Update app.py

app.py

@@ -1,41 +1,63 @@
 # app.py
-import gradio as gr
+import os
 import json
+import gradio as gr
 from transformers import pipeline
+from huggingface_hub import login
+
+# ============= 🔐 AUTHENTICATION =============
+# Hugging Face token (set in your Space settings)
+# Go to your Space → Settings → Repository secrets → add key: HF_TOKEN, value: your access token
+hf_token = os.getenv("HF_TOKEN")
+
+if hf_token:
+    login(token=hf_token)
+else:
+    raise EnvironmentError("❌ Missing HF_TOKEN environment variable. Please add it in your Space settings.")
 
-# Load labels from external JSON
+# ============= 📦 LOAD LABELS =============
 with open("labels.json", "r", encoding="utf-8") as f:
     candidate_labels = json.load(f)
 
-# Initialize the MedSigLIP pipeline
-pipe = pipeline("zero-shot-image-classification", model="google/medsiglip-448")
+# ============= 🧠 MODEL PIPELINE =============
+# Use the authenticated pipeline for MedSigLIP (gated model)
+pipe = pipeline(
+    "zero-shot-image-classification",
+    model="google/medsiglip-448",
+    use_auth_token=hf_token
+)
 
-# Define prediction function
-def classify_image(image):
-    result = pipe(image, candidate_labels=candidate_labels)
-    # Convert result list to {label: score}
-    formatted = {item["label"]: round(item["score"] * 100, 2) for item in result}
-    # Sort descending by confidence
-    sorted_result = dict(sorted(formatted.items(), key=lambda x: x[1], reverse=True))
-    return sorted_result
+# ============= ⚙️ PREDICTION FUNCTION =============
+def classify_medical_image(image):
+    """
+    Run zero-shot classification using MedSigLIP.
+    """
+    try:
+        results = pipe(image, candidate_labels=candidate_labels)
+        formatted = {r["label"]: round(r["score"] * 100, 2) for r in results}
+        sorted_result = dict(sorted(formatted.items(), key=lambda x: x[1], reverse=True))
+        return sorted_result
+    except Exception as e:
+        return {"Error": str(e)}
 
-# Build Gradio interface
+# ============= 🎨 GRADIO UI =============
 demo = gr.Interface(
-    fn=classify_image,
-    inputs=gr.Image(type="filepath", label="Upload Chest X-ray or Medical Image"),
-    outputs=gr.Label(num_top_classes=5, label="Top Predictions"),
-    title="🩻 MedSigLIP Zero-Shot Medical Image Classifier",
+    fn=classify_medical_image,
+    inputs=gr.Image(type="filepath", label="📤 Upload Medical Image (X-ray, MRI, Pathology, etc.)"),
+    outputs=gr.Label(num_top_classes=5, label="🧠 Top Predictions"),
+    title="🩻 MedSigLIP Zero-Shot Medical Classifier",
     description=(
-        "This demo uses Google's MedSigLIP (Sigmoid Loss for Language Image Pre-training) model "
-        "for zero-shot medical image classification. Upload an image and the model will estimate "
-        "its similarity with known medical conditions (loaded dynamically from labels.json)."
+        "This demo uses Google's **MedSigLIP (448x448)** model for zero-shot medical image understanding. "
+        "Upload a medical image (e.g., chest X-ray, skin lesion, fundus photo, histopathology slide) and the model "
+        "will compute similarity against known medical findings loaded from `labels.json`."
     ),
-    allow_flagging="never",
     examples=[
         ["https://storage.googleapis.com/dx-scin-public-data/dataset/images/3445096909671059178.png"],
         ["https://storage.googleapis.com/dx-scin-public-data/dataset/images/-5669089898008966381.png"]
     ],
+    allow_flagging="never",
 )
 
+# ============= 🚀 RUN APP =============
 if __name__ == "__main__":
-    demo.launch()
+    demo.launch(server_name="0.0.0.0", server_port=7860)