trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.47022947669029236,
  "best_model_checkpoint": "tiny_bert_rand_20_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": 7.0207929611206055,
      "learning_rate": 4.9e-05,
      "loss": 0.4251,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7947247706422018,
      "eval_loss": 0.47022947669029236,
      "eval_runtime": 0.2606,
      "eval_samples_per_second": 3345.676,
      "eval_steps_per_second": 15.347,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.783622741699219,
      "learning_rate": 4.8e-05,
      "loss": 0.2413,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.5579770803451538,
      "eval_runtime": 0.2835,
      "eval_samples_per_second": 3075.911,
      "eval_steps_per_second": 14.11,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.480288982391357,
      "learning_rate": 4.7e-05,
      "loss": 0.1877,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.5218008160591125,
      "eval_runtime": 0.2721,
      "eval_samples_per_second": 3204.71,
      "eval_steps_per_second": 14.701,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.568557739257812,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.156,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.5968555808067322,
      "eval_runtime": 0.2763,
      "eval_samples_per_second": 3155.805,
      "eval_steps_per_second": 14.476,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.660000801086426,
      "learning_rate": 4.5e-05,
      "loss": 0.129,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7798165137614679,
      "eval_loss": 0.6602848172187805,
      "eval_runtime": 0.2898,
      "eval_samples_per_second": 3008.708,
      "eval_steps_per_second": 13.801,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.4265594482421875,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1102,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7786697247706422,
      "eval_loss": 0.819432258605957,
      "eval_runtime": 0.2598,
      "eval_samples_per_second": 3356.857,
      "eval_steps_per_second": 15.398,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.20821162787350742,
      "train_runtime": 219.9377,
      "train_samples_per_second": 15310.928,
      "train_steps_per_second": 60.017
    }
  ],
  "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
}