trainer_state.json

{
  "best_metric": 0.8794666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0819/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": 102.16028594970703,
      "learning_rate": 0.0001,
      "loss": 2.9465,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6797333333333333,
      "eval_loss": 1.1903430223464966,
      "eval_runtime": 13.6112,
      "eval_samples_per_second": 275.509,
      "eval_steps_per_second": 4.335,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 72.56713104248047,
      "learning_rate": 0.0001,
      "loss": 0.8093,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8024,
      "eval_loss": 0.6724911332130432,
      "eval_runtime": 14.1068,
      "eval_samples_per_second": 265.83,
      "eval_steps_per_second": 4.182,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 46.76213836669922,
      "learning_rate": 0.0001,
      "loss": 0.4862,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8477333333333333,
      "eval_loss": 0.5041757822036743,
      "eval_runtime": 13.1865,
      "eval_samples_per_second": 284.382,
      "eval_steps_per_second": 4.474,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 47.30262756347656,
      "learning_rate": 0.0001,
      "loss": 0.3548,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.86,
      "eval_loss": 0.44617870450019836,
      "eval_runtime": 13.084,
      "eval_samples_per_second": 286.609,
      "eval_steps_per_second": 4.509,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 44.978214263916016,
      "learning_rate": 0.0001,
      "loss": 0.2709,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8685333333333334,
      "eval_loss": 0.4347134232521057,
      "eval_runtime": 13.126,
      "eval_samples_per_second": 285.691,
      "eval_steps_per_second": 4.495,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 67.25106811523438,
      "learning_rate": 0.0001,
      "loss": 0.2068,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8658666666666667,
      "eval_loss": 0.43697771430015564,
      "eval_runtime": 13.1824,
      "eval_samples_per_second": 284.47,
      "eval_steps_per_second": 4.476,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 81.01061248779297,
      "learning_rate": 0.0001,
      "loss": 0.1623,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8794666666666666,
      "eval_loss": 0.4111255407333374,
      "eval_runtime": 13.0645,
      "eval_samples_per_second": 287.038,
      "eval_steps_per_second": 4.516,
      "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
}