trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 3348,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.6,
      "learning_rate": 1.850657108721625e-05,
      "loss": 2.9804,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_loss": 2.4562251567840576,
      "eval_runtime": 42.4632,
      "eval_samples_per_second": 42.083,
      "eval_steps_per_second": 5.275,
      "step": 837
    },
    {
      "epoch": 1.19,
      "learning_rate": 1.7013142174432496e-05,
      "loss": 2.4595,
      "step": 1000
    },
    {
      "epoch": 1.79,
      "learning_rate": 1.5519713261648747e-05,
      "loss": 2.3818,
      "step": 1500
    },
    {
      "epoch": 2.0,
      "eval_loss": 2.373106002807617,
      "eval_runtime": 42.2281,
      "eval_samples_per_second": 42.318,
      "eval_steps_per_second": 5.305,
      "step": 1674
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.4026284348864996e-05,
      "loss": 2.3449,
      "step": 2000
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.2532855436081244e-05,
      "loss": 2.3116,
      "step": 2500
    },
    {
      "epoch": 3.0,
      "eval_loss": 2.333420515060425,
      "eval_runtime": 42.4622,
      "eval_samples_per_second": 42.084,
      "eval_steps_per_second": 5.275,
      "step": 2511
    },
    {
      "epoch": 3.58,
      "learning_rate": 1.1039426523297491e-05,
      "loss": 2.2842,
      "step": 3000
    },
    {
      "epoch": 4.0,
      "eval_loss": 2.3103713989257812,
      "eval_runtime": 42.488,
      "eval_samples_per_second": 42.059,
      "eval_steps_per_second": 5.272,
      "step": 3348
    }
  ],
  "logging_steps": 500,
  "max_steps": 6696,
  "num_train_epochs": 8,
  "save_steps": 500,
  "total_flos": 5243347206144000.0,
  "trial_name": null,
  "trial_params": null
}