Image Classification
PyTorch
timm
multiple-instance-learning
attention
interpretability
digital-humanities
historical-documents
korean
Instructions to use aronvandepol/visual-feature-discovery with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- timm
How to use aronvandepol/visual-feature-discovery with timm:
import timm model = timm.create_model("hf-hub:aronvandepol/visual-feature-discovery", pretrained=True) - Notebooks
- Google Colab
- Kaggle
File size: 716 Bytes
3a09b90 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 | {
"model_name": "convnextv2_base",
"fold": 2,
"n_folds": 5,
"split_info": {
"fold": 2,
"train_samples": 44957,
"val_samples": 8063,
"test_samples": 13254,
"train_books": 347,
"val_books": 61,
"test_books": 102,
"train_val_overlap": 0,
"train_test_overlap": 0,
"val_test_overlap": 0
},
"test_results": {
"test_loss_epoch": 0.5984542965888977,
"test_accuracy": 0.8428398966789246,
"test_precision": 0.83812016248703,
"test_recall": 0.8699315190315247,
"test_f1": 0.8505106568336487
},
"model_path": "models_book_split/convnextv2_base_fold2_20260203_1901.pth",
"checkpoint_dir": "checkpoints_book_split/convnextv2_base_fold2_20260203_1901"
} |