File size: 973 Bytes
bdff6f4
 
 
 
 
84b4a40
bdff6f4
 
0d754b8
 
bdff6f4
 
 
 
 
48efa13
84b4a40
bdff6f4
4789dac
bdff6f4
 
84b4a40
bdff6f4
 
c53e284
 
931c04e
f895a0a
b2d7a75
0a7aeac
 
cbf3da8
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
27
28
29
30
31
32
33
# AUTOGENERATED! DO NOT EDIT! File to edit: ../nbs/app.ipynb.

# %% auto 0
__all__ = ['learn', 'categories', 'image', 'label', 'examples', 'intf', 'is_cat', 'classify_image']
from fastai.vision.all import *
import gradio as gr
from pathlib import Path

project_dir = Path(__file__).parent

def is_cat(x) -> bool:
    return x[0].isupper()

learn = load_learner("model.pkl")

categories = ("Dog", "Cat")

def classify_image(img):
    img = PILImage.create(img)
    _, _, probs = learn.predict(img)
    return dict(zip(categories, [float(p) for p in probs]))

image = gr.inputs.Image(shape=(192, 192))
label = gr.outputs.Label()
examples = str((project_dir / "examples").absolute())
print(examples)
intf = gr.Interface(fn=classify_image, inputs=image, outputs=label,
                    title = "Dog/Cat Classifier",
                    description = "A dog/cat classifier.",
                    examples=examples,
                    interpretation="default")
intf.launch()