app.py

from fastapi import FastAPI
from pydantic import BaseModel
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

app = FastAPI()

model_name = "microsoft/phi-2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name)

class MessageRequest(BaseModel):
    username: str
    message: str

@app.post("/intro")
def smart_intro(req: MessageRequest):
    prompt = f"العميل اسمه {req.username}، وكتب: {req.message}\nرد عليه برد ذكي بسيط يحسسه إنك مهتم وهترد عليه بعد لحظات:"
    inputs = tokenizer(prompt, return_tensors="pt", truncation=True)
    outputs = model.generate(**inputs, max_new_tokens=50, temperature=0.8)
    response = tokenizer.decode(outputs[0], skip_special_tokens=True)
    reply = response.split(":")[-1].strip()
    return {"reply": reply}