<!-- 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. -->
isl-model
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.9812
- Accuracy: 0.7533
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: 16
- 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
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 2.6205 | 1.0 | 89 | 2.4275 | 0.5938 |
| 1.6582 | 2.0 | 178 | 1.6078 | 0.7063 |
| 1.3648 | 3.0 | 267 | 1.3754 | 0.7168 |
| 1.1069 | 4.0 | 356 | 1.2056 | 0.7323 |
| 1.1562 | 5.0 | 445 | 1.0958 | 0.7491 |
| 1.1048 | 6.0 | 534 | 1.0221 | 0.7583 |
| 0.9705 | 7.0 | 623 | 0.9831 | 0.7625 |
| 0.9059 | 8.0 | 712 | 1.0279 | 0.7386 |
| 0.9426 | 9.0 | 801 | 0.9511 | 0.7632 |
| 0.8951 | 10.0 | 890 | 0.9812 | 0.7533 |
Framework versions
- Transformers 4.28.1
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.3
