Update app.py

app.py

@@ -196,9 +196,9 @@ def save_to_cosmos_db(query, response1, response2):
     else:
         st.error("Cosmos DB is not initialized.")
 
-# 🕵️‍♂️ Search Glossary function
+
+# Add dropdowns for model and database choices
 def search_glossary(query):
-    # 🔍 Searching the glossary—uncovering secrets of the universe! 🕵️‍♂️
     st.markdown(f"### 🔍 Search Glossary for: `{query}`")
     
     # Dropdown for model selection
@@ -209,12 +209,85 @@ def search_glossary(query):
     database_options = ['Semantic Search', 'Arxiv Search - Latest - (EXPERIMENTAL)']
     database_choice = st.selectbox('📚 Select Database', options=database_options, index=0)
 
+
+    
     # Run Button with Emoji
     if st.button("🚀 Run"):
-        # 🕵️‍♂️ We have a query! Let's process it!
+        
+        # 🕵️‍♂️ Searching the glossary for: query
+        all_results = ""
+        st.markdown(f"- {query}")
+
+        # 🔍 ArXiv RAG researcher expert ~-<>-~ Paper Summary & Ask LLM
+        #database_choice Literal['Semantic Search', 'Arxiv Search - Latest - (EXPERIMENTAL)'] Default: "Semantic Search"
+        #llm_model_picked Literal['mistralai/Mixtral-8x7B-Instruct-v0.1', 'mistralai/Mistral-7B-Instruct-v0.2', 'google/gemma-7b-it', 'None'] Default: "mistralai/Mistral-7B-Instruct-v0.2"
         client = Client("awacke1/Arxiv-Paper-Search-And-QA-RAG-Pattern")
+
+        
+        # 🔍 ArXiv RAG researcher expert ~-<>-~ Paper Summary & Ask LLM - api_name: /ask_llm
+        result = client.predict(
+    		prompt=query,
+    		llm_model_picked="mistralai/Mixtral-8x7B-Instruct-v0.1",
+    		stream_outputs=True,
+    		api_name="/ask_llm"
+        )
+        st.markdown(result)
+        st.code(result, language="python", line_numbers=True)
+        save_to_cosmos_db(query, result, result) # Save both responses to Cosmos DB
+        
+        # 🔍 ArXiv RAG researcher expert ~-<>-~ Paper Summary & Ask LLM - api_name: /ask_llm
+        result2 = client.predict(
+    		prompt=query,
+    		llm_model_picked="mistralai/Mistral-7B-Instruct-v0.2",
+    		stream_outputs=True,
+    		api_name="/ask_llm"
+        )
+        st.markdown(result2)
+        st.code(result2, language="python", line_numbers=True)
+        save_to_cosmos_db(query, result2, result2) # Save both responses to Cosmos DB
+
+        # 🔍 ArXiv RAG researcher expert ~-<>-~ Paper Summary & Ask LLM - api_name: /ask_llm
+        result3 = client.predict(
+    		prompt=query,
+    		llm_model_picked="google/gemma-7b-it",
+    		stream_outputs=True,
+    		api_name="/ask_llm"
+        )
+        st.markdown(result3)
+        st.code(result3, language="python", line_numbers=True)
+        save_to_cosmos_db(query, result3, result3) # Save both responses to Cosmos DB
+
+        
+        # 🔍 ArXiv RAG researcher expert ~-<>-~ Paper Summary & Ask LLM - api_name: /update_with_rag_md
+        response2 = client.predict(
+            message=query,	# str  in 'parameter_13' Textbox component
+            llm_results_use=10,
+    		database_choice="Semantic Search",
+    		llm_model_picked="mistralai/Mistral-7B-Instruct-v0.2",
+            api_name="/update_with_rag_md"
+        )        # update_with_rag_md     Returns tuple of 2 elements   [0] str   The output value that appears in the "value_14" Markdown component.   [1] str
+
+        st.markdown(response2[0])
+        st.code(response2[0], language="python", line_numbers=True, wrap_lines=True)  
+        save_to_cosmos_db(query, response2[0], response2[0]) # Save both responses to Cosmos DB
+        
+        st.markdown(response2[1])
+        st.code(response2[1], language="python", line_numbers=True, wrap_lines=True)
+        save_to_cosmos_db(query, response2[1], response2[1]) # Save both responses to Cosmos DB
+        
         
-        # Rest of the code for processing the query...
+        # Aggregate hyperlinks and show with emojis
+        hyperlinks = extract_hyperlinks([response1, response2])
+        st.markdown("### 🔗 Aggregated Hyperlinks")
+        for link in hyperlinks:
+            st.markdown(f"🔗 [{link}]({link})")
+
+        # Show responses in a code format with line numbers
+        st.markdown("### 📜 Response Outputs with Line Numbers")
+        st.code(f"Response 1: \n{format_with_line_numbers(response1)}\n\nResponse 2: \n{format_with_line_numbers(response2)}", language="json")
+
+
+
 
 # 🎤 Function to process text input
 def process_text(text_input):