Pet Classifier

app.py

@@ -1,7 +1,21 @@
 import gradio as gr
+from fastai.vision.all import *
 
-def greet(name):
-    return "Hello " + name + "!!"
+learn = load_learner('export.pkl')
+labels = learn.dls.vocab
 
-iface = gr.Interface(fn=greet, inputs="text", outputs="text")
-iface.launch()
+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))}
+
+iface = gr.Interface(
+    fn=predict, 
+    inputs=gr.components.Image(shape=(512, 512)), 
+    outputs=gr.components.Label(num_top_classes=3),
+    description="Pet Classifier",
+    article="<p style='text-align: center'><a href='google.com' target='_blank'>Blog post</a></p>",
+    live=True,
+    enable_queue=True,
+    )
+iface.launch()

export.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:420e67bbf65e10fa2c72062d197f2827662866716ea92c08d3d2c3d15c93a9c4
+size 103071997

train.py

@@ -0,0 +1,14 @@
+from fastai.vision.all import *
+
+path = untar_data(URLs.PETS)
+dls = ImageDataLoaders.from_name_re(path, 
+                                    get_image_files(path/'images'), pat='(.+)_\d+.jpg', 
+                                    item_tfms=Resize(460), 
+                                    batch_tfms=aug_transforms(size=224, min_scale=0.75),
+                                    num_workers=0,
+                                    bs=16,
+                                    )
+learn = vision_learner(dls, models.resnet50, metrics=accuracy)
+learn.fine_tune(1)
+learn.path = Path('.')
+learn.export()