trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4194400906562805,
  "best_model_checkpoint": "bert_base_km_50_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.911167621612549,
      "learning_rate": 4.9e-05,
      "loss": 0.4224,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8096330275229358,
      "eval_loss": 0.4194400906562805,
      "eval_runtime": 0.5637,
      "eval_samples_per_second": 1547.008,
      "eval_steps_per_second": 7.096,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.654995441436768,
      "learning_rate": 4.8e-05,
      "loss": 0.2319,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8119266055045872,
      "eval_loss": 0.49020224809646606,
      "eval_runtime": 0.5809,
      "eval_samples_per_second": 1501.103,
      "eval_steps_per_second": 6.886,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.036953449249268,
      "learning_rate": 4.7e-05,
      "loss": 0.1678,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7993119266055045,
      "eval_loss": 0.6837111711502075,
      "eval_runtime": 0.5879,
      "eval_samples_per_second": 1483.323,
      "eval_steps_per_second": 6.804,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.281073808670044,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1273,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.6079096794128418,
      "eval_runtime": 0.5653,
      "eval_samples_per_second": 1542.593,
      "eval_steps_per_second": 7.076,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 11.367756843566895,
      "learning_rate": 4.5e-05,
      "loss": 0.099,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.6167923212051392,
      "eval_runtime": 0.5695,
      "eval_samples_per_second": 1531.084,
      "eval_steps_per_second": 7.023,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.729939937591553,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0746,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7958715596330275,
      "eval_loss": 0.8099145293235779,
      "eval_runtime": 0.5645,
      "eval_samples_per_second": 1544.861,
      "eval_steps_per_second": 7.087,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.1871600572508995,
      "train_runtime": 662.1873,
      "train_samples_per_second": 5085.344,
      "train_steps_per_second": 19.934
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "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": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}