resnet50_fold_2
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: 1.2330
- Accuracy: 0.9276
- F1 Score: 0.9327
- Recall: 0.9355
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 |
|---|---|---|---|---|---|---|
| 2.7704 | 1.0 | 19 | 2.7578 | 0.4 | 0.2936 | 0.3646 |
| 2.7541 | 2.0 | 38 | 2.7434 | 0.4069 | 0.2909 | 0.3420 |
| 2.7183 | 3.0 | 57 | 2.7200 | 0.4034 | 0.2708 | 0.3223 |
| 2.6750 | 4.0 | 76 | 2.6838 | 0.4103 | 0.2821 | 0.3313 |
| 2.6012 | 5.0 | 95 | 2.6482 | 0.4414 | 0.3326 | 0.3737 |
| 2.4884 | 6.0 | 114 | 2.5733 | 0.5276 | 0.4471 | 0.4836 |
| 2.3305 | 7.0 | 133 | 2.4639 | 0.6 | 0.5363 | 0.5778 |
| 2.1833 | 8.0 | 152 | 2.2702 | 0.7241 | 0.6861 | 0.7149 |
| 1.9297 | 9.0 | 171 | 1.9626 | 0.8 | 0.8048 | 0.8041 |
| 1.7430 | 10.0 | 190 | 1.6955 | 0.8483 | 0.8528 | 0.8538 |
| 1.6346 | 11.0 | 209 | 1.5015 | 0.8828 | 0.8922 | 0.8928 |
| 1.4690 | 12.0 | 228 | 1.4260 | 0.8862 | 0.8941 | 0.8979 |
| 1.3484 | 13.0 | 247 | 1.3753 | 0.8897 | 0.8963 | 0.8989 |
| 1.2518 | 14.0 | 266 | 1.3165 | 0.9 | 0.9077 | 0.9111 |
| 1.2253 | 15.0 | 285 | 1.3249 | 0.9172 | 0.9235 | 0.9271 |
| 1.2322 | 16.0 | 304 | 1.3016 | 0.8966 | 0.9041 | 0.9075 |
| 1.2039 | 17.0 | 323 | 1.3057 | 0.9034 | 0.9110 | 0.9136 |
| 1.1545 | 18.0 | 342 | 1.2971 | 0.9103 | 0.9163 | 0.9210 |
| 1.1947 | 19.0 | 361 | 1.2571 | 0.9172 | 0.9230 | 0.9258 |
| 1.2689 | 20.0 | 380 | 1.2523 | 0.9034 | 0.9113 | 0.9112 |
| 1.1614 | 21.0 | 399 | 1.2753 | 0.9069 | 0.9147 | 0.9160 |
| 1.1393 | 22.0 | 418 | 1.2788 | 0.9034 | 0.9101 | 0.9110 |
| 1.1489 | 23.0 | 437 | 1.2541 | 0.9207 | 0.9261 | 0.9258 |
| 1.1225 | 24.0 | 456 | 1.2571 | 0.9103 | 0.9175 | 0.9184 |
| 1.1244 | 25.0 | 475 | 1.2627 | 0.9172 | 0.9233 | 0.9245 |
| 1.1049 | 26.0 | 494 | 1.2678 | 0.9207 | 0.9275 | 0.9319 |
| 1.1145 | 27.0 | 513 | 1.2552 | 0.9241 | 0.9293 | 0.9294 |
| 1.1077 | 28.0 | 532 | 1.2575 | 0.9172 | 0.9239 | 0.9259 |
| 1.1031 | 29.0 | 551 | 1.2473 | 0.9172 | 0.9232 | 0.9270 |
| 1.1690 | 30.0 | 570 | 1.2503 | 0.9207 | 0.9256 | 0.9245 |
| 1.0949 | 31.0 | 589 | 1.2366 | 0.9103 | 0.9170 | 0.9161 |
| 1.0662 | 32.0 | 608 | 1.2516 | 0.9103 | 0.9169 | 0.9194 |
| 1.1033 | 33.0 | 627 | 1.2593 | 0.9138 | 0.9205 | 0.9210 |
| 1.1002 | 34.0 | 646 | 1.2317 | 0.9207 | 0.9251 | 0.9257 |
| 1.0843 | 35.0 | 665 | 1.2484 | 0.9138 | 0.9194 | 0.9196 |
| 1.0412 | 36.0 | 684 | 1.2428 | 0.9172 | 0.9236 | 0.9258 |
| 1.0517 | 37.0 | 703 | 1.2349 | 0.9138 | 0.9197 | 0.9209 |
| 1.0436 | 38.0 | 722 | 1.2566 | 0.9069 | 0.9124 | 0.9124 |
| 1.0631 | 39.0 | 741 | 1.2451 | 0.9207 | 0.9268 | 0.9283 |
| 1.0571 | 40.0 | 760 | 1.2434 | 0.9241 | 0.9293 | 0.9294 |
| 1.0431 | 41.0 | 779 | 1.2330 | 0.9276 | 0.9327 | 0.9355 |
| 1.0494 | 42.0 | 798 | 1.2347 | 0.9207 | 0.9256 | 0.9281 |
| 1.0663 | 43.0 | 817 | 1.2327 | 0.9138 | 0.9185 | 0.9196 |
| 1.0452 | 44.0 | 836 | 1.2392 | 0.9241 | 0.9299 | 0.9319 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2