trainer_state.json

{
  "best_metric": 0.9886363636363636,
  "best_model_checkpoint": "swin-large-patch4-window12-384-in22k-finetuned-batch8/checkpoint-98",
  "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": 6.88034200668335,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.643,
      "step": 10
    },
    {
      "epoch": 0.40609137055837563,
      "grad_norm": 19.021066665649414,
      "learning_rate": 4.810606060606061e-05,
      "loss": 0.3831,
      "step": 20
    },
    {
      "epoch": 0.6091370558375635,
      "grad_norm": 9.288772583007812,
      "learning_rate": 4.431818181818182e-05,
      "loss": 0.1721,
      "step": 30
    },
    {
      "epoch": 0.8121827411167513,
      "grad_norm": 17.481571197509766,
      "learning_rate": 4.053030303030303e-05,
      "loss": 0.2123,
      "step": 40
    },
    {
      "epoch": 0.9949238578680203,
      "eval_accuracy": 0.9659090909090909,
      "eval_loss": 0.09163513779640198,
      "eval_runtime": 15.1203,
      "eval_samples_per_second": 11.64,
      "eval_steps_per_second": 1.455,
      "step": 49
    },
    {
      "epoch": 1.015228426395939,
      "grad_norm": 8.028657913208008,
      "learning_rate": 3.6742424242424246e-05,
      "loss": 0.2193,
      "step": 50
    },
    {
      "epoch": 1.218274111675127,
      "grad_norm": 14.979452133178711,
      "learning_rate": 3.295454545454545e-05,
      "loss": 0.2239,
      "step": 60
    },
    {
      "epoch": 1.4213197969543148,
      "grad_norm": 8.288687705993652,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.1636,
      "step": 70
    },
    {
      "epoch": 1.6243654822335025,
      "grad_norm": 9.035058975219727,
      "learning_rate": 2.537878787878788e-05,
      "loss": 0.1848,
      "step": 80
    },
    {
      "epoch": 1.8274111675126905,
      "grad_norm": 18.96419906616211,
      "learning_rate": 2.1590909090909093e-05,
      "loss": 0.1613,
      "step": 90
    },
    {
      "epoch": 1.9898477157360406,
      "eval_accuracy": 0.9886363636363636,
      "eval_loss": 0.04295675456523895,
      "eval_runtime": 15.0722,
      "eval_samples_per_second": 11.677,
      "eval_steps_per_second": 1.46,
      "step": 98
    },
    {
      "epoch": 2.030456852791878,
      "grad_norm": 5.229619979858398,
      "learning_rate": 1.7803030303030303e-05,
      "loss": 0.1375,
      "step": 100
    },
    {
      "epoch": 2.233502538071066,
      "grad_norm": 6.713926792144775,
      "learning_rate": 1.4015151515151515e-05,
      "loss": 0.1352,
      "step": 110
    },
    {
      "epoch": 2.436548223350254,
      "grad_norm": 14.589942932128906,
      "learning_rate": 1.0227272727272729e-05,
      "loss": 0.1208,
      "step": 120
    },
    {
      "epoch": 2.6395939086294415,
      "grad_norm": 1.9984550476074219,
      "learning_rate": 6.43939393939394e-06,
      "loss": 0.107,
      "step": 130
    },
    {
      "epoch": 2.8426395939086295,
      "grad_norm": 10.917881965637207,
      "learning_rate": 2.651515151515152e-06,
      "loss": 0.116,
      "step": 140
    },
    {
      "epoch": 2.984771573604061,
      "eval_accuracy": 0.9886363636363636,
      "eval_loss": 0.034594591706991196,
      "eval_runtime": 15.0628,
      "eval_samples_per_second": 11.684,
      "eval_steps_per_second": 1.461,
      "step": 147
    },
    {
      "epoch": 2.984771573604061,
      "step": 147,
      "total_flos": 2.437168844903547e+18,
      "train_loss": 0.20754414107523808,
      "train_runtime": 1195.5632,
      "train_samples_per_second": 3.955,
      "train_steps_per_second": 0.123
    }
  ],
  "logging_steps": 10,
  "max_steps": 147,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2.437168844903547e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}