initital commit

.dockerignore

@@ -0,0 +1,30 @@
+# Git
+.git
+.gitignore
+.gitattributes
+
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+*.so
+.Python
+venv/
+ENV/
+
+# Environment
+.env
+.venv
+
+# IDE
+.vscode/
+.idea/
+
+# Chainlit
+.chainlit/
+
+# Misc
+.DS_Store
+*.log
+README.md
+LICENSE 

Dockerfile

@@ -0,0 +1,47 @@
+# Use Python 3.10 slim image as base
+FROM python:3.10-slim
+
+# Install system dependencies
+RUN apt-get update && \
+    apt-get install -y \
+    build-essential \
+    git \
+    poppler-utils \
+    && rm -rf /var/lib/apt/lists/*
+
+# Set working directory
+WORKDIR /app
+
+# Set environment variables
+ENV PYTHONUNBUFFERED=1
+ENV TRANSFORMERS_CACHE=/app/model_cache
+ENV HF_HOME=/app/model_cache
+ENV TORCH_HOME=/app/model_cache
+ENV CHAINLIT_HOST=0.0.0.0
+ENV CHAINLIT_PORT=7860
+
+# Install Python dependencies
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# Create necessary directories
+RUN mkdir -p /app/model_cache /app/vectorstore/db_faiss /app/data
+
+# Copy application files
+COPY model.py ingest.py chainlit.md download_assets.py ./
+
+# Download models and cache them
+RUN python -c "from transformers import AutoTokenizer, AutoModelForSeq2SeqLM; \
+    AutoTokenizer.from_pretrained('google/flan-t5-base'); \
+    AutoModelForSeq2SeqLM.from_pretrained('google/flan-t5-base'); \
+    from sentence_transformers import SentenceTransformer; \
+    SentenceTransformer('sentence-transformers/all-MiniLM-L6-v2')"
+
+# Download assets from Hugging Face Hub
+RUN python download_assets.py
+
+# Expose the port Chainlit runs on
+EXPOSE 7860
+
+# Run the Chainlit application
+CMD ["chainlit", "run", "model.py", "--host", "0.0.0.0", "--port", "7860"] 

README.md

@@ -0,0 +1,106 @@
+# MindMedic - AI Mental Health Assistant 🧠
+
+MindMedic is an AI-powered mental health diagnostic assistant built using FLAN-T5 and LangChain. It helps users understand potential mental health concerns by providing evidence-based information and preliminary insights based on trusted mental health resources.
+
+## Table of Contents
+
+- [MindMedic - AI Mental Health Assistant 🧠](#mindmedic---ai-mental-health-assistant-)
+  - [Table of Contents](#table-of-contents)
+  - [Introduction](#introduction)
+  - [Features](#features)
+  - [Prerequisites](#prerequisites)
+  - [Installation](#installation)
+  - [Usage](#usage)
+  - [Important Note](#important-note)
+    - [Emergency Resources:](#emergency-resources)
+  - [Contributing](#contributing)
+
+## Introduction
+
+MindMedic leverages advanced language models and vector stores to provide informative responses to mental health-related queries. It processes and understands a curated collection of mental health resources to offer reliable, evidence-based information about various mental health conditions, symptoms, and general mental wellness topics.
+
+## Features
+
+- 🤖 Powered by Google's FLAN-T5 language model
+- 📚 Knowledge base built from trusted mental health resources
+- 💡 Provides evidence-based responses with sources
+- 🔍 Semantic search capabilities for accurate information retrieval
+- 💻 User-friendly chat interface powered by Chainlit
+- 🔒 Runs locally for privacy
+
+## Prerequisites
+
+Before setting up MindMedic, ensure you have:
+
+- Python 3.6 or higher
+- pip (Python package manager)
+- 4GB+ RAM recommended
+- CPU with x86_64 architecture
+
+## Installation
+
+1. Clone this repository:
+   ```bash
+   git clone https://github.com/your-username/MindMedic.git
+   cd MindMedic
+   ```
+
+2. Create and activate a virtual environment:
+   ```bash
+   python -m venv venv
+   # On Windows
+   venv\Scripts\activate
+   # On Unix or MacOS
+   source venv/bin/activate
+   ```
+
+3. Install required packages:
+   ```bash
+   pip install -r requirements.txt
+   ```
+
+4. Prepare the knowledge base:
+   ```bash
+   python ingest.py
+   ```
+
+## Usage
+
+1. Start the MindMedic chatbot:
+   ```bash
+   chainlit run model.py -w
+   ```
+
+2. Open your web browser and navigate to `http://localhost:8000`
+
+3. Start interacting with MindMedic by asking mental health-related questions
+
+Example queries:
+- "What are the common symptoms of anxiety?"
+- "How can I tell if I'm experiencing depression?"
+- "What are some coping strategies for stress?"
+- "Can you explain what panic attacks feel like?"
+
+## Important Note
+
+⚠️ **Disclaimer**: MindMedic is an AI assistant designed to provide information and general guidance about mental health topics. It is NOT a replacement for professional mental health care. Always consult with qualified mental health professionals for diagnosis and treatment. In case of emergency, contact your local emergency services or mental health crisis hotline immediately.
+
+### Emergency Resources:
+- National Suicide Prevention Lifeline (US): 988
+- Crisis Text Line: Text HOME to 741741
+- Find local mental health resources: [NAMI HelpLine](https://www.nami.org/help)
+
+## Contributing
+
+Contributions to improve MindMedic are welcome! To contribute:
+
+1. Fork the repository
+2. Create a feature branch
+3. Make your changes
+4. Submit a pull request
+
+Please ensure your contributions align with mental health best practices and maintain the focus on providing accurate, helpful information.
+
+---
+
+Built with ❤️ for mental health awareness and support. Remember, it's okay to not be okay, and seeking help is a sign of strength.

chainlit.md

@@ -0,0 +1,11 @@
+# Welcome to MindMate! 🚀🤖
+
+Hi there, 👋 and welcome to **MindMate**, your AI-powered mental health support assistant. This bot is designed to help you get reliable, evidence-based answers to questions about mental well-being—whether it’s about managing anxiety, coping strategies for stress, or understanding depression.
+
+## Useful Links 🔗
+
+- **Knowledge Base:** All the mental health guides, fact sheets, and clinical resources we’ve ingested to power MindMate. Explore our source documents here: [Mental Health Knowledge Base](vectorstore/db_faiss) 📚  
+- **Project Repository:** View the code, contribute enhancements, or report issues on GitHub: [Llama2-Medical-Chatbot](https://github.com/AIAnytime/Llama2-Medical-Chatbot) 💻    
+
+Take care of your mind and happy chatting! 🧠😊  
+

download_assets.py

@@ -0,0 +1,51 @@
+from huggingface_hub import hf_hub_download
+import os
+
+def download_assets():
+    """Download necessary assets from Hugging Face Hub"""
+    # Create directories if they don't exist
+    os.makedirs('data', exist_ok=True)
+    os.makedirs('vectorstore/db_faiss', exist_ok=True)
+    
+    # Dataset repository ID
+    repo_id = "MoizK/mindmedic-assets"
+    
+    # Download PDF files
+    pdf_files = [
+        "71763-gale-encyclopedia-of-medicine.-vol.-1.-2nd-ed.pdf",
+        "Depression-NIM-2024.pdf",
+        "Depression-and-Other-Common-Mental-Disorders-Global-Health-Estimates.pdf",
+        "Doing-What-Matters-in-Times-of-Stress.pdf",
+        "Generalized-Anxiety-Disorder-When-Worry-Gets-Out-of-Control.pdf",
+        "WHO-mhGAP-Intervention-Guide-v2.pdf",
+        "social-anxiety-disorder-more-than-just-shyness.pdf"
+    ]
+    
+    for pdf_file in pdf_files:
+        try:
+            hf_hub_download(
+                repo_id=repo_id,
+                filename=f"data/{pdf_file}",
+                local_dir=".",
+                local_dir_use_symlinks=False
+            )
+            print(f"Downloaded {pdf_file}")
+        except Exception as e:
+            print(f"Error downloading {pdf_file}: {e}")
+    
+    # Download FAISS index files
+    index_files = ["index.faiss", "index.pkl"]
+    for index_file in index_files:
+        try:
+            hf_hub_download(
+                repo_id=repo_id,
+                filename=f"vectorstore/db_faiss/{index_file}",
+                local_dir=".",
+                local_dir_use_symlinks=False
+            )
+            print(f"Downloaded {index_file}")
+        except Exception as e:
+            print(f"Error downloading {index_file}: {e}")
+
+if __name__ == "__main__":
+    download_assets() 

ingest.py

@@ -0,0 +1,28 @@
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_community.vectorstores import FAISS
+from langchain_community.document_loaders import PyPDFLoader, DirectoryLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter 
+
+DATA_PATH = 'data/'
+DB_FAISS_PATH = 'vectorstore/db_faiss'
+
+# Create vector database
+def create_vector_db():
+    loader = DirectoryLoader(DATA_PATH,
+                             glob='*.pdf',
+                             loader_cls=PyPDFLoader)
+
+    documents = loader.load()
+    text_splitter = RecursiveCharacterTextSplitter(chunk_size=500,
+                                                   chunk_overlap=50)
+    texts = text_splitter.split_documents(documents)
+
+    embeddings = HuggingFaceEmbeddings(model_name='sentence-transformers/all-MiniLM-L6-v2',
+                                       model_kwargs={'device': 'cpu'})
+
+    db = FAISS.from_documents(texts, embeddings)
+    db.save_local(DB_FAISS_PATH)
+
+if __name__ == "__main__":
+    create_vector_db()
+

model.py

@@ -0,0 +1,128 @@
+from langchain.prompts import PromptTemplate
+from langchain_community.embeddings import HuggingFaceEmbeddings
+from langchain_community.vectorstores import FAISS
+from langchain.llms import HuggingFacePipeline
+from transformers import AutoTokenizer, AutoModelForSeq2SeqLM, pipeline
+from langchain.chains import RetrievalQA
+import chainlit as cl
+from dotenv import load_dotenv
+import torch
+import os
+
+load_dotenv()  
+
+DB_FAISS_PATH = 'vectorstore/db_faiss'
+
+# Prompt Template
+custom_prompt_template = """Use the following pieces of information to answer the user's question.
+If you don't know the answer, just say that you don't know, don't try to make up an answer.
+
+Context: {context}
+Question: {question}
+
+Only return the helpful answer below and nothing else.
+Helpful answer:
+"""
+
+def set_custom_prompt():
+    prompt = PromptTemplate(template=custom_prompt_template,
+                            input_variables=['context', 'question'])
+    return prompt
+
+# Create RetrievalQA chain
+def retrieval_qa_chain(llm, prompt, db):
+    qa_chain = RetrievalQA.from_chain_type(
+        llm=llm,
+        chain_type='stuff',
+        retriever=db.as_retriever(search_kwargs={'k': 2}),
+        return_source_documents=True,
+        chain_type_kwargs={'prompt': prompt}
+    )
+    return qa_chain
+
+# Load Hugging Face LLM
+def load_llm():
+    # Load model and tokenizer
+    tokenizer = AutoTokenizer.from_pretrained("google/flan-t5-base")
+    model = AutoModelForSeq2SeqLM.from_pretrained(
+        "google/flan-t5-base",
+        device_map="cpu",
+        torch_dtype=torch.float32
+    )
+
+    # Create text-generation pipeline without invalid parameters
+    pipe = pipeline(
+        "text2text-generation",
+        model=model,
+        tokenizer=tokenizer,
+        max_new_tokens=512,
+        repetition_penalty=1.15
+    )
+
+    # Create LangChain wrapper for the pipeline
+    llm = HuggingFacePipeline(pipeline=pipe)
+    return llm
+
+# Build full chatbot pipeline
+def qa_bot():
+    embeddings = HuggingFaceEmbeddings(
+        model_name="sentence-transformers/all-MiniLM-L6-v2",
+        model_kwargs={'device': 'cpu'}
+    )
+    db = FAISS.load_local(
+        DB_FAISS_PATH,
+        embeddings,
+        allow_dangerous_deserialization=True
+    )
+
+    llm = load_llm()
+    qa_prompt = set_custom_prompt()
+    qa = retrieval_qa_chain(llm, qa_prompt, db)
+    return qa
+
+# Run for one query (used internally)
+def final_result(query):
+    qa_result = qa_bot()
+    response = qa_result({'query': query})
+    return response
+
+# Chainlit UI - Start
+@cl.on_chat_start
+async def start():
+    chain = qa_bot()
+    msg = cl.Message(content="Starting the bot...")
+    await msg.send()
+    msg.content = "Hi, Welcome to MindMate. What is your query?"
+    await msg.update()
+    cl.user_session.set("chain", chain)
+
+# Chainlit UI - Handle messages
+@cl.on_message
+async def main(message: cl.Message):
+    chain = cl.user_session.get("chain")
+    cb = cl.AsyncLangchainCallbackHandler(
+        stream_final_answer=True, answer_prefix_tokens=["FINAL", "ANSWER"]
+    )
+    cb.answer_reached = True
+    
+    # Use invoke with proper query format
+    res = await cl.make_async(chain.invoke)(
+        {"query": message.content}, 
+        callbacks=[cb]
+    )
+    
+    # Extract result and sources from the response
+    answer = res.get("result", "No result found")
+    sources = res.get("source_documents", [])
+
+    # Format sources to show only the content
+    if sources:
+        formatted_sources = []
+        for source in sources:
+            if hasattr(source, 'page_content'):
+                formatted_sources.append(source.page_content.strip())
+        
+        if formatted_sources:
+            answer = f"{answer}\n\nBased on the following information:\n" + "\n\n".join(formatted_sources)
+
+    await cl.Message(content=answer).send()

requirements.txt

@@ -0,0 +1,12 @@
+pypdf>=3.0.0
+langchain>=0.1.0
+torch>=2.0.0
+transformers>=4.30.0
+accelerate>=0.20.0
+bitsandbytes>=0.41.0
+sentence-transformers>=2.2.0
+faiss-cpu>=1.7.0
+chainlit>=0.7.0
+huggingface-hub>=0.19.0
+langchain-community>=0.0.10
+python-dotenv>=1.0.0