trainer_state.json

{
  "best_metric": 0.8810666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0489/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": 115.9214096069336,
      "learning_rate": 3e-05,
      "loss": 3.8199,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4728,
      "eval_loss": 3.582125425338745,
      "eval_runtime": 8.672,
      "eval_samples_per_second": 432.424,
      "eval_steps_per_second": 6.803,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 85.94645690917969,
      "learning_rate": 3e-05,
      "loss": 2.8251,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6653333333333333,
      "eval_loss": 1.9038035869598389,
      "eval_runtime": 8.682,
      "eval_samples_per_second": 431.927,
      "eval_steps_per_second": 6.796,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 90.58956146240234,
      "learning_rate": 3e-05,
      "loss": 1.2747,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7922666666666667,
      "eval_loss": 0.8495551943778992,
      "eval_runtime": 8.7627,
      "eval_samples_per_second": 427.948,
      "eval_steps_per_second": 6.733,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 113.2873306274414,
      "learning_rate": 3e-05,
      "loss": 0.7384,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8336,
      "eval_loss": 0.6040249466896057,
      "eval_runtime": 8.6468,
      "eval_samples_per_second": 433.687,
      "eval_steps_per_second": 6.823,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 174.7952880859375,
      "learning_rate": 3e-05,
      "loss": 0.5213,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.47344622015953064,
      "eval_runtime": 8.69,
      "eval_samples_per_second": 431.529,
      "eval_steps_per_second": 6.789,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 323.5143737792969,
      "learning_rate": 3e-05,
      "loss": 0.3958,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8765333333333334,
      "eval_loss": 0.4226449131965637,
      "eval_runtime": 8.475,
      "eval_samples_per_second": 442.476,
      "eval_steps_per_second": 6.962,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 161.008056640625,
      "learning_rate": 3e-05,
      "loss": 0.3241,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8810666666666667,
      "eval_loss": 0.39698177576065063,
      "eval_runtime": 8.7744,
      "eval_samples_per_second": 427.378,
      "eval_steps_per_second": 6.724,
      "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
}