from flask import Flask, request, jsonify
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

app = Flask(__name__)

# Charger le modèle depuis Hugging Face
MODEL_NAME = "fatmata/psybot"  # Remplace avec le vrai nom de ton modèle
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, torch_dtype=torch.float16)

@app.route("/chat", methods=["POST"])
def chat():
    data = request.json
    user_input = data.get("message", "")

    if not user_input:
        return jsonify({"error": "Message vide"}), 400

    # Génération de la réponse
    prompt = f"<|startoftext|><|user|> {user_input} <|bot|>"
    inputs = tokenizer(prompt, return_tensors="pt").input_ids.to(model.device)

    with torch.no_grad():
        output = model.generate(inputs, max_new_tokens=100, pad_token_id=tokenizer.eos_token_id)

    response = tokenizer.decode(output[0], skip_special_tokens=True)
    if "<|bot|>" in response:
        response = response.split("<|bot|>")[-1].strip()

    return jsonify({"response": response})

if __name__ == "__main__":
    app.run(host="0.0.0.0", port=7860)