Edit model card

beit-base-patch16-224-fold1

This model is a fine-tuned version of microsoft/beit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:

  • Loss: 0.7241
  • Accuracy: 0.8481

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 32
  • seed: 42
  • gradient_accumulation_steps: 4
  • total_train_batch_size: 128
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • lr_scheduler_warmup_ratio: 0.1
  • num_epochs: 100

Training results

Training Loss Epoch Step Validation Loss Accuracy
No log 0.8571 3 0.8050 0.4557
No log 2.0 7 0.7151 0.5696
0.8103 2.8571 10 0.6822 0.5570
0.8103 4.0 14 0.6408 0.5696
0.8103 4.8571 17 0.6244 0.6709
0.6583 6.0 21 0.5893 0.6709
0.6583 6.8571 24 0.5877 0.6329
0.6583 8.0 28 0.5752 0.6835
0.5912 8.8571 31 0.5826 0.6456
0.5912 10.0 35 0.5469 0.6835
0.5912 10.8571 38 0.6173 0.6582
0.5301 12.0 42 0.5151 0.6962
0.5301 12.8571 45 0.5105 0.6962
0.5301 14.0 49 0.5489 0.7089
0.4703 14.8571 52 0.5725 0.6835
0.4703 16.0 56 0.5560 0.6962
0.4703 16.8571 59 0.5824 0.6709
0.4189 18.0 63 0.5401 0.7468
0.4189 18.8571 66 0.5147 0.7722
0.3741 20.0 70 0.4864 0.7595
0.3741 20.8571 73 0.5272 0.7342
0.3741 22.0 77 0.4914 0.7468
0.387 22.8571 80 0.5658 0.7468
0.387 24.0 84 0.4662 0.7722
0.387 24.8571 87 0.4376 0.7848
0.3502 26.0 91 0.5367 0.7722
0.3502 26.8571 94 0.5490 0.7342
0.3502 28.0 98 0.7163 0.7722
0.3148 28.8571 101 0.6005 0.7468
0.3148 30.0 105 0.6501 0.7722
0.3148 30.8571 108 0.5313 0.7975
0.2973 32.0 112 0.5466 0.7722
0.2973 32.8571 115 0.5731 0.8101
0.2973 34.0 119 0.6544 0.8101
0.2474 34.8571 122 0.6061 0.7848
0.2474 36.0 126 0.5816 0.7722
0.2474 36.8571 129 0.7161 0.7595
0.2033 38.0 133 0.6235 0.7848
0.2033 38.8571 136 0.7889 0.7595
0.2338 40.0 140 0.5943 0.7595
0.2338 40.8571 143 0.6170 0.7342
0.2338 42.0 147 0.6964 0.6962
0.2067 42.8571 150 0.7154 0.7468
0.2067 44.0 154 0.7675 0.7722
0.2067 44.8571 157 0.7766 0.7468
0.2133 46.0 161 0.9330 0.7848
0.2133 46.8571 164 0.6494 0.7975
0.2133 48.0 168 0.5709 0.7722
0.2004 48.8571 171 0.6462 0.8101
0.2004 50.0 175 0.6668 0.7722
0.2004 50.8571 178 0.6305 0.8101
0.188 52.0 182 0.7189 0.8228
0.188 52.8571 185 0.6853 0.7848
0.188 54.0 189 0.8040 0.8228
0.1623 54.8571 192 0.6958 0.8101
0.1623 56.0 196 0.6907 0.8101
0.1623 56.8571 199 0.6821 0.8101
0.1588 58.0 203 0.6534 0.8101
0.1588 58.8571 206 0.7192 0.8101
0.1607 60.0 210 0.7753 0.8228
0.1607 60.8571 213 0.8950 0.8101
0.1607 62.0 217 0.7904 0.8101
0.1767 62.8571 220 0.6973 0.8101
0.1767 64.0 224 0.6694 0.7975
0.1767 64.8571 227 0.6339 0.8101
0.1463 66.0 231 0.6530 0.8101
0.1463 66.8571 234 0.6142 0.8101
0.1463 68.0 238 0.6290 0.8228
0.1287 68.8571 241 0.6334 0.8354
0.1287 70.0 245 0.8059 0.8101
0.1287 70.8571 248 0.7241 0.8481
0.1323 72.0 252 0.6836 0.8481
0.1323 72.8571 255 0.6588 0.8228
0.1323 74.0 259 0.6598 0.8481
0.1042 74.8571 262 0.7139 0.8354
0.1042 76.0 266 0.7236 0.8354
0.1042 76.8571 269 0.6919 0.8354
0.1106 78.0 273 0.6568 0.8354
0.1106 78.8571 276 0.6556 0.8481
0.1348 80.0 280 0.6612 0.8354
0.1348 80.8571 283 0.6686 0.8228
0.1348 82.0 287 0.6705 0.8481
0.1352 82.8571 290 0.6776 0.8354
0.1352 84.0 294 0.6873 0.8354
0.1352 84.8571 297 0.6888 0.8354
0.1226 85.7143 300 0.6880 0.8354

Framework versions

  • Transformers 4.40.2
  • Pytorch 2.2.1+cu121
  • Datasets 2.19.1
  • Tokenizers 0.19.1
Downloads last month
2
Safetensors
Model size
85.8M params
Tensor type
F32
·

Finetuned from

Evaluation results