trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 14.997830802603037,
  "global_step": 3450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.43,
      "eval_loss": 3.6933705806732178,
      "eval_runtime": 193.3371,
      "eval_samples_per_second": 25.339,
      "eval_steps_per_second": 0.797,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 0.87,
      "eval_loss": 3.0763044357299805,
      "eval_runtime": 193.2997,
      "eval_samples_per_second": 25.344,
      "eval_steps_per_second": 0.797,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 1.3,
      "eval_loss": 2.9737112522125244,
      "eval_runtime": 192.6243,
      "eval_samples_per_second": 25.433,
      "eval_steps_per_second": 0.799,
      "eval_wer": 1.0,
      "step": 300
    },
    {
      "epoch": 1.74,
      "eval_loss": 2.5733721256256104,
      "eval_runtime": 192.0804,
      "eval_samples_per_second": 25.505,
      "eval_steps_per_second": 0.802,
      "eval_wer": 1.0,
      "step": 400
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.0002982,
      "loss": 5.0957,
      "step": 500
    },
    {
      "epoch": 2.17,
      "eval_loss": 1.1900373697280884,
      "eval_runtime": 192.7964,
      "eval_samples_per_second": 25.41,
      "eval_steps_per_second": 0.799,
      "eval_wer": 0.9011191003946586,
      "step": 500
    },
    {
      "epoch": 2.61,
      "eval_loss": 0.9725514054298401,
      "eval_runtime": 192.2119,
      "eval_samples_per_second": 25.487,
      "eval_steps_per_second": 0.801,
      "eval_wer": 0.7572309023084824,
      "step": 600
    },
    {
      "epoch": 3.04,
      "eval_loss": 0.8960436582565308,
      "eval_runtime": 193.7378,
      "eval_samples_per_second": 25.287,
      "eval_steps_per_second": 0.795,
      "eval_wer": 0.6209114991620263,
      "step": 700
    },
    {
      "epoch": 3.48,
      "eval_loss": 0.7850536704063416,
      "eval_runtime": 193.8724,
      "eval_samples_per_second": 25.269,
      "eval_steps_per_second": 0.794,
      "eval_wer": 0.5515488998215927,
      "step": 800
    },
    {
      "epoch": 3.91,
      "eval_loss": 0.7271264791488647,
      "eval_runtime": 195.9681,
      "eval_samples_per_second": 24.999,
      "eval_steps_per_second": 0.786,
      "eval_wer": 0.511542412283073,
      "step": 900
    },
    {
      "epoch": 4.35,
      "learning_rate": 0.00024945762711864404,
      "loss": 1.0312,
      "step": 1000
    },
    {
      "epoch": 4.35,
      "eval_loss": 0.7053087949752808,
      "eval_runtime": 192.6577,
      "eval_samples_per_second": 25.429,
      "eval_steps_per_second": 0.799,
      "eval_wer": 0.49548575444666704,
      "step": 1000
    },
    {
      "epoch": 4.78,
      "eval_loss": 0.6822740435600281,
      "eval_runtime": 193.7373,
      "eval_samples_per_second": 25.287,
      "eval_steps_per_second": 0.795,
      "eval_wer": 0.47369843758447316,
      "step": 1100
    },
    {
      "epoch": 5.22,
      "eval_loss": 0.6767629384994507,
      "eval_runtime": 192.3102,
      "eval_samples_per_second": 25.474,
      "eval_steps_per_second": 0.801,
      "eval_wer": 0.45945288425150027,
      "step": 1200
    },
    {
      "epoch": 5.65,
      "eval_loss": 0.6635140776634216,
      "eval_runtime": 192.3892,
      "eval_samples_per_second": 25.464,
      "eval_steps_per_second": 0.8,
      "eval_wer": 0.4488295399253933,
      "step": 1300
    },
    {
      "epoch": 6.09,
      "eval_loss": 0.6602210402488708,
      "eval_runtime": 191.9311,
      "eval_samples_per_second": 25.525,
      "eval_steps_per_second": 0.802,
      "eval_wer": 0.4389630750932584,
      "step": 1400
    },
    {
      "epoch": 6.52,
      "learning_rate": 0.0001986101694915254,
      "loss": 0.6815,
      "step": 1500
    },
    {
      "epoch": 6.52,
      "eval_loss": 0.6463663578033447,
      "eval_runtime": 193.1979,
      "eval_samples_per_second": 25.357,
      "eval_steps_per_second": 0.797,
      "eval_wer": 0.4310158404065524,
      "step": 1500
    },
    {
      "epoch": 6.95,
      "eval_loss": 0.6455022096633911,
      "eval_runtime": 198.417,
      "eval_samples_per_second": 24.69,
      "eval_steps_per_second": 0.776,
      "eval_wer": 0.43942260907174135,
      "step": 1600
    },
    {
      "epoch": 7.39,
      "eval_loss": 0.6629974842071533,
      "eval_runtime": 197.9313,
      "eval_samples_per_second": 24.751,
      "eval_steps_per_second": 0.778,
      "eval_wer": 0.4311780288695464,
      "step": 1700
    },
    {
      "epoch": 7.82,
      "eval_loss": 0.6521152853965759,
      "eval_runtime": 202.8016,
      "eval_samples_per_second": 24.157,
      "eval_steps_per_second": 0.759,
      "eval_wer": 0.4125804184462345,
      "step": 1800
    },
    {
      "epoch": 8.26,
      "eval_loss": 0.6281991600990295,
      "eval_runtime": 195.8017,
      "eval_samples_per_second": 25.02,
      "eval_steps_per_second": 0.787,
      "eval_wer": 0.42839379358814944,
      "step": 1900
    },
    {
      "epoch": 8.69,
      "learning_rate": 0.00014776271186440676,
      "loss": 0.544,
      "step": 2000
    },
    {
      "epoch": 8.69,
      "eval_loss": 0.6247670650482178,
      "eval_runtime": 197.0538,
      "eval_samples_per_second": 24.861,
      "eval_steps_per_second": 0.782,
      "eval_wer": 0.4178245120830405,
      "step": 2000
    },
    {
      "epoch": 9.13,
      "eval_loss": 0.6510103940963745,
      "eval_runtime": 199.9527,
      "eval_samples_per_second": 24.501,
      "eval_steps_per_second": 0.77,
      "eval_wer": 0.4103638427853165,
      "step": 2100
    },
    {
      "epoch": 9.56,
      "eval_loss": 0.6527135968208313,
      "eval_runtime": 202.4561,
      "eval_samples_per_second": 24.198,
      "eval_steps_per_second": 0.761,
      "eval_wer": 0.4012542574471536,
      "step": 2200
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.6510817408561707,
      "eval_runtime": 202.1587,
      "eval_samples_per_second": 24.233,
      "eval_steps_per_second": 0.762,
      "eval_wer": 0.40641725685246255,
      "step": 2300
    },
    {
      "epoch": 10.43,
      "eval_loss": 0.6734431982040405,
      "eval_runtime": 202.8997,
      "eval_samples_per_second": 24.145,
      "eval_steps_per_second": 0.759,
      "eval_wer": 0.4060928799264746,
      "step": 2400
    },
    {
      "epoch": 10.87,
      "learning_rate": 9.691525423728813e-05,
      "loss": 0.4478,
      "step": 2500
    },
    {
      "epoch": 10.87,
      "eval_loss": 0.6756427884101868,
      "eval_runtime": 201.7844,
      "eval_samples_per_second": 24.278,
      "eval_steps_per_second": 0.763,
      "eval_wer": 0.41452668000216253,
      "step": 2500
    },
    {
      "epoch": 11.3,
      "eval_loss": 0.6727204918861389,
      "eval_runtime": 202.413,
      "eval_samples_per_second": 24.203,
      "eval_steps_per_second": 0.761,
      "eval_wer": 0.3989565875547386,
      "step": 2600
    },
    {
      "epoch": 11.74,
      "eval_loss": 0.6618625521659851,
      "eval_runtime": 205.4327,
      "eval_samples_per_second": 23.847,
      "eval_steps_per_second": 0.75,
      "eval_wer": 0.4007406606476726,
      "step": 2700
    },
    {
      "epoch": 12.17,
      "eval_loss": 0.6613638997077942,
      "eval_runtime": 202.1353,
      "eval_samples_per_second": 24.236,
      "eval_steps_per_second": 0.762,
      "eval_wer": 0.40193004270962857,
      "step": 2800
    },
    {
      "epoch": 12.61,
      "eval_loss": 0.6695073246955872,
      "eval_runtime": 202.6466,
      "eval_samples_per_second": 24.175,
      "eval_steps_per_second": 0.76,
      "eval_wer": 0.40036222090068657,
      "step": 2900
    },
    {
      "epoch": 13.04,
      "learning_rate": 4.606779661016948e-05,
      "loss": 0.3919,
      "step": 3000
    },
    {
      "epoch": 13.04,
      "eval_loss": 0.6778153777122498,
      "eval_runtime": 202.4696,
      "eval_samples_per_second": 24.196,
      "eval_steps_per_second": 0.761,
      "eval_wer": 0.3966048548413256,
      "step": 3000
    },
    {
      "epoch": 13.48,
      "eval_loss": 0.6872451305389404,
      "eval_runtime": 201.7642,
      "eval_samples_per_second": 24.281,
      "eval_steps_per_second": 0.763,
      "eval_wer": 0.3971184516408066,
      "step": 3100
    },
    {
      "epoch": 13.91,
      "eval_loss": 0.688228964805603,
      "eval_runtime": 204.8269,
      "eval_samples_per_second": 23.918,
      "eval_steps_per_second": 0.752,
      "eval_wer": 0.39452343623290265,
      "step": 3200
    },
    {
      "epoch": 14.35,
      "eval_loss": 0.6937959790229797,
      "eval_runtime": 202.6788,
      "eval_samples_per_second": 24.171,
      "eval_steps_per_second": 0.76,
      "eval_wer": 0.39365843109693466,
      "step": 3300
    },
    {
      "epoch": 14.78,
      "eval_loss": 0.6927648782730103,
      "eval_runtime": 201.0141,
      "eval_samples_per_second": 24.371,
      "eval_steps_per_second": 0.766,
      "eval_wer": 0.3945504676434016,
      "step": 3400
    },
    {
      "epoch": 15.0,
      "step": 3450,
      "total_flos": 2.9875620378207076e+19,
      "train_loss": 1.2329752604166666,
      "train_runtime": 16157.973,
      "train_samples_per_second": 13.674,
      "train_steps_per_second": 0.214
    }
  ],
  "max_steps": 3450,
  "num_train_epochs": 15,
  "total_flos": 2.9875620378207076e+19,
  "trial_name": null,
  "trial_params": null
}