trainer_state.json

{
  "best_metric": 0.8538666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0443/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": 52.500572204589844,
      "learning_rate": 8.571428571428571e-05,
      "loss": 2.9951,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.652,
      "eval_loss": 1.2699600458145142,
      "eval_runtime": 15.8268,
      "eval_samples_per_second": 236.94,
      "eval_steps_per_second": 3.728,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 51.5329475402832,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.9641,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7837333333333333,
      "eval_loss": 0.7130506634712219,
      "eval_runtime": 15.8488,
      "eval_samples_per_second": 236.611,
      "eval_steps_per_second": 3.723,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 100.09130096435547,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.6528,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8274666666666667,
      "eval_loss": 0.5603117346763611,
      "eval_runtime": 15.6591,
      "eval_samples_per_second": 239.478,
      "eval_steps_per_second": 3.768,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 93.34806823730469,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.5096,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8421333333333333,
      "eval_loss": 0.5103253126144409,
      "eval_runtime": 15.5675,
      "eval_samples_per_second": 240.886,
      "eval_steps_per_second": 3.79,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 100.70523834228516,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.4381,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8357333333333333,
      "eval_loss": 0.49863898754119873,
      "eval_runtime": 15.5952,
      "eval_samples_per_second": 240.458,
      "eval_steps_per_second": 3.783,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 67.50264739990234,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.3855,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8453333333333334,
      "eval_loss": 0.5014702677726746,
      "eval_runtime": 15.5783,
      "eval_samples_per_second": 240.719,
      "eval_steps_per_second": 3.787,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 37.63423538208008,
      "learning_rate": 0.0,
      "loss": 0.3635,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8538666666666667,
      "eval_loss": 0.4598385989665985,
      "eval_runtime": 15.6782,
      "eval_samples_per_second": 239.186,
      "eval_steps_per_second": 3.763,
      "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
}