trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.7153232097625732,
  "best_model_checkpoint": "bert_base_rand_10_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": 2.3404905796051025,
      "learning_rate": 4.9e-05,
      "loss": 0.7378,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4693140794223827,
      "eval_loss": 0.7153232097625732,
      "eval_runtime": 0.188,
      "eval_samples_per_second": 1473.442,
      "eval_steps_per_second": 10.639,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.3382792472839355,
      "learning_rate": 4.8e-05,
      "loss": 0.6802,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.7187076807022095,
      "eval_runtime": 0.1894,
      "eval_samples_per_second": 1462.663,
      "eval_steps_per_second": 10.561,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2157424688339233,
      "learning_rate": 4.7e-05,
      "loss": 0.6332,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.516245487364621,
      "eval_loss": 0.7667179107666016,
      "eval_runtime": 0.1973,
      "eval_samples_per_second": 1403.659,
      "eval_steps_per_second": 10.135,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.7056071758270264,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5205,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 1.0242571830749512,
      "eval_runtime": 0.199,
      "eval_samples_per_second": 1391.849,
      "eval_steps_per_second": 10.049,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.9855568408966064,
      "learning_rate": 4.5e-05,
      "loss": 0.3812,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4657039711191336,
      "eval_loss": 1.1773042678833008,
      "eval_runtime": 0.1905,
      "eval_samples_per_second": 1454.383,
      "eval_steps_per_second": 10.501,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.336428642272949,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2636,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 1.4377262592315674,
      "eval_runtime": 0.2088,
      "eval_samples_per_second": 1326.849,
      "eval_steps_per_second": 9.58,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.536064600944519,
      "train_runtime": 36.94,
      "train_samples_per_second": 3370.33,
      "train_steps_per_second": 13.535
    }
  ],
  "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
}