trainer_state.json

{
  "best_metric": 0.9554666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0222/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": 0.4285023808479309,
      "learning_rate": 9.619397662556435e-05,
      "loss": 0.6086,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9146666666666666,
      "eval_loss": 0.2975988984107971,
      "eval_runtime": 5.4294,
      "eval_samples_per_second": 690.684,
      "eval_steps_per_second": 10.867,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.701775074005127,
      "learning_rate": 8.535533905932738e-05,
      "loss": 0.1143,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9405333333333333,
      "eval_loss": 0.21148596704006195,
      "eval_runtime": 5.2509,
      "eval_samples_per_second": 714.163,
      "eval_steps_per_second": 11.236,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.9087717533111572,
      "learning_rate": 6.91341716182545e-05,
      "loss": 0.0476,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9413333333333334,
      "eval_loss": 0.23115843534469604,
      "eval_runtime": 5.6157,
      "eval_samples_per_second": 667.775,
      "eval_steps_per_second": 10.506,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.036762628704309464,
      "learning_rate": 5e-05,
      "loss": 0.0165,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9437333333333333,
      "eval_loss": 0.22946017980575562,
      "eval_runtime": 5.2658,
      "eval_samples_per_second": 712.137,
      "eval_steps_per_second": 11.204,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.011106787249445915,
      "learning_rate": 3.086582838174551e-05,
      "loss": 0.0059,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9525333333333333,
      "eval_loss": 0.19271056354045868,
      "eval_runtime": 5.8894,
      "eval_samples_per_second": 636.738,
      "eval_steps_per_second": 10.018,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.01110786385834217,
      "learning_rate": 1.4644660940672627e-05,
      "loss": 0.0012,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9552,
      "eval_loss": 0.1833174079656601,
      "eval_runtime": 5.263,
      "eval_samples_per_second": 712.516,
      "eval_steps_per_second": 11.21,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.82861996937452e-06,
      "loss": 0.0009,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9554666666666667,
      "eval_loss": 0.18305395543575287,
      "eval_runtime": 5.2184,
      "eval_samples_per_second": 718.613,
      "eval_steps_per_second": 11.306,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}