File size: 967 Bytes
a1c5266
df1982d
a1c5266
 
df1982d
a1c5266
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
34
35
36
# AUTOGENERATED! DO NOT EDIT! File to edit: ../app.ipynb.

# %% auto 0
__all__ = ['learn', 'categories', 'image', 'label', 'examples', 'intf', 'classify_image']

# %% ../app.ipynb 1
import sys
import subprocess

subprocess.check_call([sys.executable, '-m', 'pip', 'install', 'gradio==3.50'])
subprocess.check_call([sys.executable, '-m', 'pip', 'install', 'fastai'])
subprocess.check_call([sys.executable, '-m', 'pip', 'install', 'ipywidgets'])

from fastai.vision.all import *
import gradio as gr 
import timm

# %% ../app.ipynb 3
learn = load_learner('model.pkl')

# %% ../app.ipynb 5
categories = learn.dls.vocab

def classify_image(img):
    pred, idx, probs = learn.predict(img)
    return dict(zip(categories, map(fload, probs)))

# %% ../app.ipynb 7
image = gr.inputs.Image(shape=(192,192))
label = gr.outputs.Label()
examples = ['basset.jpg']

# %% ../app.ipynb 8
intf = gr.Interface(fn=classify_image, inputs = image, outputs=label)
intf.launch(inline=False)