app/routes/nlp.py

from fastapi import APIRouter, Request
from fastapi.responses import JSONResponse
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch
from pydantic import BaseModel
from .tts import send_audio   
import uuid
import os

router = APIRouter()

SYSTEM_PROMPT = """you are emma an advanced AI assistant for English language learning."""

class ChatRequest(BaseModel):
    message: str

# Load NLP model
def load_model_nlp():

    model_id = "Qwen/Qwen2.5-1.5B-Instruct"

    tokenizer = AutoTokenizer.from_pretrained(model_id)
    model = AutoModelForCausalLM.from_pretrained(
        model_id,
        torch_dtype=torch.float32,   # CPU friendly
        device_map="cpu"
    )

    model.eval()
    return model, tokenizer

@router.post("/chat")
async def chat(request: Request, chat_request: ChatRequest):
    text = chat_request.message

    model = request.app.state.model_nlp
    tokenizer = request.app.state.tokenizer_nlp

    messages = [
        {"role": "system", "content": SYSTEM_PROMPT},
        {"role": "user", "content": text},
    ]

    inputs = tokenizer.apply_chat_template(
        messages,
        add_generation_prompt=True,
        tokenize=True,
        return_dict=True,
        return_tensors="pt",
    ).to(model.device)

    with torch.no_grad():
        output = model.generate(
            **inputs,
            max_new_tokens=150,
            temperature=0.7,
            top_p=0.9,
            do_sample=True,
        )

    response_text = tokenizer.decode(
        output[0][inputs["input_ids"].shape[-1]:],
        skip_special_tokens=True
    ).strip()

    # Generate audio using TTS
    audio_name = send_audio(request, response_text)
    
    
    return JSONResponse(
        {"response": response_text,
         "audio": audio_name,}
        )