{
  "best_metric": 0.05459373816847801,
  "best_model_checkpoint": "/scratch/lingjzhu_root/lingjzhu1/lingjzhu/g2p/byt5_16_layers_baseline/checkpoint-140000",
  "epoch": 9.917791237204689,
  "global_step": 140000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.07,
      "learning_rate": 0.0003,
      "loss": 7.0422,
      "step": 1000
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002999623214094654,
      "loss": 1.3957,
      "step": 2000
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.0002998493045668776,
      "loss": 0.4792,
      "step": 3000
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00029966100624977436,
      "loss": 0.304,
      "step": 4000
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00029939752105569156,
      "loss": 0.2449,
      "step": 5000
    },
    {
      "epoch": 0.35,
      "eval_cer": 0.21979775329351647,
      "eval_loss": 0.2839755415916443,
      "eval_runtime": 80.4341,
      "eval_samples_per_second": 62.163,
      "eval_steps_per_second": 0.124,
      "eval_wer": 0.5296,
      "step": 5000
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.00029905898135463893,
      "loss": 0.2141,
      "step": 6000
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0002986455572226002,
      "loss": 0.1944,
      "step": 7000
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00029815745635609,
      "loss": 0.1791,
      "step": 8000
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00029759492396781067,
      "loss": 0.1669,
      "step": 9000
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00029695824266346256,
      "loss": 0.1574,
      "step": 10000
    },
    {
      "epoch": 0.71,
      "eval_cer": 0.16565687056642014,
      "eval_loss": 0.1845117211341858,
      "eval_runtime": 79.5791,
      "eval_samples_per_second": 62.831,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.4346,
      "step": 10000
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.000296247732299768,
      "loss": 0.1486,
      "step": 11000
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.000295463749823781,
      "loss": 0.1411,
      "step": 12000
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0002946066890935644,
      "loss": 0.1339,
      "step": 13000
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.0002936769806803222,
      "loss": 0.1287,
      "step": 14000
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00029267509165208945,
      "loss": 0.1215,
      "step": 15000
    },
    {
      "epoch": 1.06,
      "eval_cer": 0.14532439250522866,
      "eval_loss": 0.1390569657087326,
      "eval_runtime": 79.7847,
      "eval_samples_per_second": 62.669,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.3858,
      "step": 15000
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00029160152533908566,
      "loss": 0.1181,
      "step": 16000
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0002904568210808509,
      "loss": 0.1137,
      "step": 17000
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00028924155395529237,
      "loss": 0.11,
      "step": 18000
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00028795633448977547,
      "loss": 0.1067,
      "step": 19000
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.0002866018083544069,
      "loss": 0.1042,
      "step": 20000
    },
    {
      "epoch": 1.42,
      "eval_cer": 0.1343495978783501,
      "eval_loss": 0.11732806265354156,
      "eval_runtime": 79.7707,
      "eval_samples_per_second": 62.68,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.3624,
      "step": 20000
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.0002851786560376614,
      "loss": 0.101,
      "step": 21000
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.00028368759250451786,
      "loss": 0.0983,
      "step": 22000
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00028212936683727404,
      "loss": 0.0954,
      "step": 23000
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.0002805047618592216,
      "loss": 0.0933,
      "step": 24000
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.0002788145937413704,
      "loss": 0.0917,
      "step": 25000
    },
    {
      "epoch": 1.77,
      "eval_cer": 0.1251644063045775,
      "eval_loss": 0.10654760897159576,
      "eval_runtime": 79.786,
      "eval_samples_per_second": 62.668,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.3488,
      "step": 25000
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00027705971159241994,
      "loss": 0.0898,
      "step": 26000
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.0002752409970321824,
      "loss": 0.0881,
      "step": 27000
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.000273359363748674,
      "loss": 0.0865,
      "step": 28000
    },
    {
      "epoch": 2.05,
      "learning_rate": 0.0002714157570390952,
      "loss": 0.0834,
      "step": 29000
    },
    {
      "epoch": 2.13,
      "learning_rate": 0.0002694111533349308,
      "loss": 0.0821,
      "step": 30000
    },
    {
      "epoch": 2.13,
      "eval_cer": 0.11837253929580198,
      "eval_loss": 0.09495185315608978,
      "eval_runtime": 79.9216,
      "eval_samples_per_second": 62.561,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.3354,
      "step": 30000
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.00026734655971140956,
      "loss": 0.0806,
      "step": 31000
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.00026522301338156835,
      "loss": 0.0791,
      "step": 32000
    },
    {
      "epoch": 2.34,
      "learning_rate": 0.00026304158117517563,
      "loss": 0.0785,
      "step": 33000
    },
    {
      "epoch": 2.41,
      "learning_rate": 0.0002608033590027766,
      "loss": 0.0773,
      "step": 34000
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.000258509471305128,
      "loss": 0.0762,
      "step": 35000
    },
    {
      "epoch": 2.48,
      "eval_cer": 0.1154617391491839,
      "eval_loss": 0.08647703379392624,
      "eval_runtime": 79.5904,
      "eval_samples_per_second": 62.822,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.3272,
      "step": 35000
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.00025616107048830045,
      "loss": 0.0747,
      "step": 36000
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.0002537593363447311,
      "loss": 0.0741,
      "step": 37000
    },
    {
      "epoch": 2.69,
      "learning_rate": 0.0002513054754605181,
      "loss": 0.0734,
      "step": 38000
    },
    {
      "epoch": 2.76,
      "learning_rate": 0.0002488007206092547,
      "loss": 0.0724,
      "step": 39000
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.00024624633013270664,
      "loss": 0.0715,
      "step": 40000
    },
    {
      "epoch": 2.83,
      "eval_cer": 0.11341339830526748,
      "eval_loss": 0.08288487792015076,
      "eval_runtime": 79.5687,
      "eval_samples_per_second": 62.839,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.3176,
      "step": 40000
    },
    {
      "epoch": 2.9,
      "learning_rate": 0.0002436435873086449,
      "loss": 0.0708,
      "step": 41000
    },
    {
      "epoch": 2.98,
      "learning_rate": 0.00024099379970615137,
      "loss": 0.0695,
      "step": 42000
    },
    {
      "epoch": 3.05,
      "learning_rate": 0.00023829829852872037,
      "loss": 0.0677,
      "step": 43000
    },
    {
      "epoch": 3.12,
      "learning_rate": 0.00023555843794548724,
      "loss": 0.0666,
      "step": 44000
    },
    {
      "epoch": 3.19,
      "learning_rate": 0.00023277559441091917,
      "loss": 0.0661,
      "step": 45000
    },
    {
      "epoch": 3.19,
      "eval_cer": 0.11099851225770284,
      "eval_loss": 0.07811947166919708,
      "eval_runtime": 79.5629,
      "eval_samples_per_second": 62.843,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.316,
      "step": 45000
    },
    {
      "epoch": 3.26,
      "learning_rate": 0.0002299511659733103,
      "loss": 0.0655,
      "step": 46000
    },
    {
      "epoch": 3.33,
      "learning_rate": 0.0002270865715724285,
      "loss": 0.0651,
      "step": 47000
    },
    {
      "epoch": 3.4,
      "learning_rate": 0.0002241832503266669,
      "loss": 0.0645,
      "step": 48000
    },
    {
      "epoch": 3.47,
      "learning_rate": 0.00022124266081005755,
      "loss": 0.0641,
      "step": 49000
    },
    {
      "epoch": 3.54,
      "learning_rate": 0.0002182662803195115,
      "loss": 0.0636,
      "step": 50000
    },
    {
      "epoch": 3.54,
      "eval_cer": 0.10931671661743461,
      "eval_loss": 0.07468874752521515,
      "eval_runtime": 80.209,
      "eval_samples_per_second": 62.337,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.3074,
      "step": 50000
    },
    {
      "epoch": 3.61,
      "learning_rate": 0.00021525560413265246,
      "loss": 0.0629,
      "step": 51000
    },
    {
      "epoch": 3.68,
      "learning_rate": 0.00021221214475661745,
      "loss": 0.0619,
      "step": 52000
    },
    {
      "epoch": 3.75,
      "learning_rate": 0.00020913743116820165,
      "loss": 0.0619,
      "step": 53000
    },
    {
      "epoch": 3.83,
      "learning_rate": 0.00020603300804572917,
      "loss": 0.0608,
      "step": 54000
    },
    {
      "epoch": 3.9,
      "learning_rate": 0.00020290043499303573,
      "loss": 0.061,
      "step": 55000
    },
    {
      "epoch": 3.9,
      "eval_cer": 0.11015761443756873,
      "eval_loss": 0.07217952609062195,
      "eval_runtime": 79.9934,
      "eval_samples_per_second": 62.505,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.3086,
      "step": 55000
    },
    {
      "epoch": 3.97,
      "learning_rate": 0.00019974128575595291,
      "loss": 0.06,
      "step": 56000
    },
    {
      "epoch": 4.04,
      "learning_rate": 0.000196557147431688,
      "loss": 0.0585,
      "step": 57000
    },
    {
      "epoch": 4.11,
      "learning_rate": 0.00019334961967149597,
      "loss": 0.0577,
      "step": 58000
    },
    {
      "epoch": 4.18,
      "learning_rate": 0.00019012031387704483,
      "loss": 0.0574,
      "step": 59000
    },
    {
      "epoch": 4.25,
      "learning_rate": 0.00018687085239087786,
      "loss": 0.0569,
      "step": 60000
    },
    {
      "epoch": 4.25,
      "eval_cer": 0.10580219495892537,
      "eval_loss": 0.06902794539928436,
      "eval_runtime": 79.9992,
      "eval_samples_per_second": 62.501,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2988,
      "step": 60000
    },
    {
      "epoch": 4.32,
      "learning_rate": 0.00018360286768137894,
      "loss": 0.0567,
      "step": 61000
    },
    {
      "epoch": 4.39,
      "learning_rate": 0.00018031800152265137,
      "loss": 0.0561,
      "step": 62000
    },
    {
      "epoch": 4.46,
      "learning_rate": 0.00017701790416972107,
      "loss": 0.0558,
      "step": 63000
    },
    {
      "epoch": 4.53,
      "learning_rate": 0.00017370423352948002,
      "loss": 0.0557,
      "step": 64000
    },
    {
      "epoch": 4.6,
      "learning_rate": 0.00017037865432778442,
      "loss": 0.0552,
      "step": 65000
    },
    {
      "epoch": 4.6,
      "eval_cer": 0.10578063347635784,
      "eval_loss": 0.06745224446058273,
      "eval_runtime": 79.5425,
      "eval_samples_per_second": 62.859,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.2978,
      "step": 65000
    },
    {
      "epoch": 4.68,
      "learning_rate": 0.00016704283727312792,
      "loss": 0.0546,
      "step": 66000
    },
    {
      "epoch": 4.75,
      "learning_rate": 0.00016369845821730933,
      "loss": 0.0544,
      "step": 67000
    },
    {
      "epoch": 4.82,
      "learning_rate": 0.00016034719731351576,
      "loss": 0.0541,
      "step": 68000
    },
    {
      "epoch": 4.89,
      "learning_rate": 0.00015699073817224542,
      "loss": 0.0537,
      "step": 69000
    },
    {
      "epoch": 4.96,
      "learning_rate": 0.00015363076701549341,
      "loss": 0.0531,
      "step": 70000
    },
    {
      "epoch": 4.96,
      "eval_cer": 0.10444382155717027,
      "eval_loss": 0.06512928009033203,
      "eval_runtime": 79.8169,
      "eval_samples_per_second": 62.643,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2962,
      "step": 70000
    },
    {
      "epoch": 5.03,
      "learning_rate": 0.00015026897182962535,
      "loss": 0.0525,
      "step": 71000
    },
    {
      "epoch": 5.1,
      "learning_rate": 0.00014690704151736484,
      "loss": 0.0509,
      "step": 72000
    },
    {
      "epoch": 5.17,
      "learning_rate": 0.0001435466650493204,
      "loss": 0.0509,
      "step": 73000
    },
    {
      "epoch": 5.24,
      "learning_rate": 0.0001401895306154785,
      "loss": 0.0504,
      "step": 74000
    },
    {
      "epoch": 5.31,
      "learning_rate": 0.0001368373247770885,
      "loss": 0.0504,
      "step": 75000
    },
    {
      "epoch": 5.31,
      "eval_cer": 0.10291295629487483,
      "eval_loss": 0.06379809230566025,
      "eval_runtime": 79.7333,
      "eval_samples_per_second": 62.709,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2864,
      "step": 75000
    },
    {
      "epoch": 5.38,
      "learning_rate": 0.00013349173161936601,
      "loss": 0.0502,
      "step": 76000
    },
    {
      "epoch": 5.45,
      "learning_rate": 0.00013015443190544018,
      "loss": 0.0495,
      "step": 77000
    },
    {
      "epoch": 5.53,
      "learning_rate": 0.00012682710223196978,
      "loss": 0.0494,
      "step": 78000
    },
    {
      "epoch": 5.6,
      "learning_rate": 0.00012351141418685273,
      "loss": 0.0496,
      "step": 79000
    },
    {
      "epoch": 5.67,
      "learning_rate": 0.00012020903350945151,
      "loss": 0.0492,
      "step": 80000
    },
    {
      "epoch": 5.67,
      "eval_cer": 0.10310700963798271,
      "eval_loss": 0.06157052516937256,
      "eval_runtime": 79.8007,
      "eval_samples_per_second": 62.656,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2888,
      "step": 80000
    },
    {
      "epoch": 5.74,
      "learning_rate": 0.00011692161925375738,
      "loss": 0.0485,
      "step": 81000
    },
    {
      "epoch": 5.81,
      "learning_rate": 0.00011365082295491233,
      "loss": 0.0484,
      "step": 82000
    },
    {
      "epoch": 5.88,
      "learning_rate": 0.00011039828779950928,
      "loss": 0.0481,
      "step": 83000
    },
    {
      "epoch": 5.95,
      "learning_rate": 0.0001071656478000858,
      "loss": 0.048,
      "step": 84000
    },
    {
      "epoch": 6.02,
      "learning_rate": 0.00010395452697422703,
      "loss": 0.0474,
      "step": 85000
    },
    {
      "epoch": 6.02,
      "eval_cer": 0.10228767330041613,
      "eval_loss": 0.06011258438229561,
      "eval_runtime": 80.0834,
      "eval_samples_per_second": 62.435,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2868,
      "step": 85000
    },
    {
      "epoch": 6.09,
      "learning_rate": 0.00010076653852868962,
      "loss": 0.0459,
      "step": 86000
    },
    {
      "epoch": 6.16,
      "learning_rate": 9.760328404895728e-05,
      "loss": 0.0457,
      "step": 87000
    },
    {
      "epoch": 6.23,
      "learning_rate": 9.446635269463389e-05,
      "loss": 0.0453,
      "step": 88000
    },
    {
      "epoch": 6.3,
      "learning_rate": 9.135732040107998e-05,
      "loss": 0.0456,
      "step": 89000
    },
    {
      "epoch": 6.38,
      "learning_rate": 8.82777490876921e-05,
      "loss": 0.0453,
      "step": 90000
    },
    {
      "epoch": 6.38,
      "eval_cer": 0.1022661118178486,
      "eval_loss": 0.05900218337774277,
      "eval_runtime": 79.8035,
      "eval_samples_per_second": 62.654,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.284,
      "step": 90000
    },
    {
      "epoch": 6.45,
      "learning_rate": 8.522918587322418e-05,
      "loss": 0.0452,
      "step": 91000
    },
    {
      "epoch": 6.52,
      "learning_rate": 8.22131622985438e-05,
      "loss": 0.0451,
      "step": 92000
    },
    {
      "epoch": 6.59,
      "learning_rate": 7.923119355721526e-05,
      "loss": 0.045,
      "step": 93000
    },
    {
      "epoch": 6.66,
      "learning_rate": 7.628477773429438e-05,
      "loss": 0.0448,
      "step": 94000
    },
    {
      "epoch": 6.73,
      "learning_rate": 7.3375395053719e-05,
      "loss": 0.0446,
      "step": 95000
    },
    {
      "epoch": 6.73,
      "eval_cer": 0.10045494728217512,
      "eval_loss": 0.05785335227847099,
      "eval_runtime": 79.8939,
      "eval_samples_per_second": 62.583,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2766,
      "step": 95000
    },
    {
      "epoch": 6.8,
      "learning_rate": 7.050450713467216e-05,
      "loss": 0.0439,
      "step": 96000
    },
    {
      "epoch": 6.87,
      "learning_rate": 6.767355625729215e-05,
      "loss": 0.0439,
      "step": 97000
    },
    {
      "epoch": 6.94,
      "learning_rate": 6.488396463809805e-05,
      "loss": 0.0438,
      "step": 98000
    },
    {
      "epoch": 7.01,
      "learning_rate": 6.213713371549497e-05,
      "loss": 0.0434,
      "step": 99000
    },
    {
      "epoch": 7.08,
      "learning_rate": 5.943444344571755e-05,
      "loss": 0.042,
      "step": 100000
    },
    {
      "epoch": 7.08,
      "eval_cer": 0.09978654132258134,
      "eval_loss": 0.05739114433526993,
      "eval_runtime": 79.8794,
      "eval_samples_per_second": 62.594,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.277,
      "step": 100000
    },
    {
      "epoch": 7.15,
      "learning_rate": 5.6777251609566e-05,
      "loss": 0.042,
      "step": 101000
    },
    {
      "epoch": 7.23,
      "learning_rate": 5.416689313028253e-05,
      "loss": 0.0421,
      "step": 102000
    },
    {
      "epoch": 7.3,
      "learning_rate": 5.1604679402911015e-05,
      "loss": 0.0419,
      "step": 103000
    },
    {
      "epoch": 7.37,
      "learning_rate": 4.9091897635476726e-05,
      "loss": 0.0418,
      "step": 104000
    },
    {
      "epoch": 7.44,
      "learning_rate": 4.6629810202317166e-05,
      "loss": 0.0417,
      "step": 105000
    },
    {
      "epoch": 7.44,
      "eval_cer": 0.09941999611893314,
      "eval_loss": 0.056080542504787445,
      "eval_runtime": 79.81,
      "eval_samples_per_second": 62.649,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2742,
      "step": 105000
    },
    {
      "epoch": 7.51,
      "learning_rate": 4.421965400988918e-05,
      "loss": 0.0416,
      "step": 106000
    },
    {
      "epoch": 7.58,
      "learning_rate": 4.1862639875370016e-05,
      "loss": 0.0417,
      "step": 107000
    },
    {
      "epoch": 7.65,
      "learning_rate": 3.955995191836581e-05,
      "loss": 0.0414,
      "step": 108000
    },
    {
      "epoch": 7.72,
      "learning_rate": 3.731274696603201e-05,
      "loss": 0.0414,
      "step": 109000
    },
    {
      "epoch": 7.79,
      "learning_rate": 3.512215397190527e-05,
      "loss": 0.0411,
      "step": 110000
    },
    {
      "epoch": 7.79,
      "eval_cer": 0.1000237176308243,
      "eval_loss": 0.0557967908680439,
      "eval_runtime": 79.8362,
      "eval_samples_per_second": 62.628,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2756,
      "step": 110000
    },
    {
      "epoch": 7.86,
      "learning_rate": 3.298927344873821e-05,
      "loss": 0.0407,
      "step": 111000
    },
    {
      "epoch": 7.93,
      "learning_rate": 3.091517691562284e-05,
      "loss": 0.0407,
      "step": 112000
    },
    {
      "epoch": 8.01,
      "learning_rate": 2.890090635967912e-05,
      "loss": 0.0412,
      "step": 113000
    },
    {
      "epoch": 8.08,
      "learning_rate": 2.694747371258042e-05,
      "loss": 0.0398,
      "step": 114000
    },
    {
      "epoch": 8.15,
      "learning_rate": 2.505586034217802e-05,
      "loss": 0.0394,
      "step": 115000
    },
    {
      "epoch": 8.15,
      "eval_cer": 0.09991591021798658,
      "eval_loss": 0.05538313463330269,
      "eval_runtime": 79.9108,
      "eval_samples_per_second": 62.57,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2758,
      "step": 115000
    },
    {
      "epoch": 8.22,
      "learning_rate": 2.3227016559480356e-05,
      "loss": 0.0396,
      "step": 116000
    },
    {
      "epoch": 8.29,
      "learning_rate": 2.1461861141234628e-05,
      "loss": 0.0396,
      "step": 117000
    },
    {
      "epoch": 8.36,
      "learning_rate": 1.9761280868350654e-05,
      "loss": 0.0397,
      "step": 118000
    },
    {
      "epoch": 8.43,
      "learning_rate": 1.8126130080398725e-05,
      "loss": 0.0395,
      "step": 119000
    },
    {
      "epoch": 8.5,
      "learning_rate": 1.655723024640553e-05,
      "loss": 0.0394,
      "step": 120000
    },
    {
      "epoch": 8.5,
      "eval_cer": 0.09993747170055413,
      "eval_loss": 0.05500805750489235,
      "eval_runtime": 79.8857,
      "eval_samples_per_second": 62.589,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.275,
      "step": 120000
    },
    {
      "epoch": 8.57,
      "learning_rate": 1.5055369552163509e-05,
      "loss": 0.0394,
      "step": 121000
    },
    {
      "epoch": 8.64,
      "learning_rate": 1.3621302504261194e-05,
      "loss": 0.0395,
      "step": 122000
    },
    {
      "epoch": 8.71,
      "learning_rate": 1.2255749551033205e-05,
      "loss": 0.0393,
      "step": 123000
    },
    {
      "epoch": 8.78,
      "learning_rate": 1.0959396720620555e-05,
      "loss": 0.0393,
      "step": 124000
    },
    {
      "epoch": 8.86,
      "learning_rate": 9.732895276323099e-06,
      "loss": 0.0392,
      "step": 125000
    },
    {
      "epoch": 8.86,
      "eval_cer": 0.09957092649690592,
      "eval_loss": 0.054800018668174744,
      "eval_runtime": 80.1483,
      "eval_samples_per_second": 62.384,
      "eval_steps_per_second": 0.125,
      "eval_wer": 0.2756,
      "step": 125000
    },
    {
      "epoch": 8.93,
      "learning_rate": 8.576861389416939e-06,
      "loss": 0.0391,
      "step": 126000
    },
    {
      "epoch": 9.0,
      "learning_rate": 7.4918758296016734e-06,
      "loss": 0.039,
      "step": 127000
    },
    {
      "epoch": 9.07,
      "learning_rate": 6.478483673232537e-06,
      "loss": 0.0388,
      "step": 128000
    },
    {
      "epoch": 9.14,
      "learning_rate": 5.537194029484349e-06,
      "loss": 0.0383,
      "step": 129000
    },
    {
      "epoch": 9.21,
      "learning_rate": 4.668479784584583e-06,
      "loss": 0.0383,
      "step": 130000
    },
    {
      "epoch": 9.21,
      "eval_cer": 0.09931218870609543,
      "eval_loss": 0.054751720279455185,
      "eval_runtime": 80.4739,
      "eval_samples_per_second": 62.132,
      "eval_steps_per_second": 0.124,
      "eval_wer": 0.2748,
      "step": 130000
    },
    {
      "epoch": 9.28,
      "learning_rate": 3.872777364244245e-06,
      "loss": 0.0387,
      "step": 131000
    },
    {
      "epoch": 9.35,
      "learning_rate": 3.150486514405798e-06,
      "loss": 0.0385,
      "step": 132000
    },
    {
      "epoch": 9.42,
      "learning_rate": 2.50197010041826e-06,
      "loss": 0.0385,
      "step": 133000
    },
    {
      "epoch": 9.49,
      "learning_rate": 1.9275539247405437e-06,
      "loss": 0.0384,
      "step": 134000
    },
    {
      "epoch": 9.56,
      "learning_rate": 1.4275265632643806e-06,
      "loss": 0.0386,
      "step": 135000
    },
    {
      "epoch": 9.56,
      "eval_cer": 0.09924750425839281,
      "eval_loss": 0.05460319668054581,
      "eval_runtime": 80.404,
      "eval_samples_per_second": 62.186,
      "eval_steps_per_second": 0.124,
      "eval_wer": 0.2752,
      "step": 135000
    },
    {
      "epoch": 9.63,
      "learning_rate": 1.0021392203391932e-06,
      "loss": 0.0383,
      "step": 136000
    },
    {
      "epoch": 9.71,
      "learning_rate": 6.516056025718585e-07,
      "loss": 0.0383,
      "step": 137000
    },
    {
      "epoch": 9.78,
      "learning_rate": 3.76101811464391e-07,
      "loss": 0.0383,
      "step": 138000
    },
    {
      "epoch": 9.85,
      "learning_rate": 1.7576625494392648e-07,
      "loss": 0.0384,
      "step": 139000
    },
    {
      "epoch": 9.92,
      "learning_rate": 5.0699577829210744e-08,
      "loss": 0.0387,
      "step": 140000
    },
    {
      "epoch": 9.92,
      "eval_cer": 0.09937687315379806,
      "eval_loss": 0.05459373816847801,
      "eval_runtime": 80.3513,
      "eval_samples_per_second": 62.227,
      "eval_steps_per_second": 0.124,
      "eval_wer": 0.2752,
      "step": 140000
    }
  ],
  "max_steps": 141160,
  "num_train_epochs": 10,
  "total_flos": 3.547844813555835e+17,
  "trial_name": null,
  "trial_params": null
}