trainer_state.json

{
  "best_metric": 0.7730666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0266/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 408.9355163574219,
      "learning_rate": 2.8514533018536286e-05,
      "loss": 3.8496,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.34346666666666664,
      "eval_loss": 3.732369899749756,
      "eval_runtime": 8.9447,
      "eval_samples_per_second": 419.241,
      "eval_steps_per_second": 6.596,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 355.156982421875,
      "learning_rate": 2.4352347027881003e-05,
      "loss": 3.2155,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5656,
      "eval_loss": 2.588406562805176,
      "eval_runtime": 8.6423,
      "eval_samples_per_second": 433.911,
      "eval_steps_per_second": 6.827,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 118.20577239990234,
      "learning_rate": 1.8337814009344716e-05,
      "loss": 1.8904,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6845333333333333,
      "eval_loss": 1.4409406185150146,
      "eval_runtime": 8.7966,
      "eval_samples_per_second": 426.303,
      "eval_steps_per_second": 6.707,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 196.27362060546875,
      "learning_rate": 1.1662185990655285e-05,
      "loss": 1.2511,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7485333333333334,
      "eval_loss": 1.0805118083953857,
      "eval_runtime": 9.0536,
      "eval_samples_per_second": 414.201,
      "eval_steps_per_second": 6.517,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 272.9058837890625,
      "learning_rate": 5.647652972118998e-06,
      "loss": 1.0455,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7704,
      "eval_loss": 0.9617211222648621,
      "eval_runtime": 8.9521,
      "eval_samples_per_second": 418.896,
      "eval_steps_per_second": 6.591,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 296.26666259765625,
      "learning_rate": 1.4854669814637145e-06,
      "loss": 0.9661,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7669333333333334,
      "eval_loss": 0.9208992123603821,
      "eval_runtime": 8.781,
      "eval_samples_per_second": 427.061,
      "eval_steps_per_second": 6.719,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 625.48876953125,
      "learning_rate": 0.0,
      "loss": 0.9424,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7730666666666667,
      "eval_loss": 0.9028689861297607,
      "eval_runtime": 9.0955,
      "eval_samples_per_second": 412.291,
      "eval_steps_per_second": 6.487,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}