trainer_state.json

{
  "best_metric": 0.8802666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0824/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": Infinity,
      "learning_rate": 9.507764069616767e-05,
      "loss": 1.5443,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7706666666666667,
      "eval_loss": 0.7734163999557495,
      "eval_runtime": 6.1709,
      "eval_samples_per_second": 607.693,
      "eval_steps_per_second": 9.561,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.127974762931168e-05,
      "loss": 0.5132,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8328,
      "eval_loss": 0.5735070109367371,
      "eval_runtime": 6.5331,
      "eval_samples_per_second": 573.997,
      "eval_steps_per_second": 9.031,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 24.359216690063477,
      "learning_rate": 6.125740150220697e-05,
      "loss": 0.2474,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.4901064336299896,
      "eval_runtime": 7.0466,
      "eval_samples_per_second": 532.175,
      "eval_steps_per_second": 8.373,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.7900772094726562,
      "learning_rate": 3.900538894462112e-05,
      "loss": 0.109,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8744,
      "eval_loss": 0.4496123492717743,
      "eval_runtime": 6.5846,
      "eval_samples_per_second": 569.513,
      "eval_steps_per_second": 8.96,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.171233132481575,
      "learning_rate": 1.893099394660879e-05,
      "loss": 0.0387,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8709333333333333,
      "eval_loss": 0.46494653820991516,
      "eval_runtime": 6.0906,
      "eval_samples_per_second": 615.706,
      "eval_steps_per_second": 9.687,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.20167602598667145,
      "learning_rate": 5.0101966338242764e-06,
      "loss": 0.0115,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8786666666666667,
      "eval_loss": 0.4543612003326416,
      "eval_runtime": 6.2823,
      "eval_samples_per_second": 596.913,
      "eval_steps_per_second": 9.391,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0845176950097084,
      "learning_rate": 1.8164143966159507e-10,
      "loss": 0.006,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8802666666666666,
      "eval_loss": 0.4529889225959778,
      "eval_runtime": 6.1126,
      "eval_samples_per_second": 613.489,
      "eval_steps_per_second": 9.652,
      "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}