Skip to content
This repository has been archived by the owner on Dec 22, 2024. It is now read-only.
/ LangChain Public archive

Implementing LangChain concepts and building meaningful stuffs

License

Notifications You must be signed in to change notification settings

arkapatra31/LangChain

Repository files navigation

LangChain



Repo for maintaining code implementation related to LangChain
Note:-
  • Install the dependencies from requirements.txt using the below command:-
    pip install -r requirements.txt
    To know more about requirements.txt visit pip freeze > requirements.txt
  • To create venv execute python -m venv .venv
  • Table of Contents

    Release Version Feature Implementation
    1.0.0
  • Ollama Integration
  • Groq Integration
  • GPT4All Integration
  • Prompt Engineering
  • LinkedIn Scraping
  • RAG Model
  • Pinecone as Vector DB
  • 1.0.3
  • ReAct Agents
  • Tools
  • Pydantic Output Parser
  • Agent Executors
  • 1.0.4
  • Flask Application
  • Re-Using ReAct Agent
  • API response from Agent Response
  • Agent Executors
  • 1.0.5
  • LangSmith Tracing
  • 1.0.6
  • Build ReAct Agent Executor from Scratch
  • Tools defining for Agent
  • ReAct Prompting
  • LLM Reasoning Engine
  • Output Parsing
  • Tool Execution
  • Handling Agent Action
  • Setting up Agent Finish
  • LLM Callback
  • 1.0.7
  • Uploading and Reading PDF Data
  • Chunking of PDF Data to list of Documents / Text
  • Implementation of both predefined and custom chat prompts
  • Pinecone Vector Store integration
  • FAISS Local Vector Store Integration
  • FAISS Vector saving local copy
  • Retrieval from saved local FAISS
  • Retrieval Chain implementation
  • Implementation of Runnable Passthrough
  • 1.0.8
  • Beautiful Soup Integration to manually scrape doc URLs
  • Using OpenAIEmbeddings for text-embedding-3-small model
  • Using RecursiveUrlLoader lib from Langchain to read docs from URL
  • Splitting and chunking data
  • Initialising the vector store using split documents and embeddings
  • Chat Retrieval QA Chain Creation
  • Streamlit Chat Components in UI/li>
  • Chat History / Context Addition to ChatLLM
  • Using Rephrase Prompt with History Aware Retriever to hold Chat History
  • Using FireCrawl to scrape and crawl websites and index them to Pinecone index
  • 1.0.9
  • Understanding of Langchain Experimental Tools PKG
  • Using the PythonREPL tool from Langchain
  • Using the tool to write Python code
  • Saving the reponse generated from the code
  • Creating a CSV Agent to answer questions related to CSV file
  • Creating a router agent to switch out to different agents based on the input prompt
  • Create a Tool Calling Agent
  • Bind Tools directly to a LLM
  • Create Tool Calling Agent on 2 LLM Providers - OpenAI and Groq
  • 1.1.0
  • Dynamic record config generation using LLM
  • Ability to change column data type, remove and associate reference value
  • Sanitize and convert raw order data to DataFrame
  • Read raw order data from a string and load it into a pandas DataFrame.
  • Strip leading/trailing whitespace from column names.
  • Ensure data consistency and handle missing/malformed data.
  • Convert column data types to the appropriate format.
  • Generate and save config files in json
  • Read config files and apply them to raw order data
  • Generate and save generated dataframe into CSV file
  • Generate Order Line Item Details
  • Adding option for the service with added validation of each columns

  • LangSmith Tracing


    Agent Workflow