{
  "best_global_step": 350,
  "best_metric": 0.6084344156610773,
  "best_model_checkpoint": "/jupyter_kernel/trained_models/lg_nss_tune_whisper_large_v1.0/checkpoint-350",
  "epoch": 1.8918918918918919,
  "eval_steps": 50,
  "global_step": 350,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_cer": 0.4712086318852988,
      "eval_loss": 1.8530521392822266,
      "eval_runtime": 883.0991,
      "eval_samples_per_second": 0.848,
      "eval_steps_per_second": 0.053,
      "eval_wer": 0.8136203891549783,
      "num_input_tokens_seen": 0,
      "step": 0
    },
    {
      "epoch": 0.02702702702702703,
      "grad_norm": 6.694199562072754,
      "learning_rate": 5e-06,
      "loss": 1.5511,
      "num_input_tokens_seen": 61440000,
      "step": 5,
      "train_runtime": 934.0126,
      "train_tokens_per_second": 65780.696
    },
    {
      "epoch": 0.05405405405405406,
      "grad_norm": 5.178955078125,
      "learning_rate": 1.125e-05,
      "loss": 1.3507,
      "num_input_tokens_seen": 122880000,
      "step": 10,
      "train_runtime": 978.6837,
      "train_tokens_per_second": 125556.396
    },
    {
      "epoch": 0.08108108108108109,
      "grad_norm": 4.756581783294678,
      "learning_rate": 1.75e-05,
      "loss": 1.179,
      "num_input_tokens_seen": 184320000,
      "step": 15,
      "train_runtime": 1023.409,
      "train_tokens_per_second": 180103.944
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 4.099735260009766,
      "learning_rate": 2.375e-05,
      "loss": 1.1463,
      "num_input_tokens_seen": 245760000,
      "step": 20,
      "train_runtime": 1068.6171,
      "train_tokens_per_second": 229979.476
    },
    {
      "epoch": 0.13513513513513514,
      "grad_norm": 4.195594310760498,
      "learning_rate": 3e-05,
      "loss": 1.0876,
      "num_input_tokens_seen": 307200000,
      "step": 25,
      "train_runtime": 1113.4632,
      "train_tokens_per_second": 275895.959
    },
    {
      "epoch": 0.16216216216216217,
      "grad_norm": 4.167407035827637,
      "learning_rate": 3.625e-05,
      "loss": 1.193,
      "num_input_tokens_seen": 368640000,
      "step": 30,
      "train_runtime": 1158.068,
      "train_tokens_per_second": 318323.272
    },
    {
      "epoch": 0.1891891891891892,
      "grad_norm": 3.8050270080566406,
      "learning_rate": 4.25e-05,
      "loss": 1.1076,
      "num_input_tokens_seen": 430080000,
      "step": 35,
      "train_runtime": 1203.1492,
      "train_tokens_per_second": 357461.888
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 3.8383731842041016,
      "learning_rate": 4.875e-05,
      "loss": 1.0734,
      "num_input_tokens_seen": 491520000,
      "step": 40,
      "train_runtime": 1248.4355,
      "train_tokens_per_second": 393708.774
    },
    {
      "epoch": 0.24324324324324326,
      "grad_norm": 3.918095350265503,
      "learning_rate": 4.9172026175368326e-05,
      "loss": 0.9229,
      "num_input_tokens_seen": 552960000,
      "step": 45,
      "train_runtime": 1293.5005,
      "train_tokens_per_second": 427491.146
    },
    {
      "epoch": 0.2702702702702703,
      "grad_norm": 3.4553701877593994,
      "learning_rate": 4.815157253826046e-05,
      "loss": 1.1542,
      "num_input_tokens_seen": 614400000,
      "step": 50,
      "train_runtime": 1338.8054,
      "train_tokens_per_second": 458916.573
    },
    {
      "epoch": 0.2702702702702703,
      "eval_cer": 0.4179175096256484,
      "eval_loss": 1.127152442932129,
      "eval_runtime": 873.5485,
      "eval_samples_per_second": 0.857,
      "eval_steps_per_second": 0.054,
      "eval_wer": 0.7243194877471738,
      "num_input_tokens_seen": 614400000,
      "step": 50
    },
    {
      "epoch": 0.2972972972972973,
      "grad_norm": 4.08140230178833,
      "learning_rate": 4.714708807666149e-05,
      "loss": 1.0003,
      "num_input_tokens_seen": 675840000,
      "step": 55,
      "train_runtime": 2304.053,
      "train_tokens_per_second": 293326.581
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 3.4815714359283447,
      "learning_rate": 4.615840531294598e-05,
      "loss": 0.9768,
      "num_input_tokens_seen": 737280000,
      "step": 60,
      "train_runtime": 2348.4672,
      "train_tokens_per_second": 313940.941
    },
    {
      "epoch": 0.35135135135135137,
      "grad_norm": 3.143886089324951,
      "learning_rate": 4.5185357652343744e-05,
      "loss": 0.921,
      "num_input_tokens_seen": 798720000,
      "step": 65,
      "train_runtime": 2393.166,
      "train_tokens_per_second": 333750.358
    },
    {
      "epoch": 0.3783783783783784,
      "grad_norm": 3.148380756378174,
      "learning_rate": 4.422777938293975e-05,
      "loss": 0.9031,
      "num_input_tokens_seen": 860160000,
      "step": 70,
      "train_runtime": 2437.7177,
      "train_tokens_per_second": 352854.636
    },
    {
      "epoch": 0.40540540540540543,
      "grad_norm": 3.1794612407684326,
      "learning_rate": 4.328550567567407e-05,
      "loss": 0.9638,
      "num_input_tokens_seen": 921600000,
      "step": 75,
      "train_runtime": 2481.9555,
      "train_tokens_per_second": 371320.115
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 3.83333158493042,
      "learning_rate": 4.2358372584342004e-05,
      "loss": 0.9354,
      "num_input_tokens_seen": 983040000,
      "step": 80,
      "train_runtime": 2526.153,
      "train_tokens_per_second": 389145.073
    },
    {
      "epoch": 0.4594594594594595,
      "grad_norm": 3.1049253940582275,
      "learning_rate": 4.144621704559403e-05,
      "loss": 0.9552,
      "num_input_tokens_seen": 1044480000,
      "step": 85,
      "train_runtime": 2570.4868,
      "train_tokens_per_second": 406335.477
    },
    {
      "epoch": 0.4864864864864865,
      "grad_norm": 3.3949313163757324,
      "learning_rate": 4.054887687893573e-05,
      "loss": 0.8676,
      "num_input_tokens_seen": 1105920000,
      "step": 90,
      "train_runtime": 2614.6645,
      "train_tokens_per_second": 422968.222
    },
    {
      "epoch": 0.5135135135135135,
      "grad_norm": 3.3787624835968018,
      "learning_rate": 3.9666190786727904e-05,
      "loss": 0.8933,
      "num_input_tokens_seen": 1167360000,
      "step": 95,
      "train_runtime": 2658.827,
      "train_tokens_per_second": 439050.749
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 3.108181953430176,
      "learning_rate": 3.879799835418654e-05,
      "loss": 0.854,
      "num_input_tokens_seen": 1228800000,
      "step": 100,
      "train_runtime": 2702.8667,
      "train_tokens_per_second": 454628.419
    },
    {
      "epoch": 0.5405405405405406,
      "eval_cer": 0.29776486627376264,
      "eval_loss": 0.9693590402603149,
      "eval_runtime": 853.632,
      "eval_samples_per_second": 0.877,
      "eval_steps_per_second": 0.055,
      "eval_wer": 0.6468414575552031,
      "num_input_tokens_seen": 1228800000,
      "step": 100
    },
    {
      "epoch": 0.5675675675675675,
      "grad_norm": 3.0080583095550537,
      "learning_rate": 3.794414004938272e-05,
      "loss": 0.9135,
      "num_input_tokens_seen": 1290240000,
      "step": 105,
      "train_runtime": 3647.5305,
      "train_tokens_per_second": 353729.737
    },
    {
      "epoch": 0.5945945945945946,
      "grad_norm": 3.331343650817871,
      "learning_rate": 3.7104457223242756e-05,
      "loss": 0.9881,
      "num_input_tokens_seen": 1351680000,
      "step": 110,
      "train_runtime": 3692.651,
      "train_tokens_per_second": 366045.961
    },
    {
      "epoch": 0.6216216216216216,
      "grad_norm": 4.435062885284424,
      "learning_rate": 3.627879210954812e-05,
      "loss": 0.9993,
      "num_input_tokens_seen": 1413120000,
      "step": 115,
      "train_runtime": 3737.2167,
      "train_tokens_per_second": 378120.973
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 3.606947422027588,
      "learning_rate": 3.546698782493543e-05,
      "loss": 1.0589,
      "num_input_tokens_seen": 1474560000,
      "step": 120,
      "train_runtime": 3782.5686,
      "train_tokens_per_second": 389830.343
    },
    {
      "epoch": 0.6756756756756757,
      "grad_norm": 2.7611401081085205,
      "learning_rate": 3.4668888368896486e-05,
      "loss": 0.7697,
      "num_input_tokens_seen": 1536000000,
      "step": 125,
      "train_runtime": 3829.5601,
      "train_tokens_per_second": 401090.456
    },
    {
      "epoch": 0.7027027027027027,
      "grad_norm": 2.7181291580200195,
      "learning_rate": 3.3884338623778245e-05,
      "loss": 0.8414,
      "num_input_tokens_seen": 1597440000,
      "step": 130,
      "train_runtime": 3875.7508,
      "train_tokens_per_second": 412162.72
    },
    {
      "epoch": 0.7297297297297297,
      "grad_norm": 3.344625234603882,
      "learning_rate": 3.311318435478286e-05,
      "loss": 0.9034,
      "num_input_tokens_seen": 1658880000,
      "step": 135,
      "train_runtime": 3920.4138,
      "train_tokens_per_second": 423139.005
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 2.8509151935577393,
      "learning_rate": 3.235527220996761e-05,
      "loss": 0.7987,
      "num_input_tokens_seen": 1720320000,
      "step": 140,
      "train_runtime": 3965.1338,
      "train_tokens_per_second": 433861.779
    },
    {
      "epoch": 0.7837837837837838,
      "grad_norm": 2.9075114727020264,
      "learning_rate": 3.1610449720244975e-05,
      "loss": 0.8147,
      "num_input_tokens_seen": 1781760000,
      "step": 145,
      "train_runtime": 4008.9754,
      "train_tokens_per_second": 444442.733
    },
    {
      "epoch": 0.8108108108108109,
      "grad_norm": 2.9646787643432617,
      "learning_rate": 3.087856529938261e-05,
      "loss": 0.855,
      "num_input_tokens_seen": 1843200000,
      "step": 150,
      "train_runtime": 4052.7288,
      "train_tokens_per_second": 454804.669
    },
    {
      "epoch": 0.8108108108108109,
      "eval_cer": 0.3384151074503347,
      "eval_loss": 0.9674391150474548,
      "eval_runtime": 871.3397,
      "eval_samples_per_second": 0.86,
      "eval_steps_per_second": 0.054,
      "eval_wer": 0.6576210232547248,
      "num_input_tokens_seen": 1843200000,
      "step": 150
    },
    {
      "epoch": 0.8378378378378378,
      "grad_norm": 3.101228952407837,
      "learning_rate": 3.015946824400329e-05,
      "loss": 0.8295,
      "num_input_tokens_seen": 1904640000,
      "step": 155,
      "train_runtime": 5021.2234,
      "train_tokens_per_second": 379317.914
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 2.8363142013549805,
      "learning_rate": 2.945300873358501e-05,
      "loss": 0.664,
      "num_input_tokens_seen": 1966080000,
      "step": 160,
      "train_runtime": 5066.1637,
      "train_tokens_per_second": 388080.63
    },
    {
      "epoch": 0.8918918918918919,
      "grad_norm": 2.5507912635803223,
      "learning_rate": 2.8759037830460915e-05,
      "loss": 0.9147,
      "num_input_tokens_seen": 2027520000,
      "step": 165,
      "train_runtime": 5110.4979,
      "train_tokens_per_second": 396736.298
    },
    {
      "epoch": 0.918918918918919,
      "grad_norm": 3.1494266986846924,
      "learning_rate": 2.8077407479819295e-05,
      "loss": 0.8447,
      "num_input_tokens_seen": 2088960000,
      "step": 170,
      "train_runtime": 5156.8332,
      "train_tokens_per_second": 405085.818
    },
    {
      "epoch": 0.9459459459459459,
      "grad_norm": 2.359222650527954,
      "learning_rate": 2.740797050970364e-05,
      "loss": 0.766,
      "num_input_tokens_seen": 2150400000,
      "step": 175,
      "train_runtime": 5200.7084,
      "train_tokens_per_second": 413482.134
    },
    {
      "epoch": 0.972972972972973,
      "grad_norm": 2.9903037548065186,
      "learning_rate": 2.6750580631012613e-05,
      "loss": 0.7747,
      "num_input_tokens_seen": 2211840000,
      "step": 180,
      "train_runtime": 5244.7543,
      "train_tokens_per_second": 421724.24
    },
    {
      "epoch": 1.0,
      "grad_norm": 9.7564115524292,
      "learning_rate": 2.61050924375e-05,
      "loss": 0.87,
      "num_input_tokens_seen": 2262912000,
      "step": 185,
      "train_runtime": 5275.773,
      "train_tokens_per_second": 428925.202
    },
    {
      "epoch": 1.027027027027027,
      "grad_norm": 1.9875379800796509,
      "learning_rate": 2.5471361405774797e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 2324352000,
      "step": 190,
      "train_runtime": 5325.1726,
      "train_tokens_per_second": 436483.88
    },
    {
      "epoch": 1.054054054054054,
      "grad_norm": 2.469710111618042,
      "learning_rate": 2.4849243895301164e-05,
      "loss": 0.515,
      "num_input_tokens_seen": 2385792000,
      "step": 195,
      "train_runtime": 5368.58,
      "train_tokens_per_second": 444399.081
    },
    {
      "epoch": 1.0810810810810811,
      "grad_norm": 1.9504098892211914,
      "learning_rate": 2.4238597148398397e-05,
      "loss": 0.4768,
      "num_input_tokens_seen": 2447232000,
      "step": 200,
      "train_runtime": 5411.6359,
      "train_tokens_per_second": 452216.674
    },
    {
      "epoch": 1.0810810810810811,
      "eval_cer": 0.338837217731344,
      "eval_loss": 0.943603515625,
      "eval_runtime": 846.9243,
      "eval_samples_per_second": 0.884,
      "eval_steps_per_second": 0.055,
      "eval_wer": 0.6542248808655325,
      "num_input_tokens_seen": 2447232000,
      "step": 200
    },
    {
      "epoch": 1.1081081081081081,
      "grad_norm": 2.1606149673461914,
      "learning_rate": 2.3639279290241e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 2508672000,
      "step": 205,
      "train_runtime": 6355.0036,
      "train_tokens_per_second": 394755.402
    },
    {
      "epoch": 1.135135135135135,
      "grad_norm": 2.4380221366882324,
      "learning_rate": 2.305114932885864e-05,
      "loss": 0.5376,
      "num_input_tokens_seen": 2570112000,
      "step": 210,
      "train_runtime": 6399.9936,
      "train_tokens_per_second": 401580.404
    },
    {
      "epoch": 1.1621621621621623,
      "grad_norm": 2.1753928661346436,
      "learning_rate": 2.2474067155136106e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 2631552000,
      "step": 215,
      "train_runtime": 6444.5402,
      "train_tokens_per_second": 408338.21
    },
    {
      "epoch": 1.1891891891891893,
      "grad_norm": 2.556577205657959,
      "learning_rate": 2.1907893542813412e-05,
      "loss": 0.5193,
      "num_input_tokens_seen": 2692992000,
      "step": 220,
      "train_runtime": 6489.0434,
      "train_tokens_per_second": 415006.011
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 2.411965847015381,
      "learning_rate": 2.1352490148485728e-05,
      "loss": 0.5359,
      "num_input_tokens_seen": 2754432000,
      "step": 225,
      "train_runtime": 6532.5362,
      "train_tokens_per_second": 421648.178
    },
    {
      "epoch": 1.2432432432432432,
      "grad_norm": 2.55607271194458,
      "learning_rate": 2.0807719511603353e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 2815872000,
      "step": 230,
      "train_runtime": 6575.6388,
      "train_tokens_per_second": 428227.898
    },
    {
      "epoch": 1.2702702702702702,
      "grad_norm": 2.326690435409546,
      "learning_rate": 2.0273445054471802e-05,
      "loss": 0.5578,
      "num_input_tokens_seen": 2877312000,
      "step": 235,
      "train_runtime": 6618.6865,
      "train_tokens_per_second": 434725.527
    },
    {
      "epoch": 1.2972972972972974,
      "grad_norm": 2.4669768810272217,
      "learning_rate": 1.974953108225174e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 2938752000,
      "step": 240,
      "train_runtime": 6662.1141,
      "train_tokens_per_second": 441114.029
    },
    {
      "epoch": 1.3243243243243243,
      "grad_norm": 2.3330671787261963,
      "learning_rate": 1.9235842782958985e-05,
      "loss": 0.5165,
      "num_input_tokens_seen": 3000192000,
      "step": 245,
      "train_runtime": 6705.2682,
      "train_tokens_per_second": 447438.033
    },
    {
      "epoch": 1.3513513513513513,
      "grad_norm": 2.568614959716797,
      "learning_rate": 1.8732246227464544e-05,
      "loss": 0.5197,
      "num_input_tokens_seen": 3061632000,
      "step": 250,
      "train_runtime": 6748.5735,
      "train_tokens_per_second": 453670.985
    },
    {
      "epoch": 1.3513513513513513,
      "eval_cer": 0.329565296789743,
      "eval_loss": 0.9339933395385742,
      "eval_runtime": 837.5193,
      "eval_samples_per_second": 0.894,
      "eval_steps_per_second": 0.056,
      "eval_wer": 0.6662925860618296,
      "num_input_tokens_seen": 3061632000,
      "step": 250
    },
    {
      "epoch": 1.3783783783783785,
      "grad_norm": 2.659301519393921,
      "learning_rate": 1.8238608369494596e-05,
      "loss": 0.5545,
      "num_input_tokens_seen": 3123072000,
      "step": 255,
      "train_runtime": 7674.6612,
      "train_tokens_per_second": 406932.882
    },
    {
      "epoch": 1.4054054054054055,
      "grad_norm": 2.174906015396118,
      "learning_rate": 1.7754797045630454e-05,
      "loss": 0.5572,
      "num_input_tokens_seen": 3184512000,
      "step": 260,
      "train_runtime": 7718.6601,
      "train_tokens_per_second": 412573.163
    },
    {
      "epoch": 1.4324324324324325,
      "grad_norm": 2.2907702922821045,
      "learning_rate": 1.728068097530864e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 3245952000,
      "step": 265,
      "train_runtime": 7761.7975,
      "train_tokens_per_second": 418195.915
    },
    {
      "epoch": 1.4594594594594594,
      "grad_norm": 2.1810221672058105,
      "learning_rate": 1.681612976082083e-05,
      "loss": 0.4723,
      "num_input_tokens_seen": 3307392000,
      "step": 270,
      "train_runtime": 7804.7904,
      "train_tokens_per_second": 423764.359
    },
    {
      "epoch": 1.4864864864864864,
      "grad_norm": 2.441494941711426,
      "learning_rate": 1.6361013887313843e-05,
      "loss": 0.5911,
      "num_input_tokens_seen": 3368832000,
      "step": 275,
      "train_runtime": 7848.1781,
      "train_tokens_per_second": 429250.199
    },
    {
      "epoch": 1.5135135135135136,
      "grad_norm": 2.4569575786590576,
      "learning_rate": 1.5915204722789705e-05,
      "loss": 0.524,
      "num_input_tokens_seen": 3430272000,
      "step": 280,
      "train_runtime": 7891.6157,
      "train_tokens_per_second": 434672.968
    },
    {
      "epoch": 1.5405405405405406,
      "grad_norm": 2.1347854137420654,
      "learning_rate": 1.547857451810559e-05,
      "loss": 0.4959,
      "num_input_tokens_seen": 3491712000,
      "step": 285,
      "train_runtime": 7934.8481,
      "train_tokens_per_second": 440047.746
    },
    {
      "epoch": 1.5675675675675675,
      "grad_norm": 2.2399299144744873,
      "learning_rate": 1.5050996406973835e-05,
      "loss": 0.5164,
      "num_input_tokens_seen": 3553152000,
      "step": 290,
      "train_runtime": 7979.2749,
      "train_tokens_per_second": 445297.605
    },
    {
      "epoch": 1.5945945945945947,
      "grad_norm": 2.3631114959716797,
      "learning_rate": 1.4632344405961962e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 3614592000,
      "step": 295,
      "train_runtime": 8023.6934,
      "train_tokens_per_second": 450489.798
    },
    {
      "epoch": 1.6216216216216215,
      "grad_norm": 1.9993412494659424,
      "learning_rate": 1.4222493414492655e-05,
      "loss": 0.5372,
      "num_input_tokens_seen": 3676032000,
      "step": 300,
      "train_runtime": 8067.0957,
      "train_tokens_per_second": 455682.209
    },
    {
      "epoch": 1.6216216216216215,
      "eval_cer": 0.3121871628496099,
      "eval_loss": 0.8871492147445679,
      "eval_runtime": 865.3356,
      "eval_samples_per_second": 0.866,
      "eval_steps_per_second": 0.054,
      "eval_wer": 0.6394179350835714,
      "num_input_tokens_seen": 3676032000,
      "step": 300
    },
    {
      "epoch": 1.6486486486486487,
      "grad_norm": 2.2491278648376465,
      "learning_rate": 1.382131921484375e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 3737472000,
      "step": 305,
      "train_runtime": 9022.1425,
      "train_tokens_per_second": 414255.482
    },
    {
      "epoch": 1.6756756756756757,
      "grad_norm": 2.2252557277679443,
      "learning_rate": 1.3428698472148269e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 3798912000,
      "step": 310,
      "train_runtime": 9065.3762,
      "train_tokens_per_second": 419057.292
    },
    {
      "epoch": 1.7027027027027026,
      "grad_norm": 2.4980313777923584,
      "learning_rate": 1.3044508734394403e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 3860352000,
      "step": 315,
      "train_runtime": 9108.6721,
      "train_tokens_per_second": 423810.621
    },
    {
      "epoch": 1.7297297297297298,
      "grad_norm": 2.2038354873657227,
      "learning_rate": 1.26686284324255e-05,
      "loss": 0.4169,
      "num_input_tokens_seen": 3921792000,
      "step": 320,
      "train_runtime": 9151.5702,
      "train_tokens_per_second": 428537.607
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 2.1892924308776855,
      "learning_rate": 1.2300936879940077e-05,
      "loss": 0.5566,
      "num_input_tokens_seen": 3983232000,
      "step": 325,
      "train_runtime": 9194.743,
      "train_tokens_per_second": 433207.539
    },
    {
      "epoch": 1.7837837837837838,
      "grad_norm": 2.390801429748535,
      "learning_rate": 1.1941314273491836e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 4044672000,
      "step": 330,
      "train_runtime": 9237.5451,
      "train_tokens_per_second": 437851.393
    },
    {
      "epoch": 1.810810810810811,
      "grad_norm": 2.390781879425049,
      "learning_rate": 1.1589641692489621e-05,
      "loss": 0.5916,
      "num_input_tokens_seen": 4106112000,
      "step": 335,
      "train_runtime": 9280.5769,
      "train_tokens_per_second": 442441.459
    },
    {
      "epoch": 1.8378378378378377,
      "grad_norm": 2.5694046020507812,
      "learning_rate": 1.1245801099197466e-05,
      "loss": 0.4809,
      "num_input_tokens_seen": 4167552000,
      "step": 340,
      "train_runtime": 9323.404,
      "train_tokens_per_second": 446998.971
    },
    {
      "epoch": 1.864864864864865,
      "grad_norm": 2.1600019931793213,
      "learning_rate": 1.0909675338734569e-05,
      "loss": 0.4057,
      "num_input_tokens_seen": 4228992000,
      "step": 345,
      "train_runtime": 9367.0963,
      "train_tokens_per_second": 451473.097
    },
    {
      "epoch": 1.8918918918918919,
      "grad_norm": 2.2605528831481934,
      "learning_rate": 1.0581148139075277e-05,
      "loss": 0.4403,
      "num_input_tokens_seen": 4290432000,
      "step": 350,
      "train_runtime": 9410.0927,
      "train_tokens_per_second": 455939.398
    },
    {
      "epoch": 1.8918918918918919,
      "eval_cer": 0.27202566552379415,
      "eval_loss": 0.8798862099647522,
      "eval_runtime": 816.6633,
      "eval_samples_per_second": 0.917,
      "eval_steps_per_second": 0.058,
      "eval_wer": 0.6084344156610773,
      "num_input_tokens_seen": 4290432000,
      "step": 350
    }
  ],
  "logging_steps": 5,
  "max_steps": 1000,
  "num_input_tokens_seen": 4290432000,
  "num_train_epochs": 6,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.796024857919488e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}