trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6933523416519165,
  "best_model_checkpoint": "bert_base_rand_5_v2_rte/checkpoint-10",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.6398528218269348,
      "learning_rate": 4.9e-05,
      "loss": 0.6922,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.6933523416519165,
      "eval_runtime": 0.1907,
      "eval_samples_per_second": 1452.804,
      "eval_steps_per_second": 10.49,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.9724056720733643,
      "learning_rate": 4.8e-05,
      "loss": 0.6718,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5306859205776173,
      "eval_loss": 0.7827130556106567,
      "eval_runtime": 0.1946,
      "eval_samples_per_second": 1423.739,
      "eval_steps_per_second": 10.28,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.9584951400756836,
      "learning_rate": 4.7e-05,
      "loss": 0.5882,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 0.8685051202774048,
      "eval_runtime": 0.1945,
      "eval_samples_per_second": 1424.282,
      "eval_steps_per_second": 10.284,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.593937397003174,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4528,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 1.0500197410583496,
      "eval_runtime": 0.2023,
      "eval_samples_per_second": 1369.573,
      "eval_steps_per_second": 9.889,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.2246103286743164,
      "learning_rate": 4.5e-05,
      "loss": 0.3018,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 1.2802542448043823,
      "eval_runtime": 0.1921,
      "eval_samples_per_second": 1441.681,
      "eval_steps_per_second": 10.409,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.388519287109375,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1965,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 1.477531909942627,
      "eval_runtime": 0.2017,
      "eval_samples_per_second": 1373.634,
      "eval_steps_per_second": 9.918,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.4838688929875692,
      "train_runtime": 38.1231,
      "train_samples_per_second": 3265.739,
      "train_steps_per_second": 13.115
    }
  ],
  "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": 1965439583539200.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}