| from fastai.learner import load_learner | |
| import gradio as gr | |
| learn = load_learner('modelcat_dog.pkl') | |
| categories = ('Cat','Dog','Lion','None','Tiger','Wolf') | |
| def classify_img(image): | |
| pred,idx,probs = learn.predict(image) | |
| return (dict(zip(categories,map(float,probs)))) | |
| image = gr.inputs.Image(shape=(192,192)) | |
| label = gr.outputs.Label() | |
| examples = ['cat.jpg','food.jpg','husky.jpg','tiger.jpg','cat_dog.jpg'] | |
| intf = gr.Interface(fn=classify_img, inputs=image, outputs=label, examples=examples) | |
| intf.launch(inline=False) |