Mike Manh

Community recipes for serving LLMs on RTX 3090/4090/5090 CUDA gpus. Multi-engine (vLLM, llama.cpp, ik_llama) and model-agnostic. Currently shipping Qwen3.6-27B Qwen3.6 35B Gemma 4 26B Gemma 4 31B configs for 1× and 2× cards.

A high-throughput and memory-efficient inference and serving engine for LLMs

Convert PDF to markdown + JSON quickly with high accuracy

a fork of llm-proxy to do WOL and sleep on a server

DDD Printable Wall Control System, 3d printable organization solutions for Wall Control pegboards

A generative world for general-purpose robotics & embodied AI learning.