LangChain Examples (Windows)

Hands-on examples exploring LangChain with Ollama local LLMs. Covers prompt templates, chat interfaces, structured outputs, output parsers, chains (sequential/conditional/parallel), runnable patterns, tools, and simple RAG flows.

Repository Structure

• 1.LLMInteraction: basic chat with Ollama
• 2.Chatbots: REPL chatbot + prompt templates
• 3.StructuredOutputs: TypedDict and Pydantic outputs
• 4.Outputparsers: parsing and fixing generated outputs
• 5.Chains: sequential, conditional, and parallel chains
• 6.Embeddings: Hugging Face embeddings example
• 7.Basic RAG: minimal retrieval-augmented generation
• 8.Runnables: sequences and parallel runnables
• 9.DocumentLoader: loading and processing documents
• 10.VectorStore: Chroma vector store with document storage and retrieval
• 11.Retrievers: document retrieval strategies
• 12.Project-1: YouTube transcript extraction and QA system using RAG
• 13.tools: tools definitions, structured tools, toolkits, and combined tool pipelines

Prerequisites

• Python 3.10+
• Ollama installed and running; pull a model once: ollama pull gemma:2b
• Optional: Google GenAI access for langchain_google_genai

Setup (PowerShell)

powershell python -m venv venv ./venv/Scripts/Activate.ps1 pip install -r requirements.txt If you hit an execution policy warning: powershell Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser

Environment

You can store API keys in a .env file (e.g., GOOGLE_API_KEY=...). This repository ignores .env and venv/ by default (see .gitignore).

Quickstart

• Basic chat: python 1.LLMInteraction/ollama_chat.py
• Chatbot REPL: python 2.Chatbots/1.chatbot.py (type exit/quit to stop)
• Prompt templates: python 2.Chatbots/2.chatprompttemp.py
• Structured outputs:
  • python 3.StructuredOutputs/1.structuredoutput.py
  • python 3.StructuredOutputs/2.deatailed_output_structured.py
  • python 3.StructuredOutputs/3.strucutred_op.py
• Output parsers:
  • python 4.Outputparsers/1.stroutparsers.py
  • python 4.Outputparsers/2.structureoutputparsers.py
• Chains:
  • python 5.Chains/1.sequential_chains.py
  • python 5.Chains/2.conditional_chains.py
  • python 5.Chains/3.parallel_chain.py
• Runnables:
  • python 8.Runnables/1.runnables_sequences.py
  • python 8.Runnables/2.runnnables_parallel.py
• Vector Store: python 10.VectorStore/1.vector_store.py
• Project 1 (YouTube QA): python 12.Project-1/1.system.py
• Tools:
  • python 13.tools/1.tools.py
  • python 13.tools/2.structured_tools.py
  • python 13.tools/3.toolkit.py
  • python 13.tools/4.tool_binding.py
  • python 13.tools/5.finaltool.py

Troubleshooting

• Missing imports (e.g., langchain_google_genai): ensure venv is active, then pip install -r requirements.txt
• Ollama model not found: start Ollama and run ollama pull gemma:2b
• venv activation permissions: use Set-ExecutionPolicy as shown above

---

Maintained as a personal learning sandbox. Contributions welcome.