import os import streamlit as st from langchain_huggingface import HuggingFaceEmbeddings from langchain_groq import ChatGroq def hf_embeddings(): """Lightweight sentence-transformer embeddings (runs on CPU, no GPU needed).""" return HuggingFaceEmbeddings( model_name="sentence-transformers/all-MiniLM-L6-v2", model_kwargs={"device": "cpu"}, ) def groq_llm(): """ Fast, free Llama-3 inference via Groq. Requires GROQ_API_KEY set as a HuggingFace Space secret. """ api_key = os.environ.get("GROQ_API_KEY", "") if not api_key: st.error( "⚠️ GROQ_API_KEY secret is not set. " "Add it in your Space → Settings → Repository secrets." ) st.stop() return ChatGroq( model="llama-3.1-8b-instant", api_key=api_key, temperature=0.2, max_tokens=1024, streaming=False, )