The latest graphrag interface is used, using the local ollama to provide the LLM interface.Support for using the pip installation

CDoc lets you chat with your documents using local LLMs, combining Ollama, ChromaDB, and LangChain for offline, secure, and efficient information extraction. Perfect for researchers, developers, and professionals seeking quick insights from their documents.

This repository contains a complete pipeline for Fine-Tuning Language Models (LLMs) using the Unsloth/MLX library. The project is structured to facilitate the loading of quantized models.

Windows Copilot is suck, that's why Optimus is created, for being your true Windows assistant

AI Model (llava from ollama) used to generate description of a image. A mere 7b Model was used. Documentation PDF : https://drive.google.com/file/d/1dHPw1Z0hPHerkl3kMPB_Inmaaovv0M1r/view?usp=drivesdk

A smart GPU-aware proxy designed for Proxmox VE that dynamically manages GPU resources between Ollama and GPU-intensive background processes.

This is a security-first NL-to-SQL solution designed for enterprises with small to medium-sized databases. It enables natural language querying of data while ensuring that all processing remains fully on-premises. No data is sent to external cloud environments, and no sensitive information is exposed to external LLM services.

Lightweight Next.js proxy for local Ollama LLM, enabling real-time AI chat with streaming responses and easy model switching.