Spaces:
Runtime error
Runtime error
Update app.py
Browse files
app.py
CHANGED
@@ -12,46 +12,52 @@ from imagebind.models.imagebind_model import ModalityType
|
|
12 |
import torch.nn as nn
|
13 |
|
14 |
|
15 |
-
device = "cuda:0" if torch.cuda.is_available() else "cpu"
|
16 |
model = imagebind_model.imagebind_huge(pretrained=True)
|
17 |
model.eval()
|
18 |
model.to(device)
|
19 |
|
20 |
-
def image_text_zeroshot(image, text_list):
|
21 |
-
|
22 |
-
|
23 |
-
|
24 |
-
|
25 |
-
|
26 |
-
|
27 |
-
|
28 |
-
|
29 |
-
|
30 |
-
|
31 |
-
|
32 |
-
|
33 |
-
|
34 |
-
|
35 |
-
|
36 |
-
|
37 |
-
|
38 |
-
|
39 |
-
|
40 |
-
|
41 |
-
|
42 |
-
|
43 |
-
def main():
|
44 |
-
|
45 |
-
|
46 |
-
|
47 |
-
|
48 |
-
|
49 |
-
|
50 |
-
|
51 |
-
|
52 |
-
|
53 |
-
|
54 |
-
|
55 |
-
|
56 |
-
|
57 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
12 |
import torch.nn as nn
|
13 |
|
14 |
|
15 |
+
device = "cpu" #"cuda:0" if torch.cuda.is_available() else "cpu"
|
16 |
model = imagebind_model.imagebind_huge(pretrained=True)
|
17 |
model.eval()
|
18 |
model.to(device)
|
19 |
|
20 |
+
# def image_text_zeroshot(image, text_list):
|
21 |
+
# image_paths = [image]
|
22 |
+
# labels = [label.strip(" ") for label in text_list.strip(" ").split("|")]
|
23 |
+
# inputs = {
|
24 |
+
# ModalityType.TEXT: data.load_and_transform_text(labels, device),
|
25 |
+
# ModalityType.VISION: data.load_and_transform_vision_data(image_paths, device),
|
26 |
+
# }
|
27 |
+
|
28 |
+
# with torch.no_grad():
|
29 |
+
# embeddings = model(inputs)
|
30 |
+
|
31 |
+
# scores = (
|
32 |
+
# torch.softmax(
|
33 |
+
# embeddings[ModalityType.VISION] @ embeddings[ModalityType.TEXT].T, dim=-1
|
34 |
+
# )
|
35 |
+
# .squeeze(0)
|
36 |
+
# .tolist()
|
37 |
+
# )
|
38 |
+
|
39 |
+
# score_dict = {label: score for label, score in zip(labels, scores)}
|
40 |
+
|
41 |
+
# return score_dict
|
42 |
+
|
43 |
+
# def main():
|
44 |
+
# inputs = [
|
45 |
+
# gr.inputs.Textbox(lines=1, label="texts"),
|
46 |
+
# gr.inputs.Image(type="filepath", label="Output image")
|
47 |
+
# ]
|
48 |
+
|
49 |
+
# iface = gr.Interface(
|
50 |
+
# image_text_zeroshot(image, text_list),
|
51 |
+
# inputs,
|
52 |
+
# "label",
|
53 |
+
# description="""...""",
|
54 |
+
# title="ImageBind",
|
55 |
+
# )
|
56 |
+
|
57 |
+
# iface.launch()
|
58 |
+
|
59 |
+
def image_classifier(inp):
|
60 |
+
return {'cat': 0.3, 'dog': 0.7}
|
61 |
+
|
62 |
+
demo = gr.Interface(fn=image_classifier, inputs="image", outputs="label")
|
63 |
+
demo.launch()
|