Overviewlangchain-models is a curated collection of Python implementations designed to master the Model I/O component of the LangChain ecosystem. This repository serves as a bridge between raw data and semantic intelligence, showcasing how to leverage industry-leading LLMs and local embedding models.From simple text completions to complex Semantic Search systems, this project provides a modular approach to understanding AI model integration.
🚀 Key Features: 🤖 LLM Mastery: Direct integration with Large Language Models for deterministic text generation. 💬 Structured Conversations: Advanced chat interfaces using System, Human, and AI message schemas. 🔍 Semantic Intelligence: Generating high-dimensional vector embeddings using OpenAI’s text-embedding-3-large. 📊 Vector Similarity: Implementing Cosine Similarity to perform semantic search across custom datasets (e.g., Cricketer profiles). 🌐 Hybrid Approach: Seamlessly switching between cloud-based (OpenAI) and local (Hugging Face) models. 🗂️ Project StructureThe architecture follows a clean, numbered organization for step-by-step learning:
📦 langchain-models: ┣ 📂 1.LLMs/ # Basic Large Language Model implementations ┣ 📂 2.ChatModels/ # Conversational schemas and Chat-based logic ┣ 📂 3.EmbeddedModels/ # The core of Semantic Search & Vector Analysis ┃ ┣ 📜 1_embedding_openai_query.py ┃ ┣ 📜 2_embedding_openai_docs.py ┃ ┣ 📜 3_embedding_hf_local.py ┃ ┗ 📜 4_document_similarity.py # Main script for Semantic Search (Cricketers Dataset) ┣ 📜 .env # Configuration & API Keys (Strictly Git-ignored) ┣ 📜 .gitignore # Protection against secret leaks and heavy venv ┣ 📜 requirements.txt # Unified dependency list ┗ 📜 README.md # Project Documentation
Created with ❤️ by @aliza-dev