Starred repositories
The best DeepSeek Harness plugin for context insight and management, with context dashboard / browser / sidebar and context command, for context statistics, composition, breakdown, evolution detail…
Agent Skills for NVIDIA products — install into Claude Code, Codex, and other coding agents to run Physical AI, robotics, simulation, CUDA, and RAG workflows end to end.
A Redmine plugin that automatically detects and switches between Textile and Markdown.
GLM-5.3-Flash EXL3 on 2x RTX PRO 6000
ReAlign models instantly, on demand. Per-request alignment for verified users and channels, or Always On. Use directly or with Hermes, OpenClaw and custom harnesses.
Serve large Qwen models fast on the GPUs you actually own. Qwen3.8-27B on a single 24 GB card with vLLM: 127 tok/s single-user (381 when the answer quotes the prompt), ~1,035 tok/s at 64 concurrent…
DeepSeek-v4-Flash 0731 recipe for 2x DGX Sparks
DeepSeek-V4.1-Flash EXL3 on one 24 GB RTX 3090 + DDR4 + NVMe: staged-DMA prefill, AVX2 CPU expert tier, elastic VRAM expert cache, Engram on disk, OpenAI API
Your GeForce runs AI at half speed. This fixes it for 20/30/40 series cards. Up to 2× FP16 / BF16 / TF32 tensor throughput
Strata fork: Qwen3.8-Flash-Next 125B MoE in NVFP4 on one RTX 20-50 card (12 GB+) and 64 GB of RAM or more. Setup installs our GPTQ quants (huihui abliterated, OrcaRouter) from Hugging Face. CPU exp…
DeepSeek Harness: Everything is a Plugin.
AI生成の日本語を自然な日本語へ推敲するAgent Skill / Agent Skill for Refining AI-Generated Japanese into Natural Japanese
Best practices, reference architectures, and examples for distributed AI training and inference on AWS.
A Collection of BM25 Algorithms in Python
LLM Inference Engine for Metal, CUDA and Vulkan.
📑 PageIndex: Document Index for Vectorless, Reasoning-based RAG
OpenShell is the safe, private runtime for autonomous AI agents.
VoiceStudio is the open-source, fully-local ElevenLabs alternative — voice cloning, voice design, video dubbing, dictation, transcription & audiobook creation in 646 languages.
Qwen3.8-Flash-Next on any consumer hardware: one-click install for Windows / Linux. Strata inference engine, OpenAI/Anthropic API on localhost, optional image input.
Tested vLLM recipe: Qwen3.8-Flash-Next (NVIDIA NVFP4) on 2x RTX PRO 6000, stock vLLM v0.30.0, TP=2, MTP n=3
verl/HybridFlow: A Flexible and Efficient RL Post-Training Framework
A coding-agent skill for multi-phase security audits with independently verified, machine-readable findings
Serving Qwen3.8-Flash-Next (Qwen4 preview) on 2x RTX PRO 6000 Blackwell SM120 with vLLM: FP8 & NVFP4, MTP spec decoding, PLE offload, 262k context. Recipe, pitfalls, benchmarks.
A unified library of SOTA model optimization techniques like quantization, distillation, pruning, neural architecture search, speculative decoding, etc. It compresses deep learning models for downs…
Fair-code workflow automation platform with native AI capabilities. Combine visual building with custom code, self-host or cloud, 400+ integrations.
A private local AI copilot for searching, understanding, and organizing your files
A privacy-first app that strips AI watermarks from content you own.
FreeToken brings datacenter-scale model serving to your desktop. Run massive models locally, fast and efficiently.
Skills for threat modeling, scanning, triage, patching, plus an autonomous scanning harness you can /customize