trainer_state.json

{
  "best_metric": 0.9869908814589666,
  "best_model_checkpoint": "../../checkpoint/unlearn/imdb/electra-base/random_label/2.0/87/checkpoint-3128",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 4692,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "df_accuracy": 0.954,
      "dt_accuracy": 0.898,
      "epoch": 1.0,
      "eval_accuracy": 0.898,
      "eval_loss": 0.36999785900115967,
      "eval_runtime": 40.7872,
      "eval_samples_per_second": 612.938,
      "eval_steps_per_second": 2.403,
      "eval_unlearn_overall_accuracy": 0.7034942528735633,
      "step": 782,
      "unlearn_overall_accuracy": 0.7034942528735633,
      "unlearn_time": null
    },
    {
      "epoch": 1.278772378516624,
      "grad_norm": 0.38743796944618225,
      "learning_rate": 3.934356351236147e-05,
      "loss": 0.1224,
      "step": 1000
    },
    {
      "df_accuracy": 0.894,
      "dt_accuracy": 0.90348,
      "epoch": 2.0,
      "eval_accuracy": 0.90348,
      "eval_loss": 0.38031086325645447,
      "eval_runtime": 39.7017,
      "eval_samples_per_second": 629.696,
      "eval_steps_per_second": 2.468,
      "eval_unlearn_overall_accuracy": 0.753000899910009,
      "step": 1564,
      "unlearn_overall_accuracy": 0.753000899910009,
      "unlearn_time": null
    },
    {
      "epoch": 2.557544757033248,
      "grad_norm": 0.1861998736858368,
      "learning_rate": 2.8687127024722932e-05,
      "loss": 0.1067,
      "step": 2000
    },
    {
      "df_accuracy": 0.83,
      "dt_accuracy": 0.90104,
      "epoch": 3.0,
      "eval_accuracy": 0.90104,
      "eval_loss": 0.41145703196525574,
      "eval_runtime": 40.4446,
      "eval_samples_per_second": 618.13,
      "eval_steps_per_second": 2.423,
      "eval_unlearn_overall_accuracy": 0.799716438389139,
      "step": 2346,
      "unlearn_overall_accuracy": 0.799716438389139,
      "unlearn_time": null
    },
    {
      "epoch": 3.836317135549872,
      "grad_norm": 0.39168480038642883,
      "learning_rate": 1.80306905370844e-05,
      "loss": 0.0911,
      "step": 3000
    },
    {
      "df_accuracy": 0.51,
      "dt_accuracy": 0.89332,
      "epoch": 4.0,
      "eval_accuracy": 0.89332,
      "eval_loss": 0.44322600960731506,
      "eval_runtime": 40.7607,
      "eval_samples_per_second": 613.337,
      "eval_steps_per_second": 2.404,
      "eval_unlearn_overall_accuracy": 0.9869908814589666,
      "step": 3128,
      "unlearn_overall_accuracy": 0.9869908814589666,
      "unlearn_time": null
    },
    {
      "df_accuracy": 0.404,
      "dt_accuracy": 0.89896,
      "epoch": 5.0,
      "eval_accuracy": 0.89896,
      "eval_loss": 0.4861825406551361,
      "eval_runtime": 40.3161,
      "eval_samples_per_second": 620.1,
      "eval_steps_per_second": 2.431,
      "eval_unlearn_overall_accuracy": 0.9448834625377579,
      "step": 3910,
      "unlearn_overall_accuracy": 0.9448834625377579,
      "unlearn_time": null
    },
    {
      "epoch": 5.115089514066496,
      "grad_norm": 0.15296652913093567,
      "learning_rate": 7.374254049445865e-06,
      "loss": 0.064,
      "step": 4000
    },
    {
      "df_accuracy": 0.38,
      "dt_accuracy": 0.89976,
      "epoch": 6.0,
      "eval_accuracy": 0.89976,
      "eval_loss": 0.5284281373023987,
      "eval_runtime": 39.6498,
      "eval_samples_per_second": 630.52,
      "eval_steps_per_second": 2.472,
      "eval_unlearn_overall_accuracy": 0,
      "step": 4692,
      "unlearn_overall_accuracy": 0,
      "unlearn_time": null
    },
    {
      "epoch": 6.0,
      "step": 4692,
      "total_flos": 9866664576000000.0,
      "train_loss": 0.08846605972459147,
      "train_runtime": 1091.0951,
      "train_samples_per_second": 137.477,
      "train_steps_per_second": 4.3
    }
  ],
  "logging_steps": 1000,
  "max_steps": 4692,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9866664576000000.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}