import streamlit as st import os import json import pandas as pd import random from os.path import join from datetime import datetime from src import ( preprocess_and_load_df, load_agent, ask_agent, decorate_with_code, show_response, get_from_user, load_smart_df, ask_question, ) from dotenv import load_dotenv from langchain_groq import ChatGroq from langchain_google_genai import ChatGoogleGenerativeAI from streamlit_feedback import streamlit_feedback from huggingface_hub import HfApi from datasets import load_dataset, get_dataset_config_info, Dataset from PIL import Image import time import uuid # Page config with beautiful theme st.set_page_config( page_title="VayuChat - AI Air Quality Assistant", page_icon="๐ŸŒฌ๏ธ", layout="wide", initial_sidebar_state="expanded" ) # Custom CSS for beautiful styling st.markdown(""" """, unsafe_allow_html=True) # Auto-scroll JavaScript st.markdown(""" """, unsafe_allow_html=True) # FORCE reload environment variables load_dotenv(override=True) # Get API keys Groq_Token = os.getenv("GROQ_API_KEY") hf_token = os.getenv("HF_TOKEN") gemini_token = os.getenv("GEMINI_TOKEN") models = { "gpt-oss-20b": "openai/gpt-oss-20b", "gpt-oss-120b": "openai/gpt-oss-120b", "llama3.1": "llama-3.1-8b-instant", "llama3.3": "llama-3.3-70b-versatile", "deepseek-R1": "deepseek-r1-distill-llama-70b", "llama4 maverik":"meta-llama/llama-4-maverick-17b-128e-instruct", "llama4 scout":"meta-llama/llama-4-scout-17b-16e-instruct", "gemini-pro": "gemini-1.5-pro" } self_path = os.path.dirname(os.path.abspath(__file__)) # Initialize session ID for this session if "session_id" not in st.session_state: st.session_state.session_id = str(uuid.uuid4()) def upload_feedback(feedback, error, output, last_prompt, code, status): """Enhanced feedback upload function with better logging and error handling""" try: if not hf_token or hf_token.strip() == "": st.warning("โš ๏ธ Cannot upload feedback - HF_TOKEN not available") return False # Create comprehensive feedback data feedback_data = { "timestamp": datetime.now().isoformat(), "session_id": st.session_state.session_id, "feedback_score": feedback.get("score", ""), "feedback_comment": feedback.get("text", ""), "user_prompt": last_prompt, "ai_output": str(output), "generated_code": code or "", "error_message": error or "", "is_image_output": status.get("is_image", False), "success": not bool(error) } # Create unique folder name with timestamp timestamp_str = datetime.now().strftime("%Y%m%d_%H%M%S") random_id = str(uuid.uuid4())[:8] folder_name = f"feedback_{timestamp_str}_{random_id}" # Create markdown feedback file markdown_content = f"""# VayuChat Feedback Report ## Session Information - **Timestamp**: {feedback_data['timestamp']} - **Session ID**: {feedback_data['session_id']} ## User Interaction **Prompt**: {feedback_data['user_prompt']} ## AI Response **Output**: {feedback_data['ai_output']} ## Generated Code ```python {feedback_data['generated_code']} ``` ## Technical Details - **Error Message**: {feedback_data['error_message']} - **Is Image Output**: {feedback_data['is_image_output']} - **Success**: {feedback_data['success']} ## User Feedback - **Score**: {feedback_data['feedback_score']} - **Comments**: {feedback_data['feedback_comment']} """ # Save markdown file locally markdown_filename = f"{folder_name}.md" markdown_local_path = f"/tmp/{markdown_filename}" with open(markdown_local_path, "w", encoding="utf-8") as f: f.write(markdown_content) # Upload to Hugging Face api = HfApi(token=hf_token) # Upload markdown feedback api.upload_file( path_or_fileobj=markdown_local_path, path_in_repo=f"data/{markdown_filename}", repo_id="SustainabilityLabIITGN/VayuChat_Feedback", repo_type="dataset", ) # Upload image if it exists and is an image output if status.get("is_image", False) and isinstance(output, str) and os.path.exists(output): try: image_filename = f"{folder_name}_plot.png" api.upload_file( path_or_fileobj=output, path_in_repo=f"data/{image_filename}", repo_id="SustainabilityLabIITGN/VayuChat_Feedback", repo_type="dataset", ) except Exception as img_error: print(f"Error uploading image: {img_error}") # Clean up local files if os.path.exists(markdown_local_path): os.remove(markdown_local_path) st.success("๐ŸŽ‰ Feedback uploaded successfully!") return True except Exception as e: st.error(f"โŒ Error uploading feedback: {e}") print(f"Feedback upload error: {e}") return False # --- MODERN HEADER & LAYOUT --- st.markdown(""" """, unsafe_allow_html=True) st.markdown("

๐ŸŒฌ๏ธ VayuChat

", unsafe_allow_html=True) st.markdown("
AI-Powered Air Quality, Funding & Population Insights
Simplifying analysis using conversational AI.
", unsafe_allow_html=True) st.markdown("
How to Use:
Select a model from the sidebar and ask questions in the chat. Use quick prompts for common queries. Switch models and rerun easily!
", unsafe_allow_html=True) os.environ["PANDASAI_API_KEY"] = "$2a$10$gbmqKotzJOnqa7iYOun8eO50TxMD/6Zw1pLI2JEoqncwsNx4XeBS2" # --- LOAD ALL DATAFRAMES --- try: df = preprocess_and_load_df(join(self_path, "Data.csv")) import pickle ncap_data = pd.read_pickle(join(self_path, "ncap_funding_data.pkl")) states_data = pd.read_pickle(join(self_path, "states_data.pkl")) st.success("โœ… Data loaded successfully!") except Exception as e: st.error(f"โŒ Error loading data: {e}") st.stop() inference_server = "https://api-inference.huggingface.co/models/mistralai/Mistral-7B-Instruct-v0.2" image_path = "IITGN_Logo.png" # --- MODERN SIDEBAR --- with st.sidebar: st.image(image_path, width=120) st.markdown("
๐Ÿค– Model Selection
", unsafe_allow_html=True) available_models = [] model_names = list(models.keys()) groq_models = [m for m in model_names if "gemini" not in m] gemini_models = [m for m in model_names if "gemini" in m] if Groq_Token and Groq_Token.strip(): available_models.extend(groq_models) if gemini_token and gemini_token.strip(): available_models.extend(gemini_models) if not available_models: st.error("โŒ No API keys available! Please set up your API keys in the .env file") st.stop() model_name = st.selectbox("Choose your AI assistant:", available_models) model_descriptions = { "llama3.1": "๐Ÿฆ™ Fast and efficient for general queries", "llama3.3": "๐Ÿฆ™ Most advanced LLaMA model for complex reasoning", "mistral": "โšก Balanced performance and speed", "gemma": "๐Ÿ’Ž Google's lightweight model", "gemini-pro": "๐Ÿง  Google's most powerful model", "gpt-oss-20b": "๐Ÿ“˜ OpenAI's compact open-weight GPT for everyday tasks", "gpt-oss-120b": "๐Ÿ“š OpenAI's massive open-weight GPT for nuanced responses", "deepseek-R1": "๐Ÿ” DeepSeek's distilled LLaMA model for efficient reasoning", "llama4 maverik": "๐Ÿš€ Meta's LLaMA 4 Maverick โ€” high-performance instruction model", "llama4 scout": "๐Ÿ›ฐ๏ธ Meta's LLaMA 4 Scout โ€” optimized for adaptive reasoning" } if model_name in model_descriptions: st.info(model_descriptions[model_name]) st.markdown("
๐Ÿ“Š Data Sources
", unsafe_allow_html=True) st.markdown("
โ€ข Data.csv: Air quality
โ€ข ncap_funding_data.pkl: Funding
โ€ข states_data.pkl: Population
", unsafe_allow_html=True) st.markdown("
๐Ÿงน Utilities
", unsafe_allow_html=True) if st.button("Clear Chat"): st.session_state.responses = [] st.session_state.processing = False st.session_state.session_id = str(uuid.uuid4()) st.rerun() st.markdown(f"
Session ID
{st.session_state.session_id[:8]}...
", unsafe_allow_html=True) # Load quick prompts questions = [] questions_file = join(self_path, "questions.txt") if os.path.exists(questions_file): try: with open(questions_file, 'r', encoding='utf-8') as f: content = f.read() questions = [q.strip() for q in content.split("\n") if q.strip()] print(f"Loaded {len(questions)} quick prompts") # Debug except Exception as e: st.error(f"Error loading questions: {e}") questions = [] # Add some default prompts if file doesn't exist or is empty if not questions: questions = [ "What is the average PM2.5 level in the dataset?", "Show me the air quality trend over time", "Which pollutant has the highest concentration?", "Create a correlation plot between different pollutants", "What are the peak pollution hours?", "Compare weekday vs weekend pollution levels" ] # --- MODERN QUICK PROMPTS --- st.markdown("
๐Ÿ’ญ Quick Prompts
", unsafe_allow_html=True) selected_prompt = None prompt_buttons = [] for i, q in enumerate(questions): if st.button(q, key=f"quick_{i}", help=q, use_container_width=True): selected_prompt = q # Initialize chat history and processing state if "responses" not in st.session_state: st.session_state.responses = [] if "processing" not in st.session_state: st.session_state.processing = False def show_custom_response(response): """Custom response display function""" role = response.get("role", "assistant") content = response.get("content", "") if role == "user": st.markdown(f"""
You
{content}
""", unsafe_allow_html=True) elif role == "assistant": st.markdown(f"""
๐Ÿค– VayuChat
{content if isinstance(content, str) else str(content)}
""", unsafe_allow_html=True) # Show generated code if available if response.get("gen_code"): with st.expander("๐Ÿ“‹ View Generated Code"): st.code(response["gen_code"], language="python") # Try to display image if content is a file path try: if isinstance(content, str) and (content.endswith('.png') or content.endswith('.jpg')): if os.path.exists(content): st.image(content) return {"is_image": True} except: pass return {"is_image": False} def show_processing_indicator(model_name, question): """Show processing indicator""" st.markdown(f"""
๐Ÿค– VayuChat โ€ข Processing with {model_name}
Question: {question}
๐Ÿ”„ Generating response...
""", unsafe_allow_html=True) # --- MODERN MAIN CHAT AREA --- st.markdown("
", unsafe_allow_html=True) for response in st.session_state.responses: if response["role"] == "user": st.markdown(f"
{response['content']}
", unsafe_allow_html=True) else: st.markdown(f"
{response['content']}
", unsafe_allow_html=True) # Show code details if available if response.get("gen_code"): with st.expander("๐Ÿ“‹ View Generated Code"): st.code(response["gen_code"], language="python") # Feedback section if "feedback" in response: feedback_data = response["feedback"] feedback_text = feedback_data.get('text', '') feedback_score = feedback_data.get('score', '') feedback_str = f"- {feedback_text}" if feedback_text else '' st.markdown(f"
๐Ÿ“ Your Feedback: {feedback_score} {feedback_str}
", unsafe_allow_html=True) else: st.markdown("
How was this response?", unsafe_allow_html=True) col1, col2 = st.columns(2) with col1: thumbs_up = st.button("๐Ÿ‘ Helpful", key=f"fb_up_{id(response)}") with col2: thumbs_down = st.button("๐Ÿ‘Ž Not Helpful", key=f"fb_down_{id(response)}") if thumbs_up or thumbs_down: thumbs = "๐Ÿ‘ Helpful" if thumbs_up else "๐Ÿ‘Ž Not Helpful" comments = st.text_area("๐Ÿ’ฌ Tell us more (optional):", key=f"fb_comments_{id(response)}", placeholder="What could be improved? Any suggestions?", max_chars=500) if st.button("๐Ÿš€ Submit Feedback", key=f"fb_submit_{id(response)}"): feedback = {"score": thumbs, "text": comments} if upload_feedback(feedback, response.get("error", ""), response.get("content", ""), response.get("last_prompt", ""), response.get("gen_code", ""), {}): response["feedback"] = feedback time.sleep(1) st.rerun() st.markdown("
", unsafe_allow_html=True) st.markdown("
", unsafe_allow_html=True) # --- MODERN CHAT INPUT & RERUN --- st.markdown("
", unsafe_allow_html=True) rerun_col, input_col = st.columns([1, 8]) with rerun_col: if st.button("๐Ÿ”„ Rerun Last", help="Rerun the last question with the selected model"): if "last_prompt" in st.session_state and st.session_state["last_prompt"]: st.session_state.processing = True st.session_state.current_model = model_name st.session_state.current_question = st.session_state["last_prompt"] st.rerun() with input_col: prompt = st.text_input("Ask about air quality, funding, or population...", value=selected_prompt or "", key="main_chat") st.markdown("
", unsafe_allow_html=True) # Handle new queries if prompt and not st.session_state.get("processing"): if "last_prompt" in st.session_state: last_prompt = st.session_state["last_prompt"] last_model_name = st.session_state.get("last_model_name", "") if (prompt == last_prompt) and (model_name == last_model_name): prompt = None if prompt: user_response = get_from_user(prompt) st.session_state.responses.append(user_response) st.session_state.processing = True st.session_state.current_model = model_name st.session_state.current_question = prompt st.rerun() # Process the question if we're in processing state if st.session_state.get("processing"): prompt = st.session_state.get("current_question") model_name = st.session_state.get("current_model") try: from src import SYSTEM_PROMPT agent = load_agent(df, SYSTEM_PROMPT, inference_server, name=model_name) response = ask_agent(agent, prompt) if not isinstance(response, dict): response = { "role": "assistant", "content": "โŒ Error: Invalid response format", "gen_code": "", "ex_code": "", "last_prompt": prompt, "error": "Invalid response format" } response.setdefault("role", "assistant") response.setdefault("content", "No content generated") response.setdefault("gen_code", "") response.setdefault("ex_code", "") response.setdefault("last_prompt", prompt) response.setdefault("error", None) except Exception as e: response = { "role": "assistant", "content": f"Sorry, I encountered an error: {str(e)}", "gen_code": "", "ex_code": "", "last_prompt": prompt, "error": str(e) } st.session_state.responses.append(response) st.session_state["last_prompt"] = prompt st.session_state["last_model_name"] = model_name st.session_state.processing = False # Clear processing state if "current_model" in st.session_state: del st.session_state.current_model if "current_question" in st.session_state: del st.session_state.current_question st.rerun() # Auto-scroll to bottom if st.session_state.responses: st.markdown("", unsafe_allow_html=True) # Beautiful sidebar footer # with st.sidebar: # st.markdown("---") # st.markdown(""" #
#

๐Ÿ“„ Paper on VayuChat

#

Learn more about VayuChat in our Research Paper.

#
# """, unsafe_allow_html=True) # Statistics (if logging is enabled) if hf_token and hf_token.strip(): st.markdown("### ๐Ÿ“ˆ Session Stats") total_interactions = len([r for r in st.session_state.get("responses", []) if r.get("role") == "assistant"]) st.metric("Interactions", total_interactions) feedbacks_given = len([r for r in st.session_state.get("responses", []) if r.get("role") == "assistant" and "feedback" in r]) st.metric("Feedbacks Given", feedbacks_given) # Footer st.markdown("""

๐ŸŒ Together for Cleaner Air

VayuChat - Empowering environmental awareness through AI

ยฉ 2025 IIT Gandhinagar Sustainability Lab
""", unsafe_allow_html=True)