trainer_state.json

{
  "best_metric": 0.8770666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0937/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": 49.83320999145508,
      "learning_rate": 5e-05,
      "loss": 3.6449,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5408,
      "eval_loss": 3.0064234733581543,
      "eval_runtime": 12.8572,
      "eval_samples_per_second": 291.665,
      "eval_steps_per_second": 4.589,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 141.43283081054688,
      "learning_rate": 5e-05,
      "loss": 1.6594,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7794666666666666,
      "eval_loss": 0.8702710866928101,
      "eval_runtime": 12.6208,
      "eval_samples_per_second": 297.129,
      "eval_steps_per_second": 4.675,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 144.1399383544922,
      "learning_rate": 5e-05,
      "loss": 0.7261,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8314666666666667,
      "eval_loss": 0.556423544883728,
      "eval_runtime": 12.355,
      "eval_samples_per_second": 303.521,
      "eval_steps_per_second": 4.775,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 151.8025360107422,
      "learning_rate": 5e-05,
      "loss": 0.507,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8506666666666667,
      "eval_loss": 0.49124833941459656,
      "eval_runtime": 12.3209,
      "eval_samples_per_second": 304.36,
      "eval_steps_per_second": 4.789,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 93.41546630859375,
      "learning_rate": 5e-05,
      "loss": 0.3959,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8682666666666666,
      "eval_loss": 0.4395199418067932,
      "eval_runtime": 12.3379,
      "eval_samples_per_second": 303.942,
      "eval_steps_per_second": 4.782,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 99.51985931396484,
      "learning_rate": 5e-05,
      "loss": 0.3219,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.3947522044181824,
      "eval_runtime": 12.7298,
      "eval_samples_per_second": 294.583,
      "eval_steps_per_second": 4.635,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 122.75646209716797,
      "learning_rate": 5e-05,
      "loss": 0.2601,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.3835596442222595,
      "eval_runtime": 18.5841,
      "eval_samples_per_second": 201.785,
      "eval_steps_per_second": 3.175,
      "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
}