trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.43391719460487366,
  "best_model_checkpoint": "distilbert_rand_50_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": 9.843780517578125,
      "learning_rate": 4.9e-05,
      "loss": 0.3909,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8153669724770642,
      "eval_loss": 0.43391719460487366,
      "eval_runtime": 0.3668,
      "eval_samples_per_second": 2377.009,
      "eval_steps_per_second": 10.904,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.787400245666504,
      "learning_rate": 4.8e-05,
      "loss": 0.2202,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8073394495412844,
      "eval_loss": 0.508598804473877,
      "eval_runtime": 0.3656,
      "eval_samples_per_second": 2384.972,
      "eval_steps_per_second": 10.94,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.8554320335388184,
      "learning_rate": 4.7e-05,
      "loss": 0.1625,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7958715596330275,
      "eval_loss": 0.5306481719017029,
      "eval_runtime": 0.3641,
      "eval_samples_per_second": 2394.624,
      "eval_steps_per_second": 10.985,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.0063581466674805,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1212,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8153669724770642,
      "eval_loss": 0.5817607641220093,
      "eval_runtime": 0.3663,
      "eval_samples_per_second": 2380.59,
      "eval_steps_per_second": 10.92,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.965244293212891,
      "learning_rate": 4.5e-05,
      "loss": 0.0919,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.6090948581695557,
      "eval_runtime": 0.367,
      "eval_samples_per_second": 2376.174,
      "eval_steps_per_second": 10.9,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 22.023202896118164,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0708,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8073394495412844,
      "eval_loss": 0.8077242374420166,
      "eval_runtime": 0.3666,
      "eval_samples_per_second": 2378.56,
      "eval_steps_per_second": 10.911,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 2.676464049624883e+16,
      "train_loss": 0.17623686790466309,
      "train_runtime": 376.0033,
      "train_samples_per_second": 8955.905,
      "train_steps_per_second": 35.106
    }
  ],
  "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": 2.676464049624883e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}