trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 17.752260397830018,
  "global_step": 2450,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.72,
      "learning_rate": 1e-05,
      "loss": 0.0194,
      "step": 100
    },
    {
      "epoch": 1.45,
      "learning_rate": 9.99927530980506e-06,
      "loss": 0.0174,
      "step": 200
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.998550619610118e-06,
      "loss": 0.0197,
      "step": 300
    },
    {
      "epoch": 2.54,
      "eval_loss": 0.2245764136314392,
      "eval_runtime": 322.1745,
      "eval_samples_per_second": 12.251,
      "eval_steps_per_second": 1.533,
      "eval_wer": 0.17368103101087395,
      "step": 350
    },
    {
      "epoch": 2.9,
      "learning_rate": 9.997825929415176e-06,
      "loss": 0.0205,
      "step": 400
    },
    {
      "epoch": 3.62,
      "learning_rate": 9.997108486122183e-06,
      "loss": 0.0199,
      "step": 500
    },
    {
      "epoch": 4.35,
      "learning_rate": 9.996383795927241e-06,
      "loss": 0.0198,
      "step": 600
    },
    {
      "epoch": 5.07,
      "learning_rate": 9.9956591057323e-06,
      "loss": 0.0193,
      "step": 700
    },
    {
      "epoch": 5.07,
      "eval_loss": 0.23122623562812805,
      "eval_runtime": 209.3421,
      "eval_samples_per_second": 18.854,
      "eval_steps_per_second": 2.36,
      "eval_wer": 0.17217076117599678,
      "step": 700
    },
    {
      "epoch": 5.8,
      "learning_rate": 9.994934415537358e-06,
      "loss": 0.0186,
      "step": 800
    },
    {
      "epoch": 6.52,
      "learning_rate": 9.994209725342417e-06,
      "loss": 0.0194,
      "step": 900
    },
    {
      "epoch": 7.25,
      "learning_rate": 9.993485035147475e-06,
      "loss": 0.0186,
      "step": 1000
    },
    {
      "epoch": 7.61,
      "eval_loss": 0.2398330122232437,
      "eval_runtime": 209.0,
      "eval_samples_per_second": 18.885,
      "eval_steps_per_second": 2.364,
      "eval_wer": 0.17232178815948448,
      "step": 1050
    },
    {
      "epoch": 7.97,
      "learning_rate": 9.992760344952534e-06,
      "loss": 0.0186,
      "step": 1100
    },
    {
      "epoch": 8.69,
      "learning_rate": 9.99203565475759e-06,
      "loss": 0.0191,
      "step": 1200
    },
    {
      "epoch": 9.42,
      "learning_rate": 9.991310964562651e-06,
      "loss": 0.0185,
      "step": 1300
    },
    {
      "epoch": 10.14,
      "learning_rate": 9.990586274367708e-06,
      "loss": 0.0171,
      "step": 1400
    },
    {
      "epoch": 10.14,
      "eval_loss": 0.24630184471607208,
      "eval_runtime": 212.2032,
      "eval_samples_per_second": 18.6,
      "eval_steps_per_second": 2.328,
      "eval_wer": 0.17020741039065646,
      "step": 1400
    },
    {
      "epoch": 10.87,
      "learning_rate": 9.989861584172766e-06,
      "loss": 0.0176,
      "step": 1500
    },
    {
      "epoch": 11.59,
      "learning_rate": 9.989136893977825e-06,
      "loss": 0.018,
      "step": 1600
    },
    {
      "epoch": 12.32,
      "learning_rate": 9.988412203782883e-06,
      "loss": 0.0172,
      "step": 1700
    },
    {
      "epoch": 12.68,
      "eval_loss": 0.24790118634700775,
      "eval_runtime": 215.9062,
      "eval_samples_per_second": 18.281,
      "eval_steps_per_second": 2.288,
      "eval_wer": 0.17081151832460734,
      "step": 1750
    },
    {
      "epoch": 13.04,
      "learning_rate": 9.987687513587942e-06,
      "loss": 0.0172,
      "step": 1800
    },
    {
      "epoch": 13.77,
      "learning_rate": 9.986962823393e-06,
      "loss": 0.0165,
      "step": 1900
    },
    {
      "epoch": 14.49,
      "learning_rate": 9.986238133198059e-06,
      "loss": 0.0176,
      "step": 2000
    },
    {
      "epoch": 15.22,
      "learning_rate": 9.985513443003117e-06,
      "loss": 0.0173,
      "step": 2100
    },
    {
      "epoch": 15.22,
      "eval_loss": 0.24780623614788055,
      "eval_runtime": 215.4375,
      "eval_samples_per_second": 18.321,
      "eval_steps_per_second": 2.293,
      "eval_wer": 0.17015706806282724,
      "step": 2100
    },
    {
      "epoch": 15.94,
      "learning_rate": 9.984788752808176e-06,
      "loss": 0.0168,
      "step": 2200
    },
    {
      "epoch": 16.67,
      "learning_rate": 9.984064062613234e-06,
      "loss": 0.0165,
      "step": 2300
    },
    {
      "epoch": 17.39,
      "learning_rate": 9.983339372418293e-06,
      "loss": 0.0169,
      "step": 2400
    },
    {
      "epoch": 17.75,
      "eval_loss": 0.24955050647258759,
      "eval_runtime": 224.6408,
      "eval_samples_per_second": 17.57,
      "eval_steps_per_second": 2.199,
      "eval_wer": 0.1689488521949255,
      "step": 2450
    }
  ],
  "max_steps": 1380000,
  "num_train_epochs": 10000,
  "total_flos": 5.57531310441053e+19,
  "trial_name": null,
  "trial_params": null
}