trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.41501814126968384,
  "best_model_checkpoint": "bert_base_km_5_v1_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": 4.982000350952148,
      "learning_rate": 4.9e-05,
      "loss": 0.3823,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.41501814126968384,
      "eval_runtime": 0.5686,
      "eval_samples_per_second": 1533.679,
      "eval_steps_per_second": 7.035,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.620168685913086,
      "learning_rate": 4.8e-05,
      "loss": 0.2144,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8211009174311926,
      "eval_loss": 0.4729911684989929,
      "eval_runtime": 0.5845,
      "eval_samples_per_second": 1491.941,
      "eval_steps_per_second": 6.844,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.695556640625,
      "learning_rate": 4.7e-05,
      "loss": 0.1506,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.571492612361908,
      "eval_runtime": 0.5942,
      "eval_samples_per_second": 1467.46,
      "eval_steps_per_second": 6.731,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.170010566711426,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1096,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8176605504587156,
      "eval_loss": 0.5840549468994141,
      "eval_runtime": 0.5741,
      "eval_samples_per_second": 1519.012,
      "eval_steps_per_second": 6.968,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.6803159713745117,
      "learning_rate": 4.5e-05,
      "loss": 0.0829,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8119266055045872,
      "eval_loss": 0.5980706810951233,
      "eval_runtime": 0.5764,
      "eval_samples_per_second": 1512.91,
      "eval_steps_per_second": 6.94,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.348969459533691,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0625,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8153669724770642,
      "eval_loss": 0.7032604813575745,
      "eval_runtime": 0.5961,
      "eval_samples_per_second": 1462.898,
      "eval_steps_per_second": 6.711,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.16702937718593713,
      "train_runtime": 670.2178,
      "train_samples_per_second": 5024.412,
      "train_steps_per_second": 19.695
    }
  ],
  "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
}