trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.7126438617706299,
  "best_model_checkpoint": "bert_base_rand_20_v2_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": 2.2691681385040283,
      "learning_rate": 4.9e-05,
      "loss": 0.7383,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4729241877256318,
      "eval_loss": 0.7126438617706299,
      "eval_runtime": 0.1905,
      "eval_samples_per_second": 1453.977,
      "eval_steps_per_second": 10.498,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.2063453197479248,
      "learning_rate": 4.8e-05,
      "loss": 0.6872,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.7162327766418457,
      "eval_runtime": 0.2088,
      "eval_samples_per_second": 1326.634,
      "eval_steps_per_second": 9.579,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.1621780395507812,
      "learning_rate": 4.7e-05,
      "loss": 0.6401,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.7891488671302795,
      "eval_runtime": 0.1956,
      "eval_samples_per_second": 1415.898,
      "eval_steps_per_second": 10.223,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0316812992095947,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5314,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 0.9643440246582031,
      "eval_runtime": 0.2113,
      "eval_samples_per_second": 1310.701,
      "eval_steps_per_second": 9.464,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.799349784851074,
      "learning_rate": 4.5e-05,
      "loss": 0.3933,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 1.0568265914916992,
      "eval_runtime": 0.2109,
      "eval_samples_per_second": 1313.513,
      "eval_steps_per_second": 9.484,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 5.7739338874816895,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2621,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.48014440433212996,
      "eval_loss": 1.448755145072937,
      "eval_runtime": 0.1948,
      "eval_samples_per_second": 1422.033,
      "eval_steps_per_second": 10.267,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 1965439583539200.0,
      "train_loss": 0.5420646985371907,
      "train_runtime": 38.9251,
      "train_samples_per_second": 3198.453,
      "train_steps_per_second": 12.845
    }
  ],
  "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
}