{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 15.0,
  "eval_steps": 500,
  "global_step": 156690,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.047865211564235115,
      "grad_norm": 4.9987640380859375,
      "learning_rate": 4.984044929478588e-05,
      "loss": 6.1091,
      "step": 500
    },
    {
      "epoch": 0.09573042312847023,
      "grad_norm": 3.691866397857666,
      "learning_rate": 4.968089858957177e-05,
      "loss": 5.4606,
      "step": 1000
    },
    {
      "epoch": 0.14359563469270534,
      "grad_norm": 3.0769104957580566,
      "learning_rate": 4.952134788435765e-05,
      "loss": 5.1824,
      "step": 1500
    },
    {
      "epoch": 0.19146084625694046,
      "grad_norm": 3.039604902267456,
      "learning_rate": 4.936179717914353e-05,
      "loss": 4.9954,
      "step": 2000
    },
    {
      "epoch": 0.23932605782117558,
      "grad_norm": 2.9064724445343018,
      "learning_rate": 4.920224647392942e-05,
      "loss": 4.8674,
      "step": 2500
    },
    {
      "epoch": 0.2871912693854107,
      "grad_norm": 2.613527536392212,
      "learning_rate": 4.90426957687153e-05,
      "loss": 4.7778,
      "step": 3000
    },
    {
      "epoch": 0.3350564809496458,
      "grad_norm": 2.9319610595703125,
      "learning_rate": 4.888314506350118e-05,
      "loss": 4.6852,
      "step": 3500
    },
    {
      "epoch": 0.3829216925138809,
      "grad_norm": 2.2705605030059814,
      "learning_rate": 4.872359435828707e-05,
      "loss": 4.6012,
      "step": 4000
    },
    {
      "epoch": 0.43078690407811604,
      "grad_norm": 2.4558980464935303,
      "learning_rate": 4.856404365307295e-05,
      "loss": 4.5279,
      "step": 4500
    },
    {
      "epoch": 0.47865211564235116,
      "grad_norm": 2.2625837326049805,
      "learning_rate": 4.840449294785883e-05,
      "loss": 4.4742,
      "step": 5000
    },
    {
      "epoch": 0.5265173272065863,
      "grad_norm": 2.0055723190307617,
      "learning_rate": 4.8244942242644715e-05,
      "loss": 4.4099,
      "step": 5500
    },
    {
      "epoch": 0.5743825387708213,
      "grad_norm": 2.514716386795044,
      "learning_rate": 4.8085391537430595e-05,
      "loss": 4.3753,
      "step": 6000
    },
    {
      "epoch": 0.6222477503350565,
      "grad_norm": 2.093000888824463,
      "learning_rate": 4.792584083221648e-05,
      "loss": 4.3379,
      "step": 6500
    },
    {
      "epoch": 0.6701129618992916,
      "grad_norm": 2.1274917125701904,
      "learning_rate": 4.776629012700236e-05,
      "loss": 4.2951,
      "step": 7000
    },
    {
      "epoch": 0.7179781734635267,
      "grad_norm": 1.9545608758926392,
      "learning_rate": 4.760673942178825e-05,
      "loss": 4.26,
      "step": 7500
    },
    {
      "epoch": 0.7658433850277618,
      "grad_norm": 2.0492897033691406,
      "learning_rate": 4.7447188716574135e-05,
      "loss": 4.2302,
      "step": 8000
    },
    {
      "epoch": 0.8137085965919969,
      "grad_norm": 1.807892918586731,
      "learning_rate": 4.7287638011360015e-05,
      "loss": 4.2007,
      "step": 8500
    },
    {
      "epoch": 0.8615738081562321,
      "grad_norm": 1.7267255783081055,
      "learning_rate": 4.7128087306145894e-05,
      "loss": 4.1619,
      "step": 9000
    },
    {
      "epoch": 0.9094390197204671,
      "grad_norm": 2.559781312942505,
      "learning_rate": 4.696853660093178e-05,
      "loss": 4.1372,
      "step": 9500
    },
    {
      "epoch": 0.9573042312847023,
      "grad_norm": 2.065793514251709,
      "learning_rate": 4.680898589571766e-05,
      "loss": 4.1102,
      "step": 10000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3191415754544042,
      "eval_loss": 4.0270867347717285,
      "eval_runtime": 359.9666,
      "eval_samples_per_second": 20.566,
      "eval_steps_per_second": 1.714,
      "step": 10446
    },
    {
      "epoch": 1.0051694428489375,
      "grad_norm": 1.8150527477264404,
      "learning_rate": 4.664943519050354e-05,
      "loss": 4.0842,
      "step": 10500
    },
    {
      "epoch": 1.0530346544131726,
      "grad_norm": 1.8704215288162231,
      "learning_rate": 4.648988448528943e-05,
      "loss": 4.0564,
      "step": 11000
    },
    {
      "epoch": 1.1008998659774076,
      "grad_norm": 1.7416319847106934,
      "learning_rate": 4.633033378007531e-05,
      "loss": 4.0353,
      "step": 11500
    },
    {
      "epoch": 1.1487650775416427,
      "grad_norm": 1.569896936416626,
      "learning_rate": 4.6170783074861194e-05,
      "loss": 4.0188,
      "step": 12000
    },
    {
      "epoch": 1.1966302891058778,
      "grad_norm": 1.760440707206726,
      "learning_rate": 4.601123236964708e-05,
      "loss": 3.9978,
      "step": 12500
    },
    {
      "epoch": 1.244495500670113,
      "grad_norm": 1.4585676193237305,
      "learning_rate": 4.585168166443296e-05,
      "loss": 3.9776,
      "step": 13000
    },
    {
      "epoch": 1.2923607122343481,
      "grad_norm": 1.6320855617523193,
      "learning_rate": 4.569213095921884e-05,
      "loss": 3.9667,
      "step": 13500
    },
    {
      "epoch": 1.3402259237985832,
      "grad_norm": 1.5580289363861084,
      "learning_rate": 4.553258025400473e-05,
      "loss": 3.9411,
      "step": 14000
    },
    {
      "epoch": 1.3880911353628183,
      "grad_norm": 1.587185025215149,
      "learning_rate": 4.537302954879061e-05,
      "loss": 3.9219,
      "step": 14500
    },
    {
      "epoch": 1.4359563469270533,
      "grad_norm": 1.592550277709961,
      "learning_rate": 4.521347884357649e-05,
      "loss": 3.9015,
      "step": 15000
    },
    {
      "epoch": 1.4838215584912886,
      "grad_norm": 1.6370365619659424,
      "learning_rate": 4.5053928138362374e-05,
      "loss": 3.8948,
      "step": 15500
    },
    {
      "epoch": 1.5316867700555237,
      "grad_norm": 1.5851680040359497,
      "learning_rate": 4.4894377433148253e-05,
      "loss": 3.8946,
      "step": 16000
    },
    {
      "epoch": 1.5795519816197587,
      "grad_norm": 1.4206801652908325,
      "learning_rate": 4.473482672793414e-05,
      "loss": 3.8669,
      "step": 16500
    },
    {
      "epoch": 1.627417193183994,
      "grad_norm": 1.5562480688095093,
      "learning_rate": 4.457527602272003e-05,
      "loss": 3.848,
      "step": 17000
    },
    {
      "epoch": 1.6752824047482289,
      "grad_norm": 1.3557647466659546,
      "learning_rate": 4.4415725317505907e-05,
      "loss": 3.8311,
      "step": 17500
    },
    {
      "epoch": 1.7231476163124642,
      "grad_norm": 1.404402732849121,
      "learning_rate": 4.4256174612291786e-05,
      "loss": 3.8065,
      "step": 18000
    },
    {
      "epoch": 1.7710128278766992,
      "grad_norm": 1.402544617652893,
      "learning_rate": 4.409662390707767e-05,
      "loss": 3.8029,
      "step": 18500
    },
    {
      "epoch": 1.8188780394409343,
      "grad_norm": 1.424489140510559,
      "learning_rate": 4.393707320186355e-05,
      "loss": 3.8062,
      "step": 19000
    },
    {
      "epoch": 1.8667432510051696,
      "grad_norm": 1.452799916267395,
      "learning_rate": 4.377752249664944e-05,
      "loss": 3.7866,
      "step": 19500
    },
    {
      "epoch": 1.9146084625694044,
      "grad_norm": 1.6548948287963867,
      "learning_rate": 4.361797179143532e-05,
      "loss": 3.7873,
      "step": 20000
    },
    {
      "epoch": 1.9624736741336397,
      "grad_norm": 1.3493238687515259,
      "learning_rate": 4.34584210862212e-05,
      "loss": 3.7697,
      "step": 20500
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3422187961367806,
      "eval_loss": 3.7410387992858887,
      "eval_runtime": 362.7524,
      "eval_samples_per_second": 20.408,
      "eval_steps_per_second": 1.701,
      "step": 20892
    },
    {
      "epoch": 2.010338885697875,
      "grad_norm": 1.430213212966919,
      "learning_rate": 4.3298870381007086e-05,
      "loss": 3.7607,
      "step": 21000
    },
    {
      "epoch": 2.05820409726211,
      "grad_norm": 1.383090615272522,
      "learning_rate": 4.313931967579297e-05,
      "loss": 3.7369,
      "step": 21500
    },
    {
      "epoch": 2.106069308826345,
      "grad_norm": 1.3713792562484741,
      "learning_rate": 4.297976897057885e-05,
      "loss": 3.7215,
      "step": 22000
    },
    {
      "epoch": 2.15393452039058,
      "grad_norm": 1.75236976146698,
      "learning_rate": 4.282021826536474e-05,
      "loss": 3.734,
      "step": 22500
    },
    {
      "epoch": 2.2017997319548153,
      "grad_norm": 1.3905434608459473,
      "learning_rate": 4.266066756015062e-05,
      "loss": 3.7114,
      "step": 23000
    },
    {
      "epoch": 2.2496649435190506,
      "grad_norm": 1.474369764328003,
      "learning_rate": 4.25011168549365e-05,
      "loss": 3.7018,
      "step": 23500
    },
    {
      "epoch": 2.2975301550832854,
      "grad_norm": 1.275120496749878,
      "learning_rate": 4.2341566149722386e-05,
      "loss": 3.6852,
      "step": 24000
    },
    {
      "epoch": 2.3453953666475207,
      "grad_norm": 1.3572947978973389,
      "learning_rate": 4.2182015444508266e-05,
      "loss": 3.6725,
      "step": 24500
    },
    {
      "epoch": 2.3932605782117555,
      "grad_norm": 1.3795580863952637,
      "learning_rate": 4.2022464739294145e-05,
      "loss": 3.6872,
      "step": 25000
    },
    {
      "epoch": 2.441125789775991,
      "grad_norm": 1.3290934562683105,
      "learning_rate": 4.186291403408003e-05,
      "loss": 3.6889,
      "step": 25500
    },
    {
      "epoch": 2.488991001340226,
      "grad_norm": 1.3350279331207275,
      "learning_rate": 4.170336332886591e-05,
      "loss": 3.6563,
      "step": 26000
    },
    {
      "epoch": 2.536856212904461,
      "grad_norm": 1.3395098447799683,
      "learning_rate": 4.15438126236518e-05,
      "loss": 3.6525,
      "step": 26500
    },
    {
      "epoch": 2.5847214244686962,
      "grad_norm": 1.2156445980072021,
      "learning_rate": 4.1384261918437685e-05,
      "loss": 3.6493,
      "step": 27000
    },
    {
      "epoch": 2.632586636032931,
      "grad_norm": 1.2468637228012085,
      "learning_rate": 4.1224711213223565e-05,
      "loss": 3.6418,
      "step": 27500
    },
    {
      "epoch": 2.6804518475971664,
      "grad_norm": 1.2122074365615845,
      "learning_rate": 4.1065160508009445e-05,
      "loss": 3.6268,
      "step": 28000
    },
    {
      "epoch": 2.7283170591614017,
      "grad_norm": 1.3240368366241455,
      "learning_rate": 4.090560980279533e-05,
      "loss": 3.6271,
      "step": 28500
    },
    {
      "epoch": 2.7761822707256365,
      "grad_norm": 1.1721067428588867,
      "learning_rate": 4.074605909758121e-05,
      "loss": 3.619,
      "step": 29000
    },
    {
      "epoch": 2.824047482289872,
      "grad_norm": 1.4006874561309814,
      "learning_rate": 4.05865083923671e-05,
      "loss": 3.6089,
      "step": 29500
    },
    {
      "epoch": 2.8719126938541066,
      "grad_norm": 1.224936604499817,
      "learning_rate": 4.042695768715298e-05,
      "loss": 3.6204,
      "step": 30000
    },
    {
      "epoch": 2.919777905418342,
      "grad_norm": 1.202871322631836,
      "learning_rate": 4.026740698193886e-05,
      "loss": 3.5843,
      "step": 30500
    },
    {
      "epoch": 2.967643116982577,
      "grad_norm": 1.3142194747924805,
      "learning_rate": 4.0107856276724745e-05,
      "loss": 3.5924,
      "step": 31000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3555951861738966,
      "eval_loss": 3.5885047912597656,
      "eval_runtime": 361.6922,
      "eval_samples_per_second": 20.468,
      "eval_steps_per_second": 1.706,
      "step": 31338
    },
    {
      "epoch": 3.015508328546812,
      "grad_norm": 1.2498036623001099,
      "learning_rate": 3.994830557151063e-05,
      "loss": 3.5975,
      "step": 31500
    },
    {
      "epoch": 3.0633735401110473,
      "grad_norm": 1.2046735286712646,
      "learning_rate": 3.978875486629651e-05,
      "loss": 3.5596,
      "step": 32000
    },
    {
      "epoch": 3.111238751675282,
      "grad_norm": 1.1853680610656738,
      "learning_rate": 3.96292041610824e-05,
      "loss": 3.5689,
      "step": 32500
    },
    {
      "epoch": 3.1591039632395175,
      "grad_norm": 1.3199185132980347,
      "learning_rate": 3.946965345586828e-05,
      "loss": 3.5572,
      "step": 33000
    },
    {
      "epoch": 3.2069691748037528,
      "grad_norm": 1.2235320806503296,
      "learning_rate": 3.931010275065416e-05,
      "loss": 3.5565,
      "step": 33500
    },
    {
      "epoch": 3.2548343863679876,
      "grad_norm": 1.0843192338943481,
      "learning_rate": 3.9150552045440044e-05,
      "loss": 3.5431,
      "step": 34000
    },
    {
      "epoch": 3.302699597932223,
      "grad_norm": 1.2282143831253052,
      "learning_rate": 3.8991001340225924e-05,
      "loss": 3.5548,
      "step": 34500
    },
    {
      "epoch": 3.3505648094964577,
      "grad_norm": 1.0993870496749878,
      "learning_rate": 3.8831450635011804e-05,
      "loss": 3.544,
      "step": 35000
    },
    {
      "epoch": 3.398430021060693,
      "grad_norm": 1.2994508743286133,
      "learning_rate": 3.867189992979769e-05,
      "loss": 3.5345,
      "step": 35500
    },
    {
      "epoch": 3.4462952326249283,
      "grad_norm": 1.1585030555725098,
      "learning_rate": 3.851234922458358e-05,
      "loss": 3.5339,
      "step": 36000
    },
    {
      "epoch": 3.494160444189163,
      "grad_norm": 1.245842456817627,
      "learning_rate": 3.835279851936946e-05,
      "loss": 3.514,
      "step": 36500
    },
    {
      "epoch": 3.5420256557533984,
      "grad_norm": 1.2206395864486694,
      "learning_rate": 3.8193247814155344e-05,
      "loss": 3.518,
      "step": 37000
    },
    {
      "epoch": 3.5898908673176333,
      "grad_norm": 1.1615595817565918,
      "learning_rate": 3.8033697108941224e-05,
      "loss": 3.4945,
      "step": 37500
    },
    {
      "epoch": 3.6377560788818686,
      "grad_norm": 1.142725944519043,
      "learning_rate": 3.7874146403727104e-05,
      "loss": 3.5211,
      "step": 38000
    },
    {
      "epoch": 3.685621290446104,
      "grad_norm": 1.1565961837768555,
      "learning_rate": 3.771459569851299e-05,
      "loss": 3.5081,
      "step": 38500
    },
    {
      "epoch": 3.733486502010339,
      "grad_norm": 1.1549856662750244,
      "learning_rate": 3.755504499329887e-05,
      "loss": 3.492,
      "step": 39000
    },
    {
      "epoch": 3.781351713574574,
      "grad_norm": 1.152442216873169,
      "learning_rate": 3.739549428808476e-05,
      "loss": 3.5004,
      "step": 39500
    },
    {
      "epoch": 3.829216925138809,
      "grad_norm": 1.1138166189193726,
      "learning_rate": 3.723594358287064e-05,
      "loss": 3.4987,
      "step": 40000
    },
    {
      "epoch": 3.877082136703044,
      "grad_norm": 1.265821099281311,
      "learning_rate": 3.7076392877656517e-05,
      "loss": 3.4859,
      "step": 40500
    },
    {
      "epoch": 3.9249473482672794,
      "grad_norm": 1.133989691734314,
      "learning_rate": 3.69168421724424e-05,
      "loss": 3.4744,
      "step": 41000
    },
    {
      "epoch": 3.9728125598315147,
      "grad_norm": 1.2658230066299438,
      "learning_rate": 3.675729146722829e-05,
      "loss": 3.4844,
      "step": 41500
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.36413086607645917,
      "eval_loss": 3.4934167861938477,
      "eval_runtime": 358.4421,
      "eval_samples_per_second": 20.653,
      "eval_steps_per_second": 1.721,
      "step": 41784
    },
    {
      "epoch": 4.02067777139575,
      "grad_norm": 1.140244483947754,
      "learning_rate": 3.659774076201417e-05,
      "loss": 3.4799,
      "step": 42000
    },
    {
      "epoch": 4.068542982959984,
      "grad_norm": 1.1685595512390137,
      "learning_rate": 3.6438190056800056e-05,
      "loss": 3.4588,
      "step": 42500
    },
    {
      "epoch": 4.11640819452422,
      "grad_norm": 1.171810269355774,
      "learning_rate": 3.6278639351585936e-05,
      "loss": 3.4544,
      "step": 43000
    },
    {
      "epoch": 4.164273406088455,
      "grad_norm": 1.2411644458770752,
      "learning_rate": 3.6119088646371816e-05,
      "loss": 3.4519,
      "step": 43500
    },
    {
      "epoch": 4.21213861765269,
      "grad_norm": 1.1433331966400146,
      "learning_rate": 3.59595379411577e-05,
      "loss": 3.4457,
      "step": 44000
    },
    {
      "epoch": 4.2600038292169256,
      "grad_norm": 1.2162096500396729,
      "learning_rate": 3.579998723594358e-05,
      "loss": 3.4489,
      "step": 44500
    },
    {
      "epoch": 4.30786904078116,
      "grad_norm": 1.1380053758621216,
      "learning_rate": 3.564043653072946e-05,
      "loss": 3.446,
      "step": 45000
    },
    {
      "epoch": 4.355734252345395,
      "grad_norm": 1.0851870775222778,
      "learning_rate": 3.5480885825515356e-05,
      "loss": 3.4392,
      "step": 45500
    },
    {
      "epoch": 4.4035994639096305,
      "grad_norm": 1.1477234363555908,
      "learning_rate": 3.5321335120301236e-05,
      "loss": 3.4287,
      "step": 46000
    },
    {
      "epoch": 4.451464675473866,
      "grad_norm": 1.3154938220977783,
      "learning_rate": 3.5161784415087116e-05,
      "loss": 3.4469,
      "step": 46500
    },
    {
      "epoch": 4.499329887038101,
      "grad_norm": 1.185604453086853,
      "learning_rate": 3.5002233709873e-05,
      "loss": 3.4226,
      "step": 47000
    },
    {
      "epoch": 4.5471950986023355,
      "grad_norm": 1.1426012516021729,
      "learning_rate": 3.484268300465888e-05,
      "loss": 3.4338,
      "step": 47500
    },
    {
      "epoch": 4.595060310166571,
      "grad_norm": 1.1751033067703247,
      "learning_rate": 3.468313229944476e-05,
      "loss": 3.4201,
      "step": 48000
    },
    {
      "epoch": 4.642925521730806,
      "grad_norm": 1.1464704275131226,
      "learning_rate": 3.452358159423065e-05,
      "loss": 3.4213,
      "step": 48500
    },
    {
      "epoch": 4.690790733295041,
      "grad_norm": 1.1576846837997437,
      "learning_rate": 3.436403088901653e-05,
      "loss": 3.4149,
      "step": 49000
    },
    {
      "epoch": 4.738655944859277,
      "grad_norm": 1.1627798080444336,
      "learning_rate": 3.4204480183802415e-05,
      "loss": 3.424,
      "step": 49500
    },
    {
      "epoch": 4.786521156423511,
      "grad_norm": 1.106953501701355,
      "learning_rate": 3.4044929478588295e-05,
      "loss": 3.4048,
      "step": 50000
    },
    {
      "epoch": 4.834386367987746,
      "grad_norm": 1.167616367340088,
      "learning_rate": 3.388537877337418e-05,
      "loss": 3.4069,
      "step": 50500
    },
    {
      "epoch": 4.882251579551982,
      "grad_norm": 1.101952075958252,
      "learning_rate": 3.372582806816006e-05,
      "loss": 3.404,
      "step": 51000
    },
    {
      "epoch": 4.930116791116217,
      "grad_norm": 1.055081844329834,
      "learning_rate": 3.356627736294595e-05,
      "loss": 3.3789,
      "step": 51500
    },
    {
      "epoch": 4.977982002680452,
      "grad_norm": 1.1978681087493896,
      "learning_rate": 3.340672665773183e-05,
      "loss": 3.4051,
      "step": 52000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.37062040183704026,
      "eval_loss": 3.4276773929595947,
      "eval_runtime": 359.7089,
      "eval_samples_per_second": 20.581,
      "eval_steps_per_second": 1.715,
      "step": 52230
    },
    {
      "epoch": 5.025847214244687,
      "grad_norm": 1.0834215879440308,
      "learning_rate": 3.3247175952517715e-05,
      "loss": 3.3835,
      "step": 52500
    },
    {
      "epoch": 5.073712425808922,
      "grad_norm": 1.174057960510254,
      "learning_rate": 3.3087625247303595e-05,
      "loss": 3.3822,
      "step": 53000
    },
    {
      "epoch": 5.121577637373157,
      "grad_norm": 1.2661558389663696,
      "learning_rate": 3.2928074542089475e-05,
      "loss": 3.3796,
      "step": 53500
    },
    {
      "epoch": 5.1694428489373925,
      "grad_norm": 1.0967440605163574,
      "learning_rate": 3.276852383687536e-05,
      "loss": 3.3675,
      "step": 54000
    },
    {
      "epoch": 5.217308060501628,
      "grad_norm": 1.079768419265747,
      "learning_rate": 3.260897313166124e-05,
      "loss": 3.3669,
      "step": 54500
    },
    {
      "epoch": 5.265173272065862,
      "grad_norm": 1.1150493621826172,
      "learning_rate": 3.244942242644713e-05,
      "loss": 3.3573,
      "step": 55000
    },
    {
      "epoch": 5.3130384836300975,
      "grad_norm": 1.1257209777832031,
      "learning_rate": 3.2289871721233015e-05,
      "loss": 3.3669,
      "step": 55500
    },
    {
      "epoch": 5.360903695194333,
      "grad_norm": 1.0811665058135986,
      "learning_rate": 3.2130321016018894e-05,
      "loss": 3.3541,
      "step": 56000
    },
    {
      "epoch": 5.408768906758568,
      "grad_norm": 1.094035029411316,
      "learning_rate": 3.1970770310804774e-05,
      "loss": 3.3606,
      "step": 56500
    },
    {
      "epoch": 5.456634118322803,
      "grad_norm": 1.1808502674102783,
      "learning_rate": 3.181121960559066e-05,
      "loss": 3.3715,
      "step": 57000
    },
    {
      "epoch": 5.504499329887038,
      "grad_norm": 1.1131441593170166,
      "learning_rate": 3.165166890037654e-05,
      "loss": 3.3658,
      "step": 57500
    },
    {
      "epoch": 5.552364541451273,
      "grad_norm": 1.1039623022079468,
      "learning_rate": 3.149211819516242e-05,
      "loss": 3.3556,
      "step": 58000
    },
    {
      "epoch": 5.600229753015508,
      "grad_norm": 1.005952000617981,
      "learning_rate": 3.133256748994831e-05,
      "loss": 3.3513,
      "step": 58500
    },
    {
      "epoch": 5.648094964579744,
      "grad_norm": 0.9689833521842957,
      "learning_rate": 3.117301678473419e-05,
      "loss": 3.3551,
      "step": 59000
    },
    {
      "epoch": 5.695960176143979,
      "grad_norm": 1.1173384189605713,
      "learning_rate": 3.101346607952007e-05,
      "loss": 3.3391,
      "step": 59500
    },
    {
      "epoch": 5.743825387708213,
      "grad_norm": 1.0669808387756348,
      "learning_rate": 3.085391537430596e-05,
      "loss": 3.3565,
      "step": 60000
    },
    {
      "epoch": 5.791690599272449,
      "grad_norm": 1.0116264820098877,
      "learning_rate": 3.069436466909184e-05,
      "loss": 3.3551,
      "step": 60500
    },
    {
      "epoch": 5.839555810836684,
      "grad_norm": 1.1355198621749878,
      "learning_rate": 3.053481396387772e-05,
      "loss": 3.331,
      "step": 61000
    },
    {
      "epoch": 5.887421022400919,
      "grad_norm": 1.000922441482544,
      "learning_rate": 3.0375263258663607e-05,
      "loss": 3.3421,
      "step": 61500
    },
    {
      "epoch": 5.935286233965154,
      "grad_norm": 1.163181185722351,
      "learning_rate": 3.0215712553449487e-05,
      "loss": 3.3229,
      "step": 62000
    },
    {
      "epoch": 5.983151445529389,
      "grad_norm": 1.1023402214050293,
      "learning_rate": 3.005616184823537e-05,
      "loss": 3.3317,
      "step": 62500
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.37539720297800067,
      "eval_loss": 3.378039836883545,
      "eval_runtime": 359.883,
      "eval_samples_per_second": 20.571,
      "eval_steps_per_second": 1.714,
      "step": 62676
    },
    {
      "epoch": 6.031016657093624,
      "grad_norm": 1.0495789051055908,
      "learning_rate": 2.9896611143021253e-05,
      "loss": 3.3059,
      "step": 63000
    },
    {
      "epoch": 6.078881868657859,
      "grad_norm": 1.0401114225387573,
      "learning_rate": 2.9737060437807133e-05,
      "loss": 3.3259,
      "step": 63500
    },
    {
      "epoch": 6.126747080222095,
      "grad_norm": 1.0515329837799072,
      "learning_rate": 2.9577509732593017e-05,
      "loss": 3.3086,
      "step": 64000
    },
    {
      "epoch": 6.17461229178633,
      "grad_norm": 1.2691503763198853,
      "learning_rate": 2.9417959027378907e-05,
      "loss": 3.3087,
      "step": 64500
    },
    {
      "epoch": 6.222477503350564,
      "grad_norm": 1.1866391897201538,
      "learning_rate": 2.9258408322164786e-05,
      "loss": 3.3125,
      "step": 65000
    },
    {
      "epoch": 6.2703427149148,
      "grad_norm": 1.164188027381897,
      "learning_rate": 2.909885761695067e-05,
      "loss": 3.3042,
      "step": 65500
    },
    {
      "epoch": 6.318207926479035,
      "grad_norm": 1.044144868850708,
      "learning_rate": 2.8939306911736553e-05,
      "loss": 3.2976,
      "step": 66000
    },
    {
      "epoch": 6.36607313804327,
      "grad_norm": 1.171011209487915,
      "learning_rate": 2.8779756206522436e-05,
      "loss": 3.3105,
      "step": 66500
    },
    {
      "epoch": 6.4139383496075055,
      "grad_norm": 1.206079363822937,
      "learning_rate": 2.8620205501308316e-05,
      "loss": 3.2922,
      "step": 67000
    },
    {
      "epoch": 6.46180356117174,
      "grad_norm": 1.114662766456604,
      "learning_rate": 2.84606547960942e-05,
      "loss": 3.3032,
      "step": 67500
    },
    {
      "epoch": 6.509668772735975,
      "grad_norm": 0.9734864830970764,
      "learning_rate": 2.8301104090880083e-05,
      "loss": 3.2974,
      "step": 68000
    },
    {
      "epoch": 6.5575339843002105,
      "grad_norm": 1.09837007522583,
      "learning_rate": 2.8141553385665963e-05,
      "loss": 3.2927,
      "step": 68500
    },
    {
      "epoch": 6.605399195864446,
      "grad_norm": 1.0099565982818604,
      "learning_rate": 2.7982002680451846e-05,
      "loss": 3.2859,
      "step": 69000
    },
    {
      "epoch": 6.653264407428681,
      "grad_norm": 1.0673879384994507,
      "learning_rate": 2.7822451975237736e-05,
      "loss": 3.3063,
      "step": 69500
    },
    {
      "epoch": 6.7011296189929155,
      "grad_norm": 1.0769966840744019,
      "learning_rate": 2.7662901270023616e-05,
      "loss": 3.3097,
      "step": 70000
    },
    {
      "epoch": 6.748994830557151,
      "grad_norm": 1.0357173681259155,
      "learning_rate": 2.75033505648095e-05,
      "loss": 3.305,
      "step": 70500
    },
    {
      "epoch": 6.796860042121386,
      "grad_norm": 1.039602518081665,
      "learning_rate": 2.7343799859595382e-05,
      "loss": 3.2874,
      "step": 71000
    },
    {
      "epoch": 6.844725253685621,
      "grad_norm": 1.0579136610031128,
      "learning_rate": 2.7184249154381266e-05,
      "loss": 3.2858,
      "step": 71500
    },
    {
      "epoch": 6.892590465249857,
      "grad_norm": 1.0391180515289307,
      "learning_rate": 2.7024698449167145e-05,
      "loss": 3.3115,
      "step": 72000
    },
    {
      "epoch": 6.940455676814091,
      "grad_norm": 1.0982191562652588,
      "learning_rate": 2.686514774395303e-05,
      "loss": 3.2869,
      "step": 72500
    },
    {
      "epoch": 6.988320888378326,
      "grad_norm": 1.0707833766937256,
      "learning_rate": 2.6705597038738912e-05,
      "loss": 3.2927,
      "step": 73000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.37949741386447516,
      "eval_loss": 3.337162971496582,
      "eval_runtime": 366.3928,
      "eval_samples_per_second": 20.205,
      "eval_steps_per_second": 1.684,
      "step": 73122
    },
    {
      "epoch": 7.036186099942562,
      "grad_norm": 1.066457748413086,
      "learning_rate": 2.6546046333524792e-05,
      "loss": 3.2695,
      "step": 73500
    },
    {
      "epoch": 7.084051311506797,
      "grad_norm": 1.0978970527648926,
      "learning_rate": 2.6386495628310675e-05,
      "loss": 3.2878,
      "step": 74000
    },
    {
      "epoch": 7.131916523071032,
      "grad_norm": 1.1022377014160156,
      "learning_rate": 2.6226944923096565e-05,
      "loss": 3.2639,
      "step": 74500
    },
    {
      "epoch": 7.1797817346352675,
      "grad_norm": 1.0848217010498047,
      "learning_rate": 2.6067394217882445e-05,
      "loss": 3.2735,
      "step": 75000
    },
    {
      "epoch": 7.227646946199502,
      "grad_norm": 1.066747784614563,
      "learning_rate": 2.5907843512668328e-05,
      "loss": 3.2672,
      "step": 75500
    },
    {
      "epoch": 7.275512157763737,
      "grad_norm": 1.0969438552856445,
      "learning_rate": 2.574829280745421e-05,
      "loss": 3.2509,
      "step": 76000
    },
    {
      "epoch": 7.3233773693279725,
      "grad_norm": 1.0943154096603394,
      "learning_rate": 2.558874210224009e-05,
      "loss": 3.2618,
      "step": 76500
    },
    {
      "epoch": 7.371242580892208,
      "grad_norm": 1.3066858053207397,
      "learning_rate": 2.5429191397025975e-05,
      "loss": 3.2706,
      "step": 77000
    },
    {
      "epoch": 7.419107792456443,
      "grad_norm": 1.0907511711120605,
      "learning_rate": 2.5269640691811858e-05,
      "loss": 3.2581,
      "step": 77500
    },
    {
      "epoch": 7.466973004020677,
      "grad_norm": 1.0295166969299316,
      "learning_rate": 2.511008998659774e-05,
      "loss": 3.2508,
      "step": 78000
    },
    {
      "epoch": 7.514838215584913,
      "grad_norm": 0.997578501701355,
      "learning_rate": 2.4950539281383625e-05,
      "loss": 3.2646,
      "step": 78500
    },
    {
      "epoch": 7.562703427149148,
      "grad_norm": 1.0320909023284912,
      "learning_rate": 2.4790988576169508e-05,
      "loss": 3.248,
      "step": 79000
    },
    {
      "epoch": 7.610568638713383,
      "grad_norm": 1.1069414615631104,
      "learning_rate": 2.463143787095539e-05,
      "loss": 3.2586,
      "step": 79500
    },
    {
      "epoch": 7.658433850277619,
      "grad_norm": 1.0554596185684204,
      "learning_rate": 2.447188716574127e-05,
      "loss": 3.2691,
      "step": 80000
    },
    {
      "epoch": 7.706299061841853,
      "grad_norm": 1.0136358737945557,
      "learning_rate": 2.4312336460527158e-05,
      "loss": 3.2546,
      "step": 80500
    },
    {
      "epoch": 7.754164273406088,
      "grad_norm": 1.0406538248062134,
      "learning_rate": 2.415278575531304e-05,
      "loss": 3.2487,
      "step": 81000
    },
    {
      "epoch": 7.802029484970324,
      "grad_norm": 1.1913374662399292,
      "learning_rate": 2.399323505009892e-05,
      "loss": 3.246,
      "step": 81500
    },
    {
      "epoch": 7.849894696534559,
      "grad_norm": 1.1145275831222534,
      "learning_rate": 2.3833684344884804e-05,
      "loss": 3.2365,
      "step": 82000
    },
    {
      "epoch": 7.897759908098794,
      "grad_norm": 1.0950984954833984,
      "learning_rate": 2.367413363967069e-05,
      "loss": 3.2418,
      "step": 82500
    },
    {
      "epoch": 7.9456251196630285,
      "grad_norm": 1.0254520177841187,
      "learning_rate": 2.351458293445657e-05,
      "loss": 3.2345,
      "step": 83000
    },
    {
      "epoch": 7.993490331227264,
      "grad_norm": 1.4270539283752441,
      "learning_rate": 2.3355032229242454e-05,
      "loss": 3.2422,
      "step": 83500
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.382483310707701,
      "eval_loss": 3.3083410263061523,
      "eval_runtime": 1443.8211,
      "eval_samples_per_second": 5.127,
      "eval_steps_per_second": 0.427,
      "step": 83568
    },
    {
      "epoch": 8.0413555427915,
      "grad_norm": 1.1160823106765747,
      "learning_rate": 2.3195481524028337e-05,
      "loss": 3.2434,
      "step": 84000
    },
    {
      "epoch": 8.089220754355734,
      "grad_norm": 1.01189386844635,
      "learning_rate": 2.303593081881422e-05,
      "loss": 3.2296,
      "step": 84500
    },
    {
      "epoch": 8.137085965919969,
      "grad_norm": 1.0501444339752197,
      "learning_rate": 2.2876380113600104e-05,
      "loss": 3.2257,
      "step": 85000
    },
    {
      "epoch": 8.184951177484205,
      "grad_norm": 1.0764137506484985,
      "learning_rate": 2.2716829408385987e-05,
      "loss": 3.2289,
      "step": 85500
    },
    {
      "epoch": 8.23281638904844,
      "grad_norm": 1.1646994352340698,
      "learning_rate": 2.255727870317187e-05,
      "loss": 3.2231,
      "step": 86000
    },
    {
      "epoch": 8.280681600612676,
      "grad_norm": 1.2481164932250977,
      "learning_rate": 2.239772799795775e-05,
      "loss": 3.2252,
      "step": 86500
    },
    {
      "epoch": 8.32854681217691,
      "grad_norm": 1.0603400468826294,
      "learning_rate": 2.2238177292743633e-05,
      "loss": 3.2168,
      "step": 87000
    },
    {
      "epoch": 8.376412023741144,
      "grad_norm": 1.0832257270812988,
      "learning_rate": 2.207862658752952e-05,
      "loss": 3.229,
      "step": 87500
    },
    {
      "epoch": 8.42427723530538,
      "grad_norm": 0.9691539406776428,
      "learning_rate": 2.19190758823154e-05,
      "loss": 3.2228,
      "step": 88000
    },
    {
      "epoch": 8.472142446869615,
      "grad_norm": 1.1980475187301636,
      "learning_rate": 2.1759525177101283e-05,
      "loss": 3.2287,
      "step": 88500
    },
    {
      "epoch": 8.520007658433851,
      "grad_norm": 1.1181550025939941,
      "learning_rate": 2.1599974471887166e-05,
      "loss": 3.229,
      "step": 89000
    },
    {
      "epoch": 8.567872869998086,
      "grad_norm": 1.074438452720642,
      "learning_rate": 2.144042376667305e-05,
      "loss": 3.2146,
      "step": 89500
    },
    {
      "epoch": 8.61573808156232,
      "grad_norm": 1.0238900184631348,
      "learning_rate": 2.1280873061458933e-05,
      "loss": 3.2244,
      "step": 90000
    },
    {
      "epoch": 8.663603293126556,
      "grad_norm": 1.0909308195114136,
      "learning_rate": 2.1121322356244816e-05,
      "loss": 3.2184,
      "step": 90500
    },
    {
      "epoch": 8.71146850469079,
      "grad_norm": 1.1594678163528442,
      "learning_rate": 2.09617716510307e-05,
      "loss": 3.2142,
      "step": 91000
    },
    {
      "epoch": 8.759333716255027,
      "grad_norm": 1.060739517211914,
      "learning_rate": 2.080222094581658e-05,
      "loss": 3.2198,
      "step": 91500
    },
    {
      "epoch": 8.807198927819261,
      "grad_norm": 1.035969853401184,
      "learning_rate": 2.0642670240602466e-05,
      "loss": 3.2107,
      "step": 92000
    },
    {
      "epoch": 8.855064139383495,
      "grad_norm": 1.0699774026870728,
      "learning_rate": 2.048311953538835e-05,
      "loss": 3.2147,
      "step": 92500
    },
    {
      "epoch": 8.902929350947732,
      "grad_norm": 1.072472333908081,
      "learning_rate": 2.032356883017423e-05,
      "loss": 3.2079,
      "step": 93000
    },
    {
      "epoch": 8.950794562511966,
      "grad_norm": 1.073217511177063,
      "learning_rate": 2.0164018124960112e-05,
      "loss": 3.209,
      "step": 93500
    },
    {
      "epoch": 8.998659774076202,
      "grad_norm": 1.025145411491394,
      "learning_rate": 2.0004467419745996e-05,
      "loss": 3.2141,
      "step": 94000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.38499121053273033,
      "eval_loss": 3.2837157249450684,
      "eval_runtime": 350.9331,
      "eval_samples_per_second": 21.095,
      "eval_steps_per_second": 1.319,
      "step": 94014
    },
    {
      "epoch": 9.046524985640437,
      "grad_norm": 0.9691001772880554,
      "learning_rate": 1.984491671453188e-05,
      "loss": 3.2058,
      "step": 94500
    },
    {
      "epoch": 9.094390197204671,
      "grad_norm": 1.0141669511795044,
      "learning_rate": 1.9685366009317762e-05,
      "loss": 3.196,
      "step": 95000
    },
    {
      "epoch": 9.142255408768907,
      "grad_norm": 1.1707212924957275,
      "learning_rate": 1.9525815304103645e-05,
      "loss": 3.1968,
      "step": 95500
    },
    {
      "epoch": 9.190120620333142,
      "grad_norm": 1.0739679336547852,
      "learning_rate": 1.936626459888953e-05,
      "loss": 3.1936,
      "step": 96000
    },
    {
      "epoch": 9.237985831897378,
      "grad_norm": 1.070914387702942,
      "learning_rate": 1.920671389367541e-05,
      "loss": 3.1985,
      "step": 96500
    },
    {
      "epoch": 9.285851043461612,
      "grad_norm": 0.9926000833511353,
      "learning_rate": 1.9047163188461295e-05,
      "loss": 3.2026,
      "step": 97000
    },
    {
      "epoch": 9.333716255025847,
      "grad_norm": 1.0520483255386353,
      "learning_rate": 1.888761248324718e-05,
      "loss": 3.1903,
      "step": 97500
    },
    {
      "epoch": 9.381581466590083,
      "grad_norm": 1.015866756439209,
      "learning_rate": 1.872806177803306e-05,
      "loss": 3.1925,
      "step": 98000
    },
    {
      "epoch": 9.429446678154317,
      "grad_norm": 1.0483351945877075,
      "learning_rate": 1.856851107281894e-05,
      "loss": 3.1846,
      "step": 98500
    },
    {
      "epoch": 9.477311889718553,
      "grad_norm": 1.0546178817749023,
      "learning_rate": 1.8408960367604825e-05,
      "loss": 3.1901,
      "step": 99000
    },
    {
      "epoch": 9.525177101282788,
      "grad_norm": 1.0416738986968994,
      "learning_rate": 1.8249409662390708e-05,
      "loss": 3.1928,
      "step": 99500
    },
    {
      "epoch": 9.573042312847022,
      "grad_norm": 1.0481783151626587,
      "learning_rate": 1.808985895717659e-05,
      "loss": 3.1961,
      "step": 100000
    },
    {
      "epoch": 9.620907524411258,
      "grad_norm": 1.0868226289749146,
      "learning_rate": 1.7930308251962475e-05,
      "loss": 3.2011,
      "step": 100500
    },
    {
      "epoch": 9.668772735975493,
      "grad_norm": 0.9941065907478333,
      "learning_rate": 1.7770757546748358e-05,
      "loss": 3.1836,
      "step": 101000
    },
    {
      "epoch": 9.716637947539729,
      "grad_norm": 1.1167347431182861,
      "learning_rate": 1.7611206841534238e-05,
      "loss": 3.1862,
      "step": 101500
    },
    {
      "epoch": 9.764503159103963,
      "grad_norm": 1.0141208171844482,
      "learning_rate": 1.7451656136320125e-05,
      "loss": 3.1814,
      "step": 102000
    },
    {
      "epoch": 9.812368370668198,
      "grad_norm": 1.0675753355026245,
      "learning_rate": 1.7292105431106008e-05,
      "loss": 3.1937,
      "step": 102500
    },
    {
      "epoch": 9.860233582232434,
      "grad_norm": 1.0234485864639282,
      "learning_rate": 1.7132554725891888e-05,
      "loss": 3.1895,
      "step": 103000
    },
    {
      "epoch": 9.908098793796668,
      "grad_norm": 1.014687180519104,
      "learning_rate": 1.697300402067777e-05,
      "loss": 3.1784,
      "step": 103500
    },
    {
      "epoch": 9.955964005360904,
      "grad_norm": 1.0305941104888916,
      "learning_rate": 1.6813453315463658e-05,
      "loss": 3.1901,
      "step": 104000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.386960637473725,
      "eval_loss": 3.2644245624542236,
      "eval_runtime": 351.4517,
      "eval_samples_per_second": 21.064,
      "eval_steps_per_second": 1.317,
      "step": 104460
    },
    {
      "epoch": 10.003829216925139,
      "grad_norm": 1.0544371604919434,
      "learning_rate": 1.6653902610249537e-05,
      "loss": 3.198,
      "step": 104500
    },
    {
      "epoch": 10.051694428489373,
      "grad_norm": 1.1083343029022217,
      "learning_rate": 1.649435190503542e-05,
      "loss": 3.1746,
      "step": 105000
    },
    {
      "epoch": 10.09955964005361,
      "grad_norm": 1.0985013246536255,
      "learning_rate": 1.6334801199821304e-05,
      "loss": 3.1735,
      "step": 105500
    },
    {
      "epoch": 10.147424851617844,
      "grad_norm": 1.114378809928894,
      "learning_rate": 1.6175250494607187e-05,
      "loss": 3.1766,
      "step": 106000
    },
    {
      "epoch": 10.19529006318208,
      "grad_norm": 1.0587921142578125,
      "learning_rate": 1.601569978939307e-05,
      "loss": 3.167,
      "step": 106500
    },
    {
      "epoch": 10.243155274746314,
      "grad_norm": 1.117253065109253,
      "learning_rate": 1.5856149084178954e-05,
      "loss": 3.1839,
      "step": 107000
    },
    {
      "epoch": 10.291020486310549,
      "grad_norm": 1.128348708152771,
      "learning_rate": 1.5696598378964837e-05,
      "loss": 3.1906,
      "step": 107500
    },
    {
      "epoch": 10.338885697874785,
      "grad_norm": 1.0678489208221436,
      "learning_rate": 1.5537047673750717e-05,
      "loss": 3.1717,
      "step": 108000
    },
    {
      "epoch": 10.38675090943902,
      "grad_norm": 1.072418451309204,
      "learning_rate": 1.53774969685366e-05,
      "loss": 3.1754,
      "step": 108500
    },
    {
      "epoch": 10.434616121003256,
      "grad_norm": 1.086338758468628,
      "learning_rate": 1.5217946263322485e-05,
      "loss": 3.1702,
      "step": 109000
    },
    {
      "epoch": 10.48248133256749,
      "grad_norm": 1.0860265493392944,
      "learning_rate": 1.5058395558108368e-05,
      "loss": 3.1755,
      "step": 109500
    },
    {
      "epoch": 10.530346544131724,
      "grad_norm": 1.0454214811325073,
      "learning_rate": 1.489884485289425e-05,
      "loss": 3.1605,
      "step": 110000
    },
    {
      "epoch": 10.57821175569596,
      "grad_norm": 1.0717874765396118,
      "learning_rate": 1.4739294147680133e-05,
      "loss": 3.1486,
      "step": 110500
    },
    {
      "epoch": 10.626076967260195,
      "grad_norm": 1.1019850969314575,
      "learning_rate": 1.4579743442466015e-05,
      "loss": 3.169,
      "step": 111000
    },
    {
      "epoch": 10.673942178824431,
      "grad_norm": 1.0683563947677612,
      "learning_rate": 1.44201927372519e-05,
      "loss": 3.1641,
      "step": 111500
    },
    {
      "epoch": 10.721807390388665,
      "grad_norm": 1.0630834102630615,
      "learning_rate": 1.4260642032037783e-05,
      "loss": 3.1684,
      "step": 112000
    },
    {
      "epoch": 10.7696726019529,
      "grad_norm": 1.0388649702072144,
      "learning_rate": 1.4101091326823665e-05,
      "loss": 3.1575,
      "step": 112500
    },
    {
      "epoch": 10.817537813517136,
      "grad_norm": 1.0644524097442627,
      "learning_rate": 1.3941540621609548e-05,
      "loss": 3.1647,
      "step": 113000
    },
    {
      "epoch": 10.86540302508137,
      "grad_norm": 1.1413675546646118,
      "learning_rate": 1.3781989916395433e-05,
      "loss": 3.1675,
      "step": 113500
    },
    {
      "epoch": 10.913268236645607,
      "grad_norm": 1.1140718460083008,
      "learning_rate": 1.3622439211181314e-05,
      "loss": 3.1592,
      "step": 114000
    },
    {
      "epoch": 10.961133448209841,
      "grad_norm": 1.0123062133789062,
      "learning_rate": 1.3462888505967198e-05,
      "loss": 3.1623,
      "step": 114500
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.38862544034815083,
      "eval_loss": 3.247739553451538,
      "eval_runtime": 351.7508,
      "eval_samples_per_second": 21.046,
      "eval_steps_per_second": 1.316,
      "step": 114906
    },
    {
      "epoch": 11.008998659774075,
      "grad_norm": 1.0549602508544922,
      "learning_rate": 1.330333780075308e-05,
      "loss": 3.1663,
      "step": 115000
    },
    {
      "epoch": 11.056863871338312,
      "grad_norm": 1.1148481369018555,
      "learning_rate": 1.3143787095538963e-05,
      "loss": 3.147,
      "step": 115500
    },
    {
      "epoch": 11.104729082902546,
      "grad_norm": 1.037546157836914,
      "learning_rate": 1.2984236390324848e-05,
      "loss": 3.1553,
      "step": 116000
    },
    {
      "epoch": 11.152594294466782,
      "grad_norm": 1.0984928607940674,
      "learning_rate": 1.2824685685110729e-05,
      "loss": 3.1557,
      "step": 116500
    },
    {
      "epoch": 11.200459506031017,
      "grad_norm": 1.0544558763504028,
      "learning_rate": 1.2665134979896612e-05,
      "loss": 3.1553,
      "step": 117000
    },
    {
      "epoch": 11.248324717595251,
      "grad_norm": 1.090158462524414,
      "learning_rate": 1.2505584274682494e-05,
      "loss": 3.1506,
      "step": 117500
    },
    {
      "epoch": 11.296189929159487,
      "grad_norm": 1.0947335958480835,
      "learning_rate": 1.2346033569468377e-05,
      "loss": 3.1463,
      "step": 118000
    },
    {
      "epoch": 11.344055140723722,
      "grad_norm": 0.9977598786354065,
      "learning_rate": 1.218648286425426e-05,
      "loss": 3.15,
      "step": 118500
    },
    {
      "epoch": 11.391920352287958,
      "grad_norm": 1.0011844635009766,
      "learning_rate": 1.2026932159040144e-05,
      "loss": 3.1558,
      "step": 119000
    },
    {
      "epoch": 11.439785563852192,
      "grad_norm": 1.05132257938385,
      "learning_rate": 1.1867381453826027e-05,
      "loss": 3.1568,
      "step": 119500
    },
    {
      "epoch": 11.487650775416427,
      "grad_norm": 0.9936347007751465,
      "learning_rate": 1.1707830748611909e-05,
      "loss": 3.1388,
      "step": 120000
    },
    {
      "epoch": 11.535515986980663,
      "grad_norm": 1.0718574523925781,
      "learning_rate": 1.1548280043397792e-05,
      "loss": 3.1674,
      "step": 120500
    },
    {
      "epoch": 11.583381198544897,
      "grad_norm": 1.2275400161743164,
      "learning_rate": 1.1388729338183675e-05,
      "loss": 3.1458,
      "step": 121000
    },
    {
      "epoch": 11.631246410109133,
      "grad_norm": 1.1291002035140991,
      "learning_rate": 1.1229178632969558e-05,
      "loss": 3.151,
      "step": 121500
    },
    {
      "epoch": 11.679111621673368,
      "grad_norm": 0.9682615995407104,
      "learning_rate": 1.1069627927755442e-05,
      "loss": 3.1471,
      "step": 122000
    },
    {
      "epoch": 11.726976833237602,
      "grad_norm": 1.033508539199829,
      "learning_rate": 1.0910077222541325e-05,
      "loss": 3.1492,
      "step": 122500
    },
    {
      "epoch": 11.774842044801838,
      "grad_norm": 1.1144520044326782,
      "learning_rate": 1.0750526517327206e-05,
      "loss": 3.1601,
      "step": 123000
    },
    {
      "epoch": 11.822707256366073,
      "grad_norm": 1.1509275436401367,
      "learning_rate": 1.059097581211309e-05,
      "loss": 3.1487,
      "step": 123500
    },
    {
      "epoch": 11.870572467930309,
      "grad_norm": 1.108750343322754,
      "learning_rate": 1.0431425106898973e-05,
      "loss": 3.141,
      "step": 124000
    },
    {
      "epoch": 11.918437679494543,
      "grad_norm": 1.1095489263534546,
      "learning_rate": 1.0271874401684856e-05,
      "loss": 3.1352,
      "step": 124500
    },
    {
      "epoch": 11.966302891058778,
      "grad_norm": 1.0765643119812012,
      "learning_rate": 1.011232369647074e-05,
      "loss": 3.1569,
      "step": 125000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.3899440254928222,
      "eval_loss": 3.2352681159973145,
      "eval_runtime": 353.863,
      "eval_samples_per_second": 20.921,
      "eval_steps_per_second": 1.308,
      "step": 125352
    },
    {
      "epoch": 12.014168102623014,
      "grad_norm": 1.058240294456482,
      "learning_rate": 9.952772991256621e-06,
      "loss": 3.1371,
      "step": 125500
    },
    {
      "epoch": 12.062033314187248,
      "grad_norm": 1.0825666189193726,
      "learning_rate": 9.793222286042506e-06,
      "loss": 3.1334,
      "step": 126000
    },
    {
      "epoch": 12.109898525751484,
      "grad_norm": 1.050256609916687,
      "learning_rate": 9.633671580828388e-06,
      "loss": 3.142,
      "step": 126500
    },
    {
      "epoch": 12.157763737315719,
      "grad_norm": 1.1124509572982788,
      "learning_rate": 9.474120875614271e-06,
      "loss": 3.1454,
      "step": 127000
    },
    {
      "epoch": 12.205628948879953,
      "grad_norm": 1.041452407836914,
      "learning_rate": 9.314570170400154e-06,
      "loss": 3.1351,
      "step": 127500
    },
    {
      "epoch": 12.25349416044419,
      "grad_norm": 1.0738871097564697,
      "learning_rate": 9.155019465186036e-06,
      "loss": 3.1227,
      "step": 128000
    },
    {
      "epoch": 12.301359372008424,
      "grad_norm": 1.072388768196106,
      "learning_rate": 8.99546875997192e-06,
      "loss": 3.1376,
      "step": 128500
    },
    {
      "epoch": 12.34922458357266,
      "grad_norm": 1.105535626411438,
      "learning_rate": 8.835918054757802e-06,
      "loss": 3.1351,
      "step": 129000
    },
    {
      "epoch": 12.397089795136894,
      "grad_norm": 1.0452914237976074,
      "learning_rate": 8.676367349543686e-06,
      "loss": 3.1384,
      "step": 129500
    },
    {
      "epoch": 12.444955006701129,
      "grad_norm": 1.1002734899520874,
      "learning_rate": 8.516816644329569e-06,
      "loss": 3.1433,
      "step": 130000
    },
    {
      "epoch": 12.492820218265365,
      "grad_norm": 1.0648058652877808,
      "learning_rate": 8.35726593911545e-06,
      "loss": 3.1307,
      "step": 130500
    },
    {
      "epoch": 12.5406854298296,
      "grad_norm": 1.2617812156677246,
      "learning_rate": 8.197715233901335e-06,
      "loss": 3.1317,
      "step": 131000
    },
    {
      "epoch": 12.588550641393836,
      "grad_norm": 1.0360037088394165,
      "learning_rate": 8.038164528687217e-06,
      "loss": 3.1296,
      "step": 131500
    },
    {
      "epoch": 12.63641585295807,
      "grad_norm": 1.1165320873260498,
      "learning_rate": 7.8786138234731e-06,
      "loss": 3.1421,
      "step": 132000
    },
    {
      "epoch": 12.684281064522306,
      "grad_norm": 1.061864972114563,
      "learning_rate": 7.719063118258983e-06,
      "loss": 3.1448,
      "step": 132500
    },
    {
      "epoch": 12.73214627608654,
      "grad_norm": 1.1287447214126587,
      "learning_rate": 7.559512413044866e-06,
      "loss": 3.136,
      "step": 133000
    },
    {
      "epoch": 12.780011487650775,
      "grad_norm": 1.1351341009140015,
      "learning_rate": 7.399961707830749e-06,
      "loss": 3.1378,
      "step": 133500
    },
    {
      "epoch": 12.827876699215011,
      "grad_norm": 1.14214289188385,
      "learning_rate": 7.2404110026166316e-06,
      "loss": 3.1375,
      "step": 134000
    },
    {
      "epoch": 12.875741910779245,
      "grad_norm": 1.096230149269104,
      "learning_rate": 7.080860297402516e-06,
      "loss": 3.1336,
      "step": 134500
    },
    {
      "epoch": 12.92360712234348,
      "grad_norm": 1.103021502494812,
      "learning_rate": 6.921309592188398e-06,
      "loss": 3.1283,
      "step": 135000
    },
    {
      "epoch": 12.971472333907716,
      "grad_norm": 1.0734912157058716,
      "learning_rate": 6.7617588869742805e-06,
      "loss": 3.132,
      "step": 135500
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.3910163497427597,
      "eval_loss": 3.226067066192627,
      "eval_runtime": 351.8351,
      "eval_samples_per_second": 21.041,
      "eval_steps_per_second": 1.316,
      "step": 135798
    },
    {
      "epoch": 13.01933754547195,
      "grad_norm": 1.1110934019088745,
      "learning_rate": 6.602208181760164e-06,
      "loss": 3.1304,
      "step": 136000
    },
    {
      "epoch": 13.067202757036187,
      "grad_norm": 1.0855175256729126,
      "learning_rate": 6.442657476546046e-06,
      "loss": 3.1294,
      "step": 136500
    },
    {
      "epoch": 13.115067968600421,
      "grad_norm": 0.9932899475097656,
      "learning_rate": 6.28310677133193e-06,
      "loss": 3.1226,
      "step": 137000
    },
    {
      "epoch": 13.162933180164657,
      "grad_norm": 1.1955283880233765,
      "learning_rate": 6.123556066117813e-06,
      "loss": 3.1255,
      "step": 137500
    },
    {
      "epoch": 13.210798391728892,
      "grad_norm": 1.1328649520874023,
      "learning_rate": 5.964005360903695e-06,
      "loss": 3.1388,
      "step": 138000
    },
    {
      "epoch": 13.258663603293126,
      "grad_norm": 1.0772650241851807,
      "learning_rate": 5.8044546556895784e-06,
      "loss": 3.1244,
      "step": 138500
    },
    {
      "epoch": 13.306528814857362,
      "grad_norm": 1.0329416990280151,
      "learning_rate": 5.644903950475462e-06,
      "loss": 3.1175,
      "step": 139000
    },
    {
      "epoch": 13.354394026421597,
      "grad_norm": 1.0305006504058838,
      "learning_rate": 5.485353245261345e-06,
      "loss": 3.1397,
      "step": 139500
    },
    {
      "epoch": 13.402259237985833,
      "grad_norm": 1.0859792232513428,
      "learning_rate": 5.3258025400472265e-06,
      "loss": 3.1285,
      "step": 140000
    },
    {
      "epoch": 13.450124449550067,
      "grad_norm": 1.013626217842102,
      "learning_rate": 5.16625183483311e-06,
      "loss": 3.138,
      "step": 140500
    },
    {
      "epoch": 13.497989661114302,
      "grad_norm": 1.0302331447601318,
      "learning_rate": 5.006701129618993e-06,
      "loss": 3.1201,
      "step": 141000
    },
    {
      "epoch": 13.545854872678538,
      "grad_norm": 1.1003705263137817,
      "learning_rate": 4.847150424404876e-06,
      "loss": 3.1195,
      "step": 141500
    },
    {
      "epoch": 13.593720084242772,
      "grad_norm": 1.0157201290130615,
      "learning_rate": 4.68759971919076e-06,
      "loss": 3.13,
      "step": 142000
    },
    {
      "epoch": 13.641585295807008,
      "grad_norm": 1.1753654479980469,
      "learning_rate": 4.528049013976642e-06,
      "loss": 3.1201,
      "step": 142500
    },
    {
      "epoch": 13.689450507371243,
      "grad_norm": 1.0680103302001953,
      "learning_rate": 4.3684983087625244e-06,
      "loss": 3.1234,
      "step": 143000
    },
    {
      "epoch": 13.737315718935477,
      "grad_norm": 1.0224796533584595,
      "learning_rate": 4.208947603548408e-06,
      "loss": 3.1262,
      "step": 143500
    },
    {
      "epoch": 13.785180930499713,
      "grad_norm": 1.0585968494415283,
      "learning_rate": 4.049396898334291e-06,
      "loss": 3.1159,
      "step": 144000
    },
    {
      "epoch": 13.833046142063948,
      "grad_norm": 1.0231329202651978,
      "learning_rate": 3.889846193120173e-06,
      "loss": 3.1197,
      "step": 144500
    },
    {
      "epoch": 13.880911353628184,
      "grad_norm": 1.1415258646011353,
      "learning_rate": 3.7302954879060567e-06,
      "loss": 3.1344,
      "step": 145000
    },
    {
      "epoch": 13.928776565192418,
      "grad_norm": 1.0792592763900757,
      "learning_rate": 3.57074478269194e-06,
      "loss": 3.1201,
      "step": 145500
    },
    {
      "epoch": 13.976641776756653,
      "grad_norm": 1.0325241088867188,
      "learning_rate": 3.4111940774778224e-06,
      "loss": 3.1211,
      "step": 146000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.3915306586891341,
      "eval_loss": 3.220797300338745,
      "eval_runtime": 352.0051,
      "eval_samples_per_second": 21.031,
      "eval_steps_per_second": 1.315,
      "step": 146244
    },
    {
      "epoch": 14.024506988320889,
      "grad_norm": 1.1887177228927612,
      "learning_rate": 3.251643372263705e-06,
      "loss": 3.1188,
      "step": 146500
    },
    {
      "epoch": 14.072372199885123,
      "grad_norm": 1.0874779224395752,
      "learning_rate": 3.0920926670495885e-06,
      "loss": 3.1226,
      "step": 147000
    },
    {
      "epoch": 14.12023741144936,
      "grad_norm": 1.055733561515808,
      "learning_rate": 2.9325419618354713e-06,
      "loss": 3.1222,
      "step": 147500
    },
    {
      "epoch": 14.168102623013594,
      "grad_norm": 1.1001226902008057,
      "learning_rate": 2.7729912566213546e-06,
      "loss": 3.1215,
      "step": 148000
    },
    {
      "epoch": 14.215967834577828,
      "grad_norm": 0.9945481419563293,
      "learning_rate": 2.6134405514072374e-06,
      "loss": 3.1183,
      "step": 148500
    },
    {
      "epoch": 14.263833046142064,
      "grad_norm": 1.1555193662643433,
      "learning_rate": 2.4538898461931203e-06,
      "loss": 3.1351,
      "step": 149000
    },
    {
      "epoch": 14.311698257706299,
      "grad_norm": 1.1480242013931274,
      "learning_rate": 2.2943391409790035e-06,
      "loss": 3.123,
      "step": 149500
    },
    {
      "epoch": 14.359563469270535,
      "grad_norm": 0.9716194272041321,
      "learning_rate": 2.134788435764886e-06,
      "loss": 3.1275,
      "step": 150000
    },
    {
      "epoch": 14.40742868083477,
      "grad_norm": 1.1331024169921875,
      "learning_rate": 1.9752377305507692e-06,
      "loss": 3.1179,
      "step": 150500
    },
    {
      "epoch": 14.455293892399004,
      "grad_norm": 1.02994966506958,
      "learning_rate": 1.815687025336652e-06,
      "loss": 3.1082,
      "step": 151000
    },
    {
      "epoch": 14.50315910396324,
      "grad_norm": 1.0279475450515747,
      "learning_rate": 1.656136320122535e-06,
      "loss": 3.117,
      "step": 151500
    },
    {
      "epoch": 14.551024315527474,
      "grad_norm": 1.064375638961792,
      "learning_rate": 1.496585614908418e-06,
      "loss": 3.1293,
      "step": 152000
    },
    {
      "epoch": 14.59888952709171,
      "grad_norm": 1.0934786796569824,
      "learning_rate": 1.337034909694301e-06,
      "loss": 3.1074,
      "step": 152500
    },
    {
      "epoch": 14.646754738655945,
      "grad_norm": 1.0857374668121338,
      "learning_rate": 1.1774842044801839e-06,
      "loss": 3.1248,
      "step": 153000
    },
    {
      "epoch": 14.69461995022018,
      "grad_norm": 1.1022164821624756,
      "learning_rate": 1.0179334992660667e-06,
      "loss": 3.1193,
      "step": 153500
    },
    {
      "epoch": 14.742485161784415,
      "grad_norm": 1.1129704713821411,
      "learning_rate": 8.583827940519499e-07,
      "loss": 3.1268,
      "step": 154000
    },
    {
      "epoch": 14.79035037334865,
      "grad_norm": 1.0595066547393799,
      "learning_rate": 6.988320888378327e-07,
      "loss": 3.1063,
      "step": 154500
    },
    {
      "epoch": 14.838215584912886,
      "grad_norm": 1.085557460784912,
      "learning_rate": 5.392813836237157e-07,
      "loss": 3.1088,
      "step": 155000
    },
    {
      "epoch": 14.88608079647712,
      "grad_norm": 1.0873228311538696,
      "learning_rate": 3.797306784095986e-07,
      "loss": 3.106,
      "step": 155500
    },
    {
      "epoch": 14.933946008041355,
      "grad_norm": 1.0385653972625732,
      "learning_rate": 2.2017997319548153e-07,
      "loss": 3.1104,
      "step": 156000
    },
    {
      "epoch": 14.981811219605591,
      "grad_norm": 1.0836658477783203,
      "learning_rate": 6.062926798136448e-08,
      "loss": 3.12,
      "step": 156500
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.3918296048905697,
      "eval_loss": 3.218470335006714,
      "eval_runtime": 351.9635,
      "eval_samples_per_second": 21.033,
      "eval_steps_per_second": 1.315,
      "step": 156690
    },
    {
      "epoch": 15.0,
      "step": 156690,
      "total_flos": 9.8251813748736e+17,
      "train_loss": 1.4742094889696864,
      "train_runtime": 76707.1992,
      "train_samples_per_second": 24.51,
      "train_steps_per_second": 2.043
    }
  ],
  "logging_steps": 500,
  "max_steps": 156690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "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": 9.8251813748736e+17,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}