{
  "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": 7.310926153142698,
      "learning_rate": 1.226993865030675e-07,
      "loss": 1.5491,
      "step": 1
    },
    {
      "epoch": 0.003078817733990148,
      "grad_norm": 7.35403837138417,
      "learning_rate": 6.134969325153375e-07,
      "loss": 1.5843,
      "step": 5
    },
    {
      "epoch": 0.006157635467980296,
      "grad_norm": 6.803167318550182,
      "learning_rate": 1.226993865030675e-06,
      "loss": 1.5309,
      "step": 10
    },
    {
      "epoch": 0.009236453201970444,
      "grad_norm": 5.835727118488878,
      "learning_rate": 1.8404907975460124e-06,
      "loss": 1.3594,
      "step": 15
    },
    {
      "epoch": 0.012315270935960592,
      "grad_norm": 3.698538754344085,
      "learning_rate": 2.45398773006135e-06,
      "loss": 1.2405,
      "step": 20
    },
    {
      "epoch": 0.01539408866995074,
      "grad_norm": 2.041108035069763,
      "learning_rate": 3.0674846625766875e-06,
      "loss": 1.1387,
      "step": 25
    },
    {
      "epoch": 0.01847290640394089,
      "grad_norm": 1.0814828345339402,
      "learning_rate": 3.680981595092025e-06,
      "loss": 1.0108,
      "step": 30
    },
    {
      "epoch": 0.021551724137931036,
      "grad_norm": 0.8687795931835679,
      "learning_rate": 4.294478527607362e-06,
      "loss": 0.9621,
      "step": 35
    },
    {
      "epoch": 0.024630541871921183,
      "grad_norm": 0.8664847350475609,
      "learning_rate": 4.9079754601227e-06,
      "loss": 1.0134,
      "step": 40
    },
    {
      "epoch": 0.02770935960591133,
      "grad_norm": 0.8801651222007403,
      "learning_rate": 5.521472392638038e-06,
      "loss": 0.9283,
      "step": 45
    },
    {
      "epoch": 0.03078817733990148,
      "grad_norm": 0.8636788681790382,
      "learning_rate": 6.134969325153375e-06,
      "loss": 0.9328,
      "step": 50
    },
    {
      "epoch": 0.033866995073891626,
      "grad_norm": 0.8253490914438462,
      "learning_rate": 6.748466257668712e-06,
      "loss": 0.9051,
      "step": 55
    },
    {
      "epoch": 0.03694581280788178,
      "grad_norm": 0.8184367283533114,
      "learning_rate": 7.36196319018405e-06,
      "loss": 1.0101,
      "step": 60
    },
    {
      "epoch": 0.04002463054187192,
      "grad_norm": 0.8465714119368565,
      "learning_rate": 7.975460122699386e-06,
      "loss": 0.9229,
      "step": 65
    },
    {
      "epoch": 0.04310344827586207,
      "grad_norm": 0.8285519257452998,
      "learning_rate": 8.588957055214725e-06,
      "loss": 0.9491,
      "step": 70
    },
    {
      "epoch": 0.046182266009852216,
      "grad_norm": 0.769823593105267,
      "learning_rate": 9.202453987730062e-06,
      "loss": 0.9726,
      "step": 75
    },
    {
      "epoch": 0.04926108374384237,
      "grad_norm": 0.8064693479221889,
      "learning_rate": 9.8159509202454e-06,
      "loss": 0.9667,
      "step": 80
    },
    {
      "epoch": 0.05233990147783251,
      "grad_norm": 1.0363791467221333,
      "learning_rate": 1.0429447852760737e-05,
      "loss": 0.9209,
      "step": 85
    },
    {
      "epoch": 0.05541871921182266,
      "grad_norm": 0.7842863481909221,
      "learning_rate": 1.1042944785276076e-05,
      "loss": 0.9398,
      "step": 90
    },
    {
      "epoch": 0.058497536945812806,
      "grad_norm": 0.7367025450271165,
      "learning_rate": 1.1656441717791411e-05,
      "loss": 0.9461,
      "step": 95
    },
    {
      "epoch": 0.06157635467980296,
      "grad_norm": 0.7689269543001077,
      "learning_rate": 1.226993865030675e-05,
      "loss": 0.9177,
      "step": 100
    },
    {
      "epoch": 0.06465517241379311,
      "grad_norm": 0.8274857270444617,
      "learning_rate": 1.2883435582822085e-05,
      "loss": 0.9525,
      "step": 105
    },
    {
      "epoch": 0.06773399014778325,
      "grad_norm": 0.7618255279710626,
      "learning_rate": 1.3496932515337424e-05,
      "loss": 0.9264,
      "step": 110
    },
    {
      "epoch": 0.0708128078817734,
      "grad_norm": 0.7980402273299848,
      "learning_rate": 1.4110429447852763e-05,
      "loss": 0.9339,
      "step": 115
    },
    {
      "epoch": 0.07389162561576355,
      "grad_norm": 0.8376644789397331,
      "learning_rate": 1.47239263803681e-05,
      "loss": 0.8859,
      "step": 120
    },
    {
      "epoch": 0.0769704433497537,
      "grad_norm": 0.8130407079299852,
      "learning_rate": 1.5337423312883436e-05,
      "loss": 0.9146,
      "step": 125
    },
    {
      "epoch": 0.08004926108374384,
      "grad_norm": 0.7956012283141837,
      "learning_rate": 1.5950920245398772e-05,
      "loss": 0.9154,
      "step": 130
    },
    {
      "epoch": 0.08312807881773399,
      "grad_norm": 0.7948481430201038,
      "learning_rate": 1.656441717791411e-05,
      "loss": 0.9576,
      "step": 135
    },
    {
      "epoch": 0.08620689655172414,
      "grad_norm": 0.8257741416380704,
      "learning_rate": 1.717791411042945e-05,
      "loss": 0.9515,
      "step": 140
    },
    {
      "epoch": 0.08928571428571429,
      "grad_norm": 0.8290236573203579,
      "learning_rate": 1.7791411042944788e-05,
      "loss": 0.9678,
      "step": 145
    },
    {
      "epoch": 0.09236453201970443,
      "grad_norm": 0.9224233612252317,
      "learning_rate": 1.8404907975460123e-05,
      "loss": 0.9644,
      "step": 150
    },
    {
      "epoch": 0.09544334975369458,
      "grad_norm": 0.8138389103383028,
      "learning_rate": 1.9018404907975462e-05,
      "loss": 0.9028,
      "step": 155
    },
    {
      "epoch": 0.09852216748768473,
      "grad_norm": 0.8149251514397249,
      "learning_rate": 1.96319018404908e-05,
      "loss": 0.9381,
      "step": 160
    },
    {
      "epoch": 0.10160098522167488,
      "grad_norm": 0.857572912046182,
      "learning_rate": 1.999990752408443e-05,
      "loss": 0.9526,
      "step": 165
    },
    {
      "epoch": 0.10467980295566502,
      "grad_norm": 0.8424271480084681,
      "learning_rate": 1.9998867189676517e-05,
      "loss": 0.9598,
      "step": 170
    },
    {
      "epoch": 0.10775862068965517,
      "grad_norm": 0.7367129010600535,
      "learning_rate": 1.999667104662347e-05,
      "loss": 0.9543,
      "step": 175
    },
    {
      "epoch": 0.11083743842364532,
      "grad_norm": 0.7601180219708648,
      "learning_rate": 1.9993319348786157e-05,
      "loss": 0.9672,
      "step": 180
    },
    {
      "epoch": 0.11391625615763547,
      "grad_norm": 0.796186238392561,
      "learning_rate": 1.9988812483600597e-05,
      "loss": 0.8957,
      "step": 185
    },
    {
      "epoch": 0.11699507389162561,
      "grad_norm": 0.7747918983305362,
      "learning_rate": 1.9983150972033186e-05,
      "loss": 0.963,
      "step": 190
    },
    {
      "epoch": 0.12007389162561577,
      "grad_norm": 0.8300480182123051,
      "learning_rate": 1.9976335468520452e-05,
      "loss": 0.9824,
      "step": 195
    },
    {
      "epoch": 0.12315270935960591,
      "grad_norm": 0.8310354348747966,
      "learning_rate": 1.9968366760893437e-05,
      "loss": 0.9692,
      "step": 200
    },
    {
      "epoch": 0.12623152709359606,
      "grad_norm": 0.8445303140068814,
      "learning_rate": 1.9959245770286602e-05,
      "loss": 0.9546,
      "step": 205
    },
    {
      "epoch": 0.12931034482758622,
      "grad_norm": 0.8979903129953134,
      "learning_rate": 1.994897355103136e-05,
      "loss": 0.9375,
      "step": 210
    },
    {
      "epoch": 0.13238916256157635,
      "grad_norm": 0.7465524065375957,
      "learning_rate": 1.9937551290534208e-05,
      "loss": 0.9057,
      "step": 215
    },
    {
      "epoch": 0.1354679802955665,
      "grad_norm": 0.7497704866674313,
      "learning_rate": 1.9924980309139455e-05,
      "loss": 0.9434,
      "step": 220
    },
    {
      "epoch": 0.13854679802955666,
      "grad_norm": 0.7076359322846222,
      "learning_rate": 1.9911262059976614e-05,
      "loss": 0.9462,
      "step": 225
    },
    {
      "epoch": 0.1416256157635468,
      "grad_norm": 0.7848088777070403,
      "learning_rate": 1.9896398128792413e-05,
      "loss": 0.9476,
      "step": 230
    },
    {
      "epoch": 0.14470443349753695,
      "grad_norm": 0.8165026350741112,
      "learning_rate": 1.988039023376751e-05,
      "loss": 0.9527,
      "step": 235
    },
    {
      "epoch": 0.1477832512315271,
      "grad_norm": 0.8498025037832465,
      "learning_rate": 1.9863240225317868e-05,
      "loss": 1.0053,
      "step": 240
    },
    {
      "epoch": 0.15086206896551724,
      "grad_norm": 0.7353554935083384,
      "learning_rate": 1.984495008588086e-05,
      "loss": 0.9222,
      "step": 245
    },
    {
      "epoch": 0.1539408866995074,
      "grad_norm": 0.7553279060128398,
      "learning_rate": 1.982552192968612e-05,
      "loss": 0.9732,
      "step": 250
    },
    {
      "epoch": 0.15701970443349753,
      "grad_norm": 0.7548683527271752,
      "learning_rate": 1.9804958002511137e-05,
      "loss": 0.9459,
      "step": 255
    },
    {
      "epoch": 0.16009852216748768,
      "grad_norm": 0.7794156820504012,
      "learning_rate": 1.9783260681421667e-05,
      "loss": 0.9573,
      "step": 260
    },
    {
      "epoch": 0.16317733990147784,
      "grad_norm": 0.7422263502441188,
      "learning_rate": 1.9760432474496963e-05,
      "loss": 0.9375,
      "step": 265
    },
    {
      "epoch": 0.16625615763546797,
      "grad_norm": 0.7259884159865896,
      "learning_rate": 1.973647602053984e-05,
      "loss": 0.9483,
      "step": 270
    },
    {
      "epoch": 0.16933497536945813,
      "grad_norm": 0.7801247475320573,
      "learning_rate": 1.9711394088771658e-05,
      "loss": 0.9663,
      "step": 275
    },
    {
      "epoch": 0.1724137931034483,
      "grad_norm": 0.8468086332749512,
      "learning_rate": 1.9685189578512206e-05,
      "loss": 0.9581,
      "step": 280
    },
    {
      "epoch": 0.17549261083743842,
      "grad_norm": 0.760222004142156,
      "learning_rate": 1.9657865518844578e-05,
      "loss": 0.9405,
      "step": 285
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.818552353689237,
      "learning_rate": 1.962942506826501e-05,
      "loss": 0.9436,
      "step": 290
    },
    {
      "epoch": 0.1816502463054187,
      "grad_norm": 0.7270444014459365,
      "learning_rate": 1.9599871514317785e-05,
      "loss": 0.9587,
      "step": 295
    },
    {
      "epoch": 0.18472906403940886,
      "grad_norm": 0.7569082732767817,
      "learning_rate": 1.9569208273215204e-05,
      "loss": 0.9092,
      "step": 300
    },
    {
      "epoch": 0.18780788177339902,
      "grad_norm": 0.8035781870618336,
      "learning_rate": 1.953743888944271e-05,
      "loss": 0.9207,
      "step": 305
    },
    {
      "epoch": 0.19088669950738915,
      "grad_norm": 0.798197452115362,
      "learning_rate": 1.950456703534915e-05,
      "loss": 0.9722,
      "step": 310
    },
    {
      "epoch": 0.1939655172413793,
      "grad_norm": 0.7911708031305178,
      "learning_rate": 1.9470596510722285e-05,
      "loss": 0.9088,
      "step": 315
    },
    {
      "epoch": 0.19704433497536947,
      "grad_norm": 0.7671171753975288,
      "learning_rate": 1.9435531242349545e-05,
      "loss": 1.0116,
      "step": 320
    },
    {
      "epoch": 0.2001231527093596,
      "grad_norm": 0.7471976542017915,
      "learning_rate": 1.9399375283564134e-05,
      "loss": 0.9443,
      "step": 325
    },
    {
      "epoch": 0.20320197044334976,
      "grad_norm": 0.7932187536453967,
      "learning_rate": 1.9362132813776472e-05,
      "loss": 0.937,
      "step": 330
    },
    {
      "epoch": 0.2062807881773399,
      "grad_norm": 0.7683161798892411,
      "learning_rate": 1.9323808137991084e-05,
      "loss": 0.9485,
      "step": 335
    },
    {
      "epoch": 0.20935960591133004,
      "grad_norm": 0.7661725970320424,
      "learning_rate": 1.9284405686308982e-05,
      "loss": 0.9133,
      "step": 340
    },
    {
      "epoch": 0.2124384236453202,
      "grad_norm": 0.7873862980673396,
      "learning_rate": 1.924393001341555e-05,
      "loss": 0.9207,
      "step": 345
    },
    {
      "epoch": 0.21551724137931033,
      "grad_norm": 0.7452478426256695,
      "learning_rate": 1.9202385798054073e-05,
      "loss": 0.9087,
      "step": 350
    },
    {
      "epoch": 0.2185960591133005,
      "grad_norm": 0.7793073146063183,
      "learning_rate": 1.9159777842484878e-05,
      "loss": 0.9345,
      "step": 355
    },
    {
      "epoch": 0.22167487684729065,
      "grad_norm": 0.8169813616813546,
      "learning_rate": 1.911611107193024e-05,
      "loss": 0.9382,
      "step": 360
    },
    {
      "epoch": 0.22475369458128078,
      "grad_norm": 0.7266251337678646,
      "learning_rate": 1.9071390534005045e-05,
      "loss": 0.9732,
      "step": 365
    },
    {
      "epoch": 0.22783251231527094,
      "grad_norm": 0.7477667607003808,
      "learning_rate": 1.9025621398133333e-05,
      "loss": 0.9518,
      "step": 370
    },
    {
      "epoch": 0.2309113300492611,
      "grad_norm": 0.9165568422039077,
      "learning_rate": 1.8978808954950722e-05,
      "loss": 0.9503,
      "step": 375
    },
    {
      "epoch": 0.23399014778325122,
      "grad_norm": 0.7529486101747109,
      "learning_rate": 1.8930958615692854e-05,
      "loss": 0.9591,
      "step": 380
    },
    {
      "epoch": 0.23706896551724138,
      "grad_norm": 0.6888920479042022,
      "learning_rate": 1.8882075911569887e-05,
      "loss": 0.88,
      "step": 385
    },
    {
      "epoch": 0.24014778325123154,
      "grad_norm": 0.7255523898786278,
      "learning_rate": 1.8832166493127128e-05,
      "loss": 0.933,
      "step": 390
    },
    {
      "epoch": 0.24322660098522167,
      "grad_norm": 0.7128259336980561,
      "learning_rate": 1.8781236129591847e-05,
      "loss": 0.9078,
      "step": 395
    },
    {
      "epoch": 0.24630541871921183,
      "grad_norm": 0.7070843217854903,
      "learning_rate": 1.8729290708206412e-05,
      "loss": 0.9636,
      "step": 400
    },
    {
      "epoch": 0.24938423645320196,
      "grad_norm": 0.718867215532029,
      "learning_rate": 1.867633623354773e-05,
      "loss": 0.9556,
      "step": 405
    },
    {
      "epoch": 0.2524630541871921,
      "grad_norm": 0.7805465337550137,
      "learning_rate": 1.8622378826833186e-05,
      "loss": 0.9685,
      "step": 410
    },
    {
      "epoch": 0.2555418719211823,
      "grad_norm": 0.7719203711056516,
      "learning_rate": 1.856742472521304e-05,
      "loss": 0.9524,
      "step": 415
    },
    {
      "epoch": 0.25862068965517243,
      "grad_norm": 0.7192834205111203,
      "learning_rate": 1.8511480281049475e-05,
      "loss": 0.9246,
      "step": 420
    },
    {
      "epoch": 0.2616995073891626,
      "grad_norm": 0.6723358874994304,
      "learning_rate": 1.8454551961182276e-05,
      "loss": 0.9173,
      "step": 425
    },
    {
      "epoch": 0.2647783251231527,
      "grad_norm": 0.6817001490385899,
      "learning_rate": 1.8396646346181327e-05,
      "loss": 0.9389,
      "step": 430
    },
    {
      "epoch": 0.26785714285714285,
      "grad_norm": 0.7207425588978127,
      "learning_rate": 1.8337770129585918e-05,
      "loss": 0.9731,
      "step": 435
    },
    {
      "epoch": 0.270935960591133,
      "grad_norm": 0.6937128223592749,
      "learning_rate": 1.8277930117131025e-05,
      "loss": 0.9666,
      "step": 440
    },
    {
      "epoch": 0.27401477832512317,
      "grad_norm": 0.7206019687649675,
      "learning_rate": 1.8217133225960597e-05,
      "loss": 0.9047,
      "step": 445
    },
    {
      "epoch": 0.2770935960591133,
      "grad_norm": 0.6589164154450292,
      "learning_rate": 1.8155386483827995e-05,
      "loss": 0.9323,
      "step": 450
    },
    {
      "epoch": 0.2801724137931034,
      "grad_norm": 0.7477503173033859,
      "learning_rate": 1.8092697028283598e-05,
      "loss": 0.9434,
      "step": 455
    },
    {
      "epoch": 0.2832512315270936,
      "grad_norm": 0.6946681549404662,
      "learning_rate": 1.8029072105849767e-05,
      "loss": 0.9097,
      "step": 460
    },
    {
      "epoch": 0.28633004926108374,
      "grad_norm": 0.7993317655881385,
      "learning_rate": 1.7964519071183188e-05,
      "loss": 0.9484,
      "step": 465
    },
    {
      "epoch": 0.2894088669950739,
      "grad_norm": 0.7061937279236051,
      "learning_rate": 1.789904538622471e-05,
      "loss": 0.9456,
      "step": 470
    },
    {
      "epoch": 0.29248768472906406,
      "grad_norm": 0.7072130152863852,
      "learning_rate": 1.7832658619336794e-05,
      "loss": 0.9524,
      "step": 475
    },
    {
      "epoch": 0.2955665024630542,
      "grad_norm": 0.714294689243862,
      "learning_rate": 1.7765366444428655e-05,
      "loss": 0.9547,
      "step": 480
    },
    {
      "epoch": 0.2986453201970443,
      "grad_norm": 0.7609301009818032,
      "learning_rate": 1.7697176640069217e-05,
      "loss": 0.9146,
      "step": 485
    },
    {
      "epoch": 0.3017241379310345,
      "grad_norm": 0.7537368897585695,
      "learning_rate": 1.762809708858793e-05,
      "loss": 0.9451,
      "step": 490
    },
    {
      "epoch": 0.30480295566502463,
      "grad_norm": 0.7477622773826048,
      "learning_rate": 1.7558135775163645e-05,
      "loss": 0.9489,
      "step": 495
    },
    {
      "epoch": 0.3078817733990148,
      "grad_norm": 0.8035379494578148,
      "learning_rate": 1.7487300786901568e-05,
      "loss": 0.887,
      "step": 500
    },
    {
      "epoch": 0.31096059113300495,
      "grad_norm": 0.8085018346753247,
      "learning_rate": 1.7415600311898436e-05,
      "loss": 0.9788,
      "step": 505
    },
    {
      "epoch": 0.31403940886699505,
      "grad_norm": 0.7199766001172444,
      "learning_rate": 1.734304263829602e-05,
      "loss": 0.9482,
      "step": 510
    },
    {
      "epoch": 0.3171182266009852,
      "grad_norm": 0.7155949547205896,
      "learning_rate": 1.726963615332308e-05,
      "loss": 0.9638,
      "step": 515
    },
    {
      "epoch": 0.32019704433497537,
      "grad_norm": 0.8023351717912234,
      "learning_rate": 1.7195389342325843e-05,
      "loss": 0.9418,
      "step": 520
    },
    {
      "epoch": 0.3232758620689655,
      "grad_norm": 0.6707946282641698,
      "learning_rate": 1.7120310787787136e-05,
      "loss": 0.9051,
      "step": 525
    },
    {
      "epoch": 0.3263546798029557,
      "grad_norm": 0.7628307879605529,
      "learning_rate": 1.7044409168334327e-05,
      "loss": 0.9301,
      "step": 530
    },
    {
      "epoch": 0.3294334975369458,
      "grad_norm": 0.7018750333060788,
      "learning_rate": 1.696769325773611e-05,
      "loss": 0.9203,
      "step": 535
    },
    {
      "epoch": 0.33251231527093594,
      "grad_norm": 0.8985483854038159,
      "learning_rate": 1.6890171923888323e-05,
      "loss": 0.942,
      "step": 540
    },
    {
      "epoch": 0.3355911330049261,
      "grad_norm": 0.7095099515570173,
      "learning_rate": 1.6811854127788857e-05,
      "loss": 0.9883,
      "step": 545
    },
    {
      "epoch": 0.33866995073891626,
      "grad_norm": 0.663301100449732,
      "learning_rate": 1.6732748922501832e-05,
      "loss": 0.9958,
      "step": 550
    },
    {
      "epoch": 0.3417487684729064,
      "grad_norm": 0.7007829684906022,
      "learning_rate": 1.6652865452111115e-05,
      "loss": 0.8977,
      "step": 555
    },
    {
      "epoch": 0.3448275862068966,
      "grad_norm": 0.7794993793101852,
      "learning_rate": 1.657221295066332e-05,
      "loss": 0.963,
      "step": 560
    },
    {
      "epoch": 0.3479064039408867,
      "grad_norm": 0.6817973499177562,
      "learning_rate": 1.6490800741100396e-05,
      "loss": 0.9448,
      "step": 565
    },
    {
      "epoch": 0.35098522167487683,
      "grad_norm": 0.7464691369559269,
      "learning_rate": 1.6408638234181975e-05,
      "loss": 0.9295,
      "step": 570
    },
    {
      "epoch": 0.354064039408867,
      "grad_norm": 0.6661854545324757,
      "learning_rate": 1.6325734927397514e-05,
      "loss": 0.943,
      "step": 575
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.6704430428924031,
      "learning_rate": 1.6242100403868472e-05,
      "loss": 0.9223,
      "step": 580
    },
    {
      "epoch": 0.3602216748768473,
      "grad_norm": 0.7450322676169298,
      "learning_rate": 1.615774433124054e-05,
      "loss": 0.9296,
      "step": 585
    },
    {
      "epoch": 0.3633004926108374,
      "grad_norm": 0.785984380526607,
      "learning_rate": 1.6072676460566136e-05,
      "loss": 0.952,
      "step": 590
    },
    {
      "epoch": 0.36637931034482757,
      "grad_norm": 0.7156872420189895,
      "learning_rate": 1.5986906625177215e-05,
      "loss": 0.9382,
      "step": 595
    },
    {
      "epoch": 0.3694581280788177,
      "grad_norm": 0.7208351590256996,
      "learning_rate": 1.590044473954863e-05,
      "loss": 0.9951,
      "step": 600
    },
    {
      "epoch": 0.3725369458128079,
      "grad_norm": 0.6980642220237895,
      "learning_rate": 1.5813300798152048e-05,
      "loss": 0.9297,
      "step": 605
    },
    {
      "epoch": 0.37561576354679804,
      "grad_norm": 0.6658042870926646,
      "learning_rate": 1.5725484874300673e-05,
      "loss": 0.9697,
      "step": 610
    },
    {
      "epoch": 0.3786945812807882,
      "grad_norm": 0.6989438524636783,
      "learning_rate": 1.5637007118984814e-05,
      "loss": 0.9356,
      "step": 615
    },
    {
      "epoch": 0.3817733990147783,
      "grad_norm": 0.6554764404296011,
      "learning_rate": 1.5547877759698498e-05,
      "loss": 0.9367,
      "step": 620
    },
    {
      "epoch": 0.38485221674876846,
      "grad_norm": 0.7212087304435114,
      "learning_rate": 1.5458107099257245e-05,
      "loss": 0.9277,
      "step": 625
    },
    {
      "epoch": 0.3879310344827586,
      "grad_norm": 0.7411621278427354,
      "learning_rate": 1.5367705514607107e-05,
      "loss": 0.9152,
      "step": 630
    },
    {
      "epoch": 0.3910098522167488,
      "grad_norm": 0.7201507693917704,
      "learning_rate": 1.527668345562516e-05,
      "loss": 0.9169,
      "step": 635
    },
    {
      "epoch": 0.39408866995073893,
      "grad_norm": 0.7894484158745393,
      "learning_rate": 1.518505144391157e-05,
      "loss": 0.946,
      "step": 640
    },
    {
      "epoch": 0.39716748768472904,
      "grad_norm": 0.7132866192747803,
      "learning_rate": 1.5092820071573358e-05,
      "loss": 0.9612,
      "step": 645
    },
    {
      "epoch": 0.4002463054187192,
      "grad_norm": 0.8036672482141785,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.936,
      "step": 650
    },
    {
      "epoch": 0.40332512315270935,
      "grad_norm": 0.730609032181912,
      "learning_rate": 1.490660195863106e-05,
      "loss": 0.9155,
      "step": 655
    },
    {
      "epoch": 0.4064039408866995,
      "grad_norm": 0.7875912602755151,
      "learning_rate": 1.4812636743715912e-05,
      "loss": 0.9289,
      "step": 660
    },
    {
      "epoch": 0.40948275862068967,
      "grad_norm": 0.7504791630648989,
      "learning_rate": 1.4718115217065766e-05,
      "loss": 0.9457,
      "step": 665
    },
    {
      "epoch": 0.4125615763546798,
      "grad_norm": 0.7348305998616946,
      "learning_rate": 1.462304830479811e-05,
      "loss": 0.9072,
      "step": 670
    },
    {
      "epoch": 0.41564039408866993,
      "grad_norm": 0.758969021564288,
      "learning_rate": 1.4527446996073714e-05,
      "loss": 0.9177,
      "step": 675
    },
    {
      "epoch": 0.4187192118226601,
      "grad_norm": 0.7012954160981419,
      "learning_rate": 1.4431322341826348e-05,
      "loss": 0.9025,
      "step": 680
    },
    {
      "epoch": 0.42179802955665024,
      "grad_norm": 0.8675884737141334,
      "learning_rate": 1.433468545348537e-05,
      "loss": 0.9515,
      "step": 685
    },
    {
      "epoch": 0.4248768472906404,
      "grad_norm": 0.6907521835103683,
      "learning_rate": 1.4237547501691298e-05,
      "loss": 0.9369,
      "step": 690
    },
    {
      "epoch": 0.42795566502463056,
      "grad_norm": 0.8760691875396653,
      "learning_rate": 1.4139919715004558e-05,
      "loss": 0.9336,
      "step": 695
    },
    {
      "epoch": 0.43103448275862066,
      "grad_norm": 0.7605520563419292,
      "learning_rate": 1.4041813378607534e-05,
      "loss": 0.9459,
      "step": 700
    },
    {
      "epoch": 0.4341133004926108,
      "grad_norm": 0.7482466819256248,
      "learning_rate": 1.3943239833000068e-05,
      "loss": 0.9281,
      "step": 705
    },
    {
      "epoch": 0.437192118226601,
      "grad_norm": 0.6853008905243693,
      "learning_rate": 1.3844210472688557e-05,
      "loss": 0.9115,
      "step": 710
    },
    {
      "epoch": 0.44027093596059114,
      "grad_norm": 0.6785408874646308,
      "learning_rate": 1.3744736744868832e-05,
      "loss": 0.9147,
      "step": 715
    },
    {
      "epoch": 0.4433497536945813,
      "grad_norm": 0.7667132134232257,
      "learning_rate": 1.3644830148102915e-05,
      "loss": 0.9454,
      "step": 720
    },
    {
      "epoch": 0.44642857142857145,
      "grad_norm": 0.7343585908583702,
      "learning_rate": 1.3544502230989868e-05,
      "loss": 0.927,
      "step": 725
    },
    {
      "epoch": 0.44950738916256155,
      "grad_norm": 0.7135467850082023,
      "learning_rate": 1.344376459083085e-05,
      "loss": 0.9248,
      "step": 730
    },
    {
      "epoch": 0.4525862068965517,
      "grad_norm": 0.7291722090004472,
      "learning_rate": 1.3342628872288505e-05,
      "loss": 0.9401,
      "step": 735
    },
    {
      "epoch": 0.45566502463054187,
      "grad_norm": 0.7200530793723301,
      "learning_rate": 1.3241106766040957e-05,
      "loss": 0.9097,
      "step": 740
    },
    {
      "epoch": 0.45874384236453203,
      "grad_norm": 0.9856265254605108,
      "learning_rate": 1.3139210007430404e-05,
      "loss": 0.9283,
      "step": 745
    },
    {
      "epoch": 0.4618226600985222,
      "grad_norm": 0.7341322171237868,
      "learning_rate": 1.3036950375106588e-05,
      "loss": 0.9333,
      "step": 750
    },
    {
      "epoch": 0.4649014778325123,
      "grad_norm": 0.7546656555280624,
      "learning_rate": 1.2934339689665274e-05,
      "loss": 0.9042,
      "step": 755
    },
    {
      "epoch": 0.46798029556650245,
      "grad_norm": 0.7226811365920998,
      "learning_rate": 1.283138981228184e-05,
      "loss": 0.9139,
      "step": 760
    },
    {
      "epoch": 0.4710591133004926,
      "grad_norm": 0.7181811040603533,
      "learning_rate": 1.2728112643340201e-05,
      "loss": 0.9264,
      "step": 765
    },
    {
      "epoch": 0.47413793103448276,
      "grad_norm": 0.8849529024867829,
      "learning_rate": 1.26245201210572e-05,
      "loss": 0.9177,
      "step": 770
    },
    {
      "epoch": 0.4772167487684729,
      "grad_norm": 0.7087551590617035,
      "learning_rate": 1.2520624220102623e-05,
      "loss": 0.921,
      "step": 775
    },
    {
      "epoch": 0.4802955665024631,
      "grad_norm": 0.6935685016158297,
      "learning_rate": 1.2416436950215001e-05,
      "loss": 0.9406,
      "step": 780
    },
    {
      "epoch": 0.4833743842364532,
      "grad_norm": 0.7473725653174663,
      "learning_rate": 1.2311970354813345e-05,
      "loss": 0.9873,
      "step": 785
    },
    {
      "epoch": 0.48645320197044334,
      "grad_norm": 0.657674984254088,
      "learning_rate": 1.220723650960502e-05,
      "loss": 0.927,
      "step": 790
    },
    {
      "epoch": 0.4895320197044335,
      "grad_norm": 0.7186327207648633,
      "learning_rate": 1.2102247521189838e-05,
      "loss": 0.9004,
      "step": 795
    },
    {
      "epoch": 0.49261083743842365,
      "grad_norm": 0.6823249047599709,
      "learning_rate": 1.199701552566064e-05,
      "loss": 0.8879,
      "step": 800
    },
    {
      "epoch": 0.4956896551724138,
      "grad_norm": 0.705573871553639,
      "learning_rate": 1.1891552687200414e-05,
      "loss": 0.9649,
      "step": 805
    },
    {
      "epoch": 0.4987684729064039,
      "grad_norm": 0.7071856320020758,
      "learning_rate": 1.1785871196676196e-05,
      "loss": 0.9731,
      "step": 810
    },
    {
      "epoch": 0.5018472906403941,
      "grad_norm": 0.6605795316984785,
      "learning_rate": 1.167998327022988e-05,
      "loss": 0.9767,
      "step": 815
    },
    {
      "epoch": 0.5049261083743842,
      "grad_norm": 0.689384383606719,
      "learning_rate": 1.1573901147866108e-05,
      "loss": 0.929,
      "step": 820
    },
    {
      "epoch": 0.5080049261083743,
      "grad_norm": 0.7016418038725151,
      "learning_rate": 1.1467637092037399e-05,
      "loss": 0.9208,
      "step": 825
    },
    {
      "epoch": 0.5110837438423645,
      "grad_norm": 0.6435950510096468,
      "learning_rate": 1.1361203386226672e-05,
      "loss": 0.933,
      "step": 830
    },
    {
      "epoch": 0.5141625615763546,
      "grad_norm": 0.6915860359426802,
      "learning_rate": 1.1254612333527368e-05,
      "loss": 0.904,
      "step": 835
    },
    {
      "epoch": 0.5172413793103449,
      "grad_norm": 0.7278487014487579,
      "learning_rate": 1.1147876255221274e-05,
      "loss": 0.926,
      "step": 840
    },
    {
      "epoch": 0.520320197044335,
      "grad_norm": 0.8402556179449276,
      "learning_rate": 1.1041007489354263e-05,
      "loss": 0.91,
      "step": 845
    },
    {
      "epoch": 0.5233990147783252,
      "grad_norm": 0.6913958208278679,
      "learning_rate": 1.093401838931009e-05,
      "loss": 0.9298,
      "step": 850
    },
    {
      "epoch": 0.5264778325123153,
      "grad_norm": 0.6942000092673813,
      "learning_rate": 1.0826921322382407e-05,
      "loss": 0.9254,
      "step": 855
    },
    {
      "epoch": 0.5295566502463054,
      "grad_norm": 0.6863019990229392,
      "learning_rate": 1.071972866834519e-05,
      "loss": 0.8875,
      "step": 860
    },
    {
      "epoch": 0.5326354679802956,
      "grad_norm": 0.7444864272324827,
      "learning_rate": 1.061245281802171e-05,
      "loss": 0.9348,
      "step": 865
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.7472300249765285,
      "learning_rate": 1.0505106171852226e-05,
      "loss": 0.8728,
      "step": 870
    },
    {
      "epoch": 0.5387931034482759,
      "grad_norm": 0.7421508787127711,
      "learning_rate": 1.039770113846056e-05,
      "loss": 0.897,
      "step": 875
    },
    {
      "epoch": 0.541871921182266,
      "grad_norm": 0.6931950229252386,
      "learning_rate": 1.0290250133219755e-05,
      "loss": 0.9215,
      "step": 880
    },
    {
      "epoch": 0.5449507389162561,
      "grad_norm": 0.7648333450579011,
      "learning_rate": 1.0182765576816916e-05,
      "loss": 0.9577,
      "step": 885
    },
    {
      "epoch": 0.5480295566502463,
      "grad_norm": 0.7006731215648692,
      "learning_rate": 1.0075259893817465e-05,
      "loss": 0.9245,
      "step": 890
    },
    {
      "epoch": 0.5511083743842364,
      "grad_norm": 0.6687572466831397,
      "learning_rate": 9.967745511228922e-06,
      "loss": 0.8813,
      "step": 895
    },
    {
      "epoch": 0.5541871921182266,
      "grad_norm": 0.7175380682948211,
      "learning_rate": 9.860234857064439e-06,
      "loss": 0.9319,
      "step": 900
    },
    {
      "epoch": 0.5572660098522167,
      "grad_norm": 0.6948450711755244,
      "learning_rate": 9.752740358906168e-06,
      "loss": 0.9341,
      "step": 905
    },
    {
      "epoch": 0.5603448275862069,
      "grad_norm": 0.6892353302488635,
      "learning_rate": 9.645274442468746e-06,
      "loss": 0.9269,
      "step": 910
    },
    {
      "epoch": 0.5634236453201971,
      "grad_norm": 0.7325769139031748,
      "learning_rate": 9.537849530162915e-06,
      "loss": 0.9228,
      "step": 915
    },
    {
      "epoch": 0.5665024630541872,
      "grad_norm": 0.7282730155939394,
      "learning_rate": 9.430478039659604e-06,
      "loss": 0.9138,
      "step": 920
    },
    {
      "epoch": 0.5695812807881774,
      "grad_norm": 0.6428816898450572,
      "learning_rate": 9.323172382454494e-06,
      "loss": 0.9274,
      "step": 925
    },
    {
      "epoch": 0.5726600985221675,
      "grad_norm": 0.6703704767112079,
      "learning_rate": 9.215944962433329e-06,
      "loss": 0.855,
      "step": 930
    },
    {
      "epoch": 0.5757389162561576,
      "grad_norm": 0.7944858406137029,
      "learning_rate": 9.108808174438128e-06,
      "loss": 0.9416,
      "step": 935
    },
    {
      "epoch": 0.5788177339901478,
      "grad_norm": 0.721760753318748,
      "learning_rate": 9.001774402834375e-06,
      "loss": 0.8895,
      "step": 940
    },
    {
      "epoch": 0.5818965517241379,
      "grad_norm": 0.7210922094960251,
      "learning_rate": 8.89485602007949e-06,
      "loss": 0.952,
      "step": 945
    },
    {
      "epoch": 0.5849753694581281,
      "grad_norm": 0.7216379603390005,
      "learning_rate": 8.788065385292637e-06,
      "loss": 0.9067,
      "step": 950
    },
    {
      "epoch": 0.5880541871921182,
      "grad_norm": 0.6880918231702351,
      "learning_rate": 8.68141484282609e-06,
      "loss": 0.9327,
      "step": 955
    },
    {
      "epoch": 0.5911330049261084,
      "grad_norm": 0.7777971176157483,
      "learning_rate": 8.574916720838293e-06,
      "loss": 0.9514,
      "step": 960
    },
    {
      "epoch": 0.5942118226600985,
      "grad_norm": 0.7227495926203028,
      "learning_rate": 8.468583329868805e-06,
      "loss": 0.9463,
      "step": 965
    },
    {
      "epoch": 0.5972906403940886,
      "grad_norm": 0.7219855521048865,
      "learning_rate": 8.362426961415274e-06,
      "loss": 0.944,
      "step": 970
    },
    {
      "epoch": 0.6003694581280788,
      "grad_norm": 0.7155939312082987,
      "learning_rate": 8.256459886512618e-06,
      "loss": 0.9108,
      "step": 975
    },
    {
      "epoch": 0.603448275862069,
      "grad_norm": 0.6993953380271827,
      "learning_rate": 8.150694354314556e-06,
      "loss": 0.9557,
      "step": 980
    },
    {
      "epoch": 0.6065270935960592,
      "grad_norm": 0.7638296699531455,
      "learning_rate": 8.0451425906777e-06,
      "loss": 0.9312,
      "step": 985
    },
    {
      "epoch": 0.6096059113300493,
      "grad_norm": 0.6831250724265701,
      "learning_rate": 7.939816796748296e-06,
      "loss": 0.8998,
      "step": 990
    },
    {
      "epoch": 0.6126847290640394,
      "grad_norm": 0.7313287861208433,
      "learning_rate": 7.834729147551858e-06,
      "loss": 0.9255,
      "step": 995
    },
    {
      "epoch": 0.6157635467980296,
      "grad_norm": 0.6590166847034968,
      "learning_rate": 7.729891790585817e-06,
      "loss": 0.8842,
      "step": 1000
    },
    {
      "epoch": 0.6188423645320197,
      "grad_norm": 0.7323284885675908,
      "learning_rate": 7.625316844415327e-06,
      "loss": 0.8812,
      "step": 1005
    },
    {
      "epoch": 0.6219211822660099,
      "grad_norm": 0.7215619643224764,
      "learning_rate": 7.521016397272436e-06,
      "loss": 0.9061,
      "step": 1010
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.7213661635662646,
      "learning_rate": 7.417002505658773e-06,
      "loss": 0.9604,
      "step": 1015
    },
    {
      "epoch": 0.6280788177339901,
      "grad_norm": 0.6664085186795183,
      "learning_rate": 7.313287192951866e-06,
      "loss": 0.9152,
      "step": 1020
    },
    {
      "epoch": 0.6311576354679803,
      "grad_norm": 0.8899370388302164,
      "learning_rate": 7.209882448015338e-06,
      "loss": 0.8997,
      "step": 1025
    },
    {
      "epoch": 0.6342364532019704,
      "grad_norm": 0.6925853341918634,
      "learning_rate": 7.1068002238130465e-06,
      "loss": 0.9186,
      "step": 1030
    },
    {
      "epoch": 0.6373152709359606,
      "grad_norm": 0.7526266211171548,
      "learning_rate": 7.004052436027397e-06,
      "loss": 0.907,
      "step": 1035
    },
    {
      "epoch": 0.6403940886699507,
      "grad_norm": 0.7116871357451918,
      "learning_rate": 6.901650961681976e-06,
      "loss": 0.9136,
      "step": 1040
    },
    {
      "epoch": 0.6434729064039408,
      "grad_norm": 0.7181997436838037,
      "learning_rate": 6.799607637768621e-06,
      "loss": 0.9165,
      "step": 1045
    },
    {
      "epoch": 0.646551724137931,
      "grad_norm": 0.7424070514212068,
      "learning_rate": 6.6979342598791395e-06,
      "loss": 0.9401,
      "step": 1050
    },
    {
      "epoch": 0.6496305418719212,
      "grad_norm": 0.7056624976371111,
      "learning_rate": 6.596642580841827e-06,
      "loss": 0.9151,
      "step": 1055
    },
    {
      "epoch": 0.6527093596059114,
      "grad_norm": 0.7679067760249104,
      "learning_rate": 6.495744309362879e-06,
      "loss": 0.8805,
      "step": 1060
    },
    {
      "epoch": 0.6557881773399015,
      "grad_norm": 0.7261332639619558,
      "learning_rate": 6.395251108672975e-06,
      "loss": 0.8987,
      "step": 1065
    },
    {
      "epoch": 0.6588669950738916,
      "grad_norm": 0.6531278111009767,
      "learning_rate": 6.295174595179041e-06,
      "loss": 0.9147,
      "step": 1070
    },
    {
      "epoch": 0.6619458128078818,
      "grad_norm": 0.74223655982831,
      "learning_rate": 6.195526337121483e-06,
      "loss": 0.9432,
      "step": 1075
    },
    {
      "epoch": 0.6650246305418719,
      "grad_norm": 0.6775090824932495,
      "learning_rate": 6.096317853236975e-06,
      "loss": 0.8558,
      "step": 1080
    },
    {
      "epoch": 0.6681034482758621,
      "grad_norm": 0.7451875244452356,
      "learning_rate": 5.997560611426947e-06,
      "loss": 0.9129,
      "step": 1085
    },
    {
      "epoch": 0.6711822660098522,
      "grad_norm": 0.775303530121011,
      "learning_rate": 5.899266027431965e-06,
      "loss": 0.9302,
      "step": 1090
    },
    {
      "epoch": 0.6742610837438424,
      "grad_norm": 0.6681173244510462,
      "learning_rate": 5.80144546351216e-06,
      "loss": 0.8755,
      "step": 1095
    },
    {
      "epoch": 0.6773399014778325,
      "grad_norm": 0.6460985572337176,
      "learning_rate": 5.704110227133792e-06,
      "loss": 0.916,
      "step": 1100
    },
    {
      "epoch": 0.6804187192118226,
      "grad_norm": 0.6903928180028339,
      "learning_rate": 5.607271569662203e-06,
      "loss": 0.8914,
      "step": 1105
    },
    {
      "epoch": 0.6834975369458128,
      "grad_norm": 0.7795810817387983,
      "learning_rate": 5.510940685061202e-06,
      "loss": 0.8998,
      "step": 1110
    },
    {
      "epoch": 0.6865763546798029,
      "grad_norm": 0.7672315116978631,
      "learning_rate": 5.41512870859912e-06,
      "loss": 0.9303,
      "step": 1115
    },
    {
      "epoch": 0.6896551724137931,
      "grad_norm": 0.7806157930588039,
      "learning_rate": 5.319846715561656e-06,
      "loss": 0.9332,
      "step": 1120
    },
    {
      "epoch": 0.6927339901477833,
      "grad_norm": 0.7003891392810607,
      "learning_rate": 5.225105719971615e-06,
      "loss": 0.8993,
      "step": 1125
    },
    {
      "epoch": 0.6958128078817734,
      "grad_norm": 0.6721487834927411,
      "learning_rate": 5.130916673315762e-06,
      "loss": 0.941,
      "step": 1130
    },
    {
      "epoch": 0.6988916256157636,
      "grad_norm": 0.6757485350567788,
      "learning_rate": 5.037290463278914e-06,
      "loss": 0.9011,
      "step": 1135
    },
    {
      "epoch": 0.7019704433497537,
      "grad_norm": 0.7048595319775736,
      "learning_rate": 4.94423791248536e-06,
      "loss": 0.8987,
      "step": 1140
    },
    {
      "epoch": 0.7050492610837439,
      "grad_norm": 0.6775506981745996,
      "learning_rate": 4.851769777247857e-06,
      "loss": 0.9616,
      "step": 1145
    },
    {
      "epoch": 0.708128078817734,
      "grad_norm": 0.7082392404353789,
      "learning_rate": 4.759896746324247e-06,
      "loss": 0.9016,
      "step": 1150
    },
    {
      "epoch": 0.7112068965517241,
      "grad_norm": 0.6819223359256046,
      "learning_rate": 4.668629439681907e-06,
      "loss": 0.9272,
      "step": 1155
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.6845551815152396,
      "learning_rate": 4.577978407270156e-06,
      "loss": 0.933,
      "step": 1160
    },
    {
      "epoch": 0.7173645320197044,
      "grad_norm": 0.6950058521766465,
      "learning_rate": 4.487954127800726e-06,
      "loss": 0.9041,
      "step": 1165
    },
    {
      "epoch": 0.7204433497536946,
      "grad_norm": 0.6867684845183903,
      "learning_rate": 4.398567007536493e-06,
      "loss": 0.9612,
      "step": 1170
    },
    {
      "epoch": 0.7235221674876847,
      "grad_norm": 0.6586394312886726,
      "learning_rate": 4.309827379088589e-06,
      "loss": 0.9089,
      "step": 1175
    },
    {
      "epoch": 0.7266009852216748,
      "grad_norm": 0.7365265974356552,
      "learning_rate": 4.221745500221992e-06,
      "loss": 0.8724,
      "step": 1180
    },
    {
      "epoch": 0.729679802955665,
      "grad_norm": 0.7378780810314269,
      "learning_rate": 4.134331552669812e-06,
      "loss": 0.9275,
      "step": 1185
    },
    {
      "epoch": 0.7327586206896551,
      "grad_norm": 0.6956027586420739,
      "learning_rate": 4.047595640956326e-06,
      "loss": 0.9094,
      "step": 1190
    },
    {
      "epoch": 0.7358374384236454,
      "grad_norm": 0.6644083331902493,
      "learning_rate": 3.961547791228963e-06,
      "loss": 0.8809,
      "step": 1195
    },
    {
      "epoch": 0.7389162561576355,
      "grad_norm": 0.6780105644457819,
      "learning_rate": 3.876197950099351e-06,
      "loss": 0.8894,
      "step": 1200
    },
    {
      "epoch": 0.7419950738916257,
      "grad_norm": 0.7238626484216509,
      "learning_rate": 3.7915559834935355e-06,
      "loss": 0.9182,
      "step": 1205
    },
    {
      "epoch": 0.7450738916256158,
      "grad_norm": 0.677354013936962,
      "learning_rate": 3.7076316755115407e-06,
      "loss": 0.9056,
      "step": 1210
    },
    {
      "epoch": 0.7481527093596059,
      "grad_norm": 0.6887868120830557,
      "learning_rate": 3.6244347272963974e-06,
      "loss": 0.9033,
      "step": 1215
    },
    {
      "epoch": 0.7512315270935961,
      "grad_norm": 0.7935141959330477,
      "learning_rate": 3.5419747559127294e-06,
      "loss": 0.9077,
      "step": 1220
    },
    {
      "epoch": 0.7543103448275862,
      "grad_norm": 0.717043682095087,
      "learning_rate": 3.4602612932351065e-06,
      "loss": 0.9383,
      "step": 1225
    },
    {
      "epoch": 0.7573891625615764,
      "grad_norm": 0.7125434144608334,
      "learning_rate": 3.3793037848461873e-06,
      "loss": 0.9025,
      "step": 1230
    },
    {
      "epoch": 0.7604679802955665,
      "grad_norm": 0.6915559899322756,
      "learning_rate": 3.2991115889448877e-06,
      "loss": 0.9113,
      "step": 1235
    },
    {
      "epoch": 0.7635467980295566,
      "grad_norm": 0.6806388512684862,
      "learning_rate": 3.2196939752646183e-06,
      "loss": 0.9104,
      "step": 1240
    },
    {
      "epoch": 0.7666256157635468,
      "grad_norm": 0.7595959827137008,
      "learning_rate": 3.141060124001776e-06,
      "loss": 0.9138,
      "step": 1245
    },
    {
      "epoch": 0.7697044334975369,
      "grad_norm": 0.7234170553960368,
      "learning_rate": 3.063219124754543e-06,
      "loss": 0.9132,
      "step": 1250
    },
    {
      "epoch": 0.7727832512315271,
      "grad_norm": 0.7023061276615833,
      "learning_rate": 2.9861799754722033e-06,
      "loss": 0.892,
      "step": 1255
    },
    {
      "epoch": 0.7758620689655172,
      "grad_norm": 0.6711943725081508,
      "learning_rate": 2.9099515814150335e-06,
      "loss": 0.887,
      "step": 1260
    },
    {
      "epoch": 0.7789408866995073,
      "grad_norm": 0.725196707404757,
      "learning_rate": 2.8345427541248993e-06,
      "loss": 0.8776,
      "step": 1265
    },
    {
      "epoch": 0.7820197044334976,
      "grad_norm": 0.6608109215361615,
      "learning_rate": 2.7599622104066937e-06,
      "loss": 0.9427,
      "step": 1270
    },
    {
      "epoch": 0.7850985221674877,
      "grad_norm": 0.6996055449820648,
      "learning_rate": 2.6862185713207467e-06,
      "loss": 0.9142,
      "step": 1275
    },
    {
      "epoch": 0.7881773399014779,
      "grad_norm": 0.7432961342047957,
      "learning_rate": 2.6133203611862554e-06,
      "loss": 0.9064,
      "step": 1280
    },
    {
      "epoch": 0.791256157635468,
      "grad_norm": 0.6236987444510391,
      "learning_rate": 2.5412760065959386e-06,
      "loss": 0.9048,
      "step": 1285
    },
    {
      "epoch": 0.7943349753694581,
      "grad_norm": 0.6220890988258023,
      "learning_rate": 2.4700938354419823e-06,
      "loss": 0.8903,
      "step": 1290
    },
    {
      "epoch": 0.7974137931034483,
      "grad_norm": 0.6892863270775752,
      "learning_rate": 2.3997820759533654e-06,
      "loss": 0.913,
      "step": 1295
    },
    {
      "epoch": 0.8004926108374384,
      "grad_norm": 1.6773601433209218,
      "learning_rate": 2.3303488557447374e-06,
      "loss": 0.8753,
      "step": 1300
    },
    {
      "epoch": 0.8035714285714286,
      "grad_norm": 0.722872865650384,
      "learning_rate": 2.26180220087692e-06,
      "loss": 0.8957,
      "step": 1305
    },
    {
      "epoch": 0.8066502463054187,
      "grad_norm": 0.6613242490598888,
      "learning_rate": 2.194150034929133e-06,
      "loss": 0.9054,
      "step": 1310
    },
    {
      "epoch": 0.8097290640394089,
      "grad_norm": 0.7330833022368549,
      "learning_rate": 2.1274001780830776e-06,
      "loss": 0.9283,
      "step": 1315
    },
    {
      "epoch": 0.812807881773399,
      "grad_norm": 0.6778392727939891,
      "learning_rate": 2.0615603462189824e-06,
      "loss": 0.9288,
      "step": 1320
    },
    {
      "epoch": 0.8158866995073891,
      "grad_norm": 0.7707106273789677,
      "learning_rate": 1.9966381500236786e-06,
      "loss": 0.875,
      "step": 1325
    },
    {
      "epoch": 0.8189655172413793,
      "grad_norm": 0.6432858185048476,
      "learning_rate": 1.932641094110855e-06,
      "loss": 0.8971,
      "step": 1330
    },
    {
      "epoch": 0.8220443349753694,
      "grad_norm": 0.6914786964712628,
      "learning_rate": 1.869576576153581e-06,
      "loss": 0.8654,
      "step": 1335
    },
    {
      "epoch": 0.8251231527093597,
      "grad_norm": 0.7215233738169529,
      "learning_rate": 1.8074518860291646e-06,
      "loss": 0.909,
      "step": 1340
    },
    {
      "epoch": 0.8282019704433498,
      "grad_norm": 0.8465873318075495,
      "learning_rate": 1.746274204976498e-06,
      "loss": 0.9038,
      "step": 1345
    },
    {
      "epoch": 0.8312807881773399,
      "grad_norm": 0.7216800410068818,
      "learning_rate": 1.6860506047659442e-06,
      "loss": 0.9222,
      "step": 1350
    },
    {
      "epoch": 0.8343596059113301,
      "grad_norm": 0.7685628442751503,
      "learning_rate": 1.6267880468818787e-06,
      "loss": 0.8895,
      "step": 1355
    },
    {
      "epoch": 0.8374384236453202,
      "grad_norm": 0.7314941509944337,
      "learning_rate": 1.5684933817180014e-06,
      "loss": 0.9106,
      "step": 1360
    },
    {
      "epoch": 0.8405172413793104,
      "grad_norm": 0.7642178298055531,
      "learning_rate": 1.5111733477854507e-06,
      "loss": 0.9374,
      "step": 1365
    },
    {
      "epoch": 0.8435960591133005,
      "grad_norm": 0.6424365034435878,
      "learning_rate": 1.454834570933884e-06,
      "loss": 0.899,
      "step": 1370
    },
    {
      "epoch": 0.8466748768472906,
      "grad_norm": 0.7054107402438561,
      "learning_rate": 1.399483563585573e-06,
      "loss": 0.8976,
      "step": 1375
    },
    {
      "epoch": 0.8497536945812808,
      "grad_norm": 0.7944345419145749,
      "learning_rate": 1.345126723982594e-06,
      "loss": 0.9645,
      "step": 1380
    },
    {
      "epoch": 0.8528325123152709,
      "grad_norm": 0.7297116413505407,
      "learning_rate": 1.2917703354472467e-06,
      "loss": 0.9013,
      "step": 1385
    },
    {
      "epoch": 0.8559113300492611,
      "grad_norm": 0.7016058795668882,
      "learning_rate": 1.2394205656557224e-06,
      "loss": 0.886,
      "step": 1390
    },
    {
      "epoch": 0.8589901477832512,
      "grad_norm": 0.6817325792524672,
      "learning_rate": 1.1880834659251706e-06,
      "loss": 0.9106,
      "step": 1395
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.7175638109399047,
      "learning_rate": 1.1377649705142012e-06,
      "loss": 0.9046,
      "step": 1400
    },
    {
      "epoch": 0.8651477832512315,
      "grad_norm": 0.7325286340177533,
      "learning_rate": 1.0884708959369116e-06,
      "loss": 0.9344,
      "step": 1405
    },
    {
      "epoch": 0.8682266009852216,
      "grad_norm": 0.6995057655558119,
      "learning_rate": 1.040206940290547e-06,
      "loss": 0.8886,
      "step": 1410
    },
    {
      "epoch": 0.8713054187192119,
      "grad_norm": 0.6821188835840444,
      "learning_rate": 9.929786825968213e-07,
      "loss": 0.8976,
      "step": 1415
    },
    {
      "epoch": 0.874384236453202,
      "grad_norm": 0.7376721682017519,
      "learning_rate": 9.467915821570228e-07,
      "loss": 0.9492,
      "step": 1420
    },
    {
      "epoch": 0.8774630541871922,
      "grad_norm": 0.9328210953314316,
      "learning_rate": 9.016509779209536e-07,
      "loss": 0.9074,
      "step": 1425
    },
    {
      "epoch": 0.8805418719211823,
      "grad_norm": 0.7340819327835291,
      "learning_rate": 8.575620878697744e-07,
      "loss": 0.8837,
      "step": 1430
    },
    {
      "epoch": 0.8836206896551724,
      "grad_norm": 0.6845434866282742,
      "learning_rate": 8.145300084128349e-07,
      "loss": 0.9185,
      "step": 1435
    },
    {
      "epoch": 0.8866995073891626,
      "grad_norm": 0.6628364676048413,
      "learning_rate": 7.725597137985741e-07,
      "loss": 0.8584,
      "step": 1440
    },
    {
      "epoch": 0.8897783251231527,
      "grad_norm": 0.7208226072354926,
      "learning_rate": 7.316560555395069e-07,
      "loss": 0.8826,
      "step": 1445
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.6773256291080844,
      "learning_rate": 6.918237618514378e-07,
      "loss": 0.856,
      "step": 1450
    },
    {
      "epoch": 0.895935960591133,
      "grad_norm": 0.7034404298614597,
      "learning_rate": 6.530674371068946e-07,
      "loss": 0.9128,
      "step": 1455
    },
    {
      "epoch": 0.8990147783251231,
      "grad_norm": 0.7210199225257817,
      "learning_rate": 6.153915613028915e-07,
      "loss": 0.9038,
      "step": 1460
    },
    {
      "epoch": 0.9020935960591133,
      "grad_norm": 0.7406834860134297,
      "learning_rate": 5.788004895430799e-07,
      "loss": 0.8682,
      "step": 1465
    },
    {
      "epoch": 0.9051724137931034,
      "grad_norm": 0.7117319738607917,
      "learning_rate": 5.43298451534312e-07,
      "loss": 0.9407,
      "step": 1470
    },
    {
      "epoch": 0.9082512315270936,
      "grad_norm": 0.6909093195401991,
      "learning_rate": 5.088895510977154e-07,
      "loss": 0.8934,
      "step": 1475
    },
    {
      "epoch": 0.9113300492610837,
      "grad_norm": 0.6395163285948796,
      "learning_rate": 4.755777656943239e-07,
      "loss": 0.9065,
      "step": 1480
    },
    {
      "epoch": 0.9144088669950738,
      "grad_norm": 0.7727497202740083,
      "learning_rate": 4.433669459652945e-07,
      "loss": 0.8946,
      "step": 1485
    },
    {
      "epoch": 0.9174876847290641,
      "grad_norm": 0.6727790753972084,
      "learning_rate": 4.1226081528680907e-07,
      "loss": 0.9206,
      "step": 1490
    },
    {
      "epoch": 0.9205665024630542,
      "grad_norm": 0.6849296173953379,
      "learning_rate": 3.822629693396651e-07,
      "loss": 0.8824,
      "step": 1495
    },
    {
      "epoch": 0.9236453201970444,
      "grad_norm": 0.7395506995675738,
      "learning_rate": 3.5337687569363734e-07,
      "loss": 0.9219,
      "step": 1500
    },
    {
      "epoch": 0.9267241379310345,
      "grad_norm": 0.7042269789976827,
      "learning_rate": 3.2560587340665694e-07,
      "loss": 0.9093,
      "step": 1505
    },
    {
      "epoch": 0.9298029556650246,
      "grad_norm": 0.7584121398933832,
      "learning_rate": 2.989531726388262e-07,
      "loss": 0.9112,
      "step": 1510
    },
    {
      "epoch": 0.9328817733990148,
      "grad_norm": 0.7779893269073751,
      "learning_rate": 2.7342185428134604e-07,
      "loss": 0.9122,
      "step": 1515
    },
    {
      "epoch": 0.9359605911330049,
      "grad_norm": 0.6361716458037133,
      "learning_rate": 2.4901486960039025e-07,
      "loss": 0.9178,
      "step": 1520
    },
    {
      "epoch": 0.9390394088669951,
      "grad_norm": 0.7272001913384153,
      "learning_rate": 2.257350398959457e-07,
      "loss": 0.8953,
      "step": 1525
    },
    {
      "epoch": 0.9421182266009852,
      "grad_norm": 0.6848577166162307,
      "learning_rate": 2.035850561756969e-07,
      "loss": 0.9155,
      "step": 1530
    },
    {
      "epoch": 0.9451970443349754,
      "grad_norm": 0.7198842919309125,
      "learning_rate": 1.8256747884395577e-07,
      "loss": 0.9037,
      "step": 1535
    },
    {
      "epoch": 0.9482758620689655,
      "grad_norm": 0.7115101379396237,
      "learning_rate": 1.6268473740569723e-07,
      "loss": 0.9647,
      "step": 1540
    },
    {
      "epoch": 0.9513546798029556,
      "grad_norm": 0.7368827581632996,
      "learning_rate": 1.4393913018572182e-07,
      "loss": 0.8805,
      "step": 1545
    },
    {
      "epoch": 0.9544334975369458,
      "grad_norm": 0.8079433748319568,
      "learning_rate": 1.2633282406298576e-07,
      "loss": 0.896,
      "step": 1550
    },
    {
      "epoch": 0.9575123152709359,
      "grad_norm": 0.6587864044048146,
      "learning_rate": 1.0986785422011593e-07,
      "loss": 0.8944,
      "step": 1555
    },
    {
      "epoch": 0.9605911330049262,
      "grad_norm": 0.6943529244689446,
      "learning_rate": 9.454612390816686e-08,
      "loss": 0.892,
      "step": 1560
    },
    {
      "epoch": 0.9636699507389163,
      "grad_norm": 0.7475181568465236,
      "learning_rate": 8.036940422660345e-08,
      "loss": 0.9109,
      "step": 1565
    },
    {
      "epoch": 0.9667487684729064,
      "grad_norm": 0.712566295810641,
      "learning_rate": 6.733933391858238e-08,
      "loss": 0.9387,
      "step": 1570
    },
    {
      "epoch": 0.9698275862068966,
      "grad_norm": 0.6953912672416684,
      "learning_rate": 5.5457419181517145e-08,
      "loss": 0.9009,
      "step": 1575
    },
    {
      "epoch": 0.9729064039408867,
      "grad_norm": 0.6683739761168683,
      "learning_rate": 4.472503349297497e-08,
      "loss": 0.8673,
      "step": 1580
    },
    {
      "epoch": 0.9759852216748769,
      "grad_norm": 0.69079632147327,
      "learning_rate": 3.5143417451907195e-08,
      "loss": 0.8563,
      "step": 1585
    },
    {
      "epoch": 0.979064039408867,
      "grad_norm": 0.708420869441777,
      "learning_rate": 2.671367863524732e-08,
      "loss": 0.9274,
      "step": 1590
    },
    {
      "epoch": 0.9821428571428571,
      "grad_norm": 0.8174030990687192,
      "learning_rate": 1.943679146988009e-08,
      "loss": 0.9223,
      "step": 1595
    },
    {
      "epoch": 0.9852216748768473,
      "grad_norm": 0.742245024367394,
      "learning_rate": 1.3313597120002686e-08,
      "loss": 0.9255,
      "step": 1600
    },
    {
      "epoch": 0.9883004926108374,
      "grad_norm": 0.6809579676933898,
      "learning_rate": 8.34480338989141e-09,
      "loss": 0.9343,
      "step": 1605
    },
    {
      "epoch": 0.9913793103448276,
      "grad_norm": 0.6818330666493024,
      "learning_rate": 4.530984642087122e-09,
      "loss": 0.8425,
      "step": 1610
    },
    {
      "epoch": 0.9944581280788177,
      "grad_norm": 0.6517282911169013,
      "learning_rate": 1.872581730997247e-09,
      "loss": 0.9258,
      "step": 1615
    },
    {
      "epoch": 0.9975369458128078,
      "grad_norm": 0.7361227785975518,
      "learning_rate": 3.699019519387559e-10,
      "loss": 0.8655,
      "step": 1620
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.9033653736114502,
      "eval_runtime": 1302.7533,
      "eval_samples_per_second": 17.739,
      "eval_steps_per_second": 0.555,
      "step": 1624
    },
    {
      "epoch": 1.0,
      "step": 1624,
      "total_flos": 165422873640960.0,
      "train_loss": 0.9319284432040059,
      "train_runtime": 12380.9765,
      "train_samples_per_second": 4.197,
      "train_steps_per_second": 0.131
    }
  ],
  "logging_steps": 5,
  "max_steps": 1624,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "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": 165422873640960.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}