trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.994840041279669,
  "global_step": 2420,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.21,
      "eval_loss": 3.4135334491729736,
      "eval_runtime": 150.5586,
      "eval_samples_per_second": 25.744,
      "eval_steps_per_second": 3.221,
      "eval_wer": 1.0,
      "step": 100
    },
    {
      "epoch": 0.41,
      "eval_loss": 2.9520680904388428,
      "eval_runtime": 111.0015,
      "eval_samples_per_second": 34.918,
      "eval_steps_per_second": 4.369,
      "eval_wer": 1.0,
      "step": 200
    },
    {
      "epoch": 0.62,
      "eval_loss": 2.333911418914795,
      "eval_runtime": 110.7137,
      "eval_samples_per_second": 35.009,
      "eval_steps_per_second": 4.381,
      "eval_wer": 0.9364516945447853,
      "step": 300
    },
    {
      "epoch": 0.83,
      "eval_loss": 1.2433439493179321,
      "eval_runtime": 111.6882,
      "eval_samples_per_second": 34.704,
      "eval_steps_per_second": 4.342,
      "eval_wer": 0.8259439063111153,
      "step": 400
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00029759999999999997,
      "loss": 3.1912,
      "step": 500
    },
    {
      "epoch": 1.03,
      "eval_loss": 0.8614437580108643,
      "eval_runtime": 110.721,
      "eval_samples_per_second": 35.007,
      "eval_steps_per_second": 4.38,
      "eval_wer": 0.6385201205623173,
      "step": 500
    },
    {
      "epoch": 1.24,
      "eval_loss": 0.7556686401367188,
      "eval_runtime": 111.0146,
      "eval_samples_per_second": 34.914,
      "eval_steps_per_second": 4.369,
      "eval_wer": 0.5611899776821664,
      "step": 600
    },
    {
      "epoch": 1.44,
      "eval_loss": 0.6781343221664429,
      "eval_runtime": 110.9557,
      "eval_samples_per_second": 34.933,
      "eval_steps_per_second": 4.371,
      "eval_wer": 0.5194533281181695,
      "step": 700
    },
    {
      "epoch": 1.65,
      "eval_loss": 0.6363114714622498,
      "eval_runtime": 111.5716,
      "eval_samples_per_second": 34.74,
      "eval_steps_per_second": 4.347,
      "eval_wer": 0.4878862480730736,
      "step": 800
    },
    {
      "epoch": 1.86,
      "eval_loss": 0.5959396362304688,
      "eval_runtime": 111.2645,
      "eval_samples_per_second": 34.836,
      "eval_steps_per_second": 4.359,
      "eval_wer": 0.4559050226629547,
      "step": 900
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.00022281249999999997,
      "loss": 0.8237,
      "step": 1000
    },
    {
      "epoch": 2.06,
      "eval_loss": 0.5430020093917847,
      "eval_runtime": 111.36,
      "eval_samples_per_second": 34.806,
      "eval_steps_per_second": 4.355,
      "eval_wer": 0.42597151600211675,
      "step": 1000
    },
    {
      "epoch": 2.27,
      "eval_loss": 0.5292537212371826,
      "eval_runtime": 111.5324,
      "eval_samples_per_second": 34.752,
      "eval_steps_per_second": 4.349,
      "eval_wer": 0.4097738306145457,
      "step": 1100
    },
    {
      "epoch": 2.48,
      "eval_loss": 0.5140984654426575,
      "eval_runtime": 111.6965,
      "eval_samples_per_second": 34.701,
      "eval_steps_per_second": 4.342,
      "eval_wer": 0.40558636081264526,
      "step": 1200
    },
    {
      "epoch": 2.68,
      "eval_loss": 0.4878837764263153,
      "eval_runtime": 110.9007,
      "eval_samples_per_second": 34.95,
      "eval_steps_per_second": 4.373,
      "eval_wer": 0.3946805328670363,
      "step": 1300
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.46965792775154114,
      "eval_runtime": 112.2618,
      "eval_samples_per_second": 34.526,
      "eval_steps_per_second": 4.32,
      "eval_wer": 0.37882796861698453,
      "step": 1400
    },
    {
      "epoch": 3.1,
      "learning_rate": 0.000145,
      "loss": 0.5625,
      "step": 1500
    },
    {
      "epoch": 3.1,
      "eval_loss": 0.4748215675354004,
      "eval_runtime": 111.3137,
      "eval_samples_per_second": 34.821,
      "eval_steps_per_second": 4.357,
      "eval_wer": 0.37799967788693833,
      "step": 1500
    },
    {
      "epoch": 3.3,
      "eval_loss": 0.48358285427093506,
      "eval_runtime": 111.983,
      "eval_samples_per_second": 34.612,
      "eval_steps_per_second": 4.331,
      "eval_wer": 0.368382302188068,
      "step": 1600
    },
    {
      "epoch": 3.51,
      "eval_loss": 0.47962862253189087,
      "eval_runtime": 112.0713,
      "eval_samples_per_second": 34.585,
      "eval_steps_per_second": 4.328,
      "eval_wer": 0.36251524285024045,
      "step": 1700
    },
    {
      "epoch": 3.72,
      "eval_loss": 0.4582051932811737,
      "eval_runtime": 112.8803,
      "eval_samples_per_second": 34.337,
      "eval_steps_per_second": 4.297,
      "eval_wer": 0.35147136644962385,
      "step": 1800
    },
    {
      "epoch": 3.92,
      "eval_loss": 0.43948230147361755,
      "eval_runtime": 112.0141,
      "eval_samples_per_second": 34.603,
      "eval_steps_per_second": 4.33,
      "eval_wer": 0.3437406529691922,
      "step": 1900
    },
    {
      "epoch": 4.13,
      "learning_rate": 6.71875e-05,
      "loss": 0.4267,
      "step": 2000
    },
    {
      "epoch": 4.13,
      "eval_loss": 0.44096100330352783,
      "eval_runtime": 111.9224,
      "eval_samples_per_second": 34.631,
      "eval_steps_per_second": 4.333,
      "eval_wer": 0.3420150472815958,
      "step": 2000
    },
    {
      "epoch": 4.33,
      "eval_loss": 0.4466901123523712,
      "eval_runtime": 112.4959,
      "eval_samples_per_second": 34.455,
      "eval_steps_per_second": 4.311,
      "eval_wer": 0.3382187147688839,
      "step": 2100
    },
    {
      "epoch": 4.54,
      "eval_loss": 0.4398203492164612,
      "eval_runtime": 113.1035,
      "eval_samples_per_second": 34.269,
      "eval_steps_per_second": 4.288,
      "eval_wer": 0.33292685732692173,
      "step": 2200
    },
    {
      "epoch": 4.75,
      "eval_loss": 0.4382670521736145,
      "eval_runtime": 112.3465,
      "eval_samples_per_second": 34.5,
      "eval_steps_per_second": 4.317,
      "eval_wer": 0.3286703632975174,
      "step": 2300
    },
    {
      "epoch": 4.95,
      "eval_loss": 0.43576258420944214,
      "eval_runtime": 111.7707,
      "eval_samples_per_second": 34.678,
      "eval_steps_per_second": 4.339,
      "eval_wer": 0.3264155718657249,
      "step": 2400
    },
    {
      "epoch": 4.99,
      "step": 2420,
      "total_flos": 8.891283011181065e+18,
      "train_loss": 1.0930316641311015,
      "train_runtime": 6038.6358,
      "train_samples_per_second": 12.837,
      "train_steps_per_second": 0.401
    }
  ],
  "max_steps": 2420,
  "num_train_epochs": 5,
  "total_flos": 8.891283011181065e+18,
  "trial_name": null,
  "trial_params": null
}