{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1624,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006157635467980296,
      "grad_norm": 10.15827904187796,
      "learning_rate": 1.226993865030675e-07,
      "loss": 1.6708,
      "step": 1
    },
    {
      "epoch": 0.003078817733990148,
      "grad_norm": 9.836680655010724,
      "learning_rate": 6.134969325153375e-07,
      "loss": 1.6715,
      "step": 5
    },
    {
      "epoch": 0.006157635467980296,
      "grad_norm": 8.545561598564545,
      "learning_rate": 1.226993865030675e-06,
      "loss": 1.6778,
      "step": 10
    },
    {
      "epoch": 0.009236453201970444,
      "grad_norm": 9.646058196195503,
      "learning_rate": 1.8404907975460124e-06,
      "loss": 1.5816,
      "step": 15
    },
    {
      "epoch": 0.012315270935960592,
      "grad_norm": 4.817089836289127,
      "learning_rate": 2.45398773006135e-06,
      "loss": 1.4522,
      "step": 20
    },
    {
      "epoch": 0.01539408866995074,
      "grad_norm": 3.170144500777755,
      "learning_rate": 3.0674846625766875e-06,
      "loss": 1.3562,
      "step": 25
    },
    {
      "epoch": 0.01847290640394089,
      "grad_norm": 1.1628885549890924,
      "learning_rate": 3.680981595092025e-06,
      "loss": 1.2687,
      "step": 30
    },
    {
      "epoch": 0.021551724137931036,
      "grad_norm": 0.7406818243288179,
      "learning_rate": 4.294478527607362e-06,
      "loss": 1.2558,
      "step": 35
    },
    {
      "epoch": 0.024630541871921183,
      "grad_norm": 0.6610765970151185,
      "learning_rate": 4.9079754601227e-06,
      "loss": 1.2318,
      "step": 40
    },
    {
      "epoch": 0.02770935960591133,
      "grad_norm": 0.5774704706258553,
      "learning_rate": 5.521472392638038e-06,
      "loss": 1.1974,
      "step": 45
    },
    {
      "epoch": 0.03078817733990148,
      "grad_norm": 0.5415942560846584,
      "learning_rate": 6.134969325153375e-06,
      "loss": 1.1922,
      "step": 50
    },
    {
      "epoch": 0.033866995073891626,
      "grad_norm": 0.5785043523116281,
      "learning_rate": 6.748466257668712e-06,
      "loss": 1.1792,
      "step": 55
    },
    {
      "epoch": 0.03694581280788178,
      "grad_norm": 0.5449653883781076,
      "learning_rate": 7.36196319018405e-06,
      "loss": 1.1733,
      "step": 60
    },
    {
      "epoch": 0.04002463054187192,
      "grad_norm": 0.5581447044816499,
      "learning_rate": 7.975460122699386e-06,
      "loss": 1.1887,
      "step": 65
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 0.5897524414645722,
      "learning_rate": 8.588957055214725e-06,
      "loss": 1.1685,
      "step": 70
    },
    {
      "epoch": 0.046182266009852216,
      "grad_norm": 0.5128162223611896,
      "learning_rate": 9.202453987730062e-06,
      "loss": 1.1435,
      "step": 75
    },
    {
      "epoch": 0.04926108374384237,
      "grad_norm": 0.5226670673543996,
      "learning_rate": 9.8159509202454e-06,
      "loss": 1.149,
      "step": 80
    },
    {
      "epoch": 0.05233990147783251,
      "grad_norm": 0.5464161772608294,
      "learning_rate": 1.0429447852760737e-05,
      "loss": 1.1476,
      "step": 85
    },
    {
      "epoch": 0.05541871921182266,
      "grad_norm": 0.5692480829600762,
      "learning_rate": 1.1042944785276076e-05,
      "loss": 1.1499,
      "step": 90
    },
    {
      "epoch": 0.058497536945812806,
      "grad_norm": 0.5432059400034088,
      "learning_rate": 1.1656441717791411e-05,
      "loss": 1.1613,
      "step": 95
    },
    {
      "epoch": 0.06157635467980296,
      "grad_norm": 0.5933853887050368,
      "learning_rate": 1.226993865030675e-05,
      "loss": 1.1383,
      "step": 100
    },
    {
      "epoch": 0.06465517241379311,
      "grad_norm": 0.5507864899694862,
      "learning_rate": 1.2883435582822085e-05,
      "loss": 1.1355,
      "step": 105
    },
    {
      "epoch": 0.06773399014778325,
      "grad_norm": 0.5409568981579986,
      "learning_rate": 1.3496932515337424e-05,
      "loss": 1.1414,
      "step": 110
    },
    {
      "epoch": 0.0708128078817734,
      "grad_norm": 0.5407180471828471,
      "learning_rate": 1.4110429447852763e-05,
      "loss": 1.1289,
      "step": 115
    },
    {
      "epoch": 0.07389162561576355,
      "grad_norm": 0.541202964180019,
      "learning_rate": 1.47239263803681e-05,
      "loss": 1.1218,
      "step": 120
    },
    {
      "epoch": 0.0769704433497537,
      "grad_norm": 0.5790619443142361,
      "learning_rate": 1.5337423312883436e-05,
      "loss": 1.1512,
      "step": 125
    },
    {
      "epoch": 0.08004926108374384,
      "grad_norm": 0.5886251561722321,
      "learning_rate": 1.5950920245398772e-05,
      "loss": 1.1167,
      "step": 130
    },
    {
      "epoch": 0.08312807881773399,
      "grad_norm": 0.5122619657316273,
      "learning_rate": 1.656441717791411e-05,
      "loss": 1.1423,
      "step": 135
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 0.5766473083784516,
      "learning_rate": 1.717791411042945e-05,
      "loss": 1.1298,
      "step": 140
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 0.556913985692295,
      "learning_rate": 1.7791411042944788e-05,
      "loss": 1.142,
      "step": 145
    },
    {
      "epoch": 0.09236453201970443,
      "grad_norm": 0.6754392620031505,
      "learning_rate": 1.8404907975460123e-05,
      "loss": 1.1337,
      "step": 150
    },
    {
      "epoch": 0.09544334975369458,
      "grad_norm": 0.5862445828380352,
      "learning_rate": 1.9018404907975462e-05,
      "loss": 1.132,
      "step": 155
    },
    {
      "epoch": 0.09852216748768473,
      "grad_norm": 0.5222926216142348,
      "learning_rate": 1.96319018404908e-05,
      "loss": 1.1359,
      "step": 160
    },
    {
      "epoch": 0.10160098522167488,
      "grad_norm": 0.6084153172489544,
      "learning_rate": 1.999990752408443e-05,
      "loss": 1.1382,
      "step": 165
    },
    {
      "epoch": 0.10467980295566502,
      "grad_norm": 0.5684108737637437,
      "learning_rate": 1.9998867189676517e-05,
      "loss": 1.1294,
      "step": 170
    },
    {
      "epoch": 0.10775862068965517,
      "grad_norm": 0.599658091716828,
      "learning_rate": 1.999667104662347e-05,
      "loss": 1.1357,
      "step": 175
    },
    {
      "epoch": 0.11083743842364532,
      "grad_norm": 0.637384965295336,
      "learning_rate": 1.9993319348786157e-05,
      "loss": 1.128,
      "step": 180
    },
    {
      "epoch": 0.11391625615763547,
      "grad_norm": 0.5706087786798316,
      "learning_rate": 1.9988812483600597e-05,
      "loss": 1.1138,
      "step": 185
    },
    {
      "epoch": 0.11699507389162561,
      "grad_norm": 0.5445052453927818,
      "learning_rate": 1.9983150972033186e-05,
      "loss": 1.0998,
      "step": 190
    },
    {
      "epoch": 0.12007389162561577,
      "grad_norm": 0.5928632486296028,
      "learning_rate": 1.9976335468520452e-05,
      "loss": 1.1345,
      "step": 195
    },
    {
      "epoch": 0.12315270935960591,
      "grad_norm": 0.5760386582392598,
      "learning_rate": 1.9968366760893437e-05,
      "loss": 1.1115,
      "step": 200
    },
    {
      "epoch": 0.12623152709359606,
      "grad_norm": 0.5270455503864016,
      "learning_rate": 1.9959245770286602e-05,
      "loss": 1.1271,
      "step": 205
    },
    {
      "epoch": 0.12931034482758622,
      "grad_norm": 0.5547168333658223,
      "learning_rate": 1.994897355103136e-05,
      "loss": 1.1393,
      "step": 210
    },
    {
      "epoch": 0.13238916256157635,
      "grad_norm": 0.5437711049467764,
      "learning_rate": 1.9937551290534208e-05,
      "loss": 1.1195,
      "step": 215
    },
    {
      "epoch": 0.1354679802955665,
      "grad_norm": 0.5041055693164401,
      "learning_rate": 1.9924980309139455e-05,
      "loss": 1.1272,
      "step": 220
    },
    {
      "epoch": 0.13854679802955666,
      "grad_norm": 0.5712625065097253,
      "learning_rate": 1.9911262059976614e-05,
      "loss": 1.1272,
      "step": 225
    },
    {
      "epoch": 0.1416256157635468,
      "grad_norm": 0.5114048371993095,
      "learning_rate": 1.9896398128792413e-05,
      "loss": 1.1341,
      "step": 230
    },
    {
      "epoch": 0.14470443349753695,
      "grad_norm": 0.5447566972074451,
      "learning_rate": 1.988039023376751e-05,
      "loss": 1.1251,
      "step": 235
    },
    {
      "epoch": 0.1477832512315271,
      "grad_norm": 0.5516499385195212,
      "learning_rate": 1.9863240225317868e-05,
      "loss": 1.1192,
      "step": 240
    },
    {
      "epoch": 0.15086206896551724,
      "grad_norm": 0.5597379666800071,
      "learning_rate": 1.984495008588086e-05,
      "loss": 1.1035,
      "step": 245
    },
    {
      "epoch": 0.1539408866995074,
      "grad_norm": 0.5358755463566468,
      "learning_rate": 1.982552192968612e-05,
      "loss": 1.1246,
      "step": 250
    },
    {
      "epoch": 0.15701970443349753,
      "grad_norm": 0.5552440543720575,
      "learning_rate": 1.9804958002511137e-05,
      "loss": 1.1329,
      "step": 255
    },
    {
      "epoch": 0.16009852216748768,
      "grad_norm": 0.5266336633257606,
      "learning_rate": 1.9783260681421667e-05,
      "loss": 1.0995,
      "step": 260
    },
    {
      "epoch": 0.16317733990147784,
      "grad_norm": 0.6190481120522302,
      "learning_rate": 1.9760432474496963e-05,
      "loss": 1.1203,
      "step": 265
    },
    {
      "epoch": 0.16625615763546797,
      "grad_norm": 0.6248703775091049,
      "learning_rate": 1.973647602053984e-05,
      "loss": 1.1224,
      "step": 270
    },
    {
      "epoch": 0.16933497536945813,
      "grad_norm": 0.5954299813641303,
      "learning_rate": 1.9711394088771658e-05,
      "loss": 1.106,
      "step": 275
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.5814897186652193,
      "learning_rate": 1.9685189578512206e-05,
      "loss": 1.1413,
      "step": 280
    },
    {
      "epoch": 0.17549261083743842,
      "grad_norm": 0.548224404648143,
      "learning_rate": 1.9657865518844578e-05,
      "loss": 1.1142,
      "step": 285
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.5526465274463118,
      "learning_rate": 1.962942506826501e-05,
      "loss": 1.1349,
      "step": 290
    },
    {
      "epoch": 0.1816502463054187,
      "grad_norm": 0.5507478495046352,
      "learning_rate": 1.9599871514317785e-05,
      "loss": 1.1024,
      "step": 295
    },
    {
      "epoch": 0.18472906403940886,
      "grad_norm": 0.5776341547415401,
      "learning_rate": 1.9569208273215204e-05,
      "loss": 1.1013,
      "step": 300
    },
    {
      "epoch": 0.18780788177339902,
      "grad_norm": 0.5914932792155793,
      "learning_rate": 1.953743888944271e-05,
      "loss": 1.1301,
      "step": 305
    },
    {
      "epoch": 0.19088669950738915,
      "grad_norm": 0.5532499476201675,
      "learning_rate": 1.950456703534915e-05,
      "loss": 1.0812,
      "step": 310
    },
    {
      "epoch": 0.1939655172413793,
      "grad_norm": 0.5378473200516649,
      "learning_rate": 1.9470596510722285e-05,
      "loss": 1.1072,
      "step": 315
    },
    {
      "epoch": 0.19704433497536947,
      "grad_norm": 0.5519689940819711,
      "learning_rate": 1.9435531242349545e-05,
      "loss": 1.1053,
      "step": 320
    },
    {
      "epoch": 0.2001231527093596,
      "grad_norm": 0.5276267926731764,
      "learning_rate": 1.9399375283564134e-05,
      "loss": 1.0888,
      "step": 325
    },
    {
      "epoch": 0.20320197044334976,
      "grad_norm": 0.5145556463022614,
      "learning_rate": 1.9362132813776472e-05,
      "loss": 1.0827,
      "step": 330
    },
    {
      "epoch": 0.2062807881773399,
      "grad_norm": 0.5708372641303956,
      "learning_rate": 1.9323808137991084e-05,
      "loss": 1.0944,
      "step": 335
    },
    {
      "epoch": 0.20935960591133004,
      "grad_norm": 0.5071088304983457,
      "learning_rate": 1.9284405686308982e-05,
      "loss": 1.0952,
      "step": 340
    },
    {
      "epoch": 0.2124384236453202,
      "grad_norm": 0.5381405634909971,
      "learning_rate": 1.924393001341555e-05,
      "loss": 1.1104,
      "step": 345
    },
    {
      "epoch": 0.21551724137931033,
      "grad_norm": 0.5289668507563129,
      "learning_rate": 1.9202385798054073e-05,
      "loss": 1.1086,
      "step": 350
    },
    {
      "epoch": 0.2185960591133005,
      "grad_norm": 0.5252703677098302,
      "learning_rate": 1.9159777842484878e-05,
      "loss": 1.1139,
      "step": 355
    },
    {
      "epoch": 0.22167487684729065,
      "grad_norm": 0.5310098852913822,
      "learning_rate": 1.911611107193024e-05,
      "loss": 1.0829,
      "step": 360
    },
    {
      "epoch": 0.22475369458128078,
      "grad_norm": 0.5126197931064982,
      "learning_rate": 1.9071390534005045e-05,
      "loss": 1.1168,
      "step": 365
    },
    {
      "epoch": 0.22783251231527094,
      "grad_norm": 0.566277348732687,
      "learning_rate": 1.9025621398133333e-05,
      "loss": 1.1034,
      "step": 370
    },
    {
      "epoch": 0.2309113300492611,
      "grad_norm": 0.5094820573732124,
      "learning_rate": 1.8978808954950722e-05,
      "loss": 1.1241,
      "step": 375
    },
    {
      "epoch": 0.23399014778325122,
      "grad_norm": 0.5197839922702459,
      "learning_rate": 1.8930958615692854e-05,
      "loss": 1.0805,
      "step": 380
    },
    {
      "epoch": 0.23706896551724138,
      "grad_norm": 0.518930427633083,
      "learning_rate": 1.8882075911569887e-05,
      "loss": 1.0994,
      "step": 385
    },
    {
      "epoch": 0.24014778325123154,
      "grad_norm": 0.514713587779145,
      "learning_rate": 1.8832166493127128e-05,
      "loss": 1.0932,
      "step": 390
    },
    {
      "epoch": 0.24322660098522167,
      "grad_norm": 0.5232409223690152,
      "learning_rate": 1.8781236129591847e-05,
      "loss": 1.1109,
      "step": 395
    },
    {
      "epoch": 0.24630541871921183,
      "grad_norm": 0.5208904781635201,
      "learning_rate": 1.8729290708206412e-05,
      "loss": 1.0873,
      "step": 400
    },
    {
      "epoch": 0.24938423645320196,
      "grad_norm": 0.5371698340804182,
      "learning_rate": 1.867633623354773e-05,
      "loss": 1.0759,
      "step": 405
    },
    {
      "epoch": 0.2524630541871921,
      "grad_norm": 0.5087807339234235,
      "learning_rate": 1.8622378826833186e-05,
      "loss": 1.0786,
      "step": 410
    },
    {
      "epoch": 0.2555418719211823,
      "grad_norm": 0.535385916088714,
      "learning_rate": 1.856742472521304e-05,
      "loss": 1.0998,
      "step": 415
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.5569084999187001,
      "learning_rate": 1.8511480281049475e-05,
      "loss": 1.0988,
      "step": 420
    },
    {
      "epoch": 0.2616995073891626,
      "grad_norm": 0.5187070981420698,
      "learning_rate": 1.8454551961182276e-05,
      "loss": 1.1246,
      "step": 425
    },
    {
      "epoch": 0.2647783251231527,
      "grad_norm": 0.5279420985835842,
      "learning_rate": 1.8396646346181327e-05,
      "loss": 1.1034,
      "step": 430
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 0.5452072365462802,
      "learning_rate": 1.8337770129585918e-05,
      "loss": 1.0763,
      "step": 435
    },
    {
      "epoch": 0.270935960591133,
      "grad_norm": 0.5389098407146274,
      "learning_rate": 1.8277930117131025e-05,
      "loss": 1.0998,
      "step": 440
    },
    {
      "epoch": 0.27401477832512317,
      "grad_norm": 0.5172070322682661,
      "learning_rate": 1.8217133225960597e-05,
      "loss": 1.0899,
      "step": 445
    },
    {
      "epoch": 0.2770935960591133,
      "grad_norm": 0.5375647663133752,
      "learning_rate": 1.8155386483827995e-05,
      "loss": 1.098,
      "step": 450
    },
    {
      "epoch": 0.2801724137931034,
      "grad_norm": 0.5547342780314599,
      "learning_rate": 1.8092697028283598e-05,
      "loss": 1.1252,
      "step": 455
    },
    {
      "epoch": 0.2832512315270936,
      "grad_norm": 0.5297060610488263,
      "learning_rate": 1.8029072105849767e-05,
      "loss": 1.0795,
      "step": 460
    },
    {
      "epoch": 0.28633004926108374,
      "grad_norm": 0.5527871147403416,
      "learning_rate": 1.7964519071183188e-05,
      "loss": 1.1054,
      "step": 465
    },
    {
      "epoch": 0.2894088669950739,
      "grad_norm": 0.512487448893779,
      "learning_rate": 1.789904538622471e-05,
      "loss": 1.1039,
      "step": 470
    },
    {
      "epoch": 0.29248768472906406,
      "grad_norm": 0.5058139154941391,
      "learning_rate": 1.7832658619336794e-05,
      "loss": 1.11,
      "step": 475
    },
    {
      "epoch": 0.2955665024630542,
      "grad_norm": 0.523771554049047,
      "learning_rate": 1.7765366444428655e-05,
      "loss": 1.0973,
      "step": 480
    },
    {
      "epoch": 0.2986453201970443,
      "grad_norm": 0.5183110774312877,
      "learning_rate": 1.7697176640069217e-05,
      "loss": 1.0942,
      "step": 485
    },
    {
      "epoch": 0.3017241379310345,
      "grad_norm": 0.5022270232188654,
      "learning_rate": 1.762809708858793e-05,
      "loss": 1.0691,
      "step": 490
    },
    {
      "epoch": 0.30480295566502463,
      "grad_norm": 0.5528980854788325,
      "learning_rate": 1.7558135775163645e-05,
      "loss": 1.0947,
      "step": 495
    },
    {
      "epoch": 0.3078817733990148,
      "grad_norm": 0.5756061513576279,
      "learning_rate": 1.7487300786901568e-05,
      "loss": 1.0652,
      "step": 500
    },
    {
      "epoch": 0.31096059113300495,
      "grad_norm": 0.5263901827150232,
      "learning_rate": 1.7415600311898436e-05,
      "loss": 1.0917,
      "step": 505
    },
    {
      "epoch": 0.31403940886699505,
      "grad_norm": 0.5229387164039296,
      "learning_rate": 1.734304263829602e-05,
      "loss": 1.0819,
      "step": 510
    },
    {
      "epoch": 0.3171182266009852,
      "grad_norm": 0.6147017579520017,
      "learning_rate": 1.726963615332308e-05,
      "loss": 1.1112,
      "step": 515
    },
    {
      "epoch": 0.32019704433497537,
      "grad_norm": 0.5800845773516132,
      "learning_rate": 1.7195389342325843e-05,
      "loss": 1.0896,
      "step": 520
    },
    {
      "epoch": 0.3232758620689655,
      "grad_norm": 0.5062578023133191,
      "learning_rate": 1.7120310787787136e-05,
      "loss": 1.0963,
      "step": 525
    },
    {
      "epoch": 0.3263546798029557,
      "grad_norm": 0.561913707064664,
      "learning_rate": 1.7044409168334327e-05,
      "loss": 1.0887,
      "step": 530
    },
    {
      "epoch": 0.3294334975369458,
      "grad_norm": 0.5584919081577276,
      "learning_rate": 1.696769325773611e-05,
      "loss": 1.0849,
      "step": 535
    },
    {
      "epoch": 0.33251231527093594,
      "grad_norm": 0.5119511999973996,
      "learning_rate": 1.6890171923888323e-05,
      "loss": 1.0885,
      "step": 540
    },
    {
      "epoch": 0.3355911330049261,
      "grad_norm": 0.54892611978747,
      "learning_rate": 1.6811854127788857e-05,
      "loss": 1.1236,
      "step": 545
    },
    {
      "epoch": 0.33866995073891626,
      "grad_norm": 0.4985099766995371,
      "learning_rate": 1.6732748922501832e-05,
      "loss": 1.0576,
      "step": 550
    },
    {
      "epoch": 0.3417487684729064,
      "grad_norm": 0.4999399844588124,
      "learning_rate": 1.6652865452111115e-05,
      "loss": 1.1033,
      "step": 555
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.5386909070925513,
      "learning_rate": 1.657221295066332e-05,
      "loss": 1.0999,
      "step": 560
    },
    {
      "epoch": 0.3479064039408867,
      "grad_norm": 0.5375529157895848,
      "learning_rate": 1.6490800741100396e-05,
      "loss": 1.1201,
      "step": 565
    },
    {
      "epoch": 0.35098522167487683,
      "grad_norm": 0.5163159007263345,
      "learning_rate": 1.6408638234181975e-05,
      "loss": 1.1109,
      "step": 570
    },
    {
      "epoch": 0.354064039408867,
      "grad_norm": 0.5264682427002149,
      "learning_rate": 1.6325734927397514e-05,
      "loss": 1.0974,
      "step": 575
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.5664009971025269,
      "learning_rate": 1.6242100403868472e-05,
      "loss": 1.0825,
      "step": 580
    },
    {
      "epoch": 0.3602216748768473,
      "grad_norm": 0.535100783370204,
      "learning_rate": 1.615774433124054e-05,
      "loss": 1.0903,
      "step": 585
    },
    {
      "epoch": 0.3633004926108374,
      "grad_norm": 0.5693571306430616,
      "learning_rate": 1.6072676460566136e-05,
      "loss": 1.0522,
      "step": 590
    },
    {
      "epoch": 0.36637931034482757,
      "grad_norm": 0.5316999528245905,
      "learning_rate": 1.5986906625177215e-05,
      "loss": 1.0885,
      "step": 595
    },
    {
      "epoch": 0.3694581280788177,
      "grad_norm": 0.5786475527569387,
      "learning_rate": 1.590044473954863e-05,
      "loss": 1.0848,
      "step": 600
    },
    {
      "epoch": 0.3725369458128079,
      "grad_norm": 0.5392315214147952,
      "learning_rate": 1.5813300798152048e-05,
      "loss": 1.0795,
      "step": 605
    },
    {
      "epoch": 0.37561576354679804,
      "grad_norm": 0.5229978056983617,
      "learning_rate": 1.5725484874300673e-05,
      "loss": 1.0786,
      "step": 610
    },
    {
      "epoch": 0.3786945812807882,
      "grad_norm": 0.5739140724063975,
      "learning_rate": 1.5637007118984814e-05,
      "loss": 1.071,
      "step": 615
    },
    {
      "epoch": 0.3817733990147783,
      "grad_norm": 0.5511410465011337,
      "learning_rate": 1.5547877759698498e-05,
      "loss": 1.082,
      "step": 620
    },
    {
      "epoch": 0.38485221674876846,
      "grad_norm": 0.5243605996854858,
      "learning_rate": 1.5458107099257245e-05,
      "loss": 1.0898,
      "step": 625
    },
    {
      "epoch": 0.3879310344827586,
      "grad_norm": 0.49837618242341025,
      "learning_rate": 1.5367705514607107e-05,
      "loss": 1.086,
      "step": 630
    },
    {
      "epoch": 0.3910098522167488,
      "grad_norm": 0.5238600479925725,
      "learning_rate": 1.527668345562516e-05,
      "loss": 1.0729,
      "step": 635
    },
    {
      "epoch": 0.39408866995073893,
      "grad_norm": 0.4885128198374361,
      "learning_rate": 1.518505144391157e-05,
      "loss": 1.0943,
      "step": 640
    },
    {
      "epoch": 0.39716748768472904,
      "grad_norm": 0.4938695351189867,
      "learning_rate": 1.5092820071573358e-05,
      "loss": 1.0859,
      "step": 645
    },
    {
      "epoch": 0.4002463054187192,
      "grad_norm": 0.5246450809840343,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 1.0907,
      "step": 650
    },
    {
      "epoch": 0.40332512315270935,
      "grad_norm": 0.5520074123860683,
      "learning_rate": 1.490660195863106e-05,
      "loss": 1.09,
      "step": 655
    },
    {
      "epoch": 0.4064039408866995,
      "grad_norm": 0.5114128859254611,
      "learning_rate": 1.4812636743715912e-05,
      "loss": 1.0845,
      "step": 660
    },
    {
      "epoch": 0.40948275862068967,
      "grad_norm": 0.5111372323638168,
      "learning_rate": 1.4718115217065766e-05,
      "loss": 1.0939,
      "step": 665
    },
    {
      "epoch": 0.4125615763546798,
      "grad_norm": 0.5130123039875292,
      "learning_rate": 1.462304830479811e-05,
      "loss": 1.0803,
      "step": 670
    },
    {
      "epoch": 0.41564039408866993,
      "grad_norm": 0.5611568580912872,
      "learning_rate": 1.4527446996073714e-05,
      "loss": 1.062,
      "step": 675
    },
    {
      "epoch": 0.4187192118226601,
      "grad_norm": 0.49715024362372484,
      "learning_rate": 1.4431322341826348e-05,
      "loss": 1.098,
      "step": 680
    },
    {
      "epoch": 0.42179802955665024,
      "grad_norm": 0.5301065034239282,
      "learning_rate": 1.433468545348537e-05,
      "loss": 1.0933,
      "step": 685
    },
    {
      "epoch": 0.4248768472906404,
      "grad_norm": 0.5275179530285631,
      "learning_rate": 1.4237547501691298e-05,
      "loss": 1.0736,
      "step": 690
    },
    {
      "epoch": 0.42795566502463056,
      "grad_norm": 0.5557717971076896,
      "learning_rate": 1.4139919715004558e-05,
      "loss": 1.0927,
      "step": 695
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.5115763802038508,
      "learning_rate": 1.4041813378607534e-05,
      "loss": 1.0573,
      "step": 700
    },
    {
      "epoch": 0.4341133004926108,
      "grad_norm": 0.5169280221898409,
      "learning_rate": 1.3943239833000068e-05,
      "loss": 1.0883,
      "step": 705
    },
    {
      "epoch": 0.437192118226601,
      "grad_norm": 0.499406239361373,
      "learning_rate": 1.3844210472688557e-05,
      "loss": 1.0849,
      "step": 710
    },
    {
      "epoch": 0.44027093596059114,
      "grad_norm": 0.519237136726592,
      "learning_rate": 1.3744736744868832e-05,
      "loss": 1.0745,
      "step": 715
    },
    {
      "epoch": 0.4433497536945813,
      "grad_norm": 0.5200708610774482,
      "learning_rate": 1.3644830148102915e-05,
      "loss": 1.0711,
      "step": 720
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 0.5364910261048887,
      "learning_rate": 1.3544502230989868e-05,
      "loss": 1.0845,
      "step": 725
    },
    {
      "epoch": 0.44950738916256155,
      "grad_norm": 0.491311081150838,
      "learning_rate": 1.344376459083085e-05,
      "loss": 1.0736,
      "step": 730
    },
    {
      "epoch": 0.4525862068965517,
      "grad_norm": 0.5421773226103814,
      "learning_rate": 1.3342628872288505e-05,
      "loss": 1.0703,
      "step": 735
    },
    {
      "epoch": 0.45566502463054187,
      "grad_norm": 0.5160402133686881,
      "learning_rate": 1.3241106766040957e-05,
      "loss": 1.0579,
      "step": 740
    },
    {
      "epoch": 0.45874384236453203,
      "grad_norm": 0.5345874554566498,
      "learning_rate": 1.3139210007430404e-05,
      "loss": 1.0896,
      "step": 745
    },
    {
      "epoch": 0.4618226600985222,
      "grad_norm": 0.5210582959671957,
      "learning_rate": 1.3036950375106588e-05,
      "loss": 1.0816,
      "step": 750
    },
    {
      "epoch": 0.4649014778325123,
      "grad_norm": 0.4900374391146739,
      "learning_rate": 1.2934339689665274e-05,
      "loss": 1.0917,
      "step": 755
    },
    {
      "epoch": 0.46798029556650245,
      "grad_norm": 0.5169473008105099,
      "learning_rate": 1.283138981228184e-05,
      "loss": 1.0927,
      "step": 760
    },
    {
      "epoch": 0.4710591133004926,
      "grad_norm": 0.5033069572412759,
      "learning_rate": 1.2728112643340201e-05,
      "loss": 1.0772,
      "step": 765
    },
    {
      "epoch": 0.47413793103448276,
      "grad_norm": 0.5192762810121773,
      "learning_rate": 1.26245201210572e-05,
      "loss": 1.1257,
      "step": 770
    },
    {
      "epoch": 0.4772167487684729,
      "grad_norm": 0.5270415802501245,
      "learning_rate": 1.2520624220102623e-05,
      "loss": 1.0644,
      "step": 775
    },
    {
      "epoch": 0.4802955665024631,
      "grad_norm": 0.49964820118171305,
      "learning_rate": 1.2416436950215001e-05,
      "loss": 1.0641,
      "step": 780
    },
    {
      "epoch": 0.4833743842364532,
      "grad_norm": 0.5200518670733051,
      "learning_rate": 1.2311970354813345e-05,
      "loss": 1.076,
      "step": 785
    },
    {
      "epoch": 0.48645320197044334,
      "grad_norm": 0.5103460892777054,
      "learning_rate": 1.220723650960502e-05,
      "loss": 1.0657,
      "step": 790
    },
    {
      "epoch": 0.4895320197044335,
      "grad_norm": 0.4923578044831893,
      "learning_rate": 1.2102247521189838e-05,
      "loss": 1.0729,
      "step": 795
    },
    {
      "epoch": 0.49261083743842365,
      "grad_norm": 0.5266682573681067,
      "learning_rate": 1.199701552566064e-05,
      "loss": 1.0928,
      "step": 800
    },
    {
      "epoch": 0.4956896551724138,
      "grad_norm": 0.5031792016986184,
      "learning_rate": 1.1891552687200414e-05,
      "loss": 1.0824,
      "step": 805
    },
    {
      "epoch": 0.4987684729064039,
      "grad_norm": 0.5018512002779485,
      "learning_rate": 1.1785871196676196e-05,
      "loss": 1.0781,
      "step": 810
    },
    {
      "epoch": 0.5018472906403941,
      "grad_norm": 0.5292904190231347,
      "learning_rate": 1.167998327022988e-05,
      "loss": 1.095,
      "step": 815
    },
    {
      "epoch": 0.5049261083743842,
      "grad_norm": 0.5176914040257699,
      "learning_rate": 1.1573901147866108e-05,
      "loss": 1.0786,
      "step": 820
    },
    {
      "epoch": 0.5080049261083743,
      "grad_norm": 0.501265325817293,
      "learning_rate": 1.1467637092037399e-05,
      "loss": 1.0869,
      "step": 825
    },
    {
      "epoch": 0.5110837438423645,
      "grad_norm": 0.5048470284561476,
      "learning_rate": 1.1361203386226672e-05,
      "loss": 1.0814,
      "step": 830
    },
    {
      "epoch": 0.5141625615763546,
      "grad_norm": 0.5213596621645071,
      "learning_rate": 1.1254612333527368e-05,
      "loss": 1.0484,
      "step": 835
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.48770105290422383,
      "learning_rate": 1.1147876255221274e-05,
      "loss": 1.0546,
      "step": 840
    },
    {
      "epoch": 0.520320197044335,
      "grad_norm": 0.4869259735699385,
      "learning_rate": 1.1041007489354263e-05,
      "loss": 1.0679,
      "step": 845
    },
    {
      "epoch": 0.5233990147783252,
      "grad_norm": 0.5377340968755271,
      "learning_rate": 1.093401838931009e-05,
      "loss": 1.0627,
      "step": 850
    },
    {
      "epoch": 0.5264778325123153,
      "grad_norm": 0.48688318184592444,
      "learning_rate": 1.0826921322382407e-05,
      "loss": 1.0537,
      "step": 855
    },
    {
      "epoch": 0.5295566502463054,
      "grad_norm": 0.5420284453273034,
      "learning_rate": 1.071972866834519e-05,
      "loss": 1.113,
      "step": 860
    },
    {
      "epoch": 0.5326354679802956,
      "grad_norm": 0.5285443463637869,
      "learning_rate": 1.061245281802171e-05,
      "loss": 1.0694,
      "step": 865
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.5317871768645912,
      "learning_rate": 1.0505106171852226e-05,
      "loss": 1.057,
      "step": 870
    },
    {
      "epoch": 0.5387931034482759,
      "grad_norm": 0.4718621057405186,
      "learning_rate": 1.039770113846056e-05,
      "loss": 1.058,
      "step": 875
    },
    {
      "epoch": 0.541871921182266,
      "grad_norm": 0.5653105369254648,
      "learning_rate": 1.0290250133219755e-05,
      "loss": 1.0521,
      "step": 880
    },
    {
      "epoch": 0.5449507389162561,
      "grad_norm": 0.5057497375728685,
      "learning_rate": 1.0182765576816916e-05,
      "loss": 1.0693,
      "step": 885
    },
    {
      "epoch": 0.5480295566502463,
      "grad_norm": 0.5232839196387389,
      "learning_rate": 1.0075259893817465e-05,
      "loss": 1.0585,
      "step": 890
    },
    {
      "epoch": 0.5511083743842364,
      "grad_norm": 0.5146311446760582,
      "learning_rate": 9.967745511228922e-06,
      "loss": 1.089,
      "step": 895
    },
    {
      "epoch": 0.5541871921182266,
      "grad_norm": 0.4902930712982599,
      "learning_rate": 9.860234857064439e-06,
      "loss": 1.0927,
      "step": 900
    },
    {
      "epoch": 0.5572660098522167,
      "grad_norm": 0.506242139362254,
      "learning_rate": 9.752740358906168e-06,
      "loss": 1.0604,
      "step": 905
    },
    {
      "epoch": 0.5603448275862069,
      "grad_norm": 0.48948810122142766,
      "learning_rate": 9.645274442468746e-06,
      "loss": 1.0643,
      "step": 910
    },
    {
      "epoch": 0.5634236453201971,
      "grad_norm": 0.5203810913782672,
      "learning_rate": 9.537849530162915e-06,
      "loss": 1.0738,
      "step": 915
    },
    {
      "epoch": 0.5665024630541872,
      "grad_norm": 0.5081111486694775,
      "learning_rate": 9.430478039659604e-06,
      "loss": 1.0363,
      "step": 920
    },
    {
      "epoch": 0.5695812807881774,
      "grad_norm": 0.5393882871495567,
      "learning_rate": 9.323172382454494e-06,
      "loss": 1.0845,
      "step": 925
    },
    {
      "epoch": 0.5726600985221675,
      "grad_norm": 0.5322827538578072,
      "learning_rate": 9.215944962433329e-06,
      "loss": 1.0886,
      "step": 930
    },
    {
      "epoch": 0.5757389162561576,
      "grad_norm": 0.5079277925791438,
      "learning_rate": 9.108808174438128e-06,
      "loss": 1.0823,
      "step": 935
    },
    {
      "epoch": 0.5788177339901478,
      "grad_norm": 0.5011485165889832,
      "learning_rate": 9.001774402834375e-06,
      "loss": 1.0684,
      "step": 940
    },
    {
      "epoch": 0.5818965517241379,
      "grad_norm": 0.5164184484743229,
      "learning_rate": 8.89485602007949e-06,
      "loss": 1.0796,
      "step": 945
    },
    {
      "epoch": 0.5849753694581281,
      "grad_norm": 0.5055296193556541,
      "learning_rate": 8.788065385292637e-06,
      "loss": 1.0593,
      "step": 950
    },
    {
      "epoch": 0.5880541871921182,
      "grad_norm": 0.5027164471297344,
      "learning_rate": 8.68141484282609e-06,
      "loss": 1.072,
      "step": 955
    },
    {
      "epoch": 0.5911330049261084,
      "grad_norm": 0.5078204866630002,
      "learning_rate": 8.574916720838293e-06,
      "loss": 1.0767,
      "step": 960
    },
    {
      "epoch": 0.5942118226600985,
      "grad_norm": 0.4825739853072064,
      "learning_rate": 8.468583329868805e-06,
      "loss": 1.068,
      "step": 965
    },
    {
      "epoch": 0.5972906403940886,
      "grad_norm": 0.48832498765314664,
      "learning_rate": 8.362426961415274e-06,
      "loss": 1.0904,
      "step": 970
    },
    {
      "epoch": 0.6003694581280788,
      "grad_norm": 0.5278557315324438,
      "learning_rate": 8.256459886512618e-06,
      "loss": 1.0822,
      "step": 975
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 0.5014309983629729,
      "learning_rate": 8.150694354314556e-06,
      "loss": 1.0642,
      "step": 980
    },
    {
      "epoch": 0.6065270935960592,
      "grad_norm": 0.5084688138427379,
      "learning_rate": 8.0451425906777e-06,
      "loss": 1.0659,
      "step": 985
    },
    {
      "epoch": 0.6096059113300493,
      "grad_norm": 0.5219658388606967,
      "learning_rate": 7.939816796748296e-06,
      "loss": 1.0466,
      "step": 990
    },
    {
      "epoch": 0.6126847290640394,
      "grad_norm": 0.5725745897010179,
      "learning_rate": 7.834729147551858e-06,
      "loss": 1.0951,
      "step": 995
    },
    {
      "epoch": 0.6157635467980296,
      "grad_norm": 0.49784928550084717,
      "learning_rate": 7.729891790585817e-06,
      "loss": 1.062,
      "step": 1000
    },
    {
      "epoch": 0.6188423645320197,
      "grad_norm": 0.5064974538353313,
      "learning_rate": 7.625316844415327e-06,
      "loss": 1.0526,
      "step": 1005
    },
    {
      "epoch": 0.6219211822660099,
      "grad_norm": 0.494143935074316,
      "learning_rate": 7.521016397272436e-06,
      "loss": 1.0488,
      "step": 1010
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.5091450180395377,
      "learning_rate": 7.417002505658773e-06,
      "loss": 1.0708,
      "step": 1015
    },
    {
      "epoch": 0.6280788177339901,
      "grad_norm": 0.5048438216104946,
      "learning_rate": 7.313287192951866e-06,
      "loss": 1.0624,
      "step": 1020
    },
    {
      "epoch": 0.6311576354679803,
      "grad_norm": 0.5289364157067876,
      "learning_rate": 7.209882448015338e-06,
      "loss": 1.0705,
      "step": 1025
    },
    {
      "epoch": 0.6342364532019704,
      "grad_norm": 0.5050920457484092,
      "learning_rate": 7.1068002238130465e-06,
      "loss": 1.0588,
      "step": 1030
    },
    {
      "epoch": 0.6373152709359606,
      "grad_norm": 0.4983375312941992,
      "learning_rate": 7.004052436027397e-06,
      "loss": 1.0646,
      "step": 1035
    },
    {
      "epoch": 0.6403940886699507,
      "grad_norm": 0.48765113672763694,
      "learning_rate": 6.901650961681976e-06,
      "loss": 1.067,
      "step": 1040
    },
    {
      "epoch": 0.6434729064039408,
      "grad_norm": 0.4859015897401179,
      "learning_rate": 6.799607637768621e-06,
      "loss": 1.0697,
      "step": 1045
    },
    {
      "epoch": 0.646551724137931,
      "grad_norm": 0.5042655482099269,
      "learning_rate": 6.6979342598791395e-06,
      "loss": 1.086,
      "step": 1050
    },
    {
      "epoch": 0.6496305418719212,
      "grad_norm": 0.5407184372280787,
      "learning_rate": 6.596642580841827e-06,
      "loss": 1.0672,
      "step": 1055
    },
    {
      "epoch": 0.6527093596059114,
      "grad_norm": 0.4974948079398,
      "learning_rate": 6.495744309362879e-06,
      "loss": 1.0975,
      "step": 1060
    },
    {
      "epoch": 0.6557881773399015,
      "grad_norm": 0.5131548369629815,
      "learning_rate": 6.395251108672975e-06,
      "loss": 1.061,
      "step": 1065
    },
    {
      "epoch": 0.6588669950738916,
      "grad_norm": 0.5172362417354719,
      "learning_rate": 6.295174595179041e-06,
      "loss": 1.0849,
      "step": 1070
    },
    {
      "epoch": 0.6619458128078818,
      "grad_norm": 0.5076165644021032,
      "learning_rate": 6.195526337121483e-06,
      "loss": 1.0703,
      "step": 1075
    },
    {
      "epoch": 0.6650246305418719,
      "grad_norm": 0.5385033591457516,
      "learning_rate": 6.096317853236975e-06,
      "loss": 1.079,
      "step": 1080
    },
    {
      "epoch": 0.6681034482758621,
      "grad_norm": 0.5018061084686454,
      "learning_rate": 5.997560611426947e-06,
      "loss": 1.083,
      "step": 1085
    },
    {
      "epoch": 0.6711822660098522,
      "grad_norm": 0.48764441502463635,
      "learning_rate": 5.899266027431965e-06,
      "loss": 1.0487,
      "step": 1090
    },
    {
      "epoch": 0.6742610837438424,
      "grad_norm": 0.5043811756950699,
      "learning_rate": 5.80144546351216e-06,
      "loss": 1.0795,
      "step": 1095
    },
    {
      "epoch": 0.6773399014778325,
      "grad_norm": 0.48269192048173676,
      "learning_rate": 5.704110227133792e-06,
      "loss": 1.0629,
      "step": 1100
    },
    {
      "epoch": 0.6804187192118226,
      "grad_norm": 0.4794572472797725,
      "learning_rate": 5.607271569662203e-06,
      "loss": 1.0405,
      "step": 1105
    },
    {
      "epoch": 0.6834975369458128,
      "grad_norm": 0.4887971073455054,
      "learning_rate": 5.510940685061202e-06,
      "loss": 1.0611,
      "step": 1110
    },
    {
      "epoch": 0.6865763546798029,
      "grad_norm": 0.5103625610202936,
      "learning_rate": 5.41512870859912e-06,
      "loss": 1.064,
      "step": 1115
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.5048114160747138,
      "learning_rate": 5.319846715561656e-06,
      "loss": 1.0751,
      "step": 1120
    },
    {
      "epoch": 0.6927339901477833,
      "grad_norm": 0.5044337541122542,
      "learning_rate": 5.225105719971615e-06,
      "loss": 1.0739,
      "step": 1125
    },
    {
      "epoch": 0.6958128078817734,
      "grad_norm": 0.5011252737864661,
      "learning_rate": 5.130916673315762e-06,
      "loss": 1.0738,
      "step": 1130
    },
    {
      "epoch": 0.6988916256157636,
      "grad_norm": 0.48301998981763894,
      "learning_rate": 5.037290463278914e-06,
      "loss": 1.048,
      "step": 1135
    },
    {
      "epoch": 0.7019704433497537,
      "grad_norm": 0.4962715345921204,
      "learning_rate": 4.94423791248536e-06,
      "loss": 1.0648,
      "step": 1140
    },
    {
      "epoch": 0.7050492610837439,
      "grad_norm": 0.46923999393856064,
      "learning_rate": 4.851769777247857e-06,
      "loss": 1.0689,
      "step": 1145
    },
    {
      "epoch": 0.708128078817734,
      "grad_norm": 0.4996406428997813,
      "learning_rate": 4.759896746324247e-06,
      "loss": 1.0702,
      "step": 1150
    },
    {
      "epoch": 0.7112068965517241,
      "grad_norm": 0.5089603766855889,
      "learning_rate": 4.668629439681907e-06,
      "loss": 1.0635,
      "step": 1155
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.5002183485428575,
      "learning_rate": 4.577978407270156e-06,
      "loss": 1.0658,
      "step": 1160
    },
    {
      "epoch": 0.7173645320197044,
      "grad_norm": 0.4966292403821584,
      "learning_rate": 4.487954127800726e-06,
      "loss": 1.0701,
      "step": 1165
    },
    {
      "epoch": 0.7204433497536946,
      "grad_norm": 0.4865973527785379,
      "learning_rate": 4.398567007536493e-06,
      "loss": 1.0494,
      "step": 1170
    },
    {
      "epoch": 0.7235221674876847,
      "grad_norm": 0.4904200048788576,
      "learning_rate": 4.309827379088589e-06,
      "loss": 1.0472,
      "step": 1175
    },
    {
      "epoch": 0.7266009852216748,
      "grad_norm": 0.5112211998495827,
      "learning_rate": 4.221745500221992e-06,
      "loss": 1.063,
      "step": 1180
    },
    {
      "epoch": 0.729679802955665,
      "grad_norm": 0.4850089638889374,
      "learning_rate": 4.134331552669812e-06,
      "loss": 1.068,
      "step": 1185
    },
    {
      "epoch": 0.7327586206896551,
      "grad_norm": 0.5006505236775846,
      "learning_rate": 4.047595640956326e-06,
      "loss": 1.0302,
      "step": 1190
    },
    {
      "epoch": 0.7358374384236454,
      "grad_norm": 0.49671559275280036,
      "learning_rate": 3.961547791228963e-06,
      "loss": 1.067,
      "step": 1195
    },
    {
      "epoch": 0.7389162561576355,
      "grad_norm": 0.49427829666952583,
      "learning_rate": 3.876197950099351e-06,
      "loss": 1.0569,
      "step": 1200
    },
    {
      "epoch": 0.7419950738916257,
      "grad_norm": 0.49703512477301603,
      "learning_rate": 3.7915559834935355e-06,
      "loss": 1.0459,
      "step": 1205
    },
    {
      "epoch": 0.7450738916256158,
      "grad_norm": 0.49152651248125223,
      "learning_rate": 3.7076316755115407e-06,
      "loss": 1.0789,
      "step": 1210
    },
    {
      "epoch": 0.7481527093596059,
      "grad_norm": 0.47060805224359553,
      "learning_rate": 3.6244347272963974e-06,
      "loss": 1.0427,
      "step": 1215
    },
    {
      "epoch": 0.7512315270935961,
      "grad_norm": 0.48993980538543624,
      "learning_rate": 3.5419747559127294e-06,
      "loss": 1.0318,
      "step": 1220
    },
    {
      "epoch": 0.7543103448275862,
      "grad_norm": 0.49390320950781325,
      "learning_rate": 3.4602612932351065e-06,
      "loss": 1.0692,
      "step": 1225
    },
    {
      "epoch": 0.7573891625615764,
      "grad_norm": 0.4833886103347225,
      "learning_rate": 3.3793037848461873e-06,
      "loss": 1.0629,
      "step": 1230
    },
    {
      "epoch": 0.7604679802955665,
      "grad_norm": 0.5214240559477047,
      "learning_rate": 3.2991115889448877e-06,
      "loss": 1.0813,
      "step": 1235
    },
    {
      "epoch": 0.7635467980295566,
      "grad_norm": 0.48023986177994704,
      "learning_rate": 3.2196939752646183e-06,
      "loss": 1.0695,
      "step": 1240
    },
    {
      "epoch": 0.7666256157635468,
      "grad_norm": 0.49010376895277485,
      "learning_rate": 3.141060124001776e-06,
      "loss": 1.0649,
      "step": 1245
    },
    {
      "epoch": 0.7697044334975369,
      "grad_norm": 0.5107873464672744,
      "learning_rate": 3.063219124754543e-06,
      "loss": 1.0663,
      "step": 1250
    },
    {
      "epoch": 0.7727832512315271,
      "grad_norm": 0.4810114620510743,
      "learning_rate": 2.9861799754722033e-06,
      "loss": 1.0592,
      "step": 1255
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.4828427013063488,
      "learning_rate": 2.9099515814150335e-06,
      "loss": 1.0723,
      "step": 1260
    },
    {
      "epoch": 0.7789408866995073,
      "grad_norm": 0.5015328858377405,
      "learning_rate": 2.8345427541248993e-06,
      "loss": 1.0614,
      "step": 1265
    },
    {
      "epoch": 0.7820197044334976,
      "grad_norm": 0.5246695338264865,
      "learning_rate": 2.7599622104066937e-06,
      "loss": 1.0693,
      "step": 1270
    },
    {
      "epoch": 0.7850985221674877,
      "grad_norm": 0.4907310416080836,
      "learning_rate": 2.6862185713207467e-06,
      "loss": 1.0588,
      "step": 1275
    },
    {
      "epoch": 0.7881773399014779,
      "grad_norm": 0.5119568107744994,
      "learning_rate": 2.6133203611862554e-06,
      "loss": 1.053,
      "step": 1280
    },
    {
      "epoch": 0.791256157635468,
      "grad_norm": 0.47620252808413777,
      "learning_rate": 2.5412760065959386e-06,
      "loss": 1.0232,
      "step": 1285
    },
    {
      "epoch": 0.7943349753694581,
      "grad_norm": 0.47617485357353967,
      "learning_rate": 2.4700938354419823e-06,
      "loss": 1.0428,
      "step": 1290
    },
    {
      "epoch": 0.7974137931034483,
      "grad_norm": 0.5075199748631211,
      "learning_rate": 2.3997820759533654e-06,
      "loss": 1.0682,
      "step": 1295
    },
    {
      "epoch": 0.8004926108374384,
      "grad_norm": 0.4923670540137769,
      "learning_rate": 2.3303488557447374e-06,
      "loss": 1.0588,
      "step": 1300
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 0.4994910294245348,
      "learning_rate": 2.26180220087692e-06,
      "loss": 1.0657,
      "step": 1305
    },
    {
      "epoch": 0.8066502463054187,
      "grad_norm": 0.49724472990867913,
      "learning_rate": 2.194150034929133e-06,
      "loss": 1.0681,
      "step": 1310
    },
    {
      "epoch": 0.8097290640394089,
      "grad_norm": 0.49138904263770483,
      "learning_rate": 2.1274001780830776e-06,
      "loss": 1.0747,
      "step": 1315
    },
    {
      "epoch": 0.812807881773399,
      "grad_norm": 0.6102881762579555,
      "learning_rate": 2.0615603462189824e-06,
      "loss": 1.0574,
      "step": 1320
    },
    {
      "epoch": 0.8158866995073891,
      "grad_norm": 0.5010292385210491,
      "learning_rate": 1.9966381500236786e-06,
      "loss": 1.0413,
      "step": 1325
    },
    {
      "epoch": 0.8189655172413793,
      "grad_norm": 0.5022580526198298,
      "learning_rate": 1.932641094110855e-06,
      "loss": 1.0637,
      "step": 1330
    },
    {
      "epoch": 0.8220443349753694,
      "grad_norm": 0.4903536984670942,
      "learning_rate": 1.869576576153581e-06,
      "loss": 1.0915,
      "step": 1335
    },
    {
      "epoch": 0.8251231527093597,
      "grad_norm": 0.48924489931400594,
      "learning_rate": 1.8074518860291646e-06,
      "loss": 1.0568,
      "step": 1340
    },
    {
      "epoch": 0.8282019704433498,
      "grad_norm": 0.49208553937624916,
      "learning_rate": 1.746274204976498e-06,
      "loss": 1.0617,
      "step": 1345
    },
    {
      "epoch": 0.8312807881773399,
      "grad_norm": 0.49254963958607983,
      "learning_rate": 1.6860506047659442e-06,
      "loss": 1.0627,
      "step": 1350
    },
    {
      "epoch": 0.8343596059113301,
      "grad_norm": 0.4943702661212073,
      "learning_rate": 1.6267880468818787e-06,
      "loss": 1.0456,
      "step": 1355
    },
    {
      "epoch": 0.8374384236453202,
      "grad_norm": 0.4874125542879152,
      "learning_rate": 1.5684933817180014e-06,
      "loss": 1.0412,
      "step": 1360
    },
    {
      "epoch": 0.8405172413793104,
      "grad_norm": 0.4785669188338194,
      "learning_rate": 1.5111733477854507e-06,
      "loss": 1.0823,
      "step": 1365
    },
    {
      "epoch": 0.8435960591133005,
      "grad_norm": 0.4968082416187376,
      "learning_rate": 1.454834570933884e-06,
      "loss": 1.0691,
      "step": 1370
    },
    {
      "epoch": 0.8466748768472906,
      "grad_norm": 0.46887456408480604,
      "learning_rate": 1.399483563585573e-06,
      "loss": 1.0436,
      "step": 1375
    },
    {
      "epoch": 0.8497536945812808,
      "grad_norm": 0.4964984955292583,
      "learning_rate": 1.345126723982594e-06,
      "loss": 1.057,
      "step": 1380
    },
    {
      "epoch": 0.8528325123152709,
      "grad_norm": 0.5048498926483475,
      "learning_rate": 1.2917703354472467e-06,
      "loss": 1.0449,
      "step": 1385
    },
    {
      "epoch": 0.8559113300492611,
      "grad_norm": 0.4826754512181378,
      "learning_rate": 1.2394205656557224e-06,
      "loss": 1.0634,
      "step": 1390
    },
    {
      "epoch": 0.8589901477832512,
      "grad_norm": 0.48716665120434605,
      "learning_rate": 1.1880834659251706e-06,
      "loss": 1.0541,
      "step": 1395
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.4844347315668472,
      "learning_rate": 1.1377649705142012e-06,
      "loss": 1.0467,
      "step": 1400
    },
    {
      "epoch": 0.8651477832512315,
      "grad_norm": 0.5019912436130637,
      "learning_rate": 1.0884708959369116e-06,
      "loss": 1.046,
      "step": 1405
    },
    {
      "epoch": 0.8682266009852216,
      "grad_norm": 0.49716156780639187,
      "learning_rate": 1.040206940290547e-06,
      "loss": 1.057,
      "step": 1410
    },
    {
      "epoch": 0.8713054187192119,
      "grad_norm": 0.4818007246635526,
      "learning_rate": 9.929786825968213e-07,
      "loss": 1.0511,
      "step": 1415
    },
    {
      "epoch": 0.874384236453202,
      "grad_norm": 0.48245117976079965,
      "learning_rate": 9.467915821570228e-07,
      "loss": 1.0499,
      "step": 1420
    },
    {
      "epoch": 0.8774630541871922,
      "grad_norm": 0.4874556324774752,
      "learning_rate": 9.016509779209536e-07,
      "loss": 1.0495,
      "step": 1425
    },
    {
      "epoch": 0.8805418719211823,
      "grad_norm": 0.5062089762290207,
      "learning_rate": 8.575620878697744e-07,
      "loss": 1.0424,
      "step": 1430
    },
    {
      "epoch": 0.8836206896551724,
      "grad_norm": 0.4667997030632563,
      "learning_rate": 8.145300084128349e-07,
      "loss": 1.0651,
      "step": 1435
    },
    {
      "epoch": 0.8866995073891626,
      "grad_norm": 0.4964295419069643,
      "learning_rate": 7.725597137985741e-07,
      "loss": 1.0693,
      "step": 1440
    },
    {
      "epoch": 0.8897783251231527,
      "grad_norm": 0.49026938173900414,
      "learning_rate": 7.316560555395069e-07,
      "loss": 1.0617,
      "step": 1445
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.5008458957189175,
      "learning_rate": 6.918237618514378e-07,
      "loss": 1.075,
      "step": 1450
    },
    {
      "epoch": 0.895935960591133,
      "grad_norm": 0.47471658876745654,
      "learning_rate": 6.530674371068946e-07,
      "loss": 1.0436,
      "step": 1455
    },
    {
      "epoch": 0.8990147783251231,
      "grad_norm": 0.49832331828043364,
      "learning_rate": 6.153915613028915e-07,
      "loss": 1.0704,
      "step": 1460
    },
    {
      "epoch": 0.9020935960591133,
      "grad_norm": 0.4790403381731389,
      "learning_rate": 5.788004895430799e-07,
      "loss": 1.0374,
      "step": 1465
    },
    {
      "epoch": 0.9051724137931034,
      "grad_norm": 0.47076541876123645,
      "learning_rate": 5.43298451534312e-07,
      "loss": 1.0486,
      "step": 1470
    },
    {
      "epoch": 0.9082512315270936,
      "grad_norm": 0.47433886715017126,
      "learning_rate": 5.088895510977154e-07,
      "loss": 1.0609,
      "step": 1475
    },
    {
      "epoch": 0.9113300492610837,
      "grad_norm": 0.4872914609613854,
      "learning_rate": 4.755777656943239e-07,
      "loss": 1.0364,
      "step": 1480
    },
    {
      "epoch": 0.9144088669950738,
      "grad_norm": 0.4623611911999158,
      "learning_rate": 4.433669459652945e-07,
      "loss": 1.0746,
      "step": 1485
    },
    {
      "epoch": 0.9174876847290641,
      "grad_norm": 0.4509604378525504,
      "learning_rate": 4.1226081528680907e-07,
      "loss": 1.056,
      "step": 1490
    },
    {
      "epoch": 0.9205665024630542,
      "grad_norm": 0.4878990399675149,
      "learning_rate": 3.822629693396651e-07,
      "loss": 1.0523,
      "step": 1495
    },
    {
      "epoch": 0.9236453201970444,
      "grad_norm": 0.48387941146785535,
      "learning_rate": 3.5337687569363734e-07,
      "loss": 1.0882,
      "step": 1500
    },
    {
      "epoch": 0.9267241379310345,
      "grad_norm": 0.48068629190435125,
      "learning_rate": 3.2560587340665694e-07,
      "loss": 1.0732,
      "step": 1505
    },
    {
      "epoch": 0.9298029556650246,
      "grad_norm": 0.5154081631468291,
      "learning_rate": 2.989531726388262e-07,
      "loss": 1.0754,
      "step": 1510
    },
    {
      "epoch": 0.9328817733990148,
      "grad_norm": 0.5155664008720566,
      "learning_rate": 2.7342185428134604e-07,
      "loss": 1.0897,
      "step": 1515
    },
    {
      "epoch": 0.9359605911330049,
      "grad_norm": 0.5096256360700303,
      "learning_rate": 2.4901486960039025e-07,
      "loss": 1.0623,
      "step": 1520
    },
    {
      "epoch": 0.9390394088669951,
      "grad_norm": 0.4741695526546047,
      "learning_rate": 2.257350398959457e-07,
      "loss": 1.0438,
      "step": 1525
    },
    {
      "epoch": 0.9421182266009852,
      "grad_norm": 0.47110987694441464,
      "learning_rate": 2.035850561756969e-07,
      "loss": 1.0583,
      "step": 1530
    },
    {
      "epoch": 0.9451970443349754,
      "grad_norm": 0.4679444204418595,
      "learning_rate": 1.8256747884395577e-07,
      "loss": 1.0425,
      "step": 1535
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 0.48869771114680993,
      "learning_rate": 1.6268473740569723e-07,
      "loss": 1.0653,
      "step": 1540
    },
    {
      "epoch": 0.9513546798029556,
      "grad_norm": 0.4815991097704461,
      "learning_rate": 1.4393913018572182e-07,
      "loss": 1.0802,
      "step": 1545
    },
    {
      "epoch": 0.9544334975369458,
      "grad_norm": 0.4912529780984026,
      "learning_rate": 1.2633282406298576e-07,
      "loss": 1.0606,
      "step": 1550
    },
    {
      "epoch": 0.9575123152709359,
      "grad_norm": 0.4796496570988276,
      "learning_rate": 1.0986785422011593e-07,
      "loss": 1.0636,
      "step": 1555
    },
    {
      "epoch": 0.9605911330049262,
      "grad_norm": 0.48236300816548217,
      "learning_rate": 9.454612390816686e-08,
      "loss": 1.0565,
      "step": 1560
    },
    {
      "epoch": 0.9636699507389163,
      "grad_norm": 0.4814920013170011,
      "learning_rate": 8.036940422660345e-08,
      "loss": 1.034,
      "step": 1565
    },
    {
      "epoch": 0.9667487684729064,
      "grad_norm": 0.4769716230526951,
      "learning_rate": 6.733933391858238e-08,
      "loss": 1.047,
      "step": 1570
    },
    {
      "epoch": 0.9698275862068966,
      "grad_norm": 0.47950879270639535,
      "learning_rate": 5.5457419181517145e-08,
      "loss": 1.0815,
      "step": 1575
    },
    {
      "epoch": 0.9729064039408867,
      "grad_norm": 0.4718983189159335,
      "learning_rate": 4.472503349297497e-08,
      "loss": 1.0689,
      "step": 1580
    },
    {
      "epoch": 0.9759852216748769,
      "grad_norm": 0.4824121053232075,
      "learning_rate": 3.5143417451907195e-08,
      "loss": 1.0797,
      "step": 1585
    },
    {
      "epoch": 0.979064039408867,
      "grad_norm": 0.4533449780416192,
      "learning_rate": 2.671367863524732e-08,
      "loss": 1.0614,
      "step": 1590
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 0.4900131839975392,
      "learning_rate": 1.943679146988009e-08,
      "loss": 1.0535,
      "step": 1595
    },
    {
      "epoch": 0.9852216748768473,
      "grad_norm": 0.4800810445856197,
      "learning_rate": 1.3313597120002686e-08,
      "loss": 1.0493,
      "step": 1600
    },
    {
      "epoch": 0.9883004926108374,
      "grad_norm": 0.47834371925070235,
      "learning_rate": 8.34480338989141e-09,
      "loss": 1.0445,
      "step": 1605
    },
    {
      "epoch": 0.9913793103448276,
      "grad_norm": 0.48573715183969157,
      "learning_rate": 4.530984642087122e-09,
      "loss": 1.0596,
      "step": 1610
    },
    {
      "epoch": 0.9944581280788177,
      "grad_norm": 0.4802714032741417,
      "learning_rate": 1.872581730997247e-09,
      "loss": 1.0529,
      "step": 1615
    },
    {
      "epoch": 0.9975369458128078,
      "grad_norm": 0.4980147880202054,
      "learning_rate": 3.699019519387559e-10,
      "loss": 1.0519,
      "step": 1620
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.0636999607086182,
      "eval_runtime": 411.5468,
      "eval_samples_per_second": 56.152,
      "eval_steps_per_second": 1.757,
      "step": 1624
    },
    {
      "epoch": 1.0,
      "step": 1624,
      "total_flos": 234525562306560.0,
      "train_loss": 1.092289737439508,
      "train_runtime": 13676.5553,
      "train_samples_per_second": 15.199,
      "train_steps_per_second": 0.119
    }
  ],
  "logging_steps": 5,
  "max_steps": 1624,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 234525562306560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}