{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.936550491510276,
  "eval_steps": 200,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 0.3254,
      "step": 100
    },
    {
      "epoch": 0.36,
      "learning_rate": 4e-05,
      "loss": 0.1748,
      "step": 200
    },
    {
      "epoch": 0.36,
      "eval_f1_score": 0.654103852596315,
      "eval_label_f1": 0.803182579564489,
      "eval_loss": 0.17060260474681854,
      "eval_runtime": 676.0932,
      "eval_samples_per_second": 1.479,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.06556235002709188,
      "step": 200
    },
    {
      "epoch": 0.54,
      "learning_rate": 6e-05,
      "loss": 0.1713,
      "step": 300
    },
    {
      "epoch": 0.71,
      "learning_rate": 8e-05,
      "loss": 0.1754,
      "step": 400
    },
    {
      "epoch": 0.71,
      "eval_f1_score": 0.7194186515946711,
      "eval_label_f1": 0.8502220427937021,
      "eval_loss": 0.17686782777309418,
      "eval_runtime": 678.74,
      "eval_samples_per_second": 1.473,
      "eval_steps_per_second": 0.184,
      "eval_wer": 0.06738137626751296,
      "step": 400
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0001,
      "loss": 0.1815,
      "step": 500
    },
    {
      "epoch": 1.07,
      "learning_rate": 9.987820251299122e-05,
      "loss": 0.1606,
      "step": 600
    },
    {
      "epoch": 1.07,
      "eval_f1_score": 0.6991492652745553,
      "eval_label_f1": 0.8406805877803557,
      "eval_loss": 0.18560311198234558,
      "eval_runtime": 678.1594,
      "eval_samples_per_second": 1.475,
      "eval_steps_per_second": 0.184,
      "eval_wer": 0.0708259153185231,
      "step": 600
    },
    {
      "epoch": 1.25,
      "learning_rate": 9.951340343707852e-05,
      "loss": 0.1282,
      "step": 700
    },
    {
      "epoch": 1.43,
      "learning_rate": 9.890738003669029e-05,
      "loss": 0.1282,
      "step": 800
    },
    {
      "epoch": 1.43,
      "eval_f1_score": 0.745497259201253,
      "eval_label_f1": 0.8723570869224745,
      "eval_loss": 0.18348248302936554,
      "eval_runtime": 676.9962,
      "eval_samples_per_second": 1.477,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.0727610496168434,
      "step": 800
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.806308479691595e-05,
      "loss": 0.1312,
      "step": 900
    },
    {
      "epoch": 1.79,
      "learning_rate": 9.698463103929542e-05,
      "loss": 0.131,
      "step": 1000
    },
    {
      "epoch": 1.79,
      "eval_f1_score": 0.7330932372949178,
      "eval_label_f1": 0.8691476590636255,
      "eval_loss": 0.1762365996837616,
      "eval_runtime": 674.9365,
      "eval_samples_per_second": 1.482,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.07132905023608638,
      "step": 1000
    },
    {
      "epoch": 1.97,
      "learning_rate": 9.567727288213005e-05,
      "loss": 0.1313,
      "step": 1100
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.414737964294636e-05,
      "loss": 0.0804,
      "step": 1200
    },
    {
      "epoch": 2.14,
      "eval_f1_score": 0.7544,
      "eval_label_f1": 0.8744,
      "eval_loss": 0.17920058965682983,
      "eval_runtime": 680.9536,
      "eval_samples_per_second": 1.469,
      "eval_steps_per_second": 0.184,
      "eval_wer": 0.06850375416053874,
      "step": 1200
    },
    {
      "epoch": 2.32,
      "learning_rate": 9.24024048078213e-05,
      "loss": 0.0681,
      "step": 1300
    },
    {
      "epoch": 2.5,
      "learning_rate": 9.045084971874738e-05,
      "loss": 0.0712,
      "step": 1400
    },
    {
      "epoch": 2.5,
      "eval_f1_score": 0.75,
      "eval_label_f1": 0.8845541401273885,
      "eval_loss": 0.1833120882511139,
      "eval_runtime": 675.2132,
      "eval_samples_per_second": 1.481,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.06908429445003483,
      "step": 1400
    },
    {
      "epoch": 2.68,
      "learning_rate": 8.83022221559489e-05,
      "loss": 0.0721,
      "step": 1500
    },
    {
      "epoch": 2.86,
      "learning_rate": 8.596699001693255e-05,
      "loss": 0.0746,
      "step": 1600
    },
    {
      "epoch": 2.86,
      "eval_f1_score": 0.7553648068669527,
      "eval_label_f1": 0.8731954740538431,
      "eval_loss": 0.18000106513500214,
      "eval_runtime": 681.7256,
      "eval_samples_per_second": 1.467,
      "eval_steps_per_second": 0.183,
      "eval_wer": 0.07380602213793637,
      "step": 1600
    },
    {
      "epoch": 3.04,
      "learning_rate": 8.345653031794292e-05,
      "loss": 0.0663,
      "step": 1700
    },
    {
      "epoch": 3.22,
      "learning_rate": 8.07830737662829e-05,
      "loss": 0.0331,
      "step": 1800
    },
    {
      "epoch": 3.22,
      "eval_f1_score": 0.7756897241103559,
      "eval_label_f1": 0.8804478208716514,
      "eval_loss": 0.19921182096004486,
      "eval_runtime": 675.6212,
      "eval_samples_per_second": 1.48,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.0702066723430606,
      "step": 1800
    },
    {
      "epoch": 3.4,
      "learning_rate": 7.795964517353735e-05,
      "loss": 0.0338,
      "step": 1900
    },
    {
      "epoch": 3.57,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.0363,
      "step": 2000
    },
    {
      "epoch": 3.57,
      "eval_f1_score": 0.7625498007968129,
      "eval_label_f1": 0.8804780876494023,
      "eval_loss": 0.19381307065486908,
      "eval_runtime": 675.5559,
      "eval_samples_per_second": 1.48,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.06877467296230358,
      "step": 2000
    },
    {
      "epoch": 3.75,
      "learning_rate": 7.191855733945387e-05,
      "loss": 0.036,
      "step": 2100
    },
    {
      "epoch": 3.93,
      "learning_rate": 6.873032967079561e-05,
      "loss": 0.037,
      "step": 2200
    },
    {
      "epoch": 3.93,
      "eval_f1_score": 0.7771474878444085,
      "eval_label_f1": 0.886547811993517,
      "eval_loss": 0.19859156012535095,
      "eval_runtime": 676.3432,
      "eval_samples_per_second": 1.479,
      "eval_steps_per_second": 0.185,
      "eval_wer": 0.0676522950692778,
      "step": 2200
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.545084971874738e-05,
      "loss": 0.0244,
      "step": 2300
    },
    {
      "epoch": 4.29,
      "learning_rate": 6.209609477998338e-05,
      "loss": 0.0153,
      "step": 2400
    },
    {
      "epoch": 4.29,
      "eval_f1_score": 0.7764609726195342,
      "eval_label_f1": 0.8794442174090723,
      "eval_loss": 0.21251419186592102,
      "eval_runtime": 47767.7011,
      "eval_samples_per_second": 0.021,
      "eval_steps_per_second": 0.003,
      "eval_wer": 0.06664602523415125,
      "step": 2400
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.868240888334653e-05,
      "loss": 0.0149,
      "step": 2500
    },
    {
      "epoch": 4.65,
      "learning_rate": 5.522642316338268e-05,
      "loss": 0.0144,
      "step": 2600
    },
    {
      "epoch": 4.65,
      "eval_f1_score": 0.7763475462590508,
      "eval_label_f1": 0.8921962992759452,
      "eval_loss": 0.2115294188261032,
      "eval_runtime": 711.4482,
      "eval_samples_per_second": 1.406,
      "eval_steps_per_second": 0.176,
      "eval_wer": 0.06811672730087467,
      "step": 2600
    },
    {
      "epoch": 4.83,
      "learning_rate": 5.174497483512506e-05,
      "loss": 0.0148,
      "step": 2700
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.825502516487497e-05,
      "loss": 0.0148,
      "step": 2800
    },
    {
      "epoch": 5.0,
      "eval_f1_score": 0.7781350482315111,
      "eval_label_f1": 0.8890675241157556,
      "eval_loss": 0.21800462901592255,
      "eval_runtime": 690.6474,
      "eval_samples_per_second": 1.448,
      "eval_steps_per_second": 0.181,
      "eval_wer": 0.06474959362179736,
      "step": 2800
    },
    {
      "epoch": 5.18,
      "learning_rate": 4.477357683661734e-05,
      "loss": 0.0062,
      "step": 2900
    },
    {
      "epoch": 5.36,
      "learning_rate": 4.131759111665349e-05,
      "loss": 0.0058,
      "step": 3000
    },
    {
      "epoch": 5.36,
      "eval_f1_score": 0.7918170878459687,
      "eval_label_f1": 0.8912956277577216,
      "eval_loss": 0.2309972196817398,
      "eval_runtime": 708.2707,
      "eval_samples_per_second": 1.412,
      "eval_steps_per_second": 0.176,
      "eval_wer": 0.06289186469540986,
      "step": 3000
    },
    {
      "epoch": 5.54,
      "learning_rate": 3.790390522001662e-05,
      "loss": 0.0056,
      "step": 3100
    },
    {
      "epoch": 5.72,
      "learning_rate": 3.4549150281252636e-05,
      "loss": 0.0058,
      "step": 3200
    },
    {
      "epoch": 5.72,
      "eval_f1_score": 0.7828200972447327,
      "eval_label_f1": 0.8938411669367909,
      "eval_loss": 0.226795494556427,
      "eval_runtime": 688.2964,
      "eval_samples_per_second": 1.453,
      "eval_steps_per_second": 0.182,
      "eval_wer": 0.06269835126557784,
      "step": 3200
    },
    {
      "epoch": 5.9,
      "learning_rate": 3.12696703292044e-05,
      "loss": 0.0053,
      "step": 3300
    },
    {
      "epoch": 6.08,
      "learning_rate": 2.8081442660546125e-05,
      "loss": 0.0036,
      "step": 3400
    },
    {
      "epoch": 6.08,
      "eval_f1_score": 0.7911111111111112,
      "eval_label_f1": 0.8937373737373737,
      "eval_loss": 0.24620352685451508,
      "eval_runtime": 710.624,
      "eval_samples_per_second": 1.407,
      "eval_steps_per_second": 0.176,
      "eval_wer": 0.062079108290115335,
      "step": 3400
    },
    {
      "epoch": 6.26,
      "learning_rate": 2.500000000000001e-05,
      "loss": 0.0021,
      "step": 3500
    },
    {
      "epoch": 6.43,
      "learning_rate": 2.2040354826462668e-05,
      "loss": 0.0019,
      "step": 3600
    },
    {
      "epoch": 6.43,
      "eval_f1_score": 0.7948303715670436,
      "eval_label_f1": 0.8949919224555736,
      "eval_loss": 0.2493194341659546,
      "eval_runtime": 697.0542,
      "eval_samples_per_second": 1.435,
      "eval_steps_per_second": 0.179,
      "eval_wer": 0.06285316200944345,
      "step": 3600
    },
    {
      "epoch": 6.61,
      "learning_rate": 1.9216926233717085e-05,
      "loss": 0.0019,
      "step": 3700
    },
    {
      "epoch": 6.79,
      "learning_rate": 1.6543469682057106e-05,
      "loss": 0.0016,
      "step": 3800
    },
    {
      "epoch": 6.79,
      "eval_f1_score": 0.7916833266693323,
      "eval_label_f1": 0.8980407836865254,
      "eval_loss": 0.25429993867874146,
      "eval_runtime": 712.0537,
      "eval_samples_per_second": 1.404,
      "eval_steps_per_second": 0.176,
      "eval_wer": 0.0631240808112083,
      "step": 3800
    },
    {
      "epoch": 6.97,
      "learning_rate": 1.4033009983067452e-05,
      "loss": 0.0016,
      "step": 3900
    },
    {
      "epoch": 7.15,
      "learning_rate": 1.1697777844051105e-05,
      "loss": 0.0009,
      "step": 4000
    },
    {
      "epoch": 7.15,
      "eval_f1_score": 0.7879999999999999,
      "eval_label_f1": 0.8944,
      "eval_loss": 0.2667454779148102,
      "eval_runtime": 705.4777,
      "eval_samples_per_second": 1.417,
      "eval_steps_per_second": 0.177,
      "eval_wer": 0.06192429754624971,
      "step": 4000
    },
    {
      "epoch": 7.33,
      "learning_rate": 9.549150281252633e-06,
      "loss": 0.0007,
      "step": 4100
    },
    {
      "epoch": 7.51,
      "learning_rate": 7.597595192178702e-06,
      "loss": 0.0007,
      "step": 4200
    },
    {
      "epoch": 7.51,
      "eval_f1_score": 0.7908653846153846,
      "eval_label_f1": 0.893429487179487,
      "eval_loss": 0.2735210061073303,
      "eval_runtime": 712.4097,
      "eval_samples_per_second": 1.404,
      "eval_steps_per_second": 0.175,
      "eval_wer": 0.06238872977784658,
      "step": 4200
    },
    {
      "epoch": 7.69,
      "learning_rate": 5.852620357053651e-06,
      "loss": 0.0007,
      "step": 4300
    },
    {
      "epoch": 7.86,
      "learning_rate": 4.322727117869951e-06,
      "loss": 0.0007,
      "step": 4400
    },
    {
      "epoch": 7.86,
      "eval_f1_score": 0.7900641025641026,
      "eval_label_f1": 0.8926282051282052,
      "eval_loss": 0.27564728260040283,
      "eval_runtime": 703.3771,
      "eval_samples_per_second": 1.422,
      "eval_steps_per_second": 0.178,
      "eval_wer": 0.06211781097608174,
      "step": 4400
    },
    {
      "epoch": 8.04,
      "learning_rate": 3.0153689607045845e-06,
      "loss": 0.0006,
      "step": 4500
    },
    {
      "epoch": 8.22,
      "learning_rate": 1.9369152030840556e-06,
      "loss": 0.0005,
      "step": 4600
    },
    {
      "epoch": 8.22,
      "eval_f1_score": 0.7913496195434521,
      "eval_label_f1": 0.893071686023228,
      "eval_loss": 0.27790385484695435,
      "eval_runtime": 808.1657,
      "eval_samples_per_second": 1.237,
      "eval_steps_per_second": 0.155,
      "eval_wer": 0.06238872977784658,
      "step": 4600
    },
    {
      "epoch": 8.4,
      "learning_rate": 1.0926199633097157e-06,
      "loss": 0.0005,
      "step": 4700
    },
    {
      "epoch": 8.58,
      "learning_rate": 4.865965629214819e-07,
      "loss": 0.0004,
      "step": 4800
    },
    {
      "epoch": 8.58,
      "eval_f1_score": 0.7908836465413833,
      "eval_label_f1": 0.8932427029188325,
      "eval_loss": 0.27951279282569885,
      "eval_runtime": 780.5493,
      "eval_samples_per_second": 1.281,
      "eval_steps_per_second": 0.16,
      "eval_wer": 0.062040405604148925,
      "step": 4800
    },
    {
      "epoch": 8.76,
      "learning_rate": 1.2179748700879012e-07,
      "loss": 0.0005,
      "step": 4900
    },
    {
      "epoch": 8.94,
      "learning_rate": 0.0,
      "loss": 0.0005,
      "step": 5000
    },
    {
      "epoch": 8.94,
      "eval_f1_score": 0.7918497802636837,
      "eval_label_f1": 0.8933280063923291,
      "eval_loss": 0.2796746790409088,
      "eval_runtime": 791.2039,
      "eval_samples_per_second": 1.264,
      "eval_steps_per_second": 0.158,
      "eval_wer": 0.06196300023221612,
      "step": 5000
    },
    {
      "epoch": 8.94,
      "step": 5000,
      "total_flos": 1.3699137246322806e+21,
      "train_loss": 0.05188751313686371,
      "train_runtime": 117179.1316,
      "train_samples_per_second": 5.462,
      "train_steps_per_second": 0.043
    }
  ],
  "logging_steps": 100,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 200,
  "total_flos": 1.3699137246322806e+21,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}