Update app.py

app.py

@@ -14,7 +14,7 @@ os.environ["OPENAI_API_KEY"] =st.secrets["OPENAI_API_KEY"]
 
 @st.cache_resource
 def load_resources():
-    llm = ChatOpenAI(model='gpt-4o', temperature=0.1)
+    llm = ChatOpenAI(model='gpt-4o', temperature=0.2)
     embeddings = OpenAIEmbeddings()
     vector_store = Chroma(embedding_function=embeddings, persist_directory="mining-rag")
     print('vector store loaded')
@@ -94,17 +94,32 @@ def get_answer(question):
     answer = chain.invoke({"question": question})
     return answer
 
+# Precompute the answers to the FAQ questions
+def precompute_faq_answers(faq_questions):
+    answers = {}
+    for faq in faq_questions:
+        answers[faq] = get_answer(faq)
+    return answers
+
+# Cache the precomputed answers to avoid recomputation
+@st.cache_data
+def get_precomputed_answers():
+    return precompute_faq_answers(faq_questions)
+
+precomputed_faq_answers = get_precomputed_answers()
+
 # Streamlit UI
 st.title('Mining Minerals Expert RAG App')
 
 # Sidebar with FAQ Section
 st.sidebar.subheader("Frequently Asked Questions")
+# Show initial 10 questions
 faq_expander = st.sidebar.expander("FAQs")
 with faq_expander:
     for i, faq in enumerate(faq_questions):
-        if st.sidebar.button(f"Q{i+1}: {faq}"):
-            answer = get_answer(faq)
-            st.sidebar.write(f"**Answer to Q{i+1}:** {answer}")
+        if st.sidebar.button(f"Q{i+1}: {faq}", key=f"faq{i+1}"):
+            st.session_state.faq_answer = f"**Answer to Q{i+1}:** {precomputed_faq_answers[faq]}"
+
 
 # Custom Question Section
 st.subheader("Ask Your Own Question")