trainer_state.json

{
  "best_metric": 0.9715909090909091,
  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-batch8-nocrop/checkpoint-147",
  "epoch": 2.984771573604061,
  "eval_steps": 500,
  "global_step": 147,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20304568527918782,
      "grad_norm": 6.709445953369141,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.6831,
      "step": 10
    },
    {
      "epoch": 0.40609137055837563,
      "grad_norm": 36.66501235961914,
      "learning_rate": 4.810606060606061e-05,
      "loss": 0.5802,
      "step": 20
    },
    {
      "epoch": 0.6091370558375635,
      "grad_norm": 7.574514865875244,
      "learning_rate": 4.431818181818182e-05,
      "loss": 0.3748,
      "step": 30
    },
    {
      "epoch": 0.8121827411167513,
      "grad_norm": 12.18085765838623,
      "learning_rate": 4.053030303030303e-05,
      "loss": 0.2052,
      "step": 40
    },
    {
      "epoch": 0.9949238578680203,
      "eval_accuracy": 0.9602272727272727,
      "eval_loss": 0.15112267434597015,
      "eval_runtime": 1.4305,
      "eval_samples_per_second": 123.032,
      "eval_steps_per_second": 15.379,
      "step": 49
    },
    {
      "epoch": 1.015228426395939,
      "grad_norm": 8.17941665649414,
      "learning_rate": 3.6742424242424246e-05,
      "loss": 0.1755,
      "step": 50
    },
    {
      "epoch": 1.218274111675127,
      "grad_norm": 7.0626115798950195,
      "learning_rate": 3.295454545454545e-05,
      "loss": 0.0786,
      "step": 60
    },
    {
      "epoch": 1.4213197969543148,
      "grad_norm": 25.362876892089844,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.1441,
      "step": 70
    },
    {
      "epoch": 1.6243654822335025,
      "grad_norm": 15.33166790008545,
      "learning_rate": 2.537878787878788e-05,
      "loss": 0.0825,
      "step": 80
    },
    {
      "epoch": 1.8274111675126905,
      "grad_norm": 2.483710527420044,
      "learning_rate": 2.1590909090909093e-05,
      "loss": 0.0852,
      "step": 90
    },
    {
      "epoch": 1.9898477157360406,
      "eval_accuracy": 0.9488636363636364,
      "eval_loss": 0.1786985695362091,
      "eval_runtime": 1.4267,
      "eval_samples_per_second": 123.365,
      "eval_steps_per_second": 15.421,
      "step": 98
    },
    {
      "epoch": 2.030456852791878,
      "grad_norm": 23.29215431213379,
      "learning_rate": 1.7803030303030303e-05,
      "loss": 0.0657,
      "step": 100
    },
    {
      "epoch": 2.233502538071066,
      "grad_norm": 4.370492458343506,
      "learning_rate": 1.4015151515151515e-05,
      "loss": 0.0307,
      "step": 110
    },
    {
      "epoch": 2.436548223350254,
      "grad_norm": 9.500088691711426,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.0276,
      "step": 120
    },
    {
      "epoch": 2.6395939086294415,
      "grad_norm": 6.448863506317139,
      "learning_rate": 6.43939393939394e-06,
      "loss": 0.0273,
      "step": 130
    },
    {
      "epoch": 2.8426395939086295,
      "grad_norm": 7.503269672393799,
      "learning_rate": 2.651515151515152e-06,
      "loss": 0.0336,
      "step": 140
    },
    {
      "epoch": 2.984771573604061,
      "eval_accuracy": 0.9715909090909091,
      "eval_loss": 0.09903481602668762,
      "eval_runtime": 1.4469,
      "eval_samples_per_second": 121.636,
      "eval_steps_per_second": 15.205,
      "step": 147
    },
    {
      "epoch": 2.984771573604061,
      "step": 147,
      "total_flos": 1.1692257218946662e+17,
      "train_loss": 0.1783457891065247,
      "train_runtime": 84.8256,
      "train_samples_per_second": 55.738,
      "train_steps_per_second": 1.733
    }
  ],
  "logging_steps": 10,
  "max_steps": 147,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 1.1692257218946662e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}