Spaces:
Running
Running
testing cuda
Browse files
hfapp.py
CHANGED
@@ -14,6 +14,7 @@ from app import (
|
|
14 |
|
15 |
@spaces.GPU
|
16 |
def run_inference(model, img):
|
|
|
17 |
img = torch.nn.functional.interpolate(img, size=64, mode="bilinear")
|
18 |
score_norms = model.scorenet(img)
|
19 |
score_norms = score_norms.square().sum(dim=(2, 3, 4)) ** 0.5
|
@@ -32,11 +33,7 @@ def localize_anomalies(input_img, preset="edm2-img64-s-fid", load_from_hub=False
|
|
32 |
img = np.array(input_img)
|
33 |
img = torch.from_numpy(img).permute(2, 0, 1).unsqueeze(0)
|
34 |
img = img.float().to(device)
|
35 |
-
|
36 |
-
model = load_model_from_hub(preset=preset, device=device)
|
37 |
-
else:
|
38 |
-
model = load_model(modeldir="models", preset=preset, device=device)
|
39 |
-
|
40 |
img_likelihood, score_norms = run_inference(model, img)
|
41 |
nll, pct, ref_nll = compute_gmm_likelihood(
|
42 |
score_norms, model_dir=f"models/{preset}"
|
@@ -49,7 +46,6 @@ def localize_anomalies(input_img, preset="edm2-img64-s-fid", load_from_hub=False
|
|
49 |
return outstr, heatmapplot, histplot
|
50 |
|
51 |
|
52 |
-
|
53 |
demo = build_demo(localize_anomalies)
|
54 |
if __name__ == "__main__":
|
55 |
demo.launch()
|
|
|
14 |
|
15 |
@spaces.GPU
|
16 |
def run_inference(model, img):
|
17 |
+
print("model on cuda:", next(model.scorenet.net.parameters()).is_cuda)
|
18 |
img = torch.nn.functional.interpolate(img, size=64, mode="bilinear")
|
19 |
score_norms = model.scorenet(img)
|
20 |
score_norms = score_norms.square().sum(dim=(2, 3, 4)) ** 0.5
|
|
|
33 |
img = np.array(input_img)
|
34 |
img = torch.from_numpy(img).permute(2, 0, 1).unsqueeze(0)
|
35 |
img = img.float().to(device)
|
36 |
+
model = load_model_from_hub(preset=preset, device=device)
|
|
|
|
|
|
|
|
|
37 |
img_likelihood, score_norms = run_inference(model, img)
|
38 |
nll, pct, ref_nll = compute_gmm_likelihood(
|
39 |
score_norms, model_dir=f"models/{preset}"
|
|
|
46 |
return outstr, heatmapplot, histplot
|
47 |
|
48 |
|
|
|
49 |
demo = build_demo(localize_anomalies)
|
50 |
if __name__ == "__main__":
|
51 |
demo.launch()
|