Stars
Fast and memory-efficient exact kmeans
Real-ESRGAN on Apple Silicon via CoreML. 1.6x faster than MLX.
All Jina AI APIs as Unix CLI commands. Search, read, embed, rerank - with pipes.
AI agents running research on single-GPU nanochat training automatically
Pure MLX implementations of UMAP, t-SNE, PaCMAP, TriMap, DREAMS, CNE, MMAE, and NNDescent for Apple Silicon. Metal GPU for computation and video rendering.
给 AI 一篇文章，自动生成一款可以玩的 RPG Maker MZ 游戏
Training neural networks on Apple Neural Engine via reverse-engineered private APIs
t-SNE in pure MLX for Apple Silicon. 12x faster than sklearn. FFT-accelerated (FIt-SNE) for O(n) scaling.
PaCMAP in pure MLX for Apple Silicon. Pure GPU, no scipy/numba.
UMAP in pure MLX for Apple Silicon. 30x faster than umap-learn.
reading the undocumented mems accelerometer + gyroscope on apple silicon macbooks via iokit hid
Embedding Inversion via Conditional Masked Diffusion: recover original text from embedding vectors using parallel denoising. Live demo + training pipeline + technical report.
MLX Local Serving (MLS) - Unified ASR, TTS, and Translation on Apple Silicon
Algorithm powering the For You feed on X
Browser automation CLI for AI agents
~950 line, minimal, extensible LLM inference engine built from scratch.
Claude Code skill implementing Manus-style persistent markdown planning — the workflow pattern behind the $2B acquisition.
MLX-VLM is a package for inference and fine-tuning of Vision Language Models (VLMs) on your Mac using MLX.
TradingAgents: Multi-Agents LLM Financial Trading Framework
Train embedding and reranker models for retrieval tasks on Apple Silicon with MLX
Renderer for the harmony response format to be used with gpt-oss