File size: 3,023 Bytes
7a70599 fdb21d8 e3862ef 6600c85 e3862ef 800cf25 fdb21d8 7a70599 fdb21d8 6600c85 fdb21d8 d3378a2 800cf25 d3378a2 800cf25 fdb21d8 7a70599 fdb21d8 |
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 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 |
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
from fastapi.responses import JSONResponse
from fastapi.staticfiles import StaticFiles
from huggingface_hub import InferenceClient
from sentence_transformers import SentenceTransformer
from fastapi import Request
import requests
import numpy as np
import argparse
import os
HOST = os.environ.get("API_URL", "0.0.0.0")
PORT = os.environ.get("PORT", 7860)
parser = argparse.ArgumentParser()
parser.add_argument("--host", default=HOST)
parser.add_argument("--port", type=int, default=PORT)
parser.add_argument("--reload", action="store_true", default=True)
parser.add_argument("--ssl_certfile")
parser.add_argument("--ssl_keyfile")
args = parser.parse_args()
app = FastAPI()
app.add_middleware(
CORSMiddleware,
allow_origins=["*"],
allow_credentials=True,
allow_methods=["*"],
allow_headers=["*"],
)
app = FastAPI()
embedder = SentenceTransformer('sentence-transformers/distiluse-base-multilingual-cased-v1')
@app.post("/api/embed")
async def embed(request: Request):
data = await request.json()
texts = data.get("texts", [])
embeddings = embedder.encode(texts)
return {"embeddings": embeddings.tolist()}
@app.get("/invert")
async def invert(text: str):
return {
"original": text,
"inverted": text[::-1],
}
HF_TOKEN = os.getenv("REACT_APP_HF_TOKEN")
if not HF_TOKEN:
raise RuntimeError("Le token Hugging Face (HF_TOKEN) n'est pas défini dans les variables d'environnement.")
# Initialisation du client HF
hf_client = InferenceClient(
provider="novita",
api_key=HF_TOKEN,
)
@app.post("/api/chat")
async def chat(request: Request):
data = await request.json()
user_message = data.get("message", "").strip()
if not user_message:
raise HTTPException(status_code=400, detail="Le champ 'message' est requis.")
try:
# Appel au modèle en mode chat
completion = hf_client.chat.completions.create(
model="mistralai/Mistral-7B-Instruct-v0.3",
messages=[
{"role": "system", "content": "Tu es un assistant médical spécialisé en schizophrénie."},
{"role": "user", "content": user_message}
],
max_tokens=512,
temperature=0.7,
)
bot_msg = completion.choices[0].message.content
return {"response": bot_msg}
except Exception as e:
# En cas d'erreur d'inférence
raise HTTPException(status_code=502, detail=f"Erreur d'inférence HF : {e}")
@app.get("/data")
async def get_data():
data = {"data": np.random.rand(100).tolist()}
return JSONResponse(data)
app.mount("/", StaticFiles(directory="static", html=True), name="static")
if __name__ == "__main__":
import uvicorn
print(args)
uvicorn.run(
"app:app",
host=args.host,
port=args.port,
reload=args.reload,
ssl_certfile=args.ssl_certfile,
ssl_keyfile=args.ssl_keyfile,
) |