trainer_state.json

{
  "best_metric": 0.9461333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0834/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": 3.225623846054077,
      "learning_rate": 2.625e-05,
      "loss": 1.1128,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9330666666666667,
      "eval_loss": 0.2762678265571594,
      "eval_runtime": 7.3386,
      "eval_samples_per_second": 510.996,
      "eval_steps_per_second": 8.04,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8658239841461182,
      "learning_rate": 2.25e-05,
      "loss": 0.1717,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9445333333333333,
      "eval_loss": 0.2081771045923233,
      "eval_runtime": 7.3722,
      "eval_samples_per_second": 508.667,
      "eval_steps_per_second": 8.003,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0862040519714355,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0807,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.944,
      "eval_loss": 0.19132301211357117,
      "eval_runtime": 7.658,
      "eval_samples_per_second": 489.685,
      "eval_steps_per_second": 7.704,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.08552885055542,
      "learning_rate": 1.5e-05,
      "loss": 0.0425,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.944,
      "eval_loss": 0.1912221610546112,
      "eval_runtime": 7.2551,
      "eval_samples_per_second": 516.881,
      "eval_steps_per_second": 8.132,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.49569204449653625,
      "learning_rate": 1.125e-05,
      "loss": 0.0225,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9429333333333333,
      "eval_loss": 0.20669427514076233,
      "eval_runtime": 7.0926,
      "eval_samples_per_second": 528.718,
      "eval_steps_per_second": 8.318,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.11636556684970856,
      "learning_rate": 7.5e-06,
      "loss": 0.0136,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9448,
      "eval_loss": 0.1974092572927475,
      "eval_runtime": 7.1389,
      "eval_samples_per_second": 525.289,
      "eval_steps_per_second": 8.265,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.7612612612612617e-06,
      "loss": 0.0086,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9453333333333334,
      "eval_loss": 0.21206238865852356,
      "eval_runtime": 7.2831,
      "eval_samples_per_second": 514.89,
      "eval_steps_per_second": 8.101,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.04288652166724205,
      "learning_rate": 1.1261261261261262e-08,
      "loss": 0.0059,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9461333333333334,
      "eval_loss": 0.21071556210517883,
      "eval_runtime": 6.8095,
      "eval_samples_per_second": 550.705,
      "eval_steps_per_second": 8.664,
      "step": 2664
    }
  ],
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}