trainer_state.json

{
  "best_metric": 0.6903626918792725,
  "best_model_checkpoint": "tinybert_train_book_v2_rte/checkpoint-20",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 70,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6140649914741516,
      "learning_rate": 4.9e-05,
      "loss": 0.7078,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 0.6915190815925598,
      "eval_runtime": 0.0896,
      "eval_samples_per_second": 3092.148,
      "eval_steps_per_second": 22.326,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6355451941490173,
      "learning_rate": 4.8e-05,
      "loss": 0.6876,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 0.6903626918792725,
      "eval_runtime": 0.1038,
      "eval_samples_per_second": 2669.849,
      "eval_steps_per_second": 19.277,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6491702198982239,
      "learning_rate": 4.7e-05,
      "loss": 0.6673,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.690405011177063,
      "eval_runtime": 0.1021,
      "eval_samples_per_second": 2714.24,
      "eval_steps_per_second": 19.597,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.069286584854126,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6411,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5595667870036101,
      "eval_loss": 0.7111313939094543,
      "eval_runtime": 0.1056,
      "eval_samples_per_second": 2623.085,
      "eval_steps_per_second": 18.939,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.2234373092651367,
      "learning_rate": 4.5e-05,
      "loss": 0.6054,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.7582637667655945,
      "eval_runtime": 0.0872,
      "eval_samples_per_second": 3176.99,
      "eval_steps_per_second": 22.939,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.08807110786438,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.5666,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 0.7696317434310913,
      "eval_runtime": 0.0892,
      "eval_samples_per_second": 3103.895,
      "eval_steps_per_second": 22.411,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.4657697677612305,
      "learning_rate": 4.3e-05,
      "loss": 0.4955,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.8306052088737488,
      "eval_runtime": 0.0891,
      "eval_samples_per_second": 3108.829,
      "eval_steps_per_second": 22.446,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 457076966492160.0,
      "train_loss": 0.6244595663888114,
      "train_runtime": 15.0091,
      "train_samples_per_second": 8294.994,
      "train_steps_per_second": 33.313
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 457076966492160.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}