Update app.py

app.py

@@ -2,9 +2,10 @@ import streamlit as st
 from streamlit_option_menu import option_menu
 
 import os
-from langchain.llms import HuggingFaceHub # for calling HuggingFace Inference API (free for our use case)
+# from langchain.llms import HuggingFaceHub # old, for calling HuggingFace Inference API (free for our use case)
+from langchain_community.llms import HuggingFaceEndpoint # for calling HuggingFace Inference API (free for our use case)
 from langchain.embeddings import HuggingFaceEmbeddings # to let program know what embeddings the vector store was embedded in earlier
-
+from langchain_community.llms import HuggingFaceEndpoint
 # to set up the agent and tools which will be used to answer questions later
 from langchain.agents import initialize_agent 
 from langchain.agents import tool # decorator so each function will be recognized as a tool
@@ -171,12 +172,19 @@ countries = [
 def get_llm():
     # This is an inference endpoint API from huggingface, the model is not run locally, it is run on huggingface
     # It is a free API that is very good for deploying online for quick testing without users having to deploy a local LLM
-    llm = HuggingFaceHub(repo_id=st.session_state['model'],
-                        model_kwargs={
-                        'temperature': st.session_state['temperature'],
-                        "max_new_tokens": st.session_state['max_new_tokens']
-                        }, 
-                        )
+    # llm = HuggingFaceHub(repo_id=st.session_state['model'],
+    #                     model_kwargs={
+    #                     'temperature': st.session_state['temperature'],
+    #                     "max_new_tokens": st.session_state['max_new_tokens']
+    #                     }, 
+    #                     )
+    llm = HuggingFaceEndpoint(
+        endpoint_url=st.session_state['model'],
+        huggingfacehub_api_token=os.environ['HUGGINGFACEHUB_API_TOKEN'],
+        task="text-generation",
+        "temperature": st.session_state['temperature'],
+        "max_new_tokens": st.session_state['max_new_tokens']
+    )
     return llm
 
 # for chromadb vectore store