trainer_state.json

{
  "best_metric": 0.8770666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0309/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": 83.48312377929688,
      "learning_rate": 5e-05,
      "loss": 3.6785,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.49173333333333336,
      "eval_loss": 3.1306865215301514,
      "eval_runtime": 8.0819,
      "eval_samples_per_second": 464.001,
      "eval_steps_per_second": 7.3,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 208.42449951171875,
      "learning_rate": 5e-05,
      "loss": 1.8251,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7258666666666667,
      "eval_loss": 1.0228182077407837,
      "eval_runtime": 7.7566,
      "eval_samples_per_second": 483.457,
      "eval_steps_per_second": 7.606,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 161.37974548339844,
      "learning_rate": 5e-05,
      "loss": 0.8365,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8064,
      "eval_loss": 0.6791761517524719,
      "eval_runtime": 7.7235,
      "eval_samples_per_second": 485.531,
      "eval_steps_per_second": 7.639,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 204.32443237304688,
      "learning_rate": 5e-05,
      "loss": 0.6056,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8338666666666666,
      "eval_loss": 0.5683948993682861,
      "eval_runtime": 7.8228,
      "eval_samples_per_second": 479.366,
      "eval_steps_per_second": 7.542,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 58.41059112548828,
      "learning_rate": 5e-05,
      "loss": 0.4866,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8474666666666667,
      "eval_loss": 0.4929230809211731,
      "eval_runtime": 7.8463,
      "eval_samples_per_second": 477.934,
      "eval_steps_per_second": 7.519,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 69.8587417602539,
      "learning_rate": 5e-05,
      "loss": 0.4074,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.4486350119113922,
      "eval_runtime": 7.8966,
      "eval_samples_per_second": 474.886,
      "eval_steps_per_second": 7.472,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 182.325927734375,
      "learning_rate": 5e-05,
      "loss": 0.3569,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.42250117659568787,
      "eval_runtime": 7.7486,
      "eval_samples_per_second": 483.958,
      "eval_steps_per_second": 7.614,
      "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
}