{
  "best_metric": 0.2446382894995737,
  "best_model_checkpoint": "/scratch/lingjzhu_root/lingjzhu1/lingjzhu/g2p/byt5_8_layers_baseline/checkpoint-5000",
  "epoch": 9.646302250803858,
  "global_step": 135000,
  "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": 2.3019,
      "step": 1000
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0002999616623572683,
      "loss": 0.7533,
      "step": 2000
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00029984666902607135,
      "loss": 0.4132,
      "step": 3000
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0002996550787873857,
      "loss": 0.3262,
      "step": 4000
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.0002993869895761197,
      "loss": 0.2811,
      "step": 5000
    },
    {
      "epoch": 0.36,
      "eval_cer": 0.2446382894995737,
      "eval_loss": 0.3433528244495392,
      "eval_runtime": 78.9346,
      "eval_samples_per_second": 62.71,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.5844444444444444,
      "step": 5000
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.000299042538431052,
      "loss": 0.2505,
      "step": 6000
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00029862190142478177,
      "loss": 0.2294,
      "step": 7000
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00029812529357372587,
      "loss": 0.2141,
      "step": 8000
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00029755296872820933,
      "loss": 0.2021,
      "step": 9000
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.0002969052194427048,
      "loss": 0.193,
      "step": 10000
    },
    {
      "epoch": 0.71,
      "eval_cer": 0.1929341291182965,
      "eval_loss": 0.23424917459487915,
      "eval_runtime": 78.954,
      "eval_samples_per_second": 62.695,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.4892929292929293,
      "step": 10000
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.0002961823768262882,
      "loss": 0.1847,
      "step": 11000
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.0002953848103733858,
      "loss": 0.1772,
      "step": 12000
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00029451292777490066,
      "loss": 0.1716,
      "step": 13000
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0002935671747098137,
      "loss": 0.165,
      "step": 14000
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00029254803461736643,
      "loss": 0.1599,
      "step": 15000
    },
    {
      "epoch": 1.07,
      "eval_cer": 0.17201198049889596,
      "eval_loss": 0.18778055906295776,
      "eval_runtime": 78.8177,
      "eval_samples_per_second": 62.803,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.43212121212121213,
      "step": 15000
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00029145602844994243,
      "loss": 0.1555,
      "step": 16000
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.0002902917144067724,
      "loss": 0.151,
      "step": 17000
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00028905568764860047,
      "loss": 0.1469,
      "step": 18000
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00028774857999345685,
      "loss": 0.144,
      "step": 19000
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.0002863710595936922,
      "loss": 0.1406,
      "step": 20000
    },
    {
      "epoch": 1.43,
      "eval_cer": 0.15764849915830437,
      "eval_loss": 0.16295142471790314,
      "eval_runtime": 79.2207,
      "eval_samples_per_second": 62.484,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.41353535353535353,
      "step": 20000
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0002849238305944389,
      "loss": 0.1368,
      "step": 21000
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00028340763277367477,
      "loss": 0.1345,
      "step": 22000
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.0002818232411640713,
      "loss": 0.132,
      "step": 23000
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00028017146565682144,
      "loss": 0.1303,
      "step": 24000
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00027845315058764886,
      "loss": 0.1261,
      "step": 25000
    },
    {
      "epoch": 1.79,
      "eval_cer": 0.14901292057453924,
      "eval_loss": 0.14625480771064758,
      "eval_runtime": 78.8658,
      "eval_samples_per_second": 62.765,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.39636363636363636,
      "step": 25000
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00027666917430520975,
      "loss": 0.1243,
      "step": 26000
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00027482044872210895,
      "loss": 0.1218,
      "step": 27000
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002729079188487587,
      "loss": 0.1196,
      "step": 28000
    },
    {
      "epoch": 2.07,
      "learning_rate": 0.00027093256231031885,
      "loss": 0.1174,
      "step": 29000
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.00026889538884696597,
      "loss": 0.116,
      "step": 30000
    },
    {
      "epoch": 2.14,
      "eval_cer": 0.1407708620275027,
      "eval_loss": 0.1319253146648407,
      "eval_runtime": 79.0974,
      "eval_samples_per_second": 62.581,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.37696969696969695,
      "step": 30000
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.0002667974397977457,
      "loss": 0.114,
      "step": 31000
    },
    {
      "epoch": 2.29,
      "learning_rate": 0.0002646397875682729,
      "loss": 0.1122,
      "step": 32000
    },
    {
      "epoch": 2.36,
      "learning_rate": 0.00026242353508255185,
      "loss": 0.111,
      "step": 33000
    },
    {
      "epoch": 2.43,
      "learning_rate": 0.0002601498152191957,
      "loss": 0.1099,
      "step": 34000
    },
    {
      "epoch": 2.5,
      "learning_rate": 0.0002578197902323352,
      "loss": 0.1089,
      "step": 35000
    },
    {
      "epoch": 2.5,
      "eval_cer": 0.13464943923394765,
      "eval_loss": 0.12354536354541779,
      "eval_runtime": 79.0393,
      "eval_samples_per_second": 62.627,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.36505050505050507,
      "step": 35000
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.00025543465115751026,
      "loss": 0.1067,
      "step": 36000
    },
    {
      "epoch": 2.64,
      "learning_rate": 0.0002529956172028505,
      "loss": 0.1058,
      "step": 37000
    },
    {
      "epoch": 2.72,
      "learning_rate": 0.0002505039351258541,
      "loss": 0.1043,
      "step": 38000
    },
    {
      "epoch": 2.79,
      "learning_rate": 0.0002479608785960846,
      "loss": 0.1033,
      "step": 39000
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.0002453677475441111,
      "loss": 0.1017,
      "step": 40000
    },
    {
      "epoch": 2.86,
      "eval_cer": 0.1299927854659933,
      "eval_loss": 0.11381864547729492,
      "eval_runtime": 79.0365,
      "eval_samples_per_second": 62.629,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.3557575757575758,
      "step": 40000
    },
    {
      "epoch": 2.93,
      "learning_rate": 0.00024272586749702474,
      "loss": 0.1001,
      "step": 41000
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0002400365889008706,
      "loss": 0.1001,
      "step": 42000
    },
    {
      "epoch": 3.07,
      "learning_rate": 0.00023730128643034235,
      "loss": 0.0972,
      "step": 43000
    },
    {
      "epoch": 3.14,
      "learning_rate": 0.00023452135828609167,
      "loss": 0.0974,
      "step": 44000
    },
    {
      "epoch": 3.22,
      "learning_rate": 0.0002316982254800121,
      "loss": 0.0963,
      "step": 45000
    },
    {
      "epoch": 3.22,
      "eval_cer": 0.12660414070527534,
      "eval_loss": 0.10989916324615479,
      "eval_runtime": 79.0323,
      "eval_samples_per_second": 62.633,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.3470707070707071,
      "step": 45000
    },
    {
      "epoch": 3.29,
      "learning_rate": 0.00022883333110886237,
      "loss": 0.0957,
      "step": 46000
    },
    {
      "epoch": 3.36,
      "learning_rate": 0.00022592813961660067,
      "loss": 0.0944,
      "step": 47000
    },
    {
      "epoch": 3.43,
      "learning_rate": 0.00022298413604580696,
      "loss": 0.0934,
      "step": 48000
    },
    {
      "epoch": 3.5,
      "learning_rate": 0.00022000282527857588,
      "loss": 0.093,
      "step": 49000
    },
    {
      "epoch": 3.57,
      "learning_rate": 0.0002169857312672683,
      "loss": 0.0919,
      "step": 50000
    },
    {
      "epoch": 3.57,
      "eval_cer": 0.12306246037471852,
      "eval_loss": 0.10183772444725037,
      "eval_runtime": 79.0109,
      "eval_samples_per_second": 62.65,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.3395959595959596,
      "step": 50000
    },
    {
      "epoch": 3.64,
      "learning_rate": 0.00021393439625551483,
      "loss": 0.0916,
      "step": 51000
    },
    {
      "epoch": 3.72,
      "learning_rate": 0.00021085037998986924,
      "loss": 0.0908,
      "step": 52000
    },
    {
      "epoch": 3.79,
      "learning_rate": 0.00020773525892251514,
      "loss": 0.0902,
      "step": 53000
    },
    {
      "epoch": 3.86,
      "learning_rate": 0.00020459062540543316,
      "loss": 0.0891,
      "step": 54000
    },
    {
      "epoch": 3.93,
      "learning_rate": 0.00020141808687644067,
      "loss": 0.0885,
      "step": 55000
    },
    {
      "epoch": 3.93,
      "eval_cer": 0.12083251349992348,
      "eval_loss": 0.09825348109006882,
      "eval_runtime": 79.07,
      "eval_samples_per_second": 62.603,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.33090909090909093,
      "step": 55000
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.00019821926503751995,
      "loss": 0.0878,
      "step": 56000
    },
    {
      "epoch": 4.07,
      "learning_rate": 0.00019499579502585537,
      "loss": 0.0867,
      "step": 57000
    },
    {
      "epoch": 4.14,
      "learning_rate": 0.00019174932457800242,
      "loss": 0.0856,
      "step": 58000
    },
    {
      "epoch": 4.22,
      "learning_rate": 0.0001884815131876167,
      "loss": 0.0855,
      "step": 59000
    },
    {
      "epoch": 4.29,
      "learning_rate": 0.00018519403125717278,
      "loss": 0.0843,
      "step": 60000
    },
    {
      "epoch": 4.29,
      "eval_cer": 0.11809973546708642,
      "eval_loss": 0.09398272633552551,
      "eval_runtime": 79.1077,
      "eval_samples_per_second": 62.573,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.32525252525252524,
      "step": 60000
    },
    {
      "epoch": 4.36,
      "learning_rate": 0.00018188855924410722,
      "loss": 0.0846,
      "step": 61000
    },
    {
      "epoch": 4.43,
      "learning_rate": 0.00017856678680182127,
      "loss": 0.0836,
      "step": 62000
    },
    {
      "epoch": 4.5,
      "learning_rate": 0.0001752304119159834,
      "loss": 0.0832,
      "step": 63000
    },
    {
      "epoch": 4.57,
      "learning_rate": 0.00017188114003657205,
      "loss": 0.0828,
      "step": 64000
    },
    {
      "epoch": 4.64,
      "learning_rate": 0.00016852068320610358,
      "loss": 0.0823,
      "step": 65000
    },
    {
      "epoch": 4.64,
      "eval_cer": 0.11665682866574845,
      "eval_loss": 0.09152530878782272,
      "eval_runtime": 79.2818,
      "eval_samples_per_second": 62.435,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.32484848484848483,
      "step": 65000
    },
    {
      "epoch": 4.72,
      "learning_rate": 0.00016515075918448972,
      "loss": 0.0818,
      "step": 66000
    },
    {
      "epoch": 4.79,
      "learning_rate": 0.00016177309057097285,
      "loss": 0.0806,
      "step": 67000
    },
    {
      "epoch": 4.86,
      "learning_rate": 0.00015838940392358722,
      "loss": 0.0807,
      "step": 68000
    },
    {
      "epoch": 4.93,
      "learning_rate": 0.00015500142887659688,
      "loss": 0.0799,
      "step": 69000
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.00015161089725636095,
      "loss": 0.0795,
      "step": 70000
    },
    {
      "epoch": 5.0,
      "eval_cer": 0.11567302857392711,
      "eval_loss": 0.08921054005622864,
      "eval_runtime": 79.061,
      "eval_samples_per_second": 62.61,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.321010101010101,
      "step": 70000
    },
    {
      "epoch": 5.07,
      "learning_rate": 0.00014821954219607845,
      "loss": 0.0783,
      "step": 71000
    },
    {
      "epoch": 5.14,
      "learning_rate": 0.0001448290972498651,
      "loss": 0.0784,
      "step": 72000
    },
    {
      "epoch": 5.22,
      "learning_rate": 0.00014144129550661485,
      "loss": 0.0778,
      "step": 73000
    },
    {
      "epoch": 5.29,
      "learning_rate": 0.0001380578687040995,
      "loss": 0.0773,
      "step": 74000
    },
    {
      "epoch": 5.36,
      "learning_rate": 0.00013468054634375843,
      "loss": 0.0772,
      "step": 75000
    },
    {
      "epoch": 5.36,
      "eval_cer": 0.11359611726897094,
      "eval_loss": 0.08604130893945694,
      "eval_runtime": 79.1221,
      "eval_samples_per_second": 62.562,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.31777777777777777,
      "step": 75000
    },
    {
      "epoch": 5.43,
      "learning_rate": 0.00013131105480663235,
      "loss": 0.0767,
      "step": 76000
    },
    {
      "epoch": 5.5,
      "learning_rate": 0.000127951116470891,
      "loss": 0.0766,
      "step": 77000
    },
    {
      "epoch": 5.57,
      "learning_rate": 0.00012460244883140783,
      "loss": 0.0757,
      "step": 78000
    },
    {
      "epoch": 5.64,
      "learning_rate": 0.0001212667636218309,
      "loss": 0.0753,
      "step": 79000
    },
    {
      "epoch": 5.72,
      "learning_rate": 0.00011794576593959775,
      "loss": 0.075,
      "step": 80000
    },
    {
      "epoch": 5.72,
      "eval_cer": 0.11416453509980105,
      "eval_loss": 0.08475763350725174,
      "eval_runtime": 79.0104,
      "eval_samples_per_second": 62.65,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.317979797979798,
      "step": 80000
    },
    {
      "epoch": 5.79,
      "learning_rate": 0.00011464115337434394,
      "loss": 0.075,
      "step": 81000
    },
    {
      "epoch": 5.86,
      "learning_rate": 0.00011135461514014796,
      "loss": 0.0742,
      "step": 82000
    },
    {
      "epoch": 5.93,
      "learning_rate": 0.00010808783121205837,
      "loss": 0.0743,
      "step": 83000
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.00010484247146734352,
      "loss": 0.0736,
      "step": 84000
    },
    {
      "epoch": 6.07,
      "learning_rate": 0.00010162019483190237,
      "loss": 0.0725,
      "step": 85000
    },
    {
      "epoch": 6.07,
      "eval_cer": 0.11407708620275027,
      "eval_loss": 0.081719771027565,
      "eval_runtime": 79.1888,
      "eval_samples_per_second": 62.509,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.31373737373737376,
      "step": 85000
    },
    {
      "epoch": 6.15,
      "learning_rate": 9.842264843227404e-05,
      "loss": 0.0727,
      "step": 86000
    },
    {
      "epoch": 6.22,
      "learning_rate": 9.52514667536784e-05,
      "loss": 0.0718,
      "step": 87000
    },
    {
      "epoch": 6.29,
      "learning_rate": 9.210827080451842e-05,
      "loss": 0.0719,
      "step": 88000
    },
    {
      "epoch": 6.36,
      "learning_rate": 8.899466728777203e-05,
      "loss": 0.0718,
      "step": 89000
    },
    {
      "epoch": 6.43,
      "learning_rate": 8.591224777969557e-05,
      "loss": 0.0713,
      "step": 90000
    },
    {
      "epoch": 6.43,
      "eval_cer": 0.11197831267353141,
      "eval_loss": 0.08096928149461746,
      "eval_runtime": 78.7279,
      "eval_samples_per_second": 62.875,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.31353535353535356,
      "step": 90000
    },
    {
      "epoch": 6.5,
      "learning_rate": 8.286258791626041e-05,
      "loss": 0.071,
      "step": 91000
    },
    {
      "epoch": 6.57,
      "learning_rate": 7.984724658773716e-05,
      "loss": 0.071,
      "step": 92000
    },
    {
      "epoch": 6.65,
      "learning_rate": 7.686776514184009e-05,
      "loss": 0.0708,
      "step": 93000
    },
    {
      "epoch": 6.72,
      "learning_rate": 7.392566659583846e-05,
      "loss": 0.0705,
      "step": 94000
    },
    {
      "epoch": 6.79,
      "learning_rate": 7.102245485803813e-05,
      "loss": 0.07,
      "step": 95000
    },
    {
      "epoch": 6.79,
      "eval_cer": 0.11123499704859972,
      "eval_loss": 0.07890674471855164,
      "eval_runtime": 79.0377,
      "eval_samples_per_second": 62.628,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.3103030303030303,
      "step": 95000
    },
    {
      "epoch": 6.86,
      "learning_rate": 6.81596139590308e-05,
      "loss": 0.0701,
      "step": 96000
    },
    {
      "epoch": 6.93,
      "learning_rate": 6.533860729310434e-05,
      "loss": 0.0698,
      "step": 97000
    },
    {
      "epoch": 7.0,
      "learning_rate": 6.256087687020127e-05,
      "loss": 0.0698,
      "step": 98000
    },
    {
      "epoch": 7.07,
      "learning_rate": 5.98278425788092e-05,
      "loss": 0.0687,
      "step": 99000
    },
    {
      "epoch": 7.15,
      "learning_rate": 5.71409014601578e-05,
      "loss": 0.0682,
      "step": 100000
    },
    {
      "epoch": 7.15,
      "eval_cer": 0.11134430816991321,
      "eval_loss": 0.07869766652584076,
      "eval_runtime": 78.9978,
      "eval_samples_per_second": 62.66,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.3088888888888889,
      "step": 100000
    },
    {
      "epoch": 7.22,
      "learning_rate": 5.4501426994095876e-05,
      "loss": 0.0682,
      "step": 101000
    },
    {
      "epoch": 7.29,
      "learning_rate": 5.191076839701103e-05,
      "loss": 0.0681,
      "step": 102000
    },
    {
      "epoch": 7.36,
      "learning_rate": 4.9370249932153075e-05,
      "loss": 0.0681,
      "step": 103000
    },
    {
      "epoch": 7.43,
      "learning_rate": 4.6881170232712164e-05,
      "loss": 0.0682,
      "step": 104000
    },
    {
      "epoch": 7.5,
      "learning_rate": 4.444480163799822e-05,
      "loss": 0.0681,
      "step": 105000
    },
    {
      "epoch": 7.5,
      "eval_cer": 0.11066657921776962,
      "eval_loss": 0.07721372693777084,
      "eval_runtime": 79.1073,
      "eval_samples_per_second": 62.573,
      "eval_steps_per_second": 0.126,
      "eval_wer": 0.30808080808080807,
      "step": 105000
    },
    {
      "epoch": 7.57,
      "learning_rate": 4.2062389543061265e-05,
      "loss": 0.0673,
      "step": 106000
    },
    {
      "epoch": 7.65,
      "learning_rate": 3.9735151762084384e-05,
      "loss": 0.0674,
      "step": 107000
    },
    {
      "epoch": 7.72,
      "learning_rate": 3.746427790587557e-05,
      "loss": 0.0666,
      "step": 108000
    },
    {
      "epoch": 7.79,
      "learning_rate": 3.525092877377602e-05,
      "loss": 0.0674,
      "step": 109000
    },
    {
      "epoch": 7.86,
      "learning_rate": 3.309623576029597e-05,
      "loss": 0.0666,
      "step": 110000
    },
    {
      "epoch": 7.86,
      "eval_cer": 0.11116941037581164,
      "eval_loss": 0.07652007043361664,
      "eval_runtime": 78.7656,
      "eval_samples_per_second": 62.845,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.30686868686868685,
      "step": 110000
    },
    {
      "epoch": 7.93,
      "learning_rate": 3.1001300276781274e-05,
      "loss": 0.0668,
      "step": 111000
    },
    {
      "epoch": 8.0,
      "learning_rate": 2.8967193188406938e-05,
      "loss": 0.0663,
      "step": 112000
    },
    {
      "epoch": 8.07,
      "learning_rate": 2.699495426678389e-05,
      "loss": 0.066,
      "step": 113000
    },
    {
      "epoch": 8.15,
      "learning_rate": 2.5085591658461056e-05,
      "loss": 0.0658,
      "step": 114000
    },
    {
      "epoch": 8.22,
      "learning_rate": 2.3240081369591984e-05,
      "loss": 0.0656,
      "step": 115000
    },
    {
      "epoch": 8.22,
      "eval_cer": 0.11088520146039658,
      "eval_loss": 0.07624918967485428,
      "eval_runtime": 78.9004,
      "eval_samples_per_second": 62.737,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.30727272727272725,
      "step": 115000
    },
    {
      "epoch": 8.29,
      "learning_rate": 2.1459366767031522e-05,
      "loss": 0.0656,
      "step": 116000
    },
    {
      "epoch": 8.36,
      "learning_rate": 1.9744358096116225e-05,
      "loss": 0.0654,
      "step": 117000
    },
    {
      "epoch": 8.43,
      "learning_rate": 1.8095932015375496e-05,
      "loss": 0.0655,
      "step": 118000
    },
    {
      "epoch": 8.5,
      "learning_rate": 1.65149311484114e-05,
      "loss": 0.0651,
      "step": 119000
    },
    {
      "epoch": 8.57,
      "learning_rate": 1.500216365317587e-05,
      "loss": 0.0655,
      "step": 120000
    },
    {
      "epoch": 8.57,
      "eval_cer": 0.11158479263680286,
      "eval_loss": 0.07576470077037811,
      "eval_runtime": 79.0474,
      "eval_samples_per_second": 62.621,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.30808080808080807,
      "step": 120000
    },
    {
      "epoch": 8.65,
      "learning_rate": 1.355840280886582e-05,
      "loss": 0.0652,
      "step": 121000
    },
    {
      "epoch": 8.72,
      "learning_rate": 1.2184386620647097e-05,
      "loss": 0.0656,
      "step": 122000
    },
    {
      "epoch": 8.79,
      "learning_rate": 1.0880817442409478e-05,
      "loss": 0.0655,
      "step": 123000
    },
    {
      "epoch": 8.86,
      "learning_rate": 9.648361617745371e-06,
      "loss": 0.0651,
      "step": 124000
    },
    {
      "epoch": 8.93,
      "learning_rate": 8.487649139335962e-06,
      "loss": 0.0652,
      "step": 125000
    },
    {
      "epoch": 8.93,
      "eval_cer": 0.1107321658905577,
      "eval_loss": 0.07536973804235458,
      "eval_runtime": 79.0386,
      "eval_samples_per_second": 62.628,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.30707070707070705,
      "step": 125000
    },
    {
      "epoch": 9.0,
      "learning_rate": 7.399273326918692e-06,
      "loss": 0.065,
      "step": 126000
    },
    {
      "epoch": 9.07,
      "learning_rate": 6.383790524001009e-06,
      "loss": 0.0645,
      "step": 127000
    },
    {
      "epoch": 9.15,
      "learning_rate": 5.441719813474849e-06,
      "loss": 0.0647,
      "step": 128000
    },
    {
      "epoch": 9.22,
      "learning_rate": 4.57354275227797e-06,
      "loss": 0.0649,
      "step": 129000
    },
    {
      "epoch": 9.29,
      "learning_rate": 3.7797031252369767e-06,
      "loss": 0.0642,
      "step": 130000
    },
    {
      "epoch": 9.29,
      "eval_cer": 0.11057913032071882,
      "eval_loss": 0.07509743422269821,
      "eval_runtime": 78.953,
      "eval_samples_per_second": 62.695,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.30747474747474746,
      "step": 130000
    },
    {
      "epoch": 9.36,
      "learning_rate": 3.0606067182186776e-06,
      "loss": 0.0642,
      "step": 131000
    },
    {
      "epoch": 9.43,
      "learning_rate": 2.4166211107049584e-06,
      "loss": 0.0644,
      "step": 132000
    },
    {
      "epoch": 9.5,
      "learning_rate": 1.8480754878977489e-06,
      "loss": 0.0644,
      "step": 133000
    },
    {
      "epoch": 9.57,
      "learning_rate": 1.3552604724498928e-06,
      "loss": 0.0648,
      "step": 134000
    },
    {
      "epoch": 9.65,
      "learning_rate": 9.384279759080127e-07,
      "loss": 0.0645,
      "step": 135000
    },
    {
      "epoch": 9.65,
      "eval_cer": 0.1107758903390831,
      "eval_loss": 0.07504051923751831,
      "eval_runtime": 78.9794,
      "eval_samples_per_second": 62.675,
      "eval_steps_per_second": 0.127,
      "eval_wer": 0.30787878787878786,
      "step": 135000
    }
  ],
  "max_steps": 139950,
  "num_train_epochs": 10,
  "total_flos": 1.1962625697651917e+17,
  "trial_name": null,
  "trial_params": null
}