Downloads · 30 days
5
5% of all-time downloads
hiera-finetuned-stroke-all-pat
BTX24/hiera-finetuned-stroke-all-pat
hiera-finetuned-stroke-all-pat is a image classification model from BTX24. Use it when you need a label for an image. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
111
Public
Parameters
50.8M
7.7 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors203 MB · 100%
At a glance
- Task
- Image Classification
- Library
- transformers
- Model type
- hiera
- Access
- Public
- Created
- Jul 9, 2025
- Updated
- Jul 9, 2025
- SHA
- 39fac09b
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
hiera-finetuned-stroke-all-pat
This model is a fine-tuned version of BTX24/hiera-finetuned-busi on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.3355
- Accuracy: 0.8798
- F1: 0.8807
- Precision: 0.8830
- Recall: 0.8798
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 64
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 32
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 0.5615 | 1.6129 | 100 | 0.4912 | 0.7664 | 0.7472 | 0.7879 | 0.7664 |
| 0.4755 | 3.2258 | 200 | 0.4393 | 0.8186 | 0.8181 | 0.8226 | 0.8186 |
| 0.3978 | 4.8387 | 300 | 0.3931 | 0.8345 | 0.8344 | 0.8356 | 0.8345 |
| 0.364 | 6.4516 | 400 | 0.3698 | 0.8413 | 0.8392 | 0.8395 | 0.8413 |
| 0.338 | 8.0645 | 500 | 0.4049 | 0.8345 | 0.8358 | 0.8421 | 0.8345 |
| 0.3049 | 9.6774 | 600 | 0.4135 | 0.8299 | 0.8268 | 0.8385 | 0.8299 |
| 0.2644 | 11.2903 | 700 | 0.3848 | 0.8413 | 0.8423 | 0.8499 | 0.8413 |
| 0.234 | 12.9032 | 800 | 0.3597 | 0.8685 | 0.8703 | 0.8770 | 0.8685 |
| 0.1951 | 14.5161 | 900 | 0.3355 | 0.8798 | 0.8807 | 0.8830 | 0.8798 |
| 0.1935 | 16.1290 | 1000 | 0.4081 | 0.8390 | 0.8410 | 0.8503 | 0.8390 |
| 0.1604 | 17.7419 | 1100 | 0.3828 | 0.8730 | 0.8737 | 0.8753 | 0.8730 |
| 0.144 | 19.3548 | 1200 | 0.3970 | 0.8571 | 0.8578 | 0.8593 | 0.8571 |
| 0.1257 | 20.9677 | 1300 | 0.3712 | 0.8730 | 0.8723 | 0.8729 | 0.8730 |
| 0.1036 | 22.5806 | 1400 | 0.3932 | 0.8685 | 0.8688 | 0.8703 | 0.8685 |
| 0.1094 | 24.1935 | 1500 | 0.3800 | 0.8798 | 0.8790 | 0.8805 | 0.8798 |
| 0.0836 | 25.8065 | 1600 | 0.3981 | 0.8662 | 0.8674 | 0.8706 | 0.8662 |
| 0.0803 | 27.4194 | 1700 | 0.4010 | 0.8707 | 0.8724 | 0.8779 | 0.8707 |
| 0.0789 | 29.0323 | 1800 | 0.3938 | 0.8753 | 0.8762 | 0.8786 | 0.8753 |
| 0.0872 | 30.6452 | 1900 | 0.4085 | 0.8662 | 0.8675 | 0.8713 | 0.8662 |
Framework versions
- Transformers 4.53.1
- Pytorch 2.7.1+cu126
- Datasets 4.0.0
- Tokenizers 0.21.2
- Task
- Image Classification
- Library
- transformers
- Type
- hiera
- Created
- Jul 9, 2025
- Updated
- Jul 9, 2025