src/title_embd/embed.py

from sentence_transformers import SentenceTransformer
from src.configs.config import EMBEDDING_MODEL, TITLE_EMBEDDINGS_FILE
import numpy as np

def generate_title_embeddings(title_data):
    model = SentenceTransformer(EMBEDDING_MODEL)
    titles = [item["title"] for item in title_data]
    embeddings = model.encode(titles, convert_to_tensor=False)
    np.save(TITLE_EMBEDDINGS_FILE, embeddings)
    return embeddings