Update app.py

app.py

@@ -39,26 +39,18 @@ def analyze(image: Image.Image, prompt: str = "What's happening with my skin?"):
             ],
         }
     ]
-
     text = processor.apply_chat_template(
         messages, tokenize=False, add_generation_prompt=True
     )
     image_inputs, _ = process_vision_info(messages)
-
     inputs = processor(
-        text=[text],
-        images=image_inputs,
-        return_tensors="pt"
+        text=[text], images=image_inputs, return_tensors="pt"
     ).to("cuda")
 
     with torch.inference_mode():
         generated_ids = model.generate(
-            **inputs,
-            max_new_tokens=256,
-            do_sample=True,
-            temperature=0.2
+            **inputs, max_new_tokens=256, do_sample=True, temperature=0.2
         )
-
     output_ids = generated_ids[0, inputs.input_ids.shape[1]:]
     return processor.decode(output_ids, skip_special_tokens=True)
 
@@ -79,7 +71,7 @@ async def analyze_endpoint(
     result = analyze(pil_image, prompt)
     return JSONResponse({"analysis": result})
 
-# ── Gradio UI ─────────────────────────────────────────────────────────────────
+# ── Gradio mounted ONCE onto FastAPI ─────────────────────────────────────────
 demo = gr.Interface(
     fn=analyze,
     inputs=[
@@ -91,5 +83,4 @@ demo = gr.Interface(
     description="Upload a skin image to get an AI-powered dermatological analysis."
 )
 
-# Mount Gradio onto FastAPI — single process, model loads once
-app = gr.mount_gradio_app(fapi, demo, path="/")
+gr.mount_gradio_app(fapi, demo, path="/ui")