resnet50_fold_1
This model is a fine-tuned version of microsoft/resnet-50 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 1.1010
- Accuracy: 0.9690
- F1 Score: 0.9708
- Recall: 0.9730
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: 0.0001
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 150
- num_epochs: 100
- label_smoothing_factor: 0.15
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Score | Recall |
|---|---|---|---|---|---|---|
| 2.7705 | 1.0 | 19 | 2.7688 | 0.2241 | 0.1833 | 0.2211 |
| 2.7598 | 2.0 | 38 | 2.7554 | 0.2793 | 0.1855 | 0.2301 |
| 2.7265 | 3.0 | 57 | 2.7304 | 0.3483 | 0.1995 | 0.2700 |
| 2.6927 | 4.0 | 76 | 2.6900 | 0.3586 | 0.1698 | 0.2627 |
| 2.6274 | 5.0 | 95 | 2.6449 | 0.3586 | 0.1505 | 0.2583 |
| 2.5275 | 6.0 | 114 | 2.5764 | 0.3931 | 0.2326 | 0.3025 |
| 2.4048 | 7.0 | 133 | 2.4666 | 0.5483 | 0.4814 | 0.5142 |
| 2.2350 | 8.0 | 152 | 2.2408 | 0.7069 | 0.6959 | 0.6968 |
| 1.9944 | 9.0 | 171 | 1.9572 | 0.8103 | 0.8143 | 0.8093 |
| 1.7608 | 10.0 | 190 | 1.6624 | 0.8724 | 0.8746 | 0.8781 |
| 1.6519 | 11.0 | 209 | 1.4591 | 0.9 | 0.9074 | 0.9084 |
| 1.5012 | 12.0 | 228 | 1.3566 | 0.9034 | 0.9111 | 0.9144 |
| 1.4065 | 13.0 | 247 | 1.3477 | 0.9034 | 0.9043 | 0.9132 |
| 1.3178 | 14.0 | 266 | 1.2477 | 0.9172 | 0.9231 | 0.9291 |
| 1.2779 | 15.0 | 285 | 1.2359 | 0.9241 | 0.9283 | 0.9329 |
| 1.2558 | 16.0 | 304 | 1.2012 | 0.9379 | 0.9427 | 0.9474 |
| 1.1987 | 17.0 | 323 | 1.2123 | 0.9345 | 0.9353 | 0.9425 |
| 1.1862 | 18.0 | 342 | 1.2075 | 0.9379 | 0.9385 | 0.9462 |
| 1.2160 | 19.0 | 361 | 1.1711 | 0.9379 | 0.9392 | 0.9449 |
| 1.3062 | 20.0 | 380 | 1.1439 | 0.9586 | 0.9585 | 0.9608 |
| 1.1480 | 21.0 | 399 | 1.1638 | 0.9517 | 0.9525 | 0.9596 |
| 1.1471 | 22.0 | 418 | 1.1646 | 0.9517 | 0.9521 | 0.9571 |
| 1.2000 | 23.0 | 437 | 1.1465 | 0.9483 | 0.9496 | 0.9534 |
| 1.1213 | 24.0 | 456 | 1.1368 | 0.9483 | 0.9481 | 0.9522 |
| 1.1640 | 25.0 | 475 | 1.1115 | 0.9621 | 0.9632 | 0.9657 |
| 1.1094 | 26.0 | 494 | 1.1468 | 0.9586 | 0.9606 | 0.9657 |
| 1.1218 | 27.0 | 513 | 1.1325 | 0.9483 | 0.9463 | 0.9522 |
| 1.1314 | 28.0 | 532 | 1.1256 | 0.9483 | 0.9480 | 0.9547 |
| 1.1161 | 29.0 | 551 | 1.1225 | 0.9552 | 0.9560 | 0.9608 |
| 1.1771 | 30.0 | 570 | 1.1137 | 0.9448 | 0.9430 | 0.9485 |
| 1.1027 | 31.0 | 589 | 1.1234 | 0.9448 | 0.9427 | 0.9473 |
| 1.1052 | 32.0 | 608 | 1.1542 | 0.9448 | 0.9428 | 0.9498 |
| 1.1163 | 33.0 | 627 | 1.1121 | 0.9552 | 0.9546 | 0.9571 |
| 1.0949 | 34.0 | 646 | 1.1226 | 0.9517 | 0.9523 | 0.9547 |
| 1.1121 | 35.0 | 665 | 1.1010 | 0.9690 | 0.9708 | 0.9730 |
| 1.0660 | 36.0 | 684 | 1.1164 | 0.9586 | 0.9600 | 0.9581 |
| 1.0682 | 37.0 | 703 | 1.1129 | 0.9655 | 0.9664 | 0.9694 |
| 1.0752 | 38.0 | 722 | 1.1153 | 0.9655 | 0.9637 | 0.9681 |
| 1.0583 | 39.0 | 741 | 1.1141 | 0.9552 | 0.9551 | 0.9596 |
| 1.0750 | 40.0 | 760 | 1.1140 | 0.9621 | 0.9639 | 0.9669 |
| 1.0351 | 41.0 | 779 | 1.1209 | 0.9655 | 0.9657 | 0.9694 |
| 1.0556 | 42.0 | 798 | 1.1155 | 0.9586 | 0.9581 | 0.9620 |
| 1.0381 | 43.0 | 817 | 1.1226 | 0.9483 | 0.9500 | 0.9559 |
| 1.0185 | 44.0 | 836 | 1.1488 | 0.9517 | 0.9537 | 0.9596 |
| 1.0324 | 45.0 | 855 | 1.1278 | 0.9552 | 0.9553 | 0.9608 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2