trainer_state.json

{
  "best_metric": 0.8610666666666666,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0391/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": 103.5033950805664,
      "learning_rate": 3e-05,
      "loss": 3.8496,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.34186666666666665,
      "eval_loss": 3.7317306995391846,
      "eval_runtime": 16.9402,
      "eval_samples_per_second": 221.367,
      "eval_steps_per_second": 3.483,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 251.27978515625,
      "learning_rate": 3e-05,
      "loss": 3.179,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5653333333333334,
      "eval_loss": 2.4721686840057373,
      "eval_runtime": 15.3718,
      "eval_samples_per_second": 243.953,
      "eval_steps_per_second": 3.838,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 568.33251953125,
      "learning_rate": 3e-05,
      "loss": 1.7867,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7152,
      "eval_loss": 1.2433525323867798,
      "eval_runtime": 15.5044,
      "eval_samples_per_second": 241.867,
      "eval_steps_per_second": 3.805,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 153.26876831054688,
      "learning_rate": 3e-05,
      "loss": 1.0164,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7978666666666666,
      "eval_loss": 0.7743657827377319,
      "eval_runtime": 15.468,
      "eval_samples_per_second": 242.436,
      "eval_steps_per_second": 3.814,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 130.37396240234375,
      "learning_rate": 3e-05,
      "loss": 0.7177,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8258666666666666,
      "eval_loss": 0.62257981300354,
      "eval_runtime": 15.5788,
      "eval_samples_per_second": 240.711,
      "eval_steps_per_second": 3.787,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 231.7349853515625,
      "learning_rate": 3e-05,
      "loss": 0.5861,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8405333333333334,
      "eval_loss": 0.5308362245559692,
      "eval_runtime": 15.424,
      "eval_samples_per_second": 243.127,
      "eval_steps_per_second": 3.825,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 112.73117065429688,
      "learning_rate": 3e-05,
      "loss": 0.4973,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8610666666666666,
      "eval_loss": 0.46945494413375854,
      "eval_runtime": 15.5252,
      "eval_samples_per_second": 241.543,
      "eval_steps_per_second": 3.8,
      "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
}