llm_vector_server/routes/buleprint.py

from quart import Blueprint, jsonify, request

from applications.config import DEFAULT_MODEL, LOCAL_MODEL_CONFIG
from applications.api import get_basic_embedding


server_bp = Blueprint("api", __name__, url_prefix="/api")


def server_routes(vector_db):

    @server_bp.route("/embed", methods=["POST"])
    async def embed():
        body = await request.get_json()
        text = body.get("text")
        model_name = body.get("model", DEFAULT_MODEL)
        if not LOCAL_MODEL_CONFIG.get(model_name):
            return jsonify({"error": "error  model"})

        embedding = await get_basic_embedding(text, model_name)
        return jsonify({"embedding": embedding})

    @server_bp.route("/search", methods=["POST"])
    async def search():
        pass

    return server_bp