{
  "best_global_step": 23500,
  "best_metric": 0.8003720045089722,
  "best_model_checkpoint": "./speecht5/urdu-tts-phonemes-finetuned-extended/checkpoint-23500",
  "epoch": 50.0,
  "eval_steps": 500,
  "global_step": 27350,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36596523330283626,
      "grad_norm": 47.381690979003906,
      "learning_rate": 1.9900000000000004e-06,
      "loss": 6.367281494140625,
      "step": 200
    },
    {
      "epoch": 0.7319304666056725,
      "grad_norm": 36.517269134521484,
      "learning_rate": 3.990000000000001e-06,
      "loss": 4.603808288574219,
      "step": 400
    },
    {
      "epoch": 0.9149130832570905,
      "eval_loss": 0.9382399320602417,
      "eval_runtime": 44.297,
      "eval_samples_per_second": 65.806,
      "eval_steps_per_second": 16.457,
      "step": 500
    },
    {
      "epoch": 1.0969807868252517,
      "grad_norm": 25.08783721923828,
      "learning_rate": 4.981564245810056e-06,
      "loss": 4.192509155273438,
      "step": 600
    },
    {
      "epoch": 1.4629460201280877,
      "grad_norm": 25.930246353149414,
      "learning_rate": 4.944320297951583e-06,
      "loss": 4.06175048828125,
      "step": 800
    },
    {
      "epoch": 1.828911253430924,
      "grad_norm": 19.781871795654297,
      "learning_rate": 4.90707635009311e-06,
      "loss": 3.972266845703125,
      "step": 1000
    },
    {
      "epoch": 1.828911253430924,
      "eval_loss": 0.8981994986534119,
      "eval_runtime": 42.3502,
      "eval_samples_per_second": 68.831,
      "eval_steps_per_second": 17.214,
      "step": 1000
    },
    {
      "epoch": 2.1939615736505034,
      "grad_norm": 32.66846466064453,
      "learning_rate": 4.869832402234637e-06,
      "loss": 3.889791259765625,
      "step": 1200
    },
    {
      "epoch": 2.5599268069533396,
      "grad_norm": 30.99549102783203,
      "learning_rate": 4.832588454376164e-06,
      "loss": 3.8795877075195313,
      "step": 1400
    },
    {
      "epoch": 2.7429094236047575,
      "eval_loss": 0.8836744427680969,
      "eval_runtime": 41.3823,
      "eval_samples_per_second": 70.441,
      "eval_steps_per_second": 17.616,
      "step": 1500
    },
    {
      "epoch": 2.9258920402561754,
      "grad_norm": 31.6311092376709,
      "learning_rate": 4.795344506517691e-06,
      "loss": 3.8096429443359376,
      "step": 1600
    },
    {
      "epoch": 3.290942360475755,
      "grad_norm": 56.990718841552734,
      "learning_rate": 4.7581005586592186e-06,
      "loss": 3.7748834228515626,
      "step": 1800
    },
    {
      "epoch": 3.656907593778591,
      "grad_norm": 43.28689956665039,
      "learning_rate": 4.720856610800745e-06,
      "loss": 3.7655291748046875,
      "step": 2000
    },
    {
      "epoch": 3.656907593778591,
      "eval_loss": 0.869123101234436,
      "eval_runtime": 42.3265,
      "eval_samples_per_second": 68.869,
      "eval_steps_per_second": 17.223,
      "step": 2000
    },
    {
      "epoch": 4.02195791399817,
      "grad_norm": 31.543128967285156,
      "learning_rate": 4.683612662942272e-06,
      "loss": 3.716641845703125,
      "step": 2200
    },
    {
      "epoch": 4.387923147301007,
      "grad_norm": 25.362565994262695,
      "learning_rate": 4.6463687150838e-06,
      "loss": 3.7278778076171877,
      "step": 2400
    },
    {
      "epoch": 4.570905763952425,
      "eval_loss": 0.8611245155334473,
      "eval_runtime": 42.4949,
      "eval_samples_per_second": 68.596,
      "eval_steps_per_second": 17.155,
      "step": 2500
    },
    {
      "epoch": 4.753888380603843,
      "grad_norm": 23.819162368774414,
      "learning_rate": 4.609124767225326e-06,
      "loss": 3.6925787353515624,
      "step": 2600
    },
    {
      "epoch": 5.1189387008234215,
      "grad_norm": 30.135427474975586,
      "learning_rate": 4.5718808193668536e-06,
      "loss": 3.6830133056640624,
      "step": 2800
    },
    {
      "epoch": 5.484903934126258,
      "grad_norm": 31.421367645263672,
      "learning_rate": 4.534636871508381e-06,
      "loss": 3.687431640625,
      "step": 3000
    },
    {
      "epoch": 5.484903934126258,
      "eval_loss": 0.8506535291671753,
      "eval_runtime": 43.0913,
      "eval_samples_per_second": 67.647,
      "eval_steps_per_second": 16.918,
      "step": 3000
    },
    {
      "epoch": 5.850869167429094,
      "grad_norm": 20.812034606933594,
      "learning_rate": 4.497392923649907e-06,
      "loss": 3.653070373535156,
      "step": 3200
    },
    {
      "epoch": 6.215919487648673,
      "grad_norm": 25.483972549438477,
      "learning_rate": 4.460148975791434e-06,
      "loss": 3.6253271484375,
      "step": 3400
    },
    {
      "epoch": 6.398902104300092,
      "eval_loss": 0.8450019955635071,
      "eval_runtime": 42.212,
      "eval_samples_per_second": 69.056,
      "eval_steps_per_second": 17.27,
      "step": 3500
    },
    {
      "epoch": 6.58188472095151,
      "grad_norm": 36.80460739135742,
      "learning_rate": 4.422905027932961e-06,
      "loss": 3.6231103515625,
      "step": 3600
    },
    {
      "epoch": 6.9478499542543455,
      "grad_norm": 29.918052673339844,
      "learning_rate": 4.385661080074488e-06,
      "loss": 3.5974215698242187,
      "step": 3800
    },
    {
      "epoch": 7.312900274473925,
      "grad_norm": 46.66632843017578,
      "learning_rate": 4.348417132216015e-06,
      "loss": 3.5747479248046874,
      "step": 4000
    },
    {
      "epoch": 7.312900274473925,
      "eval_loss": 0.8402732014656067,
      "eval_runtime": 43.2582,
      "eval_samples_per_second": 67.386,
      "eval_steps_per_second": 16.852,
      "step": 4000
    },
    {
      "epoch": 7.678865507776761,
      "grad_norm": 29.704944610595703,
      "learning_rate": 4.311173184357542e-06,
      "loss": 3.59167236328125,
      "step": 4200
    },
    {
      "epoch": 8.04391582799634,
      "grad_norm": 26.13818359375,
      "learning_rate": 4.273929236499069e-06,
      "loss": 3.5732412719726563,
      "step": 4400
    },
    {
      "epoch": 8.226898444647759,
      "eval_loss": 0.8419305086135864,
      "eval_runtime": 42.0408,
      "eval_samples_per_second": 69.337,
      "eval_steps_per_second": 17.34,
      "step": 4500
    },
    {
      "epoch": 8.409881061299176,
      "grad_norm": 60.029380798339844,
      "learning_rate": 4.236685288640596e-06,
      "loss": 3.5521734619140624,
      "step": 4600
    },
    {
      "epoch": 8.775846294602013,
      "grad_norm": 21.42950439453125,
      "learning_rate": 4.1994413407821236e-06,
      "loss": 3.5773333740234374,
      "step": 4800
    },
    {
      "epoch": 9.140896614821592,
      "grad_norm": 29.458261489868164,
      "learning_rate": 4.16219739292365e-06,
      "loss": 3.5594638061523436,
      "step": 5000
    },
    {
      "epoch": 9.140896614821592,
      "eval_loss": 0.8367183208465576,
      "eval_runtime": 43.1133,
      "eval_samples_per_second": 67.612,
      "eval_steps_per_second": 16.909,
      "step": 5000
    },
    {
      "epoch": 9.506861848124428,
      "grad_norm": 30.463623046875,
      "learning_rate": 4.124953445065177e-06,
      "loss": 3.57099365234375,
      "step": 5200
    },
    {
      "epoch": 9.872827081427264,
      "grad_norm": 20.020761489868164,
      "learning_rate": 4.087709497206705e-06,
      "loss": 3.55204833984375,
      "step": 5400
    },
    {
      "epoch": 10.054894784995426,
      "eval_loss": 0.8339092135429382,
      "eval_runtime": 42.699,
      "eval_samples_per_second": 68.269,
      "eval_steps_per_second": 17.073,
      "step": 5500
    },
    {
      "epoch": 10.237877401646843,
      "grad_norm": 34.7899055480957,
      "learning_rate": 4.050465549348231e-06,
      "loss": 3.54570068359375,
      "step": 5600
    },
    {
      "epoch": 10.60384263494968,
      "grad_norm": 43.3629035949707,
      "learning_rate": 4.0132216014897585e-06,
      "loss": 3.5393402099609377,
      "step": 5800
    },
    {
      "epoch": 10.969807868252516,
      "grad_norm": 37.639652252197266,
      "learning_rate": 3.975977653631286e-06,
      "loss": 3.5296929931640624,
      "step": 6000
    },
    {
      "epoch": 10.969807868252516,
      "eval_loss": 0.83566814661026,
      "eval_runtime": 42.2384,
      "eval_samples_per_second": 69.013,
      "eval_steps_per_second": 17.259,
      "step": 6000
    },
    {
      "epoch": 11.334858188472095,
      "grad_norm": 29.453245162963867,
      "learning_rate": 3.938733705772812e-06,
      "loss": 3.5379364013671877,
      "step": 6200
    },
    {
      "epoch": 11.700823421774931,
      "grad_norm": 31.53673553466797,
      "learning_rate": 3.901489757914339e-06,
      "loss": 3.53030029296875,
      "step": 6400
    },
    {
      "epoch": 11.88380603842635,
      "eval_loss": 0.8333375453948975,
      "eval_runtime": 42.2739,
      "eval_samples_per_second": 68.955,
      "eval_steps_per_second": 17.245,
      "step": 6500
    },
    {
      "epoch": 12.06587374199451,
      "grad_norm": 23.901918411254883,
      "learning_rate": 3.864245810055866e-06,
      "loss": 3.53107666015625,
      "step": 6600
    },
    {
      "epoch": 12.431838975297346,
      "grad_norm": 40.04535675048828,
      "learning_rate": 3.827001862197393e-06,
      "loss": 3.5215399169921877,
      "step": 6800
    },
    {
      "epoch": 12.797804208600184,
      "grad_norm": 36.95607376098633,
      "learning_rate": 3.78975791433892e-06,
      "loss": 3.505963134765625,
      "step": 7000
    },
    {
      "epoch": 12.797804208600184,
      "eval_loss": 0.8259705901145935,
      "eval_runtime": 42.7262,
      "eval_samples_per_second": 68.225,
      "eval_steps_per_second": 17.062,
      "step": 7000
    },
    {
      "epoch": 13.162854528819762,
      "grad_norm": 16.86378288269043,
      "learning_rate": 3.7525139664804474e-06,
      "loss": 3.4859466552734375,
      "step": 7200
    },
    {
      "epoch": 13.528819762122598,
      "grad_norm": 21.715669631958008,
      "learning_rate": 3.7152700186219743e-06,
      "loss": 3.4966839599609374,
      "step": 7400
    },
    {
      "epoch": 13.711802378774017,
      "eval_loss": 0.8256122469902039,
      "eval_runtime": 41.6583,
      "eval_samples_per_second": 69.974,
      "eval_steps_per_second": 17.5,
      "step": 7500
    },
    {
      "epoch": 13.894784995425434,
      "grad_norm": 35.307899475097656,
      "learning_rate": 3.678026070763501e-06,
      "loss": 3.5096063232421875,
      "step": 7600
    },
    {
      "epoch": 14.259835315645013,
      "grad_norm": 32.276493072509766,
      "learning_rate": 3.640782122905028e-06,
      "loss": 3.508584289550781,
      "step": 7800
    },
    {
      "epoch": 14.62580054894785,
      "grad_norm": 32.58161926269531,
      "learning_rate": 3.6035381750465555e-06,
      "loss": 3.502772216796875,
      "step": 8000
    },
    {
      "epoch": 14.62580054894785,
      "eval_loss": 0.8239230513572693,
      "eval_runtime": 43.0618,
      "eval_samples_per_second": 67.693,
      "eval_steps_per_second": 16.929,
      "step": 8000
    },
    {
      "epoch": 14.991765782250686,
      "grad_norm": 33.784114837646484,
      "learning_rate": 3.5662942271880824e-06,
      "loss": 3.481976318359375,
      "step": 8200
    },
    {
      "epoch": 15.356816102470265,
      "grad_norm": 46.78675842285156,
      "learning_rate": 3.5290502793296093e-06,
      "loss": 3.4785525512695314,
      "step": 8400
    },
    {
      "epoch": 15.539798719121684,
      "eval_loss": 0.8210623264312744,
      "eval_runtime": 41.56,
      "eval_samples_per_second": 70.139,
      "eval_steps_per_second": 17.541,
      "step": 8500
    },
    {
      "epoch": 15.722781335773101,
      "grad_norm": 18.62014389038086,
      "learning_rate": 3.4918063314711366e-06,
      "loss": 3.4680157470703126,
      "step": 8600
    },
    {
      "epoch": 16.08783165599268,
      "grad_norm": 47.83131790161133,
      "learning_rate": 3.4545623836126635e-06,
      "loss": 3.488092041015625,
      "step": 8800
    },
    {
      "epoch": 16.453796889295518,
      "grad_norm": 24.163806915283203,
      "learning_rate": 3.41731843575419e-06,
      "loss": 3.49182373046875,
      "step": 9000
    },
    {
      "epoch": 16.453796889295518,
      "eval_loss": 0.8208582401275635,
      "eval_runtime": 42.8306,
      "eval_samples_per_second": 68.059,
      "eval_steps_per_second": 17.021,
      "step": 9000
    },
    {
      "epoch": 16.819762122598352,
      "grad_norm": 23.606050491333008,
      "learning_rate": 3.380074487895717e-06,
      "loss": 3.4846002197265626,
      "step": 9200
    },
    {
      "epoch": 17.184812442817933,
      "grad_norm": 58.5062255859375,
      "learning_rate": 3.342830540037244e-06,
      "loss": 3.4901348876953127,
      "step": 9400
    },
    {
      "epoch": 17.36779505946935,
      "eval_loss": 0.8192538619041443,
      "eval_runtime": 42.7941,
      "eval_samples_per_second": 68.117,
      "eval_steps_per_second": 17.035,
      "step": 9500
    },
    {
      "epoch": 17.55077767612077,
      "grad_norm": 48.141632080078125,
      "learning_rate": 3.305586592178771e-06,
      "loss": 3.4592129516601564,
      "step": 9600
    },
    {
      "epoch": 17.916742909423604,
      "grad_norm": 22.53372573852539,
      "learning_rate": 3.268342644320298e-06,
      "loss": 3.4706024169921874,
      "step": 9800
    },
    {
      "epoch": 18.281793229643185,
      "grad_norm": 33.21782684326172,
      "learning_rate": 3.231098696461825e-06,
      "loss": 3.481481018066406,
      "step": 10000
    },
    {
      "epoch": 18.281793229643185,
      "eval_loss": 0.82040935754776,
      "eval_runtime": 42.6304,
      "eval_samples_per_second": 68.378,
      "eval_steps_per_second": 17.1,
      "step": 10000
    },
    {
      "epoch": 18.64775846294602,
      "grad_norm": 24.176639556884766,
      "learning_rate": 3.1938547486033524e-06,
      "loss": 3.4637158203125,
      "step": 10200
    },
    {
      "epoch": 19.0128087831656,
      "grad_norm": 26.19886016845703,
      "learning_rate": 3.1566108007448793e-06,
      "loss": 3.4687786865234376,
      "step": 10400
    },
    {
      "epoch": 19.19579139981702,
      "eval_loss": 0.8162763118743896,
      "eval_runtime": 41.6542,
      "eval_samples_per_second": 69.981,
      "eval_steps_per_second": 17.501,
      "step": 10500
    },
    {
      "epoch": 19.378774016468437,
      "grad_norm": 31.632343292236328,
      "learning_rate": 3.119366852886406e-06,
      "loss": 3.4610107421875,
      "step": 10600
    },
    {
      "epoch": 19.74473924977127,
      "grad_norm": 33.52009963989258,
      "learning_rate": 3.0821229050279335e-06,
      "loss": 3.4670669555664064,
      "step": 10800
    },
    {
      "epoch": 20.109789569990852,
      "grad_norm": 32.436702728271484,
      "learning_rate": 3.0448789571694604e-06,
      "loss": 3.442606201171875,
      "step": 11000
    },
    {
      "epoch": 20.109789569990852,
      "eval_loss": 0.8159969449043274,
      "eval_runtime": 41.6805,
      "eval_samples_per_second": 69.937,
      "eval_steps_per_second": 17.49,
      "step": 11000
    },
    {
      "epoch": 20.475754803293686,
      "grad_norm": 23.588851928710938,
      "learning_rate": 3.0076350093109874e-06,
      "loss": 3.4675115966796874,
      "step": 11200
    },
    {
      "epoch": 20.841720036596524,
      "grad_norm": 27.577604293823242,
      "learning_rate": 2.9703910614525143e-06,
      "loss": 3.4519415283203125,
      "step": 11400
    },
    {
      "epoch": 21.023787740164686,
      "eval_loss": 0.815731942653656,
      "eval_runtime": 41.6924,
      "eval_samples_per_second": 69.917,
      "eval_steps_per_second": 17.485,
      "step": 11500
    },
    {
      "epoch": 21.2067703568161,
      "grad_norm": 36.152488708496094,
      "learning_rate": 2.9331471135940416e-06,
      "loss": 3.4238546752929686,
      "step": 11600
    },
    {
      "epoch": 21.57273559011894,
      "grad_norm": 20.04086685180664,
      "learning_rate": 2.895903165735568e-06,
      "loss": 3.4361163330078126,
      "step": 11800
    },
    {
      "epoch": 21.938700823421776,
      "grad_norm": 30.39275550842285,
      "learning_rate": 2.858659217877095e-06,
      "loss": 3.451390380859375,
      "step": 12000
    },
    {
      "epoch": 21.938700823421776,
      "eval_loss": 0.8138015270233154,
      "eval_runtime": 41.6609,
      "eval_samples_per_second": 69.97,
      "eval_steps_per_second": 17.498,
      "step": 12000
    },
    {
      "epoch": 22.303751143641353,
      "grad_norm": 38.23406219482422,
      "learning_rate": 2.821415270018622e-06,
      "loss": 3.444107666015625,
      "step": 12200
    },
    {
      "epoch": 22.66971637694419,
      "grad_norm": 27.947982788085938,
      "learning_rate": 2.7841713221601493e-06,
      "loss": 3.4341265869140627,
      "step": 12400
    },
    {
      "epoch": 22.85269899359561,
      "eval_loss": 0.8147653937339783,
      "eval_runtime": 42.6887,
      "eval_samples_per_second": 68.285,
      "eval_steps_per_second": 17.077,
      "step": 12500
    },
    {
      "epoch": 23.034766697163768,
      "grad_norm": 18.88580894470215,
      "learning_rate": 2.746927374301676e-06,
      "loss": 3.4495306396484375,
      "step": 12600
    },
    {
      "epoch": 23.400731930466605,
      "grad_norm": 27.68914794921875,
      "learning_rate": 2.709683426443203e-06,
      "loss": 3.4263726806640626,
      "step": 12800
    },
    {
      "epoch": 23.766697163769443,
      "grad_norm": 26.018314361572266,
      "learning_rate": 2.67243947858473e-06,
      "loss": 3.4341275024414064,
      "step": 13000
    },
    {
      "epoch": 23.766697163769443,
      "eval_loss": 0.8114081621170044,
      "eval_runtime": 41.8968,
      "eval_samples_per_second": 69.576,
      "eval_steps_per_second": 17.4,
      "step": 13000
    },
    {
      "epoch": 24.13174748398902,
      "grad_norm": 28.45810890197754,
      "learning_rate": 2.6351955307262573e-06,
      "loss": 3.4314459228515624,
      "step": 13200
    },
    {
      "epoch": 24.497712717291858,
      "grad_norm": 55.73724365234375,
      "learning_rate": 2.5979515828677843e-06,
      "loss": 3.426131591796875,
      "step": 13400
    },
    {
      "epoch": 24.680695333943277,
      "eval_loss": 0.8129320740699768,
      "eval_runtime": 41.7307,
      "eval_samples_per_second": 69.853,
      "eval_steps_per_second": 17.469,
      "step": 13500
    },
    {
      "epoch": 24.863677950594692,
      "grad_norm": 31.70014190673828,
      "learning_rate": 2.560707635009311e-06,
      "loss": 3.4323944091796874,
      "step": 13600
    },
    {
      "epoch": 25.228728270814273,
      "grad_norm": 37.29436492919922,
      "learning_rate": 2.5234636871508385e-06,
      "loss": 3.4168301391601563,
      "step": 13800
    },
    {
      "epoch": 25.59469350411711,
      "grad_norm": 20.562915802001953,
      "learning_rate": 2.486219739292365e-06,
      "loss": 3.427581787109375,
      "step": 14000
    },
    {
      "epoch": 25.59469350411711,
      "eval_loss": 0.8111268281936646,
      "eval_runtime": 42.3026,
      "eval_samples_per_second": 68.908,
      "eval_steps_per_second": 17.233,
      "step": 14000
    },
    {
      "epoch": 25.960658737419944,
      "grad_norm": 31.984947204589844,
      "learning_rate": 2.4489757914338923e-06,
      "loss": 3.43117431640625,
      "step": 14200
    },
    {
      "epoch": 26.325709057639525,
      "grad_norm": 26.999069213867188,
      "learning_rate": 2.4117318435754193e-06,
      "loss": 3.4233984375,
      "step": 14400
    },
    {
      "epoch": 26.508691674290944,
      "eval_loss": 0.8085624575614929,
      "eval_runtime": 42.2143,
      "eval_samples_per_second": 69.053,
      "eval_steps_per_second": 17.269,
      "step": 14500
    },
    {
      "epoch": 26.69167429094236,
      "grad_norm": 29.52259635925293,
      "learning_rate": 2.374487895716946e-06,
      "loss": 3.42596435546875,
      "step": 14600
    },
    {
      "epoch": 27.05672461116194,
      "grad_norm": 20.191740036010742,
      "learning_rate": 2.337243947858473e-06,
      "loss": 3.413668212890625,
      "step": 14800
    },
    {
      "epoch": 27.422689844464777,
      "grad_norm": 24.441926956176758,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 3.4378948974609376,
      "step": 15000
    },
    {
      "epoch": 27.422689844464777,
      "eval_loss": 0.8082968592643738,
      "eval_runtime": 42.2022,
      "eval_samples_per_second": 69.072,
      "eval_steps_per_second": 17.274,
      "step": 15000
    },
    {
      "epoch": 27.78865507776761,
      "grad_norm": 32.56826400756836,
      "learning_rate": 2.262756052141527e-06,
      "loss": 3.409415283203125,
      "step": 15200
    },
    {
      "epoch": 28.153705397987192,
      "grad_norm": 18.923120498657227,
      "learning_rate": 2.2255121042830543e-06,
      "loss": 3.417109375,
      "step": 15400
    },
    {
      "epoch": 28.33668801463861,
      "eval_loss": 0.8054900169372559,
      "eval_runtime": 42.2244,
      "eval_samples_per_second": 69.036,
      "eval_steps_per_second": 17.265,
      "step": 15500
    },
    {
      "epoch": 28.519670631290026,
      "grad_norm": 21.48866081237793,
      "learning_rate": 2.188268156424581e-06,
      "loss": 3.415582580566406,
      "step": 15600
    },
    {
      "epoch": 28.885635864592864,
      "grad_norm": 35.53007125854492,
      "learning_rate": 2.151024208566108e-06,
      "loss": 3.4359011840820313,
      "step": 15800
    },
    {
      "epoch": 29.250686184812444,
      "grad_norm": 48.40018844604492,
      "learning_rate": 2.113780260707635e-06,
      "loss": 3.4082733154296876,
      "step": 16000
    },
    {
      "epoch": 29.250686184812444,
      "eval_loss": 0.8067459464073181,
      "eval_runtime": 41.9936,
      "eval_samples_per_second": 69.415,
      "eval_steps_per_second": 17.36,
      "step": 16000
    },
    {
      "epoch": 29.61665141811528,
      "grad_norm": 25.632177352905273,
      "learning_rate": 2.0765363128491623e-06,
      "loss": 3.4094720458984376,
      "step": 16200
    },
    {
      "epoch": 29.982616651418116,
      "grad_norm": 33.20957565307617,
      "learning_rate": 2.0392923649906892e-06,
      "loss": 3.422698974609375,
      "step": 16400
    },
    {
      "epoch": 30.164684354986278,
      "eval_loss": 0.8089665174484253,
      "eval_runtime": 41.8922,
      "eval_samples_per_second": 69.583,
      "eval_steps_per_second": 17.402,
      "step": 16500
    },
    {
      "epoch": 30.347666971637693,
      "grad_norm": 27.007097244262695,
      "learning_rate": 2.002048417132216e-06,
      "loss": 3.4013916015625,
      "step": 16600
    },
    {
      "epoch": 30.71363220494053,
      "grad_norm": 32.90212631225586,
      "learning_rate": 1.964804469273743e-06,
      "loss": 3.4162109375,
      "step": 16800
    },
    {
      "epoch": 31.07868252516011,
      "grad_norm": 24.65216827392578,
      "learning_rate": 1.92756052141527e-06,
      "loss": 3.4050347900390623,
      "step": 17000
    },
    {
      "epoch": 31.07868252516011,
      "eval_loss": 0.8063186407089233,
      "eval_runtime": 42.0357,
      "eval_samples_per_second": 69.346,
      "eval_steps_per_second": 17.342,
      "step": 17000
    },
    {
      "epoch": 31.444647758462946,
      "grad_norm": 49.785057067871094,
      "learning_rate": 1.8903165735567971e-06,
      "loss": 3.4042044067382813,
      "step": 17200
    },
    {
      "epoch": 31.810612991765783,
      "grad_norm": 76.18311309814453,
      "learning_rate": 1.8530726256983242e-06,
      "loss": 3.41436767578125,
      "step": 17400
    },
    {
      "epoch": 31.993595608417202,
      "eval_loss": 0.8072382211685181,
      "eval_runtime": 42.4227,
      "eval_samples_per_second": 68.713,
      "eval_steps_per_second": 17.184,
      "step": 17500
    },
    {
      "epoch": 32.17566331198536,
      "grad_norm": 38.510337829589844,
      "learning_rate": 1.8158286778398512e-06,
      "loss": 3.393199157714844,
      "step": 17600
    },
    {
      "epoch": 32.5416285452882,
      "grad_norm": 30.383955001831055,
      "learning_rate": 1.7785847299813783e-06,
      "loss": 3.4250921630859374,
      "step": 17800
    },
    {
      "epoch": 32.907593778591036,
      "grad_norm": 19.114484786987305,
      "learning_rate": 1.7413407821229052e-06,
      "loss": 3.3968218994140624,
      "step": 18000
    },
    {
      "epoch": 32.907593778591036,
      "eval_loss": 0.8027856349945068,
      "eval_runtime": 42.2162,
      "eval_samples_per_second": 69.049,
      "eval_steps_per_second": 17.268,
      "step": 18000
    },
    {
      "epoch": 33.272644098810616,
      "grad_norm": 27.924243927001953,
      "learning_rate": 1.7040968342644321e-06,
      "loss": 3.399924011230469,
      "step": 18200
    },
    {
      "epoch": 33.63860933211345,
      "grad_norm": 27.03548812866211,
      "learning_rate": 1.666852886405959e-06,
      "loss": 3.400137634277344,
      "step": 18400
    },
    {
      "epoch": 33.82159194876487,
      "eval_loss": 0.8055672645568848,
      "eval_runtime": 41.3595,
      "eval_samples_per_second": 70.48,
      "eval_steps_per_second": 17.626,
      "step": 18500
    },
    {
      "epoch": 34.00365965233303,
      "grad_norm": 24.38459587097168,
      "learning_rate": 1.6296089385474862e-06,
      "loss": 3.3994366455078127,
      "step": 18600
    },
    {
      "epoch": 34.369624885635865,
      "grad_norm": 28.692363739013672,
      "learning_rate": 1.5923649906890133e-06,
      "loss": 3.4166107177734375,
      "step": 18800
    },
    {
      "epoch": 34.7355901189387,
      "grad_norm": 21.536502838134766,
      "learning_rate": 1.5551210428305402e-06,
      "loss": 3.387282409667969,
      "step": 19000
    },
    {
      "epoch": 34.7355901189387,
      "eval_loss": 0.8031522631645203,
      "eval_runtime": 42.311,
      "eval_samples_per_second": 68.895,
      "eval_steps_per_second": 17.23,
      "step": 19000
    },
    {
      "epoch": 35.10064043915828,
      "grad_norm": 27.024391174316406,
      "learning_rate": 1.5178770949720673e-06,
      "loss": 3.409147033691406,
      "step": 19200
    },
    {
      "epoch": 35.466605672461114,
      "grad_norm": 18.235994338989258,
      "learning_rate": 1.4806331471135942e-06,
      "loss": 3.4031585693359374,
      "step": 19400
    },
    {
      "epoch": 35.649588289112536,
      "eval_loss": 0.8056690096855164,
      "eval_runtime": 41.9505,
      "eval_samples_per_second": 69.487,
      "eval_steps_per_second": 17.378,
      "step": 19500
    },
    {
      "epoch": 35.83257090576395,
      "grad_norm": 37.04148483276367,
      "learning_rate": 1.4433891992551212e-06,
      "loss": 3.4038092041015626,
      "step": 19600
    },
    {
      "epoch": 36.19762122598353,
      "grad_norm": 19.896331787109375,
      "learning_rate": 1.406145251396648e-06,
      "loss": 3.382321472167969,
      "step": 19800
    },
    {
      "epoch": 36.56358645928637,
      "grad_norm": 34.511016845703125,
      "learning_rate": 1.3689013035381752e-06,
      "loss": 3.393641357421875,
      "step": 20000
    },
    {
      "epoch": 36.56358645928637,
      "eval_loss": 0.8040483593940735,
      "eval_runtime": 42.2369,
      "eval_samples_per_second": 69.016,
      "eval_steps_per_second": 17.26,
      "step": 20000
    },
    {
      "epoch": 36.92955169258921,
      "grad_norm": 32.32827377319336,
      "learning_rate": 1.3316573556797021e-06,
      "loss": 3.4322308349609374,
      "step": 20200
    },
    {
      "epoch": 37.29460201280878,
      "grad_norm": 20.42951202392578,
      "learning_rate": 1.2944134078212292e-06,
      "loss": 3.3828802490234375,
      "step": 20400
    },
    {
      "epoch": 37.4775846294602,
      "eval_loss": 0.8036431074142456,
      "eval_runtime": 42.3918,
      "eval_samples_per_second": 68.763,
      "eval_steps_per_second": 17.197,
      "step": 20500
    },
    {
      "epoch": 37.66056724611162,
      "grad_norm": 29.195035934448242,
      "learning_rate": 1.2571694599627561e-06,
      "loss": 3.3858602905273436,
      "step": 20600
    },
    {
      "epoch": 38.0256175663312,
      "grad_norm": 28.474138259887695,
      "learning_rate": 1.219925512104283e-06,
      "loss": 3.3724441528320312,
      "step": 20800
    },
    {
      "epoch": 38.39158279963404,
      "grad_norm": 23.875751495361328,
      "learning_rate": 1.1826815642458102e-06,
      "loss": 3.392691650390625,
      "step": 21000
    },
    {
      "epoch": 38.39158279963404,
      "eval_loss": 0.8033986687660217,
      "eval_runtime": 42.8257,
      "eval_samples_per_second": 68.067,
      "eval_steps_per_second": 17.022,
      "step": 21000
    },
    {
      "epoch": 38.757548032936874,
      "grad_norm": 41.7534065246582,
      "learning_rate": 1.1454376163873373e-06,
      "loss": 3.4057916259765624,
      "step": 21200
    },
    {
      "epoch": 39.12259835315645,
      "grad_norm": 20.454809188842773,
      "learning_rate": 1.108193668528864e-06,
      "loss": 3.389534912109375,
      "step": 21400
    },
    {
      "epoch": 39.30558096980787,
      "eval_loss": 0.8037059903144836,
      "eval_runtime": 41.4943,
      "eval_samples_per_second": 70.251,
      "eval_steps_per_second": 17.569,
      "step": 21500
    },
    {
      "epoch": 39.488563586459286,
      "grad_norm": 30.62973976135254,
      "learning_rate": 1.0709497206703911e-06,
      "loss": 3.37801513671875,
      "step": 21600
    },
    {
      "epoch": 39.85452881976212,
      "grad_norm": 41.29619216918945,
      "learning_rate": 1.0337057728119183e-06,
      "loss": 3.4008929443359377,
      "step": 21800
    },
    {
      "epoch": 40.219579139981704,
      "grad_norm": 25.416288375854492,
      "learning_rate": 9.964618249534452e-07,
      "loss": 3.378941650390625,
      "step": 22000
    },
    {
      "epoch": 40.219579139981704,
      "eval_loss": 0.8027300238609314,
      "eval_runtime": 40.7902,
      "eval_samples_per_second": 71.463,
      "eval_steps_per_second": 17.872,
      "step": 22000
    },
    {
      "epoch": 40.585544373284534,
      "grad_norm": 20.888412475585938,
      "learning_rate": 9.59217877094972e-07,
      "loss": 3.3759555053710937,
      "step": 22200
    },
    {
      "epoch": 40.95150960658737,
      "grad_norm": 22.00240707397461,
      "learning_rate": 9.219739292364992e-07,
      "loss": 3.3937649536132812,
      "step": 22400
    },
    {
      "epoch": 41.13357731015554,
      "eval_loss": 0.8019812107086182,
      "eval_runtime": 41.4876,
      "eval_samples_per_second": 70.262,
      "eval_steps_per_second": 17.572,
      "step": 22500
    },
    {
      "epoch": 41.31655992680695,
      "grad_norm": 44.14933395385742,
      "learning_rate": 8.847299813780262e-07,
      "loss": 3.3816888427734373,
      "step": 22600
    },
    {
      "epoch": 41.68252516010979,
      "grad_norm": 23.328454971313477,
      "learning_rate": 8.474860335195532e-07,
      "loss": 3.400072021484375,
      "step": 22800
    },
    {
      "epoch": 42.04757548032937,
      "grad_norm": 36.49307632446289,
      "learning_rate": 8.102420856610802e-07,
      "loss": 3.390718688964844,
      "step": 23000
    },
    {
      "epoch": 42.04757548032937,
      "eval_loss": 0.801755964756012,
      "eval_runtime": 40.4509,
      "eval_samples_per_second": 72.063,
      "eval_steps_per_second": 18.022,
      "step": 23000
    },
    {
      "epoch": 42.4135407136322,
      "grad_norm": 22.958580017089844,
      "learning_rate": 7.729981378026072e-07,
      "loss": 3.3810220336914063,
      "step": 23200
    },
    {
      "epoch": 42.77950594693504,
      "grad_norm": 52.35154342651367,
      "learning_rate": 7.357541899441341e-07,
      "loss": 3.3751885986328123,
      "step": 23400
    },
    {
      "epoch": 42.96248856358646,
      "eval_loss": 0.8003720045089722,
      "eval_runtime": 41.5909,
      "eval_samples_per_second": 70.087,
      "eval_steps_per_second": 17.528,
      "step": 23500
    },
    {
      "epoch": 43.14455626715462,
      "grad_norm": 31.925992965698242,
      "learning_rate": 6.985102420856611e-07,
      "loss": 3.3631201171875,
      "step": 23600
    },
    {
      "epoch": 43.51052150045746,
      "grad_norm": 21.90959930419922,
      "learning_rate": 6.612662942271882e-07,
      "loss": 3.3778814697265624,
      "step": 23800
    },
    {
      "epoch": 43.876486733760295,
      "grad_norm": 23.6475772857666,
      "learning_rate": 6.240223463687152e-07,
      "loss": 3.37588623046875,
      "step": 24000
    },
    {
      "epoch": 43.876486733760295,
      "eval_loss": 0.8009469509124756,
      "eval_runtime": 40.6538,
      "eval_samples_per_second": 71.703,
      "eval_steps_per_second": 17.932,
      "step": 24000
    },
    {
      "epoch": 44.24153705397987,
      "grad_norm": 30.51738929748535,
      "learning_rate": 5.867783985102422e-07,
      "loss": 3.3592584228515623,
      "step": 24200
    },
    {
      "epoch": 44.607502287282706,
      "grad_norm": 21.41883087158203,
      "learning_rate": 5.495344506517691e-07,
      "loss": 3.3807147216796873,
      "step": 24400
    },
    {
      "epoch": 44.79048490393413,
      "eval_loss": 0.8014787435531616,
      "eval_runtime": 41.0155,
      "eval_samples_per_second": 71.071,
      "eval_steps_per_second": 17.774,
      "step": 24500
    },
    {
      "epoch": 44.973467520585544,
      "grad_norm": 26.60100555419922,
      "learning_rate": 5.122905027932961e-07,
      "loss": 3.387509765625,
      "step": 24600
    },
    {
      "epoch": 45.338517840805125,
      "grad_norm": 28.604902267456055,
      "learning_rate": 4.7504655493482315e-07,
      "loss": 3.35648193359375,
      "step": 24800
    },
    {
      "epoch": 45.70448307410796,
      "grad_norm": 23.216493606567383,
      "learning_rate": 4.378026070763501e-07,
      "loss": 3.3807525634765625,
      "step": 25000
    },
    {
      "epoch": 45.70448307410796,
      "eval_loss": 0.802956759929657,
      "eval_runtime": 40.4744,
      "eval_samples_per_second": 72.021,
      "eval_steps_per_second": 18.011,
      "step": 25000
    },
    {
      "epoch": 46.069533394327536,
      "grad_norm": 26.358734130859375,
      "learning_rate": 4.0055865921787714e-07,
      "loss": 3.36883056640625,
      "step": 25200
    },
    {
      "epoch": 46.43549862763037,
      "grad_norm": 39.294586181640625,
      "learning_rate": 3.633147113594041e-07,
      "loss": 3.372022705078125,
      "step": 25400
    },
    {
      "epoch": 46.618481244281796,
      "eval_loss": 0.8016297817230225,
      "eval_runtime": 40.8294,
      "eval_samples_per_second": 71.395,
      "eval_steps_per_second": 17.855,
      "step": 25500
    },
    {
      "epoch": 46.80146386093321,
      "grad_norm": 19.483142852783203,
      "learning_rate": 3.260707635009312e-07,
      "loss": 3.36287841796875,
      "step": 25600
    },
    {
      "epoch": 47.16651418115279,
      "grad_norm": 26.670150756835938,
      "learning_rate": 2.8882681564245815e-07,
      "loss": 3.370653076171875,
      "step": 25800
    },
    {
      "epoch": 47.53247941445563,
      "grad_norm": 24.604328155517578,
      "learning_rate": 2.515828677839851e-07,
      "loss": 3.3851007080078124,
      "step": 26000
    },
    {
      "epoch": 47.53247941445563,
      "eval_loss": 0.8018313050270081,
      "eval_runtime": 40.4037,
      "eval_samples_per_second": 72.147,
      "eval_steps_per_second": 18.043,
      "step": 26000
    },
    {
      "epoch": 47.89844464775846,
      "grad_norm": 32.4647216796875,
      "learning_rate": 2.1433891992551213e-07,
      "loss": 3.386794128417969,
      "step": 26200
    },
    {
      "epoch": 48.26349496797804,
      "grad_norm": 25.444473266601562,
      "learning_rate": 1.7709497206703913e-07,
      "loss": 3.3733404541015624,
      "step": 26400
    },
    {
      "epoch": 48.44647758462946,
      "eval_loss": 0.8006534576416016,
      "eval_runtime": 40.845,
      "eval_samples_per_second": 71.367,
      "eval_steps_per_second": 17.848,
      "step": 26500
    },
    {
      "epoch": 48.62946020128088,
      "grad_norm": 18.64166259765625,
      "learning_rate": 1.3985102420856612e-07,
      "loss": 3.385538330078125,
      "step": 26600
    },
    {
      "epoch": 48.995425434583716,
      "grad_norm": 14.76766300201416,
      "learning_rate": 1.0260707635009313e-07,
      "loss": 3.3784445190429686,
      "step": 26800
    },
    {
      "epoch": 49.360475754803296,
      "grad_norm": 12.997588157653809,
      "learning_rate": 6.536312849162012e-08,
      "loss": 3.3703176879882815,
      "step": 27000
    },
    {
      "epoch": 49.360475754803296,
      "eval_loss": 0.8013985753059387,
      "eval_runtime": 40.7108,
      "eval_samples_per_second": 71.603,
      "eval_steps_per_second": 17.907,
      "step": 27000
    },
    {
      "epoch": 49.72644098810613,
      "grad_norm": 51.947044372558594,
      "learning_rate": 2.8119180633147116e-08,
      "loss": 3.37430419921875,
      "step": 27200
    }
  ],
  "logging_steps": 200,
  "max_steps": 27350,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "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": 8.149791933674291e+16,
  "train_batch_size": 6,
  "trial_name": null,
  "trial_params": null
}