Update app.py

app.py

@@ -1,14 +1,27 @@
-# Demonstrates the basic usage of Streamlit
-# Requires a Hugging Face secret value : HUGGINGFACEHUB_API_TOKEN
+# Demonstrates the basic usage
+# 1. Select model
+# 2. Provide a query
+# 3. Invoke the model
 
 import streamlit as st
+from dotenv import load_dotenv
 import os
-import time
 
-from langchain_community.llms import HuggingFaceHub
+# from langchain_community.llms import HuggingFaceHub
 from langchain_community.llms import HuggingFaceEndpoint
 
 
+# Load the API keys, if running locally
+# CHANGE the path to the env file
+
+# If HF space is used then set the env var HUGGINGFACEHUB_API_TOKEN in the settings
+try:
+    load_dotenv('C:\\Users\\raj\\.jupyter\\.env')
+except:
+    print("Environment file not found !! MUST find the env var HUGGINGFACEHUB_API_TOKEN to work.")
+
+
+
 # Title
 st.title('Try out the model')
 
@@ -76,10 +89,10 @@ max_tokens = st.sidebar.number_input(
     value=50
 )
 
-# invoke the LLM
-model_kwargs={ "temperature": "0.1" }
-def invoke():
-    llm_hf = HuggingFaceEndpoint(
+# chached
+@st.cache_resource
+def get_llm(model_id):
+    return HuggingFaceEndpoint(
         repo_id=model_id, 
         temperature=temperature,
         top_k = top_k,
@@ -87,11 +100,14 @@ def invoke():
         repetition_penalty = repetition_penalty,
         max_new_tokens=max_tokens
     )
+
+# invoke the LLM
+def invoke():
+    llm_hf = get_llm(model_id)
     
     # Show spinner, while we are waiting for the response
     with st.spinner('Invoking LLM ... '):
-        time.sleep(5)
-    st.session_state['model-response'] = llm_hf.invoke(query)
-    print(query)
+        st.session_state['model-response'] = llm_hf.invoke(query)
+    
 
 st.button("Invoke", on_click=invoke)