Upload app.py with huggingface_hub

app.py

@@ -5,6 +5,7 @@ from fastapi.responses import FileResponse
 from fastapi.staticfiles import StaticFiles
 import os
 import re
+import gradio as gr
 
 try:
     from llama_cpp import Llama
@@ -174,6 +175,18 @@ async def get_frontend_file(filename: str):
         return FileResponse(file_path)
     raise HTTPException(status_code=404, detail="File not found")
 
-if __name__ == "__main__":
-    import uvicorn
-    uvicorn.run(app, host="0.0.0.0", port=7860)
+    raise HTTPException(status_code=404, detail="File not found")
+
+# ---------------------------------------------------------
+# HF ZeroGPU Gradio Wrapper (The Final Hack)
+# ---------------------------------------------------------
+# ZeroGPU REQUIRES a Gradio app to exist, so we mount a dummy one.
+@spaces.GPU(duration=120)
+def dummy_gpu_task():
+    pass
+
+demo = gr.Interface(fn=dummy_gpu_task, inputs="text", outputs="text")
+app = gr.mount_gradio_app(app, demo, path="/dummy_gradio")
+
+# ！！！注意！！！
+# uvicorn.run は絶対に呼ばないこと。Hugging FaceのGradioランチャーが自動でappを起動します。