add pretrained model

app.py

@@ -1,42 +1,13 @@
-# AUTOGENERATED! DO NOT EDIT! File to edit: cat_classifier.ipynb.
-
-# %% auto 0
-__all__ = ['path', 'files', 'dls', 'learn', 'categories', 'label_func', 'image_classifier']
-
-# %% cat_classifier.ipynb 1
-from fastai.vision.all import ImageDataLoaders, URLs, untar_data, get_image_files, Resize, \
-    vision_learner, resnet34, error_rate, default_device, load_learner, show_image, PILImage
-import torch
-torch.set_default_device('cpu')
-
-# %% cat_classifier.ipynb 3
-path = untar_data(URLs.PETS)
-path.ls()
-
-# %% cat_classifier.ipynb 5
-files = get_image_files(path/'images')
-
-files[0]
-
-# %% cat_classifier.ipynb 6
-def label_func(fn:str): 
-    return fn[0].isupper()
-
-# %% cat_classifier.ipynb 7
-dls = ImageDataLoaders.from_name_func(path/'images', fnames=files, label_func=label_func,
-                                      item_tfms=Resize(224)) 
+from fastai.vision.all import load_learner
+import gradio as gr
 
-# %% cat_classifier.ipynb 9
-learn = vision_learner(dls, resnet34, metrics=error_rate)
-learn.fine_tune(1)
+model = load_learner('./model.pkl')
 
-# %% cat_classifier.ipynb 12
-import gradio as gr
 categories = ('dog', 'cat')
+
 def image_classifier(input): 
-    pred, idx, probs = learn.predict(input) 
+    pred, idx, probs = model.predict(input) 
     result = dict(zip(categories, map(float, probs)))
-    print('result', result)
     return result
 
 
@@ -44,4 +15,4 @@ def image_classifier(input):
 gr.Interface(fn=image_classifier,
              inputs=gr.Image(type="pil"),
              outputs=gr.Label(),
-             examples=['./cat.jpeg', './dog.jpg']).launch(share=True)
+             examples=['./cat.jpeg', './dog.jpg']).launch(share=True, inline=False)

model.pkl

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:8ecdbdf838b356438d69c75f49ae2807f8b8b1ff706229b81e8a76e93ad13112
+size 87549374