Spaces:
Sleeping
Sleeping
File size: 1,373 Bytes
44d337e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 |
"""
FastAPI + Gradio combo app for Hugging Face Space
"""
import os, gradio as gr
from fastapi import FastAPI, Header, HTTPException
from pydantic import BaseModel
from rag import retrieve_info # your existing function
# -------- 1) Private key list ----------
VALID_KEYS = set(os.getenv("RAG_KEYS", "alpha,beta").split(",")) # store secret in Space > Settings > Secrets
# -------- 2) FastAPI core app ----------
app = FastAPI(title="Moffitt RAG API")
class QueryIn(BaseModel):
query: str
k: int = 5
@app.post("/v1/query")
async def rag_query(body: QueryIn, x_api_key: str = Header(None)):
"""
Secure JSON endpoint.
Caller must send: X-API-Key: <one-of-valid-keys>
"""
if x_api_key not in VALID_KEYS:
raise HTTPException(status_code=401, detail="Invalid or missing X-API-Key")
text = retrieve_info(body.query, body.k)
return {"answer": text}
# -------- 3) Public Gradio UI ----------
def run(q, k):
return retrieve_info(q, int(k))
demo = gr.Interface(
fn=run,
inputs=["text", gr.Number(label="k (Number of chunks to retrieve)")],
outputs=gr.Textbox(lines=25, label="Retrieved chunks"),
allow_flagging="never",
title="Moffitt RAG Demo",
description="Type a question; we search Chroma with E5 embeddings."
)
# Mount Gradio at ROOT path "/"
if __name__ == "__main__":
demo.launch()
|