trainer_state.json

{
  "best_metric": 0.9093333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0118/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": 122.72655487060547,
      "learning_rate": 0.000288581929876693,
      "loss": 1.5175,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8477333333333333,
      "eval_loss": 0.4902642071247101,
      "eval_runtime": 7.2741,
      "eval_samples_per_second": 515.529,
      "eval_steps_per_second": 8.111,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 65.93448638916016,
      "learning_rate": 0.00025606601717798207,
      "loss": 0.3929,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.37221813201904297,
      "eval_runtime": 7.2358,
      "eval_samples_per_second": 518.258,
      "eval_steps_per_second": 8.154,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 45.907806396484375,
      "learning_rate": 0.00020740251485476345,
      "loss": 0.2102,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8861333333333333,
      "eval_loss": 0.38064950704574585,
      "eval_runtime": 6.9863,
      "eval_samples_per_second": 536.762,
      "eval_steps_per_second": 8.445,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 38.10350799560547,
      "learning_rate": 0.00015,
      "loss": 0.1092,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9018666666666667,
      "eval_loss": 0.35700762271881104,
      "eval_runtime": 12.1254,
      "eval_samples_per_second": 309.269,
      "eval_steps_per_second": 4.866,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 30.93914031982422,
      "learning_rate": 9.259748514523653e-05,
      "loss": 0.0493,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9016,
      "eval_loss": 0.38038620352745056,
      "eval_runtime": 6.9837,
      "eval_samples_per_second": 536.968,
      "eval_steps_per_second": 8.448,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 68.55017852783203,
      "learning_rate": 4.3933982822017876e-05,
      "loss": 0.0264,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9064,
      "eval_loss": 0.38314858078956604,
      "eval_runtime": 7.0279,
      "eval_samples_per_second": 533.589,
      "eval_steps_per_second": 8.395,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 72.02960968017578,
      "learning_rate": 1.1418070123306989e-05,
      "loss": 0.0194,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9093333333333333,
      "eval_loss": 0.3754997253417969,
      "eval_runtime": 7.4519,
      "eval_samples_per_second": 503.228,
      "eval_steps_per_second": 7.917,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}