trainer_state.json

{
  "best_metric": 1.098226547241211,
  "best_model_checkpoint": "bert-base-uncased_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": 0.05347348004579544,
      "learning_rate": 0.00098,
      "loss": 1.1036,
      "step": 1534
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.0990766286849976,
      "eval_runtime": 7.11,
      "eval_samples_per_second": 1380.446,
      "eval_steps_per_second": 5.485,
      "step": 1534
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.183979332447052,
      "learning_rate": 0.00096,
      "loss": 1.0991,
      "step": 3068
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.098226547241211,
      "eval_runtime": 7.0461,
      "eval_samples_per_second": 1392.96,
      "eval_steps_per_second": 5.535,
      "step": 3068
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.04768056422472,
      "learning_rate": 0.00094,
      "loss": 1.0989,
      "step": 4602
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3273560876209883,
      "eval_loss": 1.0990049839019775,
      "eval_runtime": 7.0432,
      "eval_samples_per_second": 1393.552,
      "eval_steps_per_second": 5.537,
      "step": 4602
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0822550430893898,
      "learning_rate": 0.00092,
      "loss": 1.099,
      "step": 6136
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.1011390686035156,
      "eval_runtime": 7.0666,
      "eval_samples_per_second": 1388.934,
      "eval_steps_per_second": 5.519,
      "step": 6136
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.10713515430688858,
      "learning_rate": 0.0009000000000000001,
      "loss": 1.099,
      "step": 7670
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3544574630667346,
      "eval_loss": 1.0986531972885132,
      "eval_runtime": 7.0491,
      "eval_samples_per_second": 1392.375,
      "eval_steps_per_second": 5.533,
      "step": 7670
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05226641148328781,
      "learning_rate": 0.00088,
      "loss": 1.0989,
      "step": 9204
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.099714994430542,
      "eval_runtime": 7.064,
      "eval_samples_per_second": 1389.431,
      "eval_steps_per_second": 5.521,
      "step": 9204
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.11272826045751572,
      "learning_rate": 0.00086,
      "loss": 1.0989,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.31818644931227713,
      "eval_loss": 1.1000699996948242,
      "eval_runtime": 7.0642,
      "eval_samples_per_second": 1389.401,
      "eval_steps_per_second": 5.521,
      "step": 10738
    },
    {
      "epoch": 7.0,
      "step": 10738,
      "total_flos": 3.616380787901737e+17,
      "train_loss": 1.0996504364610495,
      "train_runtime": 4796.4824,
      "train_samples_per_second": 4093.646,
      "train_steps_per_second": 15.991
    }
  ],
  "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
}