- Bangladesh
-
02:10
- 6h ahead - @fazlulkarimweb
Highlights
Starred repositories
π€ Build voice-based LLM agents. Modular + open source.
Creepy device and browser fingerprinting
A multi-select component designed with shadcn/ui
A curated list of awesome things related to shadcn/ui.
A full-featured, hackable Next.js AI chatbot built by Vercel
A full-featured, Supabaseified Next.js AI chatbot built by Vercel Labs & Supabase
Simple Streamlit in Snowpark Container Services
Draw a mockup and generate html for it
node-http-proxy plus a REST API
llama3 implementation one matrix multiplication at a time
LLM based autonomous agent that conducts deep local and web research on any topic and generates a long report with citations.
Ready to use docker images for Gitpod workspaces
Codespaces but open-source, client-only and unopinionated: Works with any IDE and lets you use any cloud, kubernetes or just localhost docker.
CLI for adding customizable and re-usable chat components to your applications. Build beautiful chat interfaces in minutes.
A starting point for building an iOS, Android, and Progressive Web App with Tailwind CSS, React w/ Next.js, Ionic Framework, and Capacitor
π streaming logs to the browser
π» Bear necessities for state management in React
A set of beautifully-designed, accessible components and a code distribution platform. Works with your favorite frameworks. Open Source. Open Code.
AI companions with memory: a lightweight stack to create and host your own AI companions
Semantic cache for LLMs. Fully integrated with LangChain and llama_index.
Retrieval Augmented Generation (RAG) framework and context engine powered by Pinecone
A super fast Graph Database uses GraphBLAS under the hood for its sparse adjacency matrix graph representation. Our goal is to provide the best Knowledge Graph for LLM (GraphRAG).
Extract core logic from qdrant and make it available as a library.
High Available Datastore for Kubernetes
Code for the paper "QMoE: Practical Sub-1-Bit Compression of Trillion-Parameter Models".
Unified framework for building enterprise RAG pipelines with small, specialized models
Tools for merging pretrained large language models.
An Extensible Toolkit for Finetuning and Inference of Large Foundation Models. Large Models for All.
Accessible large language models via k-bit quantization for PyTorch.