{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 3500,
  "global_step": 27876,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007174630506528914,
      "grad_norm": 1.9597077369689941,
      "learning_rate": 5.9139784946236566e-06,
      "loss": 3.6869268798828125,
      "step": 100
    },
    {
      "epoch": 0.014349261013057828,
      "grad_norm": 0.9890213012695312,
      "learning_rate": 1.1887694145758663e-05,
      "loss": 0.7186312866210938,
      "step": 200
    },
    {
      "epoch": 0.02152389151958674,
      "grad_norm": 1.899065375328064,
      "learning_rate": 1.7861409796893668e-05,
      "loss": 0.36630226135253907,
      "step": 300
    },
    {
      "epoch": 0.028698522026115655,
      "grad_norm": 2.6302435398101807,
      "learning_rate": 2.3835125448028675e-05,
      "loss": 0.303447265625,
      "step": 400
    },
    {
      "epoch": 0.035873152532644566,
      "grad_norm": 1.8647853136062622,
      "learning_rate": 2.9808841099163682e-05,
      "loss": 0.29018028259277345,
      "step": 500
    },
    {
      "epoch": 0.04304778303917348,
      "grad_norm": 2.053941249847412,
      "learning_rate": 3.578255675029869e-05,
      "loss": 0.2563763618469238,
      "step": 600
    },
    {
      "epoch": 0.05022241354570239,
      "grad_norm": 2.2911510467529297,
      "learning_rate": 4.1756272401433697e-05,
      "loss": 0.22638999938964843,
      "step": 700
    },
    {
      "epoch": 0.05739704405223131,
      "grad_norm": 2.0702409744262695,
      "learning_rate": 4.77299880525687e-05,
      "loss": 0.2134855270385742,
      "step": 800
    },
    {
      "epoch": 0.06457167455876023,
      "grad_norm": 2.6445889472961426,
      "learning_rate": 4.999935135061931e-05,
      "loss": 0.19725006103515624,
      "step": 900
    },
    {
      "epoch": 0.07174630506528913,
      "grad_norm": 2.1372897624969482,
      "learning_rate": 4.999557161150297e-05,
      "loss": 0.19304241180419923,
      "step": 1000
    },
    {
      "epoch": 0.07892093557181805,
      "grad_norm": 2.9511172771453857,
      "learning_rate": 4.998841762685811e-05,
      "loss": 0.1907896041870117,
      "step": 1100
    },
    {
      "epoch": 0.08609556607834697,
      "grad_norm": 2.054408550262451,
      "learning_rate": 4.997789036242782e-05,
      "loss": 0.17207706451416016,
      "step": 1200
    },
    {
      "epoch": 0.09327019658487588,
      "grad_norm": 1.5231208801269531,
      "learning_rate": 4.9963991239326835e-05,
      "loss": 0.1723944091796875,
      "step": 1300
    },
    {
      "epoch": 0.10044482709140479,
      "grad_norm": 2.8132405281066895,
      "learning_rate": 4.994672213384968e-05,
      "loss": 0.162562255859375,
      "step": 1400
    },
    {
      "epoch": 0.1076194575979337,
      "grad_norm": 1.6805394887924194,
      "learning_rate": 4.992608537721735e-05,
      "loss": 0.15793445587158203,
      "step": 1500
    },
    {
      "epoch": 0.11479408810446262,
      "grad_norm": 1.9100033044815063,
      "learning_rate": 4.990208375526269e-05,
      "loss": 0.15448735237121583,
      "step": 1600
    },
    {
      "epoch": 0.12196871861099154,
      "grad_norm": 2.079188346862793,
      "learning_rate": 4.987472050805424e-05,
      "loss": 0.14562548637390138,
      "step": 1700
    },
    {
      "epoch": 0.12914334911752046,
      "grad_norm": 2.061568021774292,
      "learning_rate": 4.98439993294589e-05,
      "loss": 0.15155485153198242,
      "step": 1800
    },
    {
      "epoch": 0.13631797962404937,
      "grad_norm": 1.8127555847167969,
      "learning_rate": 4.980992436664327e-05,
      "loss": 0.14210834503173828,
      "step": 1900
    },
    {
      "epoch": 0.14349261013057826,
      "grad_norm": 2.088500738143921,
      "learning_rate": 4.977250021951378e-05,
      "loss": 0.14464487075805665,
      "step": 2000
    },
    {
      "epoch": 0.15066724063710718,
      "grad_norm": 1.117219090461731,
      "learning_rate": 4.9731731940095774e-05,
      "loss": 0.12635778427124023,
      "step": 2100
    },
    {
      "epoch": 0.1578418711436361,
      "grad_norm": 1.9483405351638794,
      "learning_rate": 4.9687625031851485e-05,
      "loss": 0.1291893196105957,
      "step": 2200
    },
    {
      "epoch": 0.16501650165016502,
      "grad_norm": 2.7950611114501953,
      "learning_rate": 4.964018544893714e-05,
      "loss": 0.12796722412109374,
      "step": 2300
    },
    {
      "epoch": 0.17219113215669393,
      "grad_norm": 1.4632441997528076,
      "learning_rate": 4.9589419595399125e-05,
      "loss": 0.13129316329956053,
      "step": 2400
    },
    {
      "epoch": 0.17936576266322285,
      "grad_norm": 1.621394157409668,
      "learning_rate": 4.9535334324309535e-05,
      "loss": 0.12220394134521484,
      "step": 2500
    },
    {
      "epoch": 0.18654039316975177,
      "grad_norm": 1.1649906635284424,
      "learning_rate": 4.947793693684105e-05,
      "loss": 0.12823129653930665,
      "step": 2600
    },
    {
      "epoch": 0.19371502367628068,
      "grad_norm": 1.6868650913238525,
      "learning_rate": 4.9417235181281274e-05,
      "loss": 0.12181013107299804,
      "step": 2700
    },
    {
      "epoch": 0.20088965418280957,
      "grad_norm": 1.9284274578094482,
      "learning_rate": 4.9353237251986815e-05,
      "loss": 0.11575795173645019,
      "step": 2800
    },
    {
      "epoch": 0.2080642846893385,
      "grad_norm": 1.866542100906372,
      "learning_rate": 4.928595178827707e-05,
      "loss": 0.11822829246520997,
      "step": 2900
    },
    {
      "epoch": 0.2152389151958674,
      "grad_norm": 2.062695026397705,
      "learning_rate": 4.921538787326799e-05,
      "loss": 0.11779312133789062,
      "step": 3000
    },
    {
      "epoch": 0.22241354570239633,
      "grad_norm": 1.6521011590957642,
      "learning_rate": 4.914155503264589e-05,
      "loss": 0.12041742324829102,
      "step": 3100
    },
    {
      "epoch": 0.22958817620892524,
      "grad_norm": 1.9350285530090332,
      "learning_rate": 4.906446323338159e-05,
      "loss": 0.11249205589294434,
      "step": 3200
    },
    {
      "epoch": 0.23676280671545416,
      "grad_norm": 1.6741434335708618,
      "learning_rate": 4.898412288238484e-05,
      "loss": 0.11555561065673828,
      "step": 3300
    },
    {
      "epoch": 0.24393743722198308,
      "grad_norm": 1.971616268157959,
      "learning_rate": 4.89005448250996e-05,
      "loss": 0.11045341491699219,
      "step": 3400
    },
    {
      "epoch": 0.25111206772851197,
      "grad_norm": 1.606021523475647,
      "learning_rate": 4.881374034403983e-05,
      "loss": 0.11493149757385254,
      "step": 3500
    },
    {
      "epoch": 0.25111206772851197,
      "eval_loss": 0.10866887867450714,
      "eval_runtime": 149.2661,
      "eval_samples_per_second": 66.988,
      "eval_steps_per_second": 2.097,
      "step": 3500
    },
    {
      "epoch": 0.2582866982350409,
      "grad_norm": 2.926435947418213,
      "learning_rate": 4.87237211572665e-05,
      "loss": 0.10716718673706055,
      "step": 3600
    },
    {
      "epoch": 0.2654613287415698,
      "grad_norm": 1.6348823308944702,
      "learning_rate": 4.8630499416805694e-05,
      "loss": 0.10849567413330079,
      "step": 3700
    },
    {
      "epoch": 0.27263595924809875,
      "grad_norm": 2.4208922386169434,
      "learning_rate": 4.85340877070082e-05,
      "loss": 0.09899539947509765,
      "step": 3800
    },
    {
      "epoch": 0.27981058975462764,
      "grad_norm": 1.559791088104248,
      "learning_rate": 4.8434499042850654e-05,
      "loss": 0.11465280532836913,
      "step": 3900
    },
    {
      "epoch": 0.2869852202611565,
      "grad_norm": 1.2442468404769897,
      "learning_rate": 4.8331746868178626e-05,
      "loss": 0.10655641555786133,
      "step": 4000
    },
    {
      "epoch": 0.29415985076768547,
      "grad_norm": 2.1395716667175293,
      "learning_rate": 4.8225845053891785e-05,
      "loss": 0.113049955368042,
      "step": 4100
    },
    {
      "epoch": 0.30133448127421436,
      "grad_norm": 0.8998376131057739,
      "learning_rate": 4.8116807896071404e-05,
      "loss": 0.10541413307189941,
      "step": 4200
    },
    {
      "epoch": 0.3085091117807433,
      "grad_norm": 1.936471700668335,
      "learning_rate": 4.800465011405049e-05,
      "loss": 0.10073881149291992,
      "step": 4300
    },
    {
      "epoch": 0.3156837422872722,
      "grad_norm": 1.1070330142974854,
      "learning_rate": 4.7889386848426755e-05,
      "loss": 0.1081316089630127,
      "step": 4400
    },
    {
      "epoch": 0.32285837279380114,
      "grad_norm": 1.4007771015167236,
      "learning_rate": 4.7771033659018756e-05,
      "loss": 0.10138772010803222,
      "step": 4500
    },
    {
      "epoch": 0.33003300330033003,
      "grad_norm": 1.3453152179718018,
      "learning_rate": 4.7649606522765365e-05,
      "loss": 0.09714742660522462,
      "step": 4600
    },
    {
      "epoch": 0.3372076338068589,
      "grad_norm": 1.3695811033248901,
      "learning_rate": 4.7525121831569054e-05,
      "loss": 0.1012276554107666,
      "step": 4700
    },
    {
      "epoch": 0.34438226431338786,
      "grad_norm": 1.5470471382141113,
      "learning_rate": 4.739759639008301e-05,
      "loss": 0.09525470733642578,
      "step": 4800
    },
    {
      "epoch": 0.35155689481991675,
      "grad_norm": 1.3758277893066406,
      "learning_rate": 4.726704741344268e-05,
      "loss": 0.10572370529174804,
      "step": 4900
    },
    {
      "epoch": 0.3587315253264457,
      "grad_norm": 1.269905686378479,
      "learning_rate": 4.71334925249418e-05,
      "loss": 0.10359272956848145,
      "step": 5000
    },
    {
      "epoch": 0.3659061558329746,
      "grad_norm": 1.5478543043136597,
      "learning_rate": 4.6996949753653384e-05,
      "loss": 0.0949331283569336,
      "step": 5100
    },
    {
      "epoch": 0.37308078633950353,
      "grad_norm": 1.3514728546142578,
      "learning_rate": 4.685743753199587e-05,
      "loss": 0.09247732162475586,
      "step": 5200
    },
    {
      "epoch": 0.3802554168460324,
      "grad_norm": 1.3794364929199219,
      "learning_rate": 4.6714974693244925e-05,
      "loss": 0.0969287109375,
      "step": 5300
    },
    {
      "epoch": 0.38743004735256137,
      "grad_norm": 1.268118977546692,
      "learning_rate": 4.656958046899099e-05,
      "loss": 0.09732013702392578,
      "step": 5400
    },
    {
      "epoch": 0.39460467785909026,
      "grad_norm": 2.6610381603240967,
      "learning_rate": 4.642127448654322e-05,
      "loss": 0.09799112319946289,
      "step": 5500
    },
    {
      "epoch": 0.40177930836561915,
      "grad_norm": 0.5645221471786499,
      "learning_rate": 4.627007676627986e-05,
      "loss": 0.08652667999267578,
      "step": 5600
    },
    {
      "epoch": 0.4089539388721481,
      "grad_norm": 1.276166319847107,
      "learning_rate": 4.611600771894562e-05,
      "loss": 0.09176987648010254,
      "step": 5700
    },
    {
      "epoch": 0.416128569378677,
      "grad_norm": 1.6661149263381958,
      "learning_rate": 4.595908814289641e-05,
      "loss": 0.0958226203918457,
      "step": 5800
    },
    {
      "epoch": 0.4233031998852059,
      "grad_norm": 2.3982605934143066,
      "learning_rate": 4.579933922129163e-05,
      "loss": 0.09369464874267579,
      "step": 5900
    },
    {
      "epoch": 0.4304778303917348,
      "grad_norm": 1.6622225046157837,
      "learning_rate": 4.5636782519234585e-05,
      "loss": 0.09838510513305664,
      "step": 6000
    },
    {
      "epoch": 0.43765246089826376,
      "grad_norm": 1.0257179737091064,
      "learning_rate": 4.547143998086139e-05,
      "loss": 0.09203416824340821,
      "step": 6100
    },
    {
      "epoch": 0.44482709140479265,
      "grad_norm": 1.2619930505752563,
      "learning_rate": 4.530333392637856e-05,
      "loss": 0.09255922317504883,
      "step": 6200
    },
    {
      "epoch": 0.45200172191132154,
      "grad_norm": 2.0735652446746826,
      "learning_rate": 4.513248704904999e-05,
      "loss": 0.0923232650756836,
      "step": 6300
    },
    {
      "epoch": 0.4591763524178505,
      "grad_norm": 1.2585643529891968,
      "learning_rate": 4.4958922412133483e-05,
      "loss": 0.08465845108032227,
      "step": 6400
    },
    {
      "epoch": 0.4663509829243794,
      "grad_norm": 1.9748129844665527,
      "learning_rate": 4.478266344576734e-05,
      "loss": 0.08935056686401367,
      "step": 6500
    },
    {
      "epoch": 0.4735256134309083,
      "grad_norm": 1.265183925628662,
      "learning_rate": 4.460373394380747e-05,
      "loss": 0.08919876098632812,
      "step": 6600
    },
    {
      "epoch": 0.4807002439374372,
      "grad_norm": 1.4090503454208374,
      "learning_rate": 4.4422158060615346e-05,
      "loss": 0.09309642791748046,
      "step": 6700
    },
    {
      "epoch": 0.48787487444396616,
      "grad_norm": 0.7966151237487793,
      "learning_rate": 4.423796030779731e-05,
      "loss": 0.08681303024291992,
      "step": 6800
    },
    {
      "epoch": 0.49504950495049505,
      "grad_norm": 1.725354790687561,
      "learning_rate": 4.405116555089571e-05,
      "loss": 0.0832568359375,
      "step": 6900
    },
    {
      "epoch": 0.5022241354570239,
      "grad_norm": 1.0632050037384033,
      "learning_rate": 4.386179900603212e-05,
      "loss": 0.08722116470336914,
      "step": 7000
    },
    {
      "epoch": 0.5022241354570239,
      "eval_loss": 0.0860578715801239,
      "eval_runtime": 149.0191,
      "eval_samples_per_second": 67.099,
      "eval_steps_per_second": 2.1,
      "step": 7000
    },
    {
      "epoch": 0.5093987659635528,
      "grad_norm": 1.5935779809951782,
      "learning_rate": 4.366988623650342e-05,
      "loss": 0.08583742141723633,
      "step": 7100
    },
    {
      "epoch": 0.5165733964700818,
      "grad_norm": 1.6912461519241333,
      "learning_rate": 4.347545314933083e-05,
      "loss": 0.0800228214263916,
      "step": 7200
    },
    {
      "epoch": 0.5237480269766107,
      "grad_norm": 1.405499815940857,
      "learning_rate": 4.327852599176267e-05,
      "loss": 0.08579876899719238,
      "step": 7300
    },
    {
      "epoch": 0.5309226574831396,
      "grad_norm": 1.492091417312622,
      "learning_rate": 4.3079131347731145e-05,
      "loss": 0.08771139144897461,
      "step": 7400
    },
    {
      "epoch": 0.5380972879896685,
      "grad_norm": 1.7305445671081543,
      "learning_rate": 4.2877296134263646e-05,
      "loss": 0.09152100563049316,
      "step": 7500
    },
    {
      "epoch": 0.5452719184961975,
      "grad_norm": 0.8501316905021667,
      "learning_rate": 4.267304759784916e-05,
      "loss": 0.07824550151824951,
      "step": 7600
    },
    {
      "epoch": 0.5524465490027264,
      "grad_norm": 1.1813442707061768,
      "learning_rate": 4.246641331076015e-05,
      "loss": 0.09139251708984375,
      "step": 7700
    },
    {
      "epoch": 0.5596211795092553,
      "grad_norm": 1.1864808797836304,
      "learning_rate": 4.2257421167330455e-05,
      "loss": 0.08820032119750977,
      "step": 7800
    },
    {
      "epoch": 0.5667958100157842,
      "grad_norm": 0.9204437732696533,
      "learning_rate": 4.2046099380189764e-05,
      "loss": 0.0799638557434082,
      "step": 7900
    },
    {
      "epoch": 0.573970440522313,
      "grad_norm": 0.8114761114120483,
      "learning_rate": 4.183247647645504e-05,
      "loss": 0.08707637786865234,
      "step": 8000
    },
    {
      "epoch": 0.581145071028842,
      "grad_norm": 1.7211333513259888,
      "learning_rate": 4.16165812938796e-05,
      "loss": 0.084498291015625,
      "step": 8100
    },
    {
      "epoch": 0.5883197015353709,
      "grad_norm": 1.3709509372711182,
      "learning_rate": 4.1398442976960125e-05,
      "loss": 0.07480822563171387,
      "step": 8200
    },
    {
      "epoch": 0.5954943320418998,
      "grad_norm": 0.9520241618156433,
      "learning_rate": 4.117809097300244e-05,
      "loss": 0.08572841644287109,
      "step": 8300
    },
    {
      "epoch": 0.6026689625484287,
      "grad_norm": 1.6982910633087158,
      "learning_rate": 4.095555502814616e-05,
      "loss": 0.08172292709350586,
      "step": 8400
    },
    {
      "epoch": 0.6098435930549577,
      "grad_norm": 1.441841959953308,
      "learning_rate": 4.07308651833493e-05,
      "loss": 0.08631341934204101,
      "step": 8500
    },
    {
      "epoch": 0.6170182235614866,
      "grad_norm": 1.2558258771896362,
      "learning_rate": 4.050405177033285e-05,
      "loss": 0.07842582702636719,
      "step": 8600
    },
    {
      "epoch": 0.6241928540680155,
      "grad_norm": 1.5181397199630737,
      "learning_rate": 4.027514540748621e-05,
      "loss": 0.07550670146942139,
      "step": 8700
    },
    {
      "epoch": 0.6313674845745444,
      "grad_norm": 1.7649136781692505,
      "learning_rate": 4.004417699573387e-05,
      "loss": 0.08165624618530273,
      "step": 8800
    },
    {
      "epoch": 0.6385421150810733,
      "grad_norm": 1.27147376537323,
      "learning_rate": 3.9811177714364026e-05,
      "loss": 0.08272171020507812,
      "step": 8900
    },
    {
      "epoch": 0.6457167455876023,
      "grad_norm": 1.3053189516067505,
      "learning_rate": 3.957617901681952e-05,
      "loss": 0.08126266479492188,
      "step": 9000
    },
    {
      "epoch": 0.6528913760941312,
      "grad_norm": 1.7846314907073975,
      "learning_rate": 3.933921262645186e-05,
      "loss": 0.07549744129180908,
      "step": 9100
    },
    {
      "epoch": 0.6600660066006601,
      "grad_norm": 1.3431315422058105,
      "learning_rate": 3.910031053223877e-05,
      "loss": 0.07644423484802246,
      "step": 9200
    },
    {
      "epoch": 0.667240637107189,
      "grad_norm": 1.2648569345474243,
      "learning_rate": 3.885950498446584e-05,
      "loss": 0.07915946960449219,
      "step": 9300
    },
    {
      "epoch": 0.6744152676137178,
      "grad_norm": 1.6945096254348755,
      "learning_rate": 3.861682849037297e-05,
      "loss": 0.0817841911315918,
      "step": 9400
    },
    {
      "epoch": 0.6815898981202468,
      "grad_norm": 1.2732951641082764,
      "learning_rate": 3.8372313809766094e-05,
      "loss": 0.07225199222564697,
      "step": 9500
    },
    {
      "epoch": 0.6887645286267757,
      "grad_norm": 0.1465909630060196,
      "learning_rate": 3.812599395059479e-05,
      "loss": 0.08223398208618164,
      "step": 9600
    },
    {
      "epoch": 0.6959391591333046,
      "grad_norm": 1.6093218326568604,
      "learning_rate": 3.787790216449643e-05,
      "loss": 0.0781820011138916,
      "step": 9700
    },
    {
      "epoch": 0.7031137896398335,
      "grad_norm": 1.4999920129776,
      "learning_rate": 3.7628071942307464e-05,
      "loss": 0.07415190696716309,
      "step": 9800
    },
    {
      "epoch": 0.7102884201463625,
      "grad_norm": 1.3299458026885986,
      "learning_rate": 3.737653700954226e-05,
      "loss": 0.07337451934814453,
      "step": 9900
    },
    {
      "epoch": 0.7174630506528914,
      "grad_norm": 1.0557976961135864,
      "learning_rate": 3.712333132184045e-05,
      "loss": 0.07776775360107421,
      "step": 10000
    },
    {
      "epoch": 0.7246376811594203,
      "grad_norm": 1.1716148853302002,
      "learning_rate": 3.6868489060383127e-05,
      "loss": 0.0788364315032959,
      "step": 10100
    },
    {
      "epoch": 0.7318123116659492,
      "grad_norm": 1.0893652439117432,
      "learning_rate": 3.661204462727857e-05,
      "loss": 0.07858343124389648,
      "step": 10200
    },
    {
      "epoch": 0.7389869421724781,
      "grad_norm": 1.5502619743347168,
      "learning_rate": 3.635403264091821e-05,
      "loss": 0.07226887226104736,
      "step": 10300
    },
    {
      "epoch": 0.7461615726790071,
      "grad_norm": 0.7753734588623047,
      "learning_rate": 3.609448793130331e-05,
      "loss": 0.0813060188293457,
      "step": 10400
    },
    {
      "epoch": 0.753336203185536,
      "grad_norm": 1.374515175819397,
      "learning_rate": 3.583344553534319e-05,
      "loss": 0.07035542488098144,
      "step": 10500
    },
    {
      "epoch": 0.753336203185536,
      "eval_loss": 0.07575151324272156,
      "eval_runtime": 148.3569,
      "eval_samples_per_second": 67.398,
      "eval_steps_per_second": 2.11,
      "step": 10500
    },
    {
      "epoch": 0.7605108336920648,
      "grad_norm": 1.334509015083313,
      "learning_rate": 3.5570940692125446e-05,
      "loss": 0.07447919368743897,
      "step": 10600
    },
    {
      "epoch": 0.7676854641985937,
      "grad_norm": 1.2184432744979858,
      "learning_rate": 3.5307008838158876e-05,
      "loss": 0.0723937702178955,
      "step": 10700
    },
    {
      "epoch": 0.7748600947051227,
      "grad_norm": 0.9506904482841492,
      "learning_rate": 3.504168560258979e-05,
      "loss": 0.07003112316131592,
      "step": 10800
    },
    {
      "epoch": 0.7820347252116516,
      "grad_norm": 1.3243482112884521,
      "learning_rate": 3.477500680239229e-05,
      "loss": 0.07373955726623535,
      "step": 10900
    },
    {
      "epoch": 0.7892093557181805,
      "grad_norm": 1.0208053588867188,
      "learning_rate": 3.450700843753324e-05,
      "loss": 0.06773545265197754,
      "step": 11000
    },
    {
      "epoch": 0.7963839862247094,
      "grad_norm": 0.8675241470336914,
      "learning_rate": 3.423772668611241e-05,
      "loss": 0.07413051605224609,
      "step": 11100
    },
    {
      "epoch": 0.8035586167312383,
      "grad_norm": 1.206604242324829,
      "learning_rate": 3.3967197899478755e-05,
      "loss": 0.07240287780761719,
      "step": 11200
    },
    {
      "epoch": 0.8107332472377673,
      "grad_norm": 1.6686036586761475,
      "learning_rate": 3.369545859732315e-05,
      "loss": 0.07362545967102051,
      "step": 11300
    },
    {
      "epoch": 0.8179078777442962,
      "grad_norm": 1.1721315383911133,
      "learning_rate": 3.342254546274849e-05,
      "loss": 0.06512235641479493,
      "step": 11400
    },
    {
      "epoch": 0.8250825082508251,
      "grad_norm": 1.0817862749099731,
      "learning_rate": 3.3148495337317674e-05,
      "loss": 0.07151537895202636,
      "step": 11500
    },
    {
      "epoch": 0.832257138757354,
      "grad_norm": 1.1494413614273071,
      "learning_rate": 3.287334521608028e-05,
      "loss": 0.07197807788848877,
      "step": 11600
    },
    {
      "epoch": 0.839431769263883,
      "grad_norm": 1.1358568668365479,
      "learning_rate": 3.2597132242578365e-05,
      "loss": 0.06177080154418945,
      "step": 11700
    },
    {
      "epoch": 0.8466063997704119,
      "grad_norm": 1.3649406433105469,
      "learning_rate": 3.2319893703832435e-05,
      "loss": 0.06873297691345215,
      "step": 11800
    },
    {
      "epoch": 0.8537810302769407,
      "grad_norm": 1.5489788055419922,
      "learning_rate": 3.204166702530786e-05,
      "loss": 0.07243677616119384,
      "step": 11900
    },
    {
      "epoch": 0.8609556607834696,
      "grad_norm": 0.9062689542770386,
      "learning_rate": 3.1762489765862684e-05,
      "loss": 0.07555675983428956,
      "step": 12000
    },
    {
      "epoch": 0.8681302912899985,
      "grad_norm": 1.188604474067688,
      "learning_rate": 3.1482399612677416e-05,
      "loss": 0.0659021806716919,
      "step": 12100
    },
    {
      "epoch": 0.8753049217965275,
      "grad_norm": 1.053107738494873,
      "learning_rate": 3.12014343761675e-05,
      "loss": 0.06588594913482666,
      "step": 12200
    },
    {
      "epoch": 0.8824795523030564,
      "grad_norm": 1.7348968982696533,
      "learning_rate": 3.0919631984879156e-05,
      "loss": 0.0732758092880249,
      "step": 12300
    },
    {
      "epoch": 0.8896541828095853,
      "grad_norm": 1.1809568405151367,
      "learning_rate": 3.063703048036925e-05,
      "loss": 0.07246605396270751,
      "step": 12400
    },
    {
      "epoch": 0.8968288133161142,
      "grad_norm": 1.3747830390930176,
      "learning_rate": 3.0353668012069904e-05,
      "loss": 0.07011266231536865,
      "step": 12500
    },
    {
      "epoch": 0.9040034438226431,
      "grad_norm": 0.8150450587272644,
      "learning_rate": 3.00695828321386e-05,
      "loss": 0.07035545825958252,
      "step": 12600
    },
    {
      "epoch": 0.9111780743291721,
      "grad_norm": 0.7044129967689514,
      "learning_rate": 2.978481329029437e-05,
      "loss": 0.06655097007751465,
      "step": 12700
    },
    {
      "epoch": 0.918352704835701,
      "grad_norm": 1.1079609394073486,
      "learning_rate": 2.9499397828640812e-05,
      "loss": 0.06627664089202881,
      "step": 12800
    },
    {
      "epoch": 0.9255273353422299,
      "grad_norm": 0.9070701003074646,
      "learning_rate": 2.921337497647662e-05,
      "loss": 0.06677388191223145,
      "step": 12900
    },
    {
      "epoch": 0.9327019658487588,
      "grad_norm": 1.0263826847076416,
      "learning_rate": 2.892678334509445e-05,
      "loss": 0.07018782615661621,
      "step": 13000
    },
    {
      "epoch": 0.9398765963552878,
      "grad_norm": 1.2620368003845215,
      "learning_rate": 2.863966162256855e-05,
      "loss": 0.06896811485290527,
      "step": 13100
    },
    {
      "epoch": 0.9470512268618166,
      "grad_norm": 1.3818238973617554,
      "learning_rate": 2.8352048568532174e-05,
      "loss": 0.07202363967895507,
      "step": 13200
    },
    {
      "epoch": 0.9542258573683455,
      "grad_norm": 1.0401277542114258,
      "learning_rate": 2.806398300894522e-05,
      "loss": 0.06861380577087402,
      "step": 13300
    },
    {
      "epoch": 0.9614004878748744,
      "grad_norm": 1.2165743112564087,
      "learning_rate": 2.777550383085304e-05,
      "loss": 0.06642416000366211,
      "step": 13400
    },
    {
      "epoch": 0.9685751183814033,
      "grad_norm": 0.9264984726905823,
      "learning_rate": 2.7486649977136865e-05,
      "loss": 0.06832408428192138,
      "step": 13500
    },
    {
      "epoch": 0.9757497488879323,
      "grad_norm": 1.5410667657852173,
      "learning_rate": 2.7197460441256795e-05,
      "loss": 0.06425261497497559,
      "step": 13600
    },
    {
      "epoch": 0.9829243793944612,
      "grad_norm": 0.5595086216926575,
      "learning_rate": 2.690797426198791e-05,
      "loss": 0.07268239974975586,
      "step": 13700
    },
    {
      "epoch": 0.9900990099009901,
      "grad_norm": 1.125222086906433,
      "learning_rate": 2.6618230518150316e-05,
      "loss": 0.06679293632507324,
      "step": 13800
    },
    {
      "epoch": 0.997273640407519,
      "grad_norm": 1.3064922094345093,
      "learning_rate": 2.632826832333368e-05,
      "loss": 0.06414957523345947,
      "step": 13900
    },
    {
      "epoch": 1.0044482709140479,
      "grad_norm": 1.1672838926315308,
      "learning_rate": 2.6038126820617227e-05,
      "loss": 0.059987783432006836,
      "step": 14000
    },
    {
      "epoch": 1.0044482709140479,
      "eval_loss": 0.06724809110164642,
      "eval_runtime": 147.0466,
      "eval_samples_per_second": 67.999,
      "eval_steps_per_second": 2.129,
      "step": 14000
    },
    {
      "epoch": 1.0116229014205769,
      "grad_norm": 1.3335286378860474,
      "learning_rate": 2.5747845177285586e-05,
      "loss": 0.053536911010742184,
      "step": 14100
    },
    {
      "epoch": 1.0187975319271056,
      "grad_norm": 1.2691210508346558,
      "learning_rate": 2.545746257954151e-05,
      "loss": 0.05340280532836914,
      "step": 14200
    },
    {
      "epoch": 1.0259721624336346,
      "grad_norm": 1.1815472841262817,
      "learning_rate": 2.5167018227215962e-05,
      "loss": 0.051975679397583005,
      "step": 14300
    },
    {
      "epoch": 1.0331467929401636,
      "grad_norm": 1.6674917936325073,
      "learning_rate": 2.4876551328476386e-05,
      "loss": 0.055572891235351564,
      "step": 14400
    },
    {
      "epoch": 1.0403214234466924,
      "grad_norm": 1.1760724782943726,
      "learning_rate": 2.4586101094533852e-05,
      "loss": 0.05527249336242676,
      "step": 14500
    },
    {
      "epoch": 1.0474960539532214,
      "grad_norm": 0.7432792782783508,
      "learning_rate": 2.4295706734349798e-05,
      "loss": 0.05280819892883301,
      "step": 14600
    },
    {
      "epoch": 1.0546706844597504,
      "grad_norm": 1.031694769859314,
      "learning_rate": 2.4005407449343025e-05,
      "loss": 0.05506425380706787,
      "step": 14700
    },
    {
      "epoch": 1.0618453149662792,
      "grad_norm": 1.248627781867981,
      "learning_rate": 2.37152424280978e-05,
      "loss": 0.05392072677612305,
      "step": 14800
    },
    {
      "epoch": 1.0690199454728082,
      "grad_norm": 1.211077094078064,
      "learning_rate": 2.3425250841073644e-05,
      "loss": 0.05422006607055664,
      "step": 14900
    },
    {
      "epoch": 1.076194575979337,
      "grad_norm": 1.2443134784698486,
      "learning_rate": 2.3135471835317495e-05,
      "loss": 0.048745040893554685,
      "step": 15000
    },
    {
      "epoch": 1.083369206485866,
      "grad_norm": 2.079069137573242,
      "learning_rate": 2.2845944529179193e-05,
      "loss": 0.05105120658874512,
      "step": 15100
    },
    {
      "epoch": 1.090543836992395,
      "grad_norm": 1.8662784099578857,
      "learning_rate": 2.2556708007030704e-05,
      "loss": 0.04701868534088135,
      "step": 15200
    },
    {
      "epoch": 1.0977184674989238,
      "grad_norm": 1.5977221727371216,
      "learning_rate": 2.2267801313989947e-05,
      "loss": 0.05204158782958984,
      "step": 15300
    },
    {
      "epoch": 1.1048930980054528,
      "grad_norm": 1.2682287693023682,
      "learning_rate": 2.1979263450650022e-05,
      "loss": 0.05470413208007813,
      "step": 15400
    },
    {
      "epoch": 1.1120677285119815,
      "grad_norm": 0.9770413637161255,
      "learning_rate": 2.169113336781429e-05,
      "loss": 0.05012000560760498,
      "step": 15500
    },
    {
      "epoch": 1.1192423590185105,
      "grad_norm": 0.8554680347442627,
      "learning_rate": 2.1403449961238318e-05,
      "loss": 0.05378786087036133,
      "step": 15600
    },
    {
      "epoch": 1.1264169895250395,
      "grad_norm": 0.9937000274658203,
      "learning_rate": 2.111625206637919e-05,
      "loss": 0.05037656307220459,
      "step": 15700
    },
    {
      "epoch": 1.1335916200315683,
      "grad_norm": 1.2181024551391602,
      "learning_rate": 2.082957845315292e-05,
      "loss": 0.05694090843200684,
      "step": 15800
    },
    {
      "epoch": 1.1407662505380973,
      "grad_norm": 1.92032790184021,
      "learning_rate": 2.0543467820700818e-05,
      "loss": 0.057715382575988766,
      "step": 15900
    },
    {
      "epoch": 1.147940881044626,
      "grad_norm": 2.086174249649048,
      "learning_rate": 2.0257958792165304e-05,
      "loss": 0.05871417999267578,
      "step": 16000
    },
    {
      "epoch": 1.155115511551155,
      "grad_norm": 0.8641619682312012,
      "learning_rate": 1.9973089909476027e-05,
      "loss": 0.04839668273925781,
      "step": 16100
    },
    {
      "epoch": 1.162290142057684,
      "grad_norm": 1.3958889245986938,
      "learning_rate": 1.9688899628146977e-05,
      "loss": 0.048901238441467286,
      "step": 16200
    },
    {
      "epoch": 1.1694647725642129,
      "grad_norm": 1.048177719116211,
      "learning_rate": 1.9405426312085217e-05,
      "loss": 0.04399550437927246,
      "step": 16300
    },
    {
      "epoch": 1.1766394030707419,
      "grad_norm": 0.913404107093811,
      "learning_rate": 1.9122708228411956e-05,
      "loss": 0.0568333625793457,
      "step": 16400
    },
    {
      "epoch": 1.1838140335772707,
      "grad_norm": 1.2032625675201416,
      "learning_rate": 1.884078354229681e-05,
      "loss": 0.053453073501586915,
      "step": 16500
    },
    {
      "epoch": 1.1909886640837997,
      "grad_norm": 0.983741283416748,
      "learning_rate": 1.8559690311805666e-05,
      "loss": 0.05076848030090332,
      "step": 16600
    },
    {
      "epoch": 1.1981632945903287,
      "grad_norm": 0.9596076607704163,
      "learning_rate": 1.827946648276313e-05,
      "loss": 0.04985082626342773,
      "step": 16700
    },
    {
      "epoch": 1.2053379250968574,
      "grad_norm": 1.3797062635421753,
      "learning_rate": 1.800014988363005e-05,
      "loss": 0.05314798355102539,
      "step": 16800
    },
    {
      "epoch": 1.2125125556033864,
      "grad_norm": 0.9229831099510193,
      "learning_rate": 1.7721778220396906e-05,
      "loss": 0.055236310958862306,
      "step": 16900
    },
    {
      "epoch": 1.2196871861099154,
      "grad_norm": 1.1550999879837036,
      "learning_rate": 1.7444389071493794e-05,
      "loss": 0.057296619415283204,
      "step": 17000
    },
    {
      "epoch": 1.2268618166164442,
      "grad_norm": 1.525516152381897,
      "learning_rate": 1.71680198827175e-05,
      "loss": 0.050670623779296875,
      "step": 17100
    },
    {
      "epoch": 1.2340364471229732,
      "grad_norm": 2.2197818756103516,
      "learning_rate": 1.689270796217659e-05,
      "loss": 0.04868894577026367,
      "step": 17200
    },
    {
      "epoch": 1.241211077629502,
      "grad_norm": 1.0739034414291382,
      "learning_rate": 1.6618490475255048e-05,
      "loss": 0.04822113513946533,
      "step": 17300
    },
    {
      "epoch": 1.248385708136031,
      "grad_norm": 0.9117043018341064,
      "learning_rate": 1.6345404439595187e-05,
      "loss": 0.05410886287689209,
      "step": 17400
    },
    {
      "epoch": 1.2555603386425598,
      "grad_norm": 1.5217725038528442,
      "learning_rate": 1.6073486720100476e-05,
      "loss": 0.051833877563476564,
      "step": 17500
    },
    {
      "epoch": 1.2555603386425598,
      "eval_loss": 0.06531932950019836,
      "eval_runtime": 149.0603,
      "eval_samples_per_second": 67.08,
      "eval_steps_per_second": 2.1,
      "step": 17500
    },
    {
      "epoch": 1.2627349691490888,
      "grad_norm": 0.882745087146759,
      "learning_rate": 1.580277402395902e-05,
      "loss": 0.046353416442871095,
      "step": 17600
    },
    {
      "epoch": 1.2699095996556178,
      "grad_norm": 1.408157467842102,
      "learning_rate": 1.553330289568834e-05,
      "loss": 0.048478012084960935,
      "step": 17700
    },
    {
      "epoch": 1.2770842301621466,
      "grad_norm": 1.1981309652328491,
      "learning_rate": 1.5265109712202065e-05,
      "loss": 0.05171732902526856,
      "step": 17800
    },
    {
      "epoch": 1.2842588606686756,
      "grad_norm": 1.35702383518219,
      "learning_rate": 1.4998230677899284e-05,
      "loss": 0.05537195205688476,
      "step": 17900
    },
    {
      "epoch": 1.2914334911752046,
      "grad_norm": 1.3542907238006592,
      "learning_rate": 1.4732701819777178e-05,
      "loss": 0.0555442476272583,
      "step": 18000
    },
    {
      "epoch": 1.2986081216817333,
      "grad_norm": 1.0912405252456665,
      "learning_rate": 1.4468558982567632e-05,
      "loss": 0.051614513397216795,
      "step": 18100
    },
    {
      "epoch": 1.3057827521882623,
      "grad_norm": 0.49335864186286926,
      "learning_rate": 1.4205837823898343e-05,
      "loss": 0.056519789695739744,
      "step": 18200
    },
    {
      "epoch": 1.3129573826947913,
      "grad_norm": 1.6202818155288696,
      "learning_rate": 1.394457380947938e-05,
      "loss": 0.04868396759033203,
      "step": 18300
    },
    {
      "epoch": 1.3201320132013201,
      "grad_norm": 1.1292474269866943,
      "learning_rate": 1.3684802208315406e-05,
      "loss": 0.04953648090362549,
      "step": 18400
    },
    {
      "epoch": 1.3273066437078491,
      "grad_norm": 1.542555809020996,
      "learning_rate": 1.3426558087944683e-05,
      "loss": 0.055932865142822266,
      "step": 18500
    },
    {
      "epoch": 1.334481274214378,
      "grad_norm": 1.0680972337722778,
      "learning_rate": 1.316987630970513e-05,
      "loss": 0.051041364669799805,
      "step": 18600
    },
    {
      "epoch": 1.341655904720907,
      "grad_norm": 1.8201940059661865,
      "learning_rate": 1.2914791524028269e-05,
      "loss": 0.05028592109680176,
      "step": 18700
    },
    {
      "epoch": 1.3488305352274357,
      "grad_norm": 1.8100615739822388,
      "learning_rate": 1.2661338165761638e-05,
      "loss": 0.049348244667053225,
      "step": 18800
    },
    {
      "epoch": 1.3560051657339647,
      "grad_norm": 0.9206331372261047,
      "learning_rate": 1.2409550449520287e-05,
      "loss": 0.05312205314636231,
      "step": 18900
    },
    {
      "epoch": 1.3631797962404937,
      "grad_norm": 1.585785984992981,
      "learning_rate": 1.2159462365068033e-05,
      "loss": 0.0519862174987793,
      "step": 19000
    },
    {
      "epoch": 1.3703544267470225,
      "grad_norm": 0.9856590628623962,
      "learning_rate": 1.1911107672729085e-05,
      "loss": 0.0482769775390625,
      "step": 19100
    },
    {
      "epoch": 1.3775290572535515,
      "grad_norm": 1.1745953559875488,
      "learning_rate": 1.1664519898830523e-05,
      "loss": 0.04800943374633789,
      "step": 19200
    },
    {
      "epoch": 1.3847036877600805,
      "grad_norm": 1.2835592031478882,
      "learning_rate": 1.1419732331176536e-05,
      "loss": 0.04779685974121094,
      "step": 19300
    },
    {
      "epoch": 1.3918783182666092,
      "grad_norm": 1.6603394746780396,
      "learning_rate": 1.1176778014554787e-05,
      "loss": 0.04568360328674317,
      "step": 19400
    },
    {
      "epoch": 1.3990529487731382,
      "grad_norm": 1.2128928899765015,
      "learning_rate": 1.0935689746275474e-05,
      "loss": 0.04861447334289551,
      "step": 19500
    },
    {
      "epoch": 1.406227579279667,
      "grad_norm": 1.5717813968658447,
      "learning_rate": 1.0696500071743996e-05,
      "loss": 0.046369132995605467,
      "step": 19600
    },
    {
      "epoch": 1.413402209786196,
      "grad_norm": 0.9322656989097595,
      "learning_rate": 1.0459241280067514e-05,
      "loss": 0.04947755813598633,
      "step": 19700
    },
    {
      "epoch": 1.4205768402927248,
      "grad_norm": 1.0949965715408325,
      "learning_rate": 1.022394539969605e-05,
      "loss": 0.04265530586242676,
      "step": 19800
    },
    {
      "epoch": 1.4277514707992538,
      "grad_norm": 1.7212597131729126,
      "learning_rate": 9.99064419409897e-06,
      "loss": 0.05672057151794434,
      "step": 19900
    },
    {
      "epoch": 1.4349261013057828,
      "grad_norm": 0.6662020087242126,
      "learning_rate": 9.759369157476974e-06,
      "loss": 0.046635923385620115,
      "step": 20000
    },
    {
      "epoch": 1.4421007318123116,
      "grad_norm": 0.8181309700012207,
      "learning_rate": 9.530151510510688e-06,
      "loss": 0.050442686080932615,
      "step": 20100
    },
    {
      "epoch": 1.4492753623188406,
      "grad_norm": 1.605729103088379,
      "learning_rate": 9.303022196146041e-06,
      "loss": 0.04867252349853515,
      "step": 20200
    },
    {
      "epoch": 1.4564499928253696,
      "grad_norm": 1.7314263582229614,
      "learning_rate": 9.078011875417098e-06,
      "loss": 0.05573555946350098,
      "step": 20300
    },
    {
      "epoch": 1.4636246233318984,
      "grad_norm": 0.6051461100578308,
      "learning_rate": 8.855150923307115e-06,
      "loss": 0.051524810791015625,
      "step": 20400
    },
    {
      "epoch": 1.4707992538384274,
      "grad_norm": 1.4942373037338257,
      "learning_rate": 8.634469424648032e-06,
      "loss": 0.04892472743988037,
      "step": 20500
    },
    {
      "epoch": 1.4779738843449564,
      "grad_norm": 0.7416260838508606,
      "learning_rate": 8.4159971700592e-06,
      "loss": 0.047811784744262696,
      "step": 20600
    },
    {
      "epoch": 1.4851485148514851,
      "grad_norm": 1.162663459777832,
      "learning_rate": 8.199763651925926e-06,
      "loss": 0.046135540008544924,
      "step": 20700
    },
    {
      "epoch": 1.4923231453580141,
      "grad_norm": 1.2641029357910156,
      "learning_rate": 7.98579806041809e-06,
      "loss": 0.04572174549102783,
      "step": 20800
    },
    {
      "epoch": 1.499497775864543,
      "grad_norm": 1.0548759698867798,
      "learning_rate": 7.774129279549668e-06,
      "loss": 0.0437416410446167,
      "step": 20900
    },
    {
      "epoch": 1.506672406371072,
      "grad_norm": 1.6857285499572754,
      "learning_rate": 7.5647858832796485e-06,
      "loss": 0.04923810958862305,
      "step": 21000
    },
    {
      "epoch": 1.506672406371072,
      "eval_loss": 0.06312502175569534,
      "eval_runtime": 148.9513,
      "eval_samples_per_second": 67.129,
      "eval_steps_per_second": 2.101,
      "step": 21000
    },
    {
      "epoch": 1.5138470368776007,
      "grad_norm": 1.4431586265563965,
      "learning_rate": 7.357796131654627e-06,
      "loss": 0.049534149169921875,
      "step": 21100
    },
    {
      "epoch": 1.5210216673841297,
      "grad_norm": 0.6273611187934875,
      "learning_rate": 7.153187966993937e-06,
      "loss": 0.05015117645263672,
      "step": 21200
    },
    {
      "epoch": 1.5281962978906587,
      "grad_norm": 1.2017526626586914,
      "learning_rate": 6.950989010117606e-06,
      "loss": 0.04353977203369141,
      "step": 21300
    },
    {
      "epoch": 1.5353709283971875,
      "grad_norm": 0.8428612947463989,
      "learning_rate": 6.751226556617673e-06,
      "loss": 0.04460501670837402,
      "step": 21400
    },
    {
      "epoch": 1.5425455589037165,
      "grad_norm": 0.5900961756706238,
      "learning_rate": 6.553927573173541e-06,
      "loss": 0.047952399253845215,
      "step": 21500
    },
    {
      "epoch": 1.5497201894102455,
      "grad_norm": 1.2832746505737305,
      "learning_rate": 6.359118693911561e-06,
      "loss": 0.050580711364746095,
      "step": 21600
    },
    {
      "epoch": 1.5568948199167743,
      "grad_norm": 0.7824345827102661,
      "learning_rate": 6.166826216809612e-06,
      "loss": 0.04548311710357666,
      "step": 21700
    },
    {
      "epoch": 1.5640694504233033,
      "grad_norm": 1.127432942390442,
      "learning_rate": 5.977076100147094e-06,
      "loss": 0.04847379207611084,
      "step": 21800
    },
    {
      "epoch": 1.5712440809298323,
      "grad_norm": 0.7834171056747437,
      "learning_rate": 5.789893959000642e-06,
      "loss": 0.04097845554351807,
      "step": 21900
    },
    {
      "epoch": 1.578418711436361,
      "grad_norm": 2.322270154953003,
      "learning_rate": 5.605305061786303e-06,
      "loss": 0.04809154510498047,
      "step": 22000
    },
    {
      "epoch": 1.5855933419428898,
      "grad_norm": 1.1508792638778687,
      "learning_rate": 5.4233343268484446e-06,
      "loss": 0.04745537757873535,
      "step": 22100
    },
    {
      "epoch": 1.5927679724494188,
      "grad_norm": 0.9398218989372253,
      "learning_rate": 5.244006319095931e-06,
      "loss": 0.043388500213623046,
      "step": 22200
    },
    {
      "epoch": 1.5999426029559478,
      "grad_norm": 2.1089656352996826,
      "learning_rate": 5.06734524668602e-06,
      "loss": 0.04768535137176513,
      "step": 22300
    },
    {
      "epoch": 1.6071172334624766,
      "grad_norm": 1.4436591863632202,
      "learning_rate": 4.893374957756411e-06,
      "loss": 0.0458716344833374,
      "step": 22400
    },
    {
      "epoch": 1.6142918639690056,
      "grad_norm": 1.0683164596557617,
      "learning_rate": 4.722118937205891e-06,
      "loss": 0.04412764549255371,
      "step": 22500
    },
    {
      "epoch": 1.6214664944755346,
      "grad_norm": 1.6844348907470703,
      "learning_rate": 4.553600303524036e-06,
      "loss": 0.05096036911010742,
      "step": 22600
    },
    {
      "epoch": 1.6286411249820634,
      "grad_norm": 1.3367619514465332,
      "learning_rate": 4.387841805670334e-06,
      "loss": 0.04533133506774902,
      "step": 22700
    },
    {
      "epoch": 1.6358157554885924,
      "grad_norm": 1.6877501010894775,
      "learning_rate": 4.2248658200032485e-06,
      "loss": 0.04637275695800781,
      "step": 22800
    },
    {
      "epoch": 1.6429903859951214,
      "grad_norm": 1.3961207866668701,
      "learning_rate": 4.0646943472595234e-06,
      "loss": 0.04682679176330566,
      "step": 22900
    },
    {
      "epoch": 1.6501650165016502,
      "grad_norm": 1.3452887535095215,
      "learning_rate": 3.907349009584238e-06,
      "loss": 0.04511795997619629,
      "step": 23000
    },
    {
      "epoch": 1.657339647008179,
      "grad_norm": 1.2425464391708374,
      "learning_rate": 3.7528510476119467e-06,
      "loss": 0.04864484786987305,
      "step": 23100
    },
    {
      "epoch": 1.6645142775147082,
      "grad_norm": 1.0512237548828125,
      "learning_rate": 3.6012213175993292e-06,
      "loss": 0.04659308910369873,
      "step": 23200
    },
    {
      "epoch": 1.671688908021237,
      "grad_norm": 1.0648466348648071,
      "learning_rate": 3.452480288609722e-06,
      "loss": 0.0448155403137207,
      "step": 23300
    },
    {
      "epoch": 1.6788635385277657,
      "grad_norm": 0.40967854857444763,
      "learning_rate": 3.3066480397499296e-06,
      "loss": 0.04373236656188965,
      "step": 23400
    },
    {
      "epoch": 1.6860381690342947,
      "grad_norm": 1.0134800672531128,
      "learning_rate": 3.163744257459661e-06,
      "loss": 0.04981218814849853,
      "step": 23500
    },
    {
      "epoch": 1.6932127995408237,
      "grad_norm": 1.8333792686462402,
      "learning_rate": 3.0237882328539996e-06,
      "loss": 0.04597721576690674,
      "step": 23600
    },
    {
      "epoch": 1.7003874300473525,
      "grad_norm": 0.918961763381958,
      "learning_rate": 2.8867988591192023e-06,
      "loss": 0.0487159252166748,
      "step": 23700
    },
    {
      "epoch": 1.7075620605538815,
      "grad_norm": 1.5625665187835693,
      "learning_rate": 2.752794628962263e-06,
      "loss": 0.046385250091552734,
      "step": 23800
    },
    {
      "epoch": 1.7147366910604105,
      "grad_norm": 0.5554918050765991,
      "learning_rate": 2.6217936321144916e-06,
      "loss": 0.044139723777771,
      "step": 23900
    },
    {
      "epoch": 1.7219113215669393,
      "grad_norm": 0.8911906480789185,
      "learning_rate": 2.4938135528895123e-06,
      "loss": 0.04505068778991699,
      "step": 24000
    },
    {
      "epoch": 1.7290859520734683,
      "grad_norm": 1.5743203163146973,
      "learning_rate": 2.3688716677960273e-06,
      "loss": 0.04274391651153565,
      "step": 24100
    },
    {
      "epoch": 1.7362605825799973,
      "grad_norm": 0.848380446434021,
      "learning_rate": 2.2469848432055436e-06,
      "loss": 0.04197495460510254,
      "step": 24200
    },
    {
      "epoch": 1.743435213086526,
      "grad_norm": 1.9898673295974731,
      "learning_rate": 2.128169533075558e-06,
      "loss": 0.04518091678619385,
      "step": 24300
    },
    {
      "epoch": 1.7506098435930548,
      "grad_norm": 1.210407018661499,
      "learning_rate": 2.012441776728363e-06,
      "loss": 0.05085911750793457,
      "step": 24400
    },
    {
      "epoch": 1.7577844740995838,
      "grad_norm": 1.1385959386825562,
      "learning_rate": 1.899817196685838e-06,
      "loss": 0.04531978607177734,
      "step": 24500
    },
    {
      "epoch": 1.7577844740995838,
      "eval_loss": 0.06059190258383751,
      "eval_runtime": 150.0007,
      "eval_samples_per_second": 66.66,
      "eval_steps_per_second": 2.087,
      "step": 24500
    },
    {
      "epoch": 1.7649591046061128,
      "grad_norm": 1.4091085195541382,
      "learning_rate": 1.790310996560507e-06,
      "loss": 0.04507139682769776,
      "step": 24600
    },
    {
      "epoch": 1.7721337351126416,
      "grad_norm": 1.1383436918258667,
      "learning_rate": 1.6839379590031457e-06,
      "loss": 0.04703217029571533,
      "step": 24700
    },
    {
      "epoch": 1.7793083656191706,
      "grad_norm": 0.31827428936958313,
      "learning_rate": 1.580712443707219e-06,
      "loss": 0.04364439964294434,
      "step": 24800
    },
    {
      "epoch": 1.7864829961256996,
      "grad_norm": 0.7810333967208862,
      "learning_rate": 1.4806483854704206e-06,
      "loss": 0.04452046871185303,
      "step": 24900
    },
    {
      "epoch": 1.7936576266322284,
      "grad_norm": 1.9447439908981323,
      "learning_rate": 1.38375929231353e-06,
      "loss": 0.047868146896362304,
      "step": 25000
    },
    {
      "epoch": 1.8008322571387574,
      "grad_norm": 2.4073574542999268,
      "learning_rate": 1.290058243656958e-06,
      "loss": 0.046537981033325196,
      "step": 25100
    },
    {
      "epoch": 1.8080068876452864,
      "grad_norm": 1.283727765083313,
      "learning_rate": 1.1995578885550861e-06,
      "loss": 0.045397920608520506,
      "step": 25200
    },
    {
      "epoch": 1.8151815181518152,
      "grad_norm": 1.7792607545852661,
      "learning_rate": 1.1122704439887055e-06,
      "loss": 0.04585344314575195,
      "step": 25300
    },
    {
      "epoch": 1.822356148658344,
      "grad_norm": 1.0718834400177002,
      "learning_rate": 1.0282076932158352e-06,
      "loss": 0.048363561630249026,
      "step": 25400
    },
    {
      "epoch": 1.8295307791648732,
      "grad_norm": 0.8162381052970886,
      "learning_rate": 9.473809841810427e-07,
      "loss": 0.04825215339660645,
      "step": 25500
    },
    {
      "epoch": 1.836705409671402,
      "grad_norm": 1.5586066246032715,
      "learning_rate": 8.698012279835188e-07,
      "loss": 0.049013128280639646,
      "step": 25600
    },
    {
      "epoch": 1.8438800401779307,
      "grad_norm": 0.5607184171676636,
      "learning_rate": 7.954788974042005e-07,
      "loss": 0.04248374938964844,
      "step": 25700
    },
    {
      "epoch": 1.8510546706844597,
      "grad_norm": 1.474769115447998,
      "learning_rate": 7.244240254919488e-07,
      "loss": 0.047613582611083984,
      "step": 25800
    },
    {
      "epoch": 1.8582293011909887,
      "grad_norm": 0.7129104137420654,
      "learning_rate": 6.566462042091964e-07,
      "loss": 0.052542953491210936,
      "step": 25900
    },
    {
      "epoch": 1.8654039316975175,
      "grad_norm": 1.426215648651123,
      "learning_rate": 5.921545831370862e-07,
      "loss": 0.055969982147216796,
      "step": 26000
    },
    {
      "epoch": 1.8725785622040465,
      "grad_norm": 1.2196667194366455,
      "learning_rate": 5.309578682403004e-07,
      "loss": 0.04260244369506836,
      "step": 26100
    },
    {
      "epoch": 1.8797531927105755,
      "grad_norm": 1.5276991128921509,
      "learning_rate": 4.7306432069185723e-07,
      "loss": 0.051242551803588866,
      "step": 26200
    },
    {
      "epoch": 1.8869278232171043,
      "grad_norm": 1.1367182731628418,
      "learning_rate": 4.1848175575787965e-07,
      "loss": 0.04391949653625488,
      "step": 26300
    },
    {
      "epoch": 1.8941024537236333,
      "grad_norm": 1.011906385421753,
      "learning_rate": 3.672175417425733e-07,
      "loss": 0.045631518363952635,
      "step": 26400
    },
    {
      "epoch": 1.9012770842301623,
      "grad_norm": 0.9637262225151062,
      "learning_rate": 3.1927859899358313e-07,
      "loss": 0.04020712375640869,
      "step": 26500
    },
    {
      "epoch": 1.908451714736691,
      "grad_norm": 0.7773815393447876,
      "learning_rate": 2.746713989677602e-07,
      "loss": 0.0514189863204956,
      "step": 26600
    },
    {
      "epoch": 1.9156263452432198,
      "grad_norm": 0.5528236627578735,
      "learning_rate": 2.334019633575657e-07,
      "loss": 0.045413846969604495,
      "step": 26700
    },
    {
      "epoch": 1.9228009757497488,
      "grad_norm": 1.6739938259124756,
      "learning_rate": 1.9547586327818833e-07,
      "loss": 0.04651668071746826,
      "step": 26800
    },
    {
      "epoch": 1.9299756062562778,
      "grad_norm": 1.3062328100204468,
      "learning_rate": 1.608982185154706e-07,
      "loss": 0.045240769386291506,
      "step": 26900
    },
    {
      "epoch": 1.9371502367628066,
      "grad_norm": 1.0647732019424438,
      "learning_rate": 1.2967369683476173e-07,
      "loss": 0.04900087833404541,
      "step": 27000
    },
    {
      "epoch": 1.9443248672693356,
      "grad_norm": 1.496711254119873,
      "learning_rate": 1.018065133508217e-07,
      "loss": 0.04883632183074951,
      "step": 27100
    },
    {
      "epoch": 1.9514994977758646,
      "grad_norm": 2.0054495334625244,
      "learning_rate": 7.730042995878206e-08,
      "loss": 0.04609965324401855,
      "step": 27200
    },
    {
      "epoch": 1.9586741282823934,
      "grad_norm": 1.0895386934280396,
      "learning_rate": 5.6158754826329815e-08,
      "loss": 0.040398755073547364,
      "step": 27300
    },
    {
      "epoch": 1.9658487587889224,
      "grad_norm": 1.0245835781097412,
      "learning_rate": 3.838434194712304e-08,
      "loss": 0.04322698593139648,
      "step": 27400
    },
    {
      "epoch": 1.9730233892954514,
      "grad_norm": 1.6254724264144897,
      "learning_rate": 2.397959075551015e-08,
      "loss": 0.04525808811187744,
      "step": 27500
    },
    {
      "epoch": 1.9801980198019802,
      "grad_norm": 1.7849334478378296,
      "learning_rate": 1.2946445802633489e-08,
      "loss": 0.04429808616638184,
      "step": 27600
    },
    {
      "epoch": 1.987372650308509,
      "grad_norm": 1.2970390319824219,
      "learning_rate": 5.2863964939225965e-09,
      "loss": 0.048364996910095215,
      "step": 27700
    },
    {
      "epoch": 1.9945472808150382,
      "grad_norm": 2.868112564086914,
      "learning_rate": 1.0004768880272907e-09,
      "loss": 0.04765310287475586,
      "step": 27800
    },
    {
      "epoch": 2.0,
      "step": 27876,
      "total_flos": 1.391742478201127e+19,
      "train_loss": 0.09068071038227954,
      "train_runtime": 36452.8937,
      "train_samples_per_second": 24.469,
      "train_steps_per_second": 0.765
    }
  ],
  "logging_steps": 100,
  "max_steps": 27876,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 3500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.391742478201127e+19,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}