trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4518633186817169,
  "best_model_checkpoint": "bert_base_rand_50_v1_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 9.048876762390137,
      "learning_rate": 4.9e-05,
      "loss": 0.3835,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8153669724770642,
      "eval_loss": 0.4518633186817169,
      "eval_runtime": 0.5613,
      "eval_samples_per_second": 1553.491,
      "eval_steps_per_second": 7.126,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.388989448547363,
      "learning_rate": 4.8e-05,
      "loss": 0.2172,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.801605504587156,
      "eval_loss": 0.5851306915283203,
      "eval_runtime": 0.5679,
      "eval_samples_per_second": 1535.435,
      "eval_steps_per_second": 7.043,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.672374725341797,
      "learning_rate": 4.7e-05,
      "loss": 0.1602,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.801605504587156,
      "eval_loss": 0.5361992120742798,
      "eval_runtime": 0.5612,
      "eval_samples_per_second": 1553.924,
      "eval_steps_per_second": 7.128,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.452095985412598,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1207,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8038990825688074,
      "eval_loss": 0.5897943377494812,
      "eval_runtime": 0.5849,
      "eval_samples_per_second": 1490.974,
      "eval_steps_per_second": 6.839,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 12.441170692443848,
      "learning_rate": 4.5e-05,
      "loss": 0.0927,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.6799452304840088,
      "eval_runtime": 0.5598,
      "eval_samples_per_second": 1557.596,
      "eval_steps_per_second": 7.145,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.0967302322387695,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0722,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8084862385321101,
      "eval_loss": 0.7554370760917664,
      "eval_runtime": 0.5651,
      "eval_samples_per_second": 1543.192,
      "eval_steps_per_second": 7.079,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.17441612662691058,
      "train_runtime": 662.508,
      "train_samples_per_second": 5082.882,
      "train_steps_per_second": 19.924
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}