trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6951009631156921,
  "best_model_checkpoint": "bert_base_rand_100_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": 1.0093417167663574,
      "learning_rate": 4.9e-05,
      "loss": 0.7123,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.48014440433212996,
      "eval_loss": 0.6951009631156921,
      "eval_runtime": 0.1883,
      "eval_samples_per_second": 1471.171,
      "eval_steps_per_second": 10.622,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7576552629470825,
      "learning_rate": 4.8e-05,
      "loss": 0.6921,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.7024692296981812,
      "eval_runtime": 0.1948,
      "eval_samples_per_second": 1422.322,
      "eval_steps_per_second": 10.269,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.316249370574951,
      "learning_rate": 4.7e-05,
      "loss": 0.6395,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.7600758671760559,
      "eval_runtime": 0.19,
      "eval_samples_per_second": 1458.18,
      "eval_steps_per_second": 10.528,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.0747175216674805,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5336,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.9793204069137573,
      "eval_runtime": 0.2043,
      "eval_samples_per_second": 1355.837,
      "eval_steps_per_second": 9.789,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.384427547454834,
      "learning_rate": 4.5e-05,
      "loss": 0.4135,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.47653429602888087,
      "eval_loss": 1.0974644422531128,
      "eval_runtime": 0.1915,
      "eval_samples_per_second": 1446.746,
      "eval_steps_per_second": 10.446,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.4029014110565186,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2812,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 1.3327354192733765,
      "eval_runtime": 0.1895,
      "eval_samples_per_second": 1461.505,
      "eval_steps_per_second": 10.552,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.5453740040461222,
      "train_runtime": 39.1458,
      "train_samples_per_second": 3180.415,
      "train_steps_per_second": 12.773
    }
  ],
  "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
}