trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.7135111093521118,
  "best_model_checkpoint": "bert_base_rand_20_v1_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": 2.4074349403381348,
      "learning_rate": 4.9e-05,
      "loss": 0.7385,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.7135111093521118,
      "eval_runtime": 0.1915,
      "eval_samples_per_second": 1446.362,
      "eval_steps_per_second": 10.443,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2212562561035156,
      "learning_rate": 4.8e-05,
      "loss": 0.6868,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.7156969308853149,
      "eval_runtime": 0.2151,
      "eval_samples_per_second": 1287.553,
      "eval_steps_per_second": 9.296,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.1496970653533936,
      "learning_rate": 4.7e-05,
      "loss": 0.6384,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.7885266542434692,
      "eval_runtime": 0.1899,
      "eval_samples_per_second": 1458.941,
      "eval_steps_per_second": 10.534,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.1219680309295654,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5308,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 0.9607929587364197,
      "eval_runtime": 0.1907,
      "eval_samples_per_second": 1452.912,
      "eval_steps_per_second": 10.49,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.706192970275879,
      "learning_rate": 4.5e-05,
      "loss": 0.3905,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 1.0572311878204346,
      "eval_runtime": 0.1903,
      "eval_samples_per_second": 1455.427,
      "eval_steps_per_second": 10.508,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.4368133544921875,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2645,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 1.4440592527389526,
      "eval_runtime": 0.2345,
      "eval_samples_per_second": 1181.147,
      "eval_steps_per_second": 8.528,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.5415834585825602,
      "train_runtime": 37.8354,
      "train_samples_per_second": 3290.567,
      "train_steps_per_second": 13.215
    }
  ],
  "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
}