vit-base-patch16-224-finetuned-face-match-4360
This model is a fine-tuned version of google/vit-base-patch16-224 on the imagefolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.3192
- Accuracy: 0.8624
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- 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 |
---|---|---|---|---|
0.325 | 1.0 | 109 | 0.2983 | 0.8498 |
0.3048 | 2.0 | 218 | 0.3009 | 0.8521 |
0.2623 | 3.0 | 327 | 0.3098 | 0.8532 |
0.2224 | 4.0 | 436 | 0.3225 | 0.8440 |
0.2224 | 5.0 | 545 | 0.2950 | 0.8693 |
0.265 | 6.0 | 654 | 0.3331 | 0.8509 |
0.1922 | 7.0 | 763 | 0.3079 | 0.8555 |
0.2083 | 8.0 | 872 | 0.3580 | 0.8612 |
0.1871 | 9.0 | 981 | 0.3510 | 0.8578 |
0.1784 | 10.0 | 1090 | 0.3538 | 0.8452 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
- Downloads last month
- 17