Updated AI-Travel-Agent sample with latest changes

AI-Travel-Agent/AI_Travel_Agent_streamlit.py

@@ -2,13 +2,13 @@
 # streamlit run AI_Travel_Agent_streamlit.py
 
 # Importing necessary modules
-
 import os
 import re
 import time
 import streamlit as st
 from amadeus import Client
 from dotenv import load_dotenv
+from huggingface_hub import hf_hub_download
 from langchain_community.llms import LlamaCpp
 from langchain_core.callbacks import CallbackManager, StreamingStdOutCallbackHandler
 from langchain_community.utilities import GoogleSerperAPIWrapper
@@ -27,10 +27,10 @@
 def create_llm():
     """
     Create and initialize the LlamaCpp with the selected model.
-    Parameters can be changed based on the end user requirements.
-    Here we are using Meta Llama 3.1(Q4_K_S) model which is configured using some hyperparameters.
-    For example, GPU Layers to be offloaded on all the available layers for inference.
-    Context Length of 4096 tokens. Temperature set as 0 for deterministic output.
+    Parameters can be changed based on the end user's requirements.
+    Here we are using the  Meta Llama 3.1(Q4_K_S) model, which is configured using some hyperparameters.
+    For example, GPU Layers are to be offloaded to all the available layers for inference.
+    Context Length of 4096 tokens. The temperature is set as 0 for deterministic output.
     Top-P Sampling as 0.95 for controlled randomness, Batch Size as 512 for parallel processing
 
     Returns:
@@ -40,10 +40,12 @@ def create_llm():
         Exception: If there is any error during model loading, a Streamlit error is displayed.
     """
     try:
+        model_path = hf_hub_download(repo_id="bartowski/Meta-Llama-3.1-8B-Instruct-GGUF", filename="Meta-Llama-3.1-8B-Instruct-Q4_K_S.gguf")  # Downloading the model here
+
         llm = LlamaCpp(
             # Path to the Llama model file
-            model_path="./models/Meta-Llama-3.1-8B-Instruct-Q4_K_S.gguf",
-            # Number of layers to be loaded into gpu memory (default: 0)
+            model_path=model_path,
+            # Number of layers to be loaded into GPU memory (default: 0)
             n_gpu_layers=-1,
             # Random number generator (RNG) seed (default: -1, -1 = random seed)
             seed=512,
@@ -112,7 +114,8 @@ def create_prompt_template():
 
     FORMAT_INSTRUCTIONS = """Use a json blob to specify a tool by providing an action key (tool name) and an action_input key (tool input).
 
-    Use the closest_airport tool and single_flight_search tool for any flight related queries. Give all the flight details including Flight Number, Carrier, Departure time, Arrival time and Terminal details to the human.
+    Use the closest_airport tool and single_flight_search tool for any flight related queries.
+    Give all the flight details including Flight Number, Carrier, Departure time, Arrival time and Terminal details to the human.
     Use the Google Search tool and knowledge base for any itinerary-related queries. Give all the detailed information on tourist attractions, must-visit places, and hotels with ratings to the human.
     Use the Google Search tool for distance calculations. Give all the web results to the human.
     Always consider the traveler's preferences, budget constraints, and any specific requirements mentioned in their query.
@@ -144,7 +147,8 @@ def create_prompt_template():
     }}}}
     ```[/INST]"""
 
-    SUFFIX = """Begin! Reminder to ALWAYS respond with a valid json blob of a single action. Use tools if necessary. Respond directly if appropriate. Format is Action:```$JSON_BLOB```then Observation:.
+    SUFFIX = """Begin! Reminder to ALWAYS respond with a valid json blob of a single action. Use tools if necessary. Respond directly if appropriate.
+    Format is Action:```$JSON_BLOB```then Observation:.
     Thought:[INST]"""
 
     HUMAN_MESSAGE_TEMPLATE = "{input}\n\n{agent_scratchpad}"
@@ -251,7 +255,8 @@ def streamlit_UI():
         Exception: If there is any error during the fetching or streaming the response, a streamlit error is displayed.
     """
     st.title(":earth_africa::airplane: AI Travel Agent")
-    st.write("This Langchain-powered AI Travel Agent is designed to assist you with quick travel-related queries. You can request **flight details** for a specific day or find **nearby airports** by location. For other questions, we use **Google Search** for the latest information.")
+    st.write("This Langchain-powered AI Travel Agent is designed to assist you with quick travel-related queries. \
+    You can request **flight details** for a specific day or find **nearby airports** by location. For other questions, we use **Google Search** for the latest information.")
 
     # Sidebar with questions on the UI
     st.sidebar.title(":bulb: Example Queries")
@@ -296,14 +301,10 @@ def streamlit_UI():
             try:
                 with st.spinner("Generating answer..."):
                     with st.expander("Agent Execution", expanded=True):
-                        # Create placeholder for streaming output
-                        # response_placeholder = st.empty()
-
                         chunks = []
                         for chunk in agent_executor.stream({"input": question}):
                             chunks.append(chunk)
                             st.write(chunk)
-                    # response = agent_executor.invoke({"input": question})
 
                     placeholder = st.empty()
                     content = chunks[2]['output']
@@ -345,8 +346,11 @@ def streamlit_UI():
 PREFIX, FORMAT_INSTRUCTIONS, SUFFIX, HUMAN_MESSAGE_TEMPLATE = create_prompt_template()
 
 # calling the create_agent function here
-agent = create_agent(llm, tools, PREFIX, SUFFIX,
-                     HUMAN_MESSAGE_TEMPLATE, FORMAT_INSTRUCTIONS)
+try:
+    agent = create_agent(llm, tools, PREFIX, SUFFIX,
+                         HUMAN_MESSAGE_TEMPLATE, FORMAT_INSTRUCTIONS)
+except Exception as e:
+    st.error(f"Error loading the agent with tools : {e}")
 
 # calling the run_agent function here
 agent_executor = run_agent(agent, tools)