trainer_state.json

{
  "best_metric": 0.8706666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0321/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": 455.7652587890625,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 3.6448,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.588,
      "eval_loss": 3.076093912124634,
      "eval_runtime": 7.6111,
      "eval_samples_per_second": 492.699,
      "eval_steps_per_second": 7.752,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 216.0070037841797,
      "learning_rate": 3.571428571428572e-05,
      "loss": 1.9565,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7605333333333333,
      "eval_loss": 1.0637831687927246,
      "eval_runtime": 7.6815,
      "eval_samples_per_second": 488.185,
      "eval_steps_per_second": 7.681,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 121.97456359863281,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.8311,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8304,
      "eval_loss": 0.6215339303016663,
      "eval_runtime": 7.6883,
      "eval_samples_per_second": 487.755,
      "eval_steps_per_second": 7.674,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 157.86314392089844,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.584,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8474666666666667,
      "eval_loss": 0.512190043926239,
      "eval_runtime": 7.7359,
      "eval_samples_per_second": 484.756,
      "eval_steps_per_second": 7.627,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 135.15408325195312,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.4867,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.46384692192077637,
      "eval_runtime": 7.6122,
      "eval_samples_per_second": 492.63,
      "eval_steps_per_second": 7.751,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 195.56370544433594,
      "learning_rate": 7.142857142857143e-06,
      "loss": 0.4246,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8645333333333334,
      "eval_loss": 0.44718268513679504,
      "eval_runtime": 7.543,
      "eval_samples_per_second": 497.148,
      "eval_steps_per_second": 7.822,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 363.8642272949219,
      "learning_rate": 0.0,
      "loss": 0.4058,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8706666666666667,
      "eval_loss": 0.4354051351547241,
      "eval_runtime": 7.6049,
      "eval_samples_per_second": 493.106,
      "eval_steps_per_second": 7.758,
      "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
}