Spaces:
Running
Running
File size: 1,529 Bytes
76cbdff |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 |
import streamlit as st
from langchain_community.embeddings import HuggingFaceBgeEmbeddings
from langchain_pinecone import PineconeVectorStore
from langchain_community.vectorstores.utils import DistanceStrategy
def load_bge_embeddings():
model_name = "BAAI/bge-small-en-v1.5"
model_kwargs = {"device": "cpu"}
encode_kwargs = {"normalize_embeddings": True}
emb_fn = HuggingFaceBgeEmbeddings(
model_name=model_name,
model_kwargs=model_kwargs,
encode_kwargs=encode_kwargs,
query_instruction="Represent this question for searching relevant passages: ",
)
return emb_fn
def load_pinecone_vectorstore():
emb_fn = load_bge_embeddings()
vectorstore = PineconeVectorStore(
embedding=emb_fn,
text_key="text",
distance_strategy=DistanceStrategy.COSINE,
pinecone_api_key=st.secrets["pinecone_api_key"],
index_name=st.secrets["pinecone_index_name"],
)
return vectorstore
def get_vectorstore_filter(ret_config: dict) -> dict:
vs_filter = {}
if ret_config["filter_legis_id"] != "":
vs_filter["legis_id"] = ret_config["filter_legis_id"]
if ret_config["filter_bioguide_id"] != "":
vs_filter["sponsor_bioguide_id"] = ret_config["filter_bioguide_id"]
vs_filter = {
**vs_filter,
"congress_num": {"$in": ret_config["filter_congress_nums"]},
}
vs_filter = {
**vs_filter,
"sponsor_party": {"$in": ret_config["filter_sponsor_parties"]},
}
return vs_filter
|