trainer_state.json

{
  "best_metric": 0.4962666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0784/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0004850126564054637,
      "loss": 3.377,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1832,
      "eval_loss": 3.067507743835449,
      "eval_runtime": 7.4686,
      "eval_samples_per_second": 502.105,
      "eval_steps_per_second": 7.9,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00044184758936602033,
      "loss": 2.8298,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2432,
      "eval_loss": 2.7184760570526123,
      "eval_runtime": 7.6726,
      "eval_samples_per_second": 488.75,
      "eval_steps_per_second": 7.69,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 8.609707832336426,
      "learning_rate": 0.0003754536287075859,
      "loss": 2.5506,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.304,
      "eval_loss": 2.534919500350952,
      "eval_runtime": 7.3825,
      "eval_samples_per_second": 507.956,
      "eval_steps_per_second": 7.992,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00029418606892278544,
      "loss": 2.2862,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.35386666666666666,
      "eval_loss": 2.2957489490509033,
      "eval_runtime": 7.3132,
      "eval_samples_per_second": 512.775,
      "eval_steps_per_second": 8.068,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 19.02993392944336,
      "learning_rate": 0.0002073624094068776,
      "loss": 1.9931,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4056,
      "eval_loss": 2.105799436569214,
      "eval_runtime": 7.2646,
      "eval_samples_per_second": 516.202,
      "eval_steps_per_second": 8.122,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 12.135807991027832,
      "learning_rate": 0.00012568147258031897,
      "loss": 1.684,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.44,
      "eval_loss": 2.002918243408203,
      "eval_runtime": 6.836,
      "eval_samples_per_second": 548.565,
      "eval_steps_per_second": 8.631,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 21.180992126464844,
      "learning_rate": 5.899518490663266e-05,
      "loss": 1.3675,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.47173333333333334,
      "eval_loss": 1.8673189878463745,
      "eval_runtime": 7.4249,
      "eval_samples_per_second": 505.055,
      "eval_steps_per_second": 7.946,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.2181549072265625,
      "learning_rate": 1.534689686405272e-05,
      "loss": 1.0868,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.49466666666666664,
      "eval_loss": 1.8258147239685059,
      "eval_runtime": 6.9923,
      "eval_samples_per_second": 536.308,
      "eval_steps_per_second": 8.438,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 23.34157371520996,
      "learning_rate": 1.2361706385188854e-09,
      "loss": 0.9381,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4962666666666667,
      "eval_loss": 1.821169376373291,
      "eval_runtime": 6.8757,
      "eval_samples_per_second": 545.399,
      "eval_steps_per_second": 8.581,
      "step": 2997
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}