trainer_state.json

{
  "best_metric": 0.4842231273651123,
  "best_model_checkpoint": "bert_base_lda_20_v1_book_mnli/checkpoint-3068",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 10738,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 2.756409168243408,
      "learning_rate": 4.9e-05,
      "loss": 0.6384,
      "step": 1534
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7886907794192562,
      "eval_loss": 0.540008544921875,
      "eval_runtime": 6.4664,
      "eval_samples_per_second": 1517.849,
      "eval_steps_per_second": 6.031,
      "step": 1534
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.385667562484741,
      "learning_rate": 4.8e-05,
      "loss": 0.4424,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8124299541518084,
      "eval_loss": 0.4842231273651123,
      "eval_runtime": 6.3433,
      "eval_samples_per_second": 1547.312,
      "eval_steps_per_second": 6.148,
      "step": 3068
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.59613037109375,
      "learning_rate": 4.7e-05,
      "loss": 0.3282,
      "step": 4602
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8086602139582272,
      "eval_loss": 0.5306477546691895,
      "eval_runtime": 6.4065,
      "eval_samples_per_second": 1532.046,
      "eval_steps_per_second": 6.088,
      "step": 4602
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.2465038299560547,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.2371,
      "step": 6136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8161996943453897,
      "eval_loss": 0.5694966912269592,
      "eval_runtime": 6.2971,
      "eval_samples_per_second": 1558.663,
      "eval_steps_per_second": 6.193,
      "step": 6136
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.7642693519592285,
      "learning_rate": 4.5e-05,
      "loss": 0.1756,
      "step": 7670
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8067244014263881,
      "eval_loss": 0.6396401524543762,
      "eval_runtime": 6.3159,
      "eval_samples_per_second": 1554.021,
      "eval_steps_per_second": 6.175,
      "step": 7670
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.754429340362549,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1332,
      "step": 9204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8104941416199695,
      "eval_loss": 0.7667672634124756,
      "eval_runtime": 6.3302,
      "eval_samples_per_second": 1550.505,
      "eval_steps_per_second": 6.161,
      "step": 9204
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.643418788909912,
      "learning_rate": 4.3e-05,
      "loss": 0.1049,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8100866021395823,
      "eval_loss": 0.7942864894866943,
      "eval_runtime": 6.3413,
      "eval_samples_per_second": 1547.78,
      "eval_steps_per_second": 6.15,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "step": 10738,
      "total_flos": 3.616380787901737e+17,
      "train_loss": 0.2942728265345618,
      "train_runtime": 4517.9626,
      "train_samples_per_second": 4346.008,
      "train_steps_per_second": 16.977
    }
  ],
  "logging_steps": 1,
  "max_steps": 76700,
  "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": 3.616380787901737e+17,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}