embedding_config.py

from langchain_community.embeddings import HuggingFaceEmbeddings
from utils.vector_store import get_vector_store
import os


EMBEDDING_CONFIGS = {
    "General-purpose (bge-large-en)": HuggingFaceEmbeddings(
        model_name="BAAI/bge-large-en",
        model_kwargs={"device": "cpu"},   # or "cuda" if you have GPU
        encode_kwargs={"normalize_embeddings": True}
    ),
    "Fast & lightweight (bge-small-en)": HuggingFaceEmbeddings(
        model_name="BAAI/bge-small-en",
        model_kwargs={"device": "cpu"},
        encode_kwargs={"normalize_embeddings": True}
    ),
    "QA optimized (e5-large-v2)": HuggingFaceEmbeddings(
        model_name="intfloat/e5-large-v2",
        model_kwargs={"device": "cpu"},
        encode_kwargs={"normalize_embeddings": True}
    ),
    "Instruction-tuned (instructor-large)": HuggingFaceEmbeddings(
        model_name="hkunlp/instructor-large",
        model_kwargs={"device": "cpu"},
        encode_kwargs={"normalize_embeddings": True}
    ),
}