Auto-tuned launcher for GGUF models on llama.cpp / ik_llama.cpp — OpenAI-compatible server with multi-GPU tensor-split, MoE expert placement, measured flag tuning (AI Tune), hardware-matched HuggingFace downloads, and crash recovery. An Ollama alternative for multi-GPU rigs.

A command line tool for all things anime

Pure-Go, no-cgo local LLM inference — run Gemma, Qwen, Llama and friends from safetensors or GGUF in a single static binary.

A sample architecture that mimics MoE (Mixture of Experts) using Go.

CLI for gang.moe written in Go

nlp tagger and MoE model learning

Single-kernel Codex runtime that distills skills, agents, plugins, and workflows into atomic capabilities and routes them through one main chain. / 面向 Codex 的单核运行时：将技能、Agent、插件与工作流蒸馏为原子能力，并由唯一主链统一调度。

CacheGuard（缓存卫士）— a drop-in proxy that keeps DeepSeek's server-side prefix-cache stable in front of any coding agent, so cache-hit pricing never silently breaks.