vit-large-patch16-384-limb-person-crop
This model is a fine-tuned version of google/vit-large-patch16-384 on the c14kevincardenas/beta_caller_284_person_crop dataset. It achieves the following results on the evaluation set:
- Loss: 0.7555
- Accuracy: 0.6816
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: 32
- eval_batch_size: 32
- seed: 2014
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 10.0
- mixed_precision_training: Native AMP
- label_smoothing_factor: 0.005
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy |
---|---|---|---|---|
1.3935 | 1.0 | 214 | 1.4174 | 0.2935 |
1.2459 | 2.0 | 428 | 1.2275 | 0.4353 |
1.0906 | 3.0 | 642 | 0.9373 | 0.6028 |
1.0267 | 4.0 | 856 | 0.8975 | 0.6103 |
0.9659 | 5.0 | 1070 | 0.9388 | 0.6128 |
0.84 | 6.0 | 1284 | 0.7555 | 0.6816 |
0.8287 | 7.0 | 1498 | 0.7678 | 0.6957 |
0.6833 | 8.0 | 1712 | 0.7987 | 0.7048 |
0.6592 | 9.0 | 1926 | 0.8045 | 0.7156 |
0.6604 | 10.0 | 2140 | 0.8059 | 0.7197 |
Framework versions
- Transformers 4.41.0.dev0
- Pytorch 2.0.1+cu117
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 1