trainer_state.json

{
  "best_metric": 0.9450666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0300/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 5.7958083152771,
      "learning_rate": 8e-05,
      "loss": 0.6205,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9210666666666667,
      "eval_loss": 0.2606699764728546,
      "eval_runtime": 5.8373,
      "eval_samples_per_second": 642.422,
      "eval_steps_per_second": 10.107,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3059037923812866,
      "learning_rate": 7.000000000000001e-05,
      "loss": 0.124,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9325333333333333,
      "eval_loss": 0.23118914663791656,
      "eval_runtime": 6.9272,
      "eval_samples_per_second": 541.348,
      "eval_steps_per_second": 8.517,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.02100568637251854,
      "learning_rate": 6e-05,
      "loss": 0.0499,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9322666666666667,
      "eval_loss": 0.2462834268808365,
      "eval_runtime": 5.7036,
      "eval_samples_per_second": 657.474,
      "eval_steps_per_second": 10.344,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.07997412979602814,
      "learning_rate": 5e-05,
      "loss": 0.0207,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.2651498317718506,
      "eval_runtime": 6.2787,
      "eval_samples_per_second": 597.256,
      "eval_steps_per_second": 9.397,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.02619319222867489,
      "learning_rate": 4e-05,
      "loss": 0.0069,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9416,
      "eval_loss": 0.2357293665409088,
      "eval_runtime": 6.3144,
      "eval_samples_per_second": 593.884,
      "eval_steps_per_second": 9.344,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.02882780134677887,
      "learning_rate": 3e-05,
      "loss": 0.0025,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9429333333333333,
      "eval_loss": 0.24376992881298065,
      "eval_runtime": 5.7999,
      "eval_samples_per_second": 646.562,
      "eval_steps_per_second": 10.173,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.0030030030030034e-05,
      "loss": 0.0021,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9445333333333333,
      "eval_loss": 0.24343667924404144,
      "eval_runtime": 5.9591,
      "eval_samples_per_second": 629.287,
      "eval_steps_per_second": 9.901,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.01545003056526184,
      "learning_rate": 1.0030030030030031e-05,
      "loss": 0.0014,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9450666666666667,
      "eval_loss": 0.2438563108444214,
      "eval_runtime": 7.3592,
      "eval_samples_per_second": 509.564,
      "eval_steps_per_second": 8.017,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}