app.py

from fastapi import FastAPI, Request
from pydantic import BaseModel
from transformers import AutoTokenizer, AutoModel
import torch

app = FastAPI()

# Load model
model_name = "AITeamVN/Vietnamese_Embedding_v2"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModel.from_pretrained(model_name)

class InputText(BaseModel):
    text: str

@app.get("/")
def root():
    return {"message": "AITeamVN/Vietnamese_Embedding_v2 embedding API is running."}

@app.post("/embed")
def get_embedding(data: InputText):
    inputs = tokenizer(data.text, return_tensors="pt", padding=True, truncation=True)
    with torch.no_grad():
        outputs = model(**inputs)
        # Get CLS token or use pooling method
        embedding = outputs.last_hidden_state[:, 0, :].squeeze().tolist()
    return {"embedding": embedding}