File size: 576 Bytes
62ead1b
 
60c56bd
36f6e50
593ed83
62ead1b
 
60c56bd
 
36f6e50
2b49a10
444179e
593ed83
 
 
 
bf8d52c
444179e
 
 
 
ea6d7fc
444179e
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
import pathlib
import platform
if platform.system()=='Linux': pathlib.WindowsPath=pathlib.PosixPath

import dill
import gradio as gr

with open('model_dill.pkl','rb') as f: learn=dill.load(f)

labels=['Dog','Cat']

def predict(img):
    from fastai.vision.all import PILImage
    img=PILImage.create(img)
    pred,pred_idx,probs=learn.predict(img)
    return {labels[i]:float(probs[i]) for i in range(len(labels))}

gr.Interface(
    fn=predict,
    inputs=gr.Image(type="pil"),
    outputs=gr.Label(num_top_classes=2),
    examples=['dog.jpg','cat.jpg','dcat.jpg']
).launch()