How to use LangChain to chat with your PDFs

You can install Docker for any PC, Mac, or Linux at home or in any cloud provider that you wish. It will just take a few moments. If you are on Linux, just:

apt-get update && sudo apt-get upgrade && curl -fsSL https://get.docker.com -o get-docker.sh
sh get-docker.sh
#sudo apt install docker-compose -y

And install also Docker-compose with:

apt install docker-compose -y

When the process finishes, you can use it to self-host other services as well. You should see the versions with:

docker --version
docker-compose --version
#sudo systemctl status docker #and the status

docker build -t chat_multiple_pdf .

#export DOCKER_BUILDKIT=1
#docker build --no-cache -t chat_multiple_pdf .

#docker build --no-cache -t chat_multiple_pdf . > build_log.txt 2>&1

#docker run -p 8501:8501 chat_multiple_pdf:latest
#docker exec -it chat_multiple_pdf /bin/bash

#sudo docker run -it -p 8502:8501 chat_multiple_pdf:latest /bin/bash

You can make this build manually, use Github Actions, or your can even combine Gitea and Jenkins to do it for you.

You just need this at: .github/workflows:

name: CI/CD Pipeline

on:
  push:
    branches:
      - main

jobs:
  build-and-push:
    runs-on: ubuntu-latest

    steps:
    - name: Checkout repository
      uses: actions/checkout@v2

    - name: Set up Docker Buildx
      uses: docker/setup-buildx-action@v1 #it will produce x86 and ARM64 arch images

    - name: Login to GitHub Container Registry
      uses: docker/login-action@v1
      with:
        registry: ghcr.io
        username: ${{ github.actor }}
        password: ${{ secrets.CICD_TOKEN_AskPDF }} #Settings -> Dev Settings -> PAT's -> Tokens +++ Repo Settings -> Secrets & variables -> Actions -> New repo secret 

    - name: Build and push Docker image
      uses: docker/build-push-action@v2
      with:
        context: .
        push: true
        tags: ghcr.io/yourghuser/ask-multiple-pdfs:v1.0

Copy it to your project and get the most out of GH Actions

Examples of Embedded Models:

• NLP (Natural Language Processing): Models like BERT or GPT-3 used for tasks like text summarization, sentiment analysis, chatbots.
• Computer Vision: CNNs for object recognition, facial recognition, or medical image analysis.
• Speech Recognition: Used in transcribing speech for voice-controlled systems.

How Embedded Models Enhance Semantic Search:

1. Natural Language Understanding: Grasp the semantics and relationships in user queries.
2. Intent Recognition: Identify different user intents (informational, transactional, etc.).
3. Query Expansion: Suggest related terms or synonyms to improve results.
4. Contextual Search: Take into account the query and content context for relevant results.
5. Entity Recognition: Identify entities (people, places, etc.) in queries.
6. Question Answering: Answer questions using context from the query and data.
7. Relevance Ranking: Rank search results by semantic similarity.
8. Personalization: Tailor results to individual user preferences.

How Components Work Together:

1. Embedded Models: Convert text into dense vector representations (embeddings).
2. Vector Store: Stores embeddings with metadata for efficient retrieval.
3. Chroma: Scalable vector similarity search library (developed by Spotify).
4. FAISS: High-performance similarity search library (developed by Facebook AI).

Typical Workflow for Semantic Search:

• Text is processed by embedded models into vectors.
• These vectors are stored in the vector store.
• User queries are converted into vectors.
• Libraries like Chroma/FAISS search for the most similar vectors.
• Results are ranked and presented as recommendations or search results.

Applications

• Semantic search
• Recommendation systems
• Content retrieval
• Personalized content delivery
• Question-answering systems

• Provides a high-level interface for building RAG systems
• Supports various retrieval methods, including vector databases and search engines
• Offers a wide range of pre-built components and utilities for text processing and generation
• Integrates well with popular language models like OpenAI’s GPT series

• Designed specifically for building index-based retrieval systems
• Provides a simple and intuitive API for indexing and querying documents
• Supports various indexing techniques, including vector-based and keyword-based methods
• Offers built-in support for common document formats (e.g., PDF, HTML)
• Lightweight and easy to integrate into existing projects

• LangFlow - Pros:
  • Offers a visual programming interface for building RAG pipelines
  • Allows for easy experimentation and prototyping without extensive coding
  • Provides a library of pre-built nodes for various tasks (e.g., retrieval, generation)
  • Supports integration with popular language models and libraries
  • Enables rapid development and iteration of RAG systems
• Cons:
  • May have limitations in terms of customization and fine-grained control compared to code-based approaches
  • Visual interface may not be suitable for complex or large-scale projects
  • Dependency on the LangFlow platform and its ecosystem