cleaning up

app.py

@@ -1,80 +1,43 @@
 import gradio as gr
-from fastai.vision.all import *
-import matplotlib.pyplot as plt
-import numpy as np
-from transformers import pipeline 
-from huggingface_hub import from_pretrained_fastai
-
-pipe = pipeline(
-    task="image-classification",
-    model ="ethanmac/dr-maclboomber-retina-classifier",
-    description="Retinal Condition Classifier",
-    examples=['sick-eye.jpeg', 'healthy.jpg']
-    ).launch()
-
-
-# from huggingface_hub import from_pretrained_fastai
-
-# # repo_id = "YOUR_USERNAME/YOUR_LEARNER_NAME"
-# repo_id = "espejelomar/identify-my-cat"
-
-# learner = from_pretrained_fastai(repo_id)
-
-
-
-
-
-
-# # Load your trained FastAI model
-# learn = load_learner('export.pkl') 
-
-# def predict_and_plot(img):
-#     # Get predictions
-#     pred, pred_idx, probs = learn.predict(img)
-#     # Get class names
-#     class_names = ['Normal',
-#                    'Hollenhorst Emboli',
-#                    'Hypertensive Retinopathy',
-#                    'Coat\'s',
-#                    'Macroaneurism',
-#                    'Choroidal Neovascularization',
-#                    'Other',
-#                    'Branch Retinal Artery Occlusion',
-#                    'Cilio-Retinal Artery Occlusion',
-#                    'Branch Retinal Vein Occlusion',
-#                    'Central Retinal Vein Occlusion',
-#                    'Hemi-Central Retinal Vein Occlusion',
-#                    'Background Diabetic Retinopathy',
-#                    'Proliferative Diabetic Retinopathy',
-#                    'Arteriosclerotic Retinopathy']
-#     probs_np = probs.numpy()
-
-#     threshold = 0.5
-
-#     present = probs_np > threshold
-#     fig, ax = plt.subplots(figsize=(10, 6))
-#     y_pos = np.arange(len(class_names))
-
-#     colors = ['green' if is_present else 'red' for is_present in present]
-#     ax.barh(y_pos, probs_np, color=colors)
-#     ax.set_yticks(y_pos)
-#     ax.set_yticklabels(class_names)
-#     ax.invert_yaxis()
-#     ax.set_xlabel('Probability')
-#     ax.set_xlim(0, 1)
-#     ax.set_title('Predicted Probabilities for Each Condition')
-#     plt.tight_layout()
-#     return fig
-
-# # Create the Gradio interface
-# interface = gr.Interface(
-#     fn=predict_and_plot,
-#     inputs=gr.Image(type='pil'),
-#     outputs=gr.Plot(),
-#     title="Dr. Macloomber",
-#     description="Upload an image of a retina to predict the probabilities of various eye conditions."
-# )
-
-# # Launch the app
-# interface.launch()
-
+from huggingface_hub import hf_hub_download
+from fastai.learner import load_learner
+
+repo_id = "ethanmac/dr-macbloomber-retina-condition-classifier"
+
+model = load_learner(
+    hf_hub_download(repo_id, "model.pkl")
+)
+
+class_names = [
+'Normal',
+'Hollenhorst Emboli',
+'Hypertensive Retinopathy',
+'Coat\'s',
+'Macroaneurism',
+'Choroidal Neovascularization',
+'Other',
+'Branch Retinal Artery Occlusion',
+'Cilio-Retinal Artery Occlusion',
+'Branch Retinal Vein Occlusion',
+'Central Retinal Vein Occlusion',
+'Hemi-Central Retinal Vein Occlusion',
+'Background Diabetic Retinopathy',
+'Proliferative Diabetic Retinopathy',
+'Arteriosclerotic Retinopathy'
+]
+
+categories = [c.replace('_', ' ').title() for c in class_names]
+
+def classify_image(img):
+    pred, idx, probs = model.predict(img)
+    out = dict(zip(categories, map(float, probs)))
+    return out
+
+intf = gr.Interface(
+    fn=classify_image,
+    inputs=gr.Image(),
+    outputs=gr.Label(num_top_classes=5),
+    examples=['healthy.jpg', 'macroaneurism.jpg']
+)
+
+intf.launch(inline=False)

export.pkl

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:77458acaeca030340762d1abaeb82968614bac52b777813e37b4b0576e61d7b0
-size 47019902