IRSANet_IDC v0.1.0 Pretrained Weights (Seed1)
Model Details
| Field |
Value |
| Model |
IRSANet_IDC |
| Parameters |
347,775 |
| Modules |
124 |
Weight Details
| Field |
Value |
| Experiment |
Validation |
| Run |
Run1 |
| Seed |
1 |
| Fold |
5 |
| Epoch |
67 |
| Checkpoint |
bestAUROC |
Evaluation Details
| Field |
Value |
| Dataset |
IDC |
| Test samples |
42,637 |
| Evaluated |
2026-06-03 06:54 |
| TTA |
False |
Key Metrics (Threshold = 0.461)
| Metric |
Test |
Val |
| AUC (ROC) |
0.959 |
0.946 |
| Average Precision |
0.910 |
0.863 |
| F1 Score |
0.833 |
— |
| Balanced Accuracy |
0.897 |
— |
| Recall (Sensitivity) |
0.924 |
— |
Usage
Reconstruction
from irsanet import load_checkpoint
model, ckpt = load_checkpoint("IRSANet_IDC_seed1_fold5_bestAUROC.pt")Evaluation
from irsanet import evaluate_model
from irsanet_idc import build_datasets
_,_,test_loader,_,_ = build_datasets([],
cross_val = True,
eval_mode = True,
seed = 1,
)
results = evaluate_model(
weights = "IRSANet_IDC_seed1_fold5_bestAUROC.pt",
test_loader = test_loader,
)
print(results["test_results"])Including a full visual dashboard:
from irsanet.plotting import plot_eval_panel
from irsanet_idc import build_datasets
_,_,test_loader,_,_ = build_datasets([],
cross_val = True,
eval_mode = True,
seed = 1,
)
results = plot_eval_panel(
weights = "IRSANet_IDC_seed1_fold5_bestAUROC.pt",
test_loader = test_loader,
)
print(results["test_results"])