File size: 883 Bytes
0c055f3 40c38d1 3cad0c7 7ef6ec1 4a04f4f 0ce8c49 7102374 7ef6ec1 4a04f4f 7ef6ec1 2b52a97 0c055f3 0ce8c49 29196b3 0ce8c49 |
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 |
import gradio as gr
from fastbook import load_learner
from fastai.vision.all import *
from PIL import Image
js_func = """
function refresh() {
const url = new URL(window.location);
if (url.searchParams.get('__theme') !== 'dark') {
url.searchParams.set('__theme', 'dark');
window.location.href = url.href;
}
}
"""
def predict(im):
resized_im = im['composite'].resize((28,28))
pred,idx,probs = model.predict(resized_im)
return dict(zip(categories, map(float,probs)))
categories = ('0', '1', '2', '3', '4', '5', '6', '7', '8', '9')
model = load_learner('./dr_model.pkl')
labels = model.dls.vocab
with gr.Blocks(js=js_func) as demo:
demo = gr.Interface(
fn=predict,
inputs=gr.Sketchpad(image_mode='L', brush=gr.Brush(default_color="FFFFFFFF"), type='pil'),
outputs = "label",
theme=gr.themes.Monochrome())
demo.launch() |