trainer_state.json

{
  "best_metric": 0.8701333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0531/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": 91.16040802001953,
      "learning_rate": 0.0001,
      "loss": 2.9002,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.676,
      "eval_loss": 1.2175014019012451,
      "eval_runtime": 5.6198,
      "eval_samples_per_second": 667.281,
      "eval_steps_per_second": 10.499,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 129.60702514648438,
      "learning_rate": 0.0001,
      "loss": 0.878,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7989333333333334,
      "eval_loss": 0.6494498252868652,
      "eval_runtime": 5.5197,
      "eval_samples_per_second": 679.389,
      "eval_steps_per_second": 10.689,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 87.67269897460938,
      "learning_rate": 0.0001,
      "loss": 0.582,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8330666666666666,
      "eval_loss": 0.5319434404373169,
      "eval_runtime": 5.5017,
      "eval_samples_per_second": 681.609,
      "eval_steps_per_second": 10.724,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 101.76753234863281,
      "learning_rate": 0.0001,
      "loss": 0.4486,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8456,
      "eval_loss": 0.4868689775466919,
      "eval_runtime": 5.5896,
      "eval_samples_per_second": 670.887,
      "eval_steps_per_second": 10.555,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 35.98136901855469,
      "learning_rate": 0.0001,
      "loss": 0.3398,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8608,
      "eval_loss": 0.4397899806499481,
      "eval_runtime": 5.564,
      "eval_samples_per_second": 673.972,
      "eval_steps_per_second": 10.604,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 110.56379699707031,
      "learning_rate": 0.0001,
      "loss": 0.2669,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.864,
      "eval_loss": 0.4263385534286499,
      "eval_runtime": 5.5407,
      "eval_samples_per_second": 676.806,
      "eval_steps_per_second": 10.648,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 32.488765716552734,
      "learning_rate": 0.0001,
      "loss": 0.2103,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8701333333333333,
      "eval_loss": 0.423382043838501,
      "eval_runtime": 5.4431,
      "eval_samples_per_second": 688.949,
      "eval_steps_per_second": 10.839,
      "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
}