<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
swin-food101-jpqd-1to2r1.5-epo10-finetuned-student
This model is a fine-tuned version of skylord/swin-finetuned-food101 on the food101 dataset. It achieves the following results on the evaluation set:
- Loss: 0.2391
- Accuracy: 0.9184
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: 16
- eval_batch_size: 128
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10.0
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.3011 | 0.42 | 500 | 0.1951 | 0.9124 |
| 0.2613 | 0.84 | 1000 | 0.1897 | 0.9139 |
| 100.1552 | 1.27 | 1500 | 99.5975 | 0.7445 |
| 162.0751 | 1.69 | 2000 | 162.5020 | 0.3512 |
| 1.061 | 2.11 | 2500 | 0.7523 | 0.8550 |
| 0.9728 | 2.54 | 3000 | 0.5263 | 0.8767 |
| 0.5851 | 2.96 | 3500 | 0.4599 | 0.8892 |
| 0.4668 | 3.38 | 4000 | 0.4064 | 0.8938 |
| 0.6967 | 3.8 | 4500 | 0.3814 | 0.8986 |
| 0.4928 | 4.23 | 5000 | 0.3522 | 0.9036 |
| 0.4893 | 4.65 | 5500 | 0.3562 | 0.9026 |
| 0.5421 | 5.07 | 6000 | 0.3182 | 0.9049 |
| 0.4405 | 5.49 | 6500 | 0.3112 | 0.9071 |
| 0.4423 | 5.92 | 7000 | 0.3012 | 0.9092 |
| 0.4143 | 6.34 | 7500 | 0.2958 | 0.9095 |
| 0.4997 | 6.76 | 8000 | 0.2796 | 0.9126 |
| 0.2448 | 7.19 | 8500 | 0.2747 | 0.9124 |
| 0.4468 | 7.61 | 9000 | 0.2699 | 0.9144 |
| 0.4163 | 8.03 | 9500 | 0.2583 | 0.9166 |
| 0.3651 | 8.45 | 10000 | 0.2567 | 0.9165 |
| 0.3946 | 8.88 | 10500 | 0.2489 | 0.9176 |
| 0.3196 | 9.3 | 11000 | 0.2444 | 0.9180 |
| 0.312 | 9.72 | 11500 | 0.2402 | 0.9172 |
Framework versions
- Transformers 4.26.0
- Pytorch 1.13.1+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2
