init

app.py

@@ -0,0 +1,43 @@
+# AUTOGENERATED! DO NOT EDIT! File to edit: app.ipynb.
+
+# %% auto 0
+__all__ = ['labels', 'title', 'description', 'article', 'image', 'label', 'examples', 'interpretation', 'enable_queue', 'intf',
+           'classify_image']
+
+# %% app.ipynb 1
+import nbdev
+from fastai.vision.all import *
+import scipy
+import gradio as gr
+
+# %% app.ipynb 3
+labels = learner.dls.vocab
+
+def classify_image(img):
+    """Gradio need a f-n that returns a dict of each class and its probability.
+    It also does not accept tensors.
+    """
+    img = PILImage.create(img)
+    pred, pred_idx, probs = learner.predict(img)
+    return dict(zip(labels, map(float, probs)))
+
+# %% app.ipynb 4
+title = "Cat & Dog Breed Classifier"
+description = "A pet breed classifier trained on the Oxford Pets dataset with fastai. Created as a demo for Gradio and HuggingFace Spaces."
+article="<p style='text-align: center'><a href='https://tmabraham.github.io/blog/gradio_hf_spaces_tutorial' target='_blank'>Blog post</a></p>"
+
+image = gr.components.Image(shape=(512,512))
+label = gr.components.Label(num_top_classes=3)
+examples = ['shiba.jpeg', 'yorkshire_terrier.jpeg']
+interpretation = 'default'
+enable_queue=True
+
+intf = gr.Interface(classify_image,
+                    inputs = image,
+                    outputs=label,
+                    examples=examples,
+                    title=title,
+                    description=description,
+                    interpretation=interpretation)
+
+intf.launch(enable_queue=enable_queue)

requirements.txt

@@ -0,0 +1,2 @@
+fastai
+scikit-image