from fastapi import FastAPI
from transformers import AutoTokenizer, AutoModel
import torch

# เปลี่ยน cache directory เป็น /tmp
MODEL_NAME = "sentence-transformers/paraphrase-multilingual-mpnet-base-v2"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME, cache_dir="/tmp")
model = AutoModel.from_pretrained(MODEL_NAME, cache_dir="/tmp")

app = FastAPI()

def get_embedding(text):
    inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True)
    with torch.no_grad():
        outputs = model(**inputs)
    embedding = outputs.last_hidden_state.mean(dim=1)
    return embedding.squeeze().tolist()

@app.get("/")
def home():
    return "หวัดดี คุณกำลังเข้าสู่บริการรับฝากหัวใจ API ใช้ post .../embed"

@app.post("/embed")
async def embed_text(data: dict):
    text = data.get("input", "")
    if not text:
        return {"error": "No text provided"}
    vector = get_embedding(text)
    return {"input": text, "embedding": vector}