trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.43735307455062866,
  "best_model_checkpoint": "tiny_bert_rand_10_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": 6.992720127105713,
      "learning_rate": 4.9e-05,
      "loss": 0.4215,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8107798165137615,
      "eval_loss": 0.43735307455062866,
      "eval_runtime": 0.3093,
      "eval_samples_per_second": 2818.957,
      "eval_steps_per_second": 12.931,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.604894638061523,
      "learning_rate": 4.8e-05,
      "loss": 0.2391,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.5441927313804626,
      "eval_runtime": 0.2672,
      "eval_samples_per_second": 3263.418,
      "eval_steps_per_second": 14.97,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.7732110023498535,
      "learning_rate": 4.7e-05,
      "loss": 0.1868,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.4935286343097687,
      "eval_runtime": 0.2626,
      "eval_samples_per_second": 3320.059,
      "eval_steps_per_second": 15.23,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.498906135559082,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1554,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8119266055045872,
      "eval_loss": 0.5800129771232605,
      "eval_runtime": 0.2616,
      "eval_samples_per_second": 3333.379,
      "eval_steps_per_second": 15.291,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.418341636657715,
      "learning_rate": 4.5e-05,
      "loss": 0.1287,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5927637815475464,
      "eval_runtime": 0.2708,
      "eval_samples_per_second": 3219.943,
      "eval_steps_per_second": 14.77,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.71963119506836,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1108,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7947247706422018,
      "eval_loss": 0.8025162816047668,
      "eval_runtime": 0.2752,
      "eval_samples_per_second": 3168.161,
      "eval_steps_per_second": 14.533,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.20706956555144956,
      "train_runtime": 218.205,
      "train_samples_per_second": 15432.509,
      "train_steps_per_second": 60.494
    }
  ],
  "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": 1.0596790573590528e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}