Fix prediction code

app.py

@@ -4,22 +4,29 @@ __all__ = ['is_cat', 'learn', 'classify_image', 'categories', 'image', 'label',
 from fastai.vision.all import *
 import gradio as gr
 
-def is_cat(x): return x[0].isupper()
-
 # Cell
 learn = load_learner('model.pkl')
 
 # Cell
-categories = ('Toronto, Canada', 'Denver, Colorado, USA', 'Emeryville, California, USA','Lehi, Utah, USA', 'Los Angeles, California, USA', 'College Park, Maryland, USA','Mexico City, Mexico', 'Arden Hills, Minnesota, USA', 'São Paulo, Brazil','New York, New York, USA', 'Portland, Oregon, USA', 'San Francisco, California, USA','Seattle, Washington, USA', 'Shanghai, China', 'Guangzhou, China', 'Hong Kong, China','Noida, India', 'Dublin, Ireland', 'Milan, Italy', 'Rome, Italy', 'Chisinau, Moldova','Amsterdam, Netherlands', 'Austin, Texas, USA', 'McLean, Virginia, USA','Washington D.C., USA', 'Atlanta, Georgia, USA', 'Waltham, Massachusetts, USA','Cambridge, Massachusetts, USA', 'San Jose, California, USA', 'Chicago, Illinois, USA','Ottawa, Canada', 'Bangalore, India', 'Mumbai, India', 'Gurgaon, India','Tokyo, Japan', 'Seoul, South Korea', 'Singapore, Singapore', 'Shenzhen, China','Bangkok, Thailand', 'Taipei, Taiwan', 'Melbourne, Australia', 'Sydney, Australia','Canberra, Australia', 'Yerevan, Armenia', 'Diegem, Belgium', 'Copenhagen, Denmark','Paris, France', 'Munich, Germany', 'Berlin, Germany', 'Hamburg, Germany','Basel, Switzerland', 'Zurich, Switzerland', 'Warsaw, Poland', 'Bucharest, Romania','Johannesburg, South Africa', 'Madrid, Spain', 'Barcelona, Spain', 'Stockholm, Sweden','Maidenhead, United Kingdom', 'London, United Kingdom', 'Edinburgh, Scotland')
+# categories = ('Toronto, Canada', 'Denver, Colorado, USA', 'Emeryville, California, USA','Lehi, Utah, USA', 'Los Angeles, California, USA', 'College Park, Maryland, USA','Mexico City, Mexico', 'Arden Hills, Minnesota, USA', 'São Paulo, Brazil','New York, New York, USA', 'Portland, Oregon, USA', 'San Francisco, California, USA','Seattle, Washington, USA', 'Shanghai, China', 'Guangzhou, China', 'Hong Kong, China','Noida, India', 'Dublin, Ireland', 'Milan, Italy', 'Rome, Italy', 'Chisinau, Moldova','Amsterdam, Netherlands', 'Austin, Texas, USA', 'McLean, Virginia, USA','Washington D.C., USA', 'Atlanta, Georgia, USA', 'Waltham, Massachusetts, USA','Cambridge, Massachusetts, USA', 'San Jose, California, USA', 'Chicago, Illinois, USA','Ottawa, Canada', 'Bangalore, India', 'Mumbai, India', 'Gurgaon, India','Tokyo, Japan', 'Seoul, South Korea', 'Singapore, Singapore', 'Shenzhen, China','Bangkok, Thailand', 'Taipei, Taiwan', 'Melbourne, Australia', 'Sydney, Australia','Canberra, Australia', 'Yerevan, Armenia', 'Diegem, Belgium', 'Copenhagen, Denmark','Paris, France', 'Munich, Germany', 'Berlin, Germany', 'Hamburg, Germany','Basel, Switzerland', 'Zurich, Switzerland', 'Warsaw, Poland', 'Bucharest, Romania','Johannesburg, South Africa', 'Madrid, Spain', 'Barcelona, Spain', 'Stockholm, Sweden','Maidenhead, United Kingdom', 'London, United Kingdom', 'Edinburgh, Scotland')
+
 
-def classify_image(img):
-    pred,idx,probs = learn.predict(img)
-    return dict(zip(categories, map(float,probs)))
+labels = learn.dls.vocab
+def predict(img):
+    img = PILImage.create(img)
+    pred,pred_idx,probs = learn.predict(img)
+    return {labels[i]: float(probs[i]) for i in range(len(labels))}
+
+# def classify_image(img):
+#     pred,idx,probs = learn.predict(img)
+#     return dict(zip(categories, map(float,probs)))
 
 # Cell
-image = gr.inputs.Image(shape=(192, 192))
-label = gr.outputs.Label()
-examples = []
+# image = gr.inputs.Image(shape=(192, 192))
+# label = gr.outputs.Label()
+# examples = []
+
+# intf = gr.Interface(fn=classify_image, inputs=image, outputs=label, examples=examples)
+# intf.launch(inline=False)
 
-intf = gr.Interface(fn=classify_image, inputs=image, outputs=label, examples=examples)
-intf.launch(inline=False)
+gr.Interface(fn=predict, inputs=gr.inputs.Image(shape=(512, 512)), outputs=gr.outputs.Label(num_top_classes=3)).launch(share=True)