trainer_state.json

{
  "best_metric": 0.9002666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0466/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": 77.10993957519531,
      "learning_rate": 7.714285714285714e-05,
      "loss": 2.7333,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7656,
      "eval_loss": 0.9132193922996521,
      "eval_runtime": 14.1641,
      "eval_samples_per_second": 264.754,
      "eval_steps_per_second": 4.165,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 140.9851531982422,
      "learning_rate": 6.428571428571429e-05,
      "loss": 0.676,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8634666666666667,
      "eval_loss": 0.4987955093383789,
      "eval_runtime": 13.9381,
      "eval_samples_per_second": 269.046,
      "eval_steps_per_second": 4.233,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 68.31900024414062,
      "learning_rate": 5.142857142857143e-05,
      "loss": 0.3922,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8845333333333333,
      "eval_loss": 0.39569273591041565,
      "eval_runtime": 14.1698,
      "eval_samples_per_second": 264.648,
      "eval_steps_per_second": 4.164,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 86.45220947265625,
      "learning_rate": 3.857142857142857e-05,
      "loss": 0.2647,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8936,
      "eval_loss": 0.3662910759449005,
      "eval_runtime": 14.0589,
      "eval_samples_per_second": 266.735,
      "eval_steps_per_second": 4.197,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 66.13668060302734,
      "learning_rate": 2.5714285714285714e-05,
      "loss": 0.1923,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.35647982358932495,
      "eval_runtime": 14.0457,
      "eval_samples_per_second": 266.985,
      "eval_steps_per_second": 4.201,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 55.57662582397461,
      "learning_rate": 1.2857142857142857e-05,
      "loss": 0.1489,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.35168787837028503,
      "eval_runtime": 14.0007,
      "eval_samples_per_second": 267.843,
      "eval_steps_per_second": 4.214,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 177.43081665039062,
      "learning_rate": 0.0,
      "loss": 0.1288,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9002666666666667,
      "eval_loss": 0.33683711290359497,
      "eval_runtime": 14.068,
      "eval_samples_per_second": 266.562,
      "eval_steps_per_second": 4.194,
      "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
}