 1.0191602119853241e-08,
+      "loss": 0.9438,
+      "step": 117720
+    },
+    {
+      "epoch": 7.998709063731486,
+      "grad_norm": 7.15625,
+      "learning_rate": 8.06835167821715e-09,
+      "loss": 0.8443,
+      "step": 117725
+    },
+    {
+      "epoch": 7.999048783802147,
+      "grad_norm": 7.75,
+      "learning_rate": 5.945101236581058e-09,
+      "loss": 0.8795,
+      "step": 117730
+    },
+    {
+      "epoch": 7.999388503872809,
+      "grad_norm": 7.625,
+      "learning_rate": 3.821850794944965e-09,
+      "loss": 0.7991,
+      "step": 117735
+    },
+    {
+      "epoch": 7.999728223943471,
+      "grad_norm": 8.6875,
+      "learning_rate": 1.6986003533088736e-09,
+      "loss": 0.8081,
+      "step": 117740
+    },
+    {
+      "epoch": 8.0,
+      "eval_loss": 1.7691917419433594,
+      "eval_runtime": 146.918,
+      "eval_samples_per_second": 70.236,
+      "eval_steps_per_second": 8.78,
+      "step": 117744
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 117744,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 8,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5533960899744236e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}