Image Classification
timm
ONNX
PyTorch
English
medical
histopathology
cancer-classification
oncotree
resnet50
int8
Eval Results (legacy)
Instructions to use AegisOSS/stage-2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- timm
How to use AegisOSS/stage-2 with timm:
import timm model = timm.create_model("hf_hub:AegisOSS/stage-2", pretrained=True) - Notebooks
- Google Colab
- Kaggle
| { | |
| "best": { | |
| "accuracy": 0.8615384615384616, | |
| "balanced_accuracy": 0.8831168831168832, | |
| "macro_f1": 0.8830379746835443, | |
| "weighted_f1": 0.8614449853943525, | |
| "train_loss": 0.10038254000460797 | |
| }, | |
| "history": [ | |
| { | |
| "epoch": 1, | |
| "accuracy": 0.6192307692307693, | |
| "balanced_accuracy": 0.6270423125261835, | |
| "macro_f1": 0.6320931302169766, | |
| "weighted_f1": 0.604872716650196, | |
| "train_loss": 1.204805181667074 | |
| }, | |
| { | |
| "epoch": 2, | |
| "accuracy": 0.7961538461538461, | |
| "balanced_accuracy": 0.827922077922078, | |
| "macro_f1": 0.8081030062584729, | |
| "weighted_f1": 0.7911598121935696, | |
| "train_loss": 0.4501433116839181 | |
| }, | |
| { | |
| "epoch": 3, | |
| "accuracy": 0.8384615384615385, | |
| "balanced_accuracy": 0.8612012987012987, | |
| "macro_f1": 0.8589727645815197, | |
| "weighted_f1": 0.835821582658108, | |
| "train_loss": 0.19666762697765253 | |
| }, | |
| { | |
| "epoch": 4, | |
| "accuracy": 0.8538461538461538, | |
| "balanced_accuracy": 0.8766233766233766, | |
| "macro_f1": 0.874506776462515, | |
| "weighted_f1": 0.8513387967325178, | |
| "train_loss": 0.14478116510045655 | |
| }, | |
| { | |
| "epoch": 5, | |
| "accuracy": 0.85, | |
| "balanced_accuracy": 0.8733766233766234, | |
| "macro_f1": 0.8733285527481335, | |
| "weighted_f1": 0.8499430547939428, | |
| "train_loss": 0.11779297081387817 | |
| }, | |
| { | |
| "epoch": 6, | |
| "accuracy": 0.8576923076923076, | |
| "balanced_accuracy": 0.8798701298701299, | |
| "macro_f1": 0.8778423522321237, | |
| "weighted_f1": 0.8567162730936462, | |
| "train_loss": 0.1143842127446572 | |
| }, | |
| { | |
| "epoch": 7, | |
| "accuracy": 0.8538461538461538, | |
| "balanced_accuracy": 0.8766233766233766, | |
| "macro_f1": 0.8765400843881856, | |
| "weighted_f1": 0.8537474845829277, | |
| "train_loss": 0.10838874956576082 | |
| }, | |
| { | |
| "epoch": 8, | |
| "accuracy": 0.8615384615384616, | |
| "balanced_accuracy": 0.8831168831168832, | |
| "macro_f1": 0.8830379746835443, | |
| "weighted_f1": 0.8614449853943525, | |
| "train_loss": 0.10038254000460797 | |
| }, | |
| { | |
| "epoch": 9, | |
| "accuracy": 0.8307692307692308, | |
| "balanced_accuracy": 0.8571428571428572, | |
| "macro_f1": 0.8567562996786742, | |
| "weighted_f1": 0.8303113088501216, | |
| "train_loss": 0.09510761173709847 | |
| }, | |
| { | |
| "epoch": 10, | |
| "accuracy": 0.8615384615384616, | |
| "balanced_accuracy": 0.8831168831168832, | |
| "macro_f1": 0.8811241512666561, | |
| "weighted_f1": 0.8606039427191693, | |
| "train_loss": 0.11316575840935907 | |
| } | |
| ], | |
| "data_summary": { | |
| "total": 1743, | |
| "per_subtype": { | |
| "LUAD": 517, | |
| "LUSC": 517, | |
| "CRC": 294, | |
| "BRCA": 415 | |
| }, | |
| "per_dataset": { | |
| "lc25000": 900, | |
| "lung_cancer_dorsar": 428, | |
| "breast_histopathology_patches": 415 | |
| }, | |
| "num_classes_present": 4, | |
| "trainable_oncotree": [ | |
| "BRCA", | |
| "CRC", | |
| "LUAD", | |
| "LUSC" | |
| ] | |
| } | |
| } |