Popular repositories Loading

1. asus-gx10-qwen35-speed-hack asus-gx10-qwen35-speed-hack Public

   4-5x faster Qwen3.5 on ASUS GX10 / DGX Spark — Hybrid INT4+FP8 + MTP via one shell script

   Shell 31 2

2. qwen3.6-27b-speedhack-gx10-dgx-spark qwen3.6-27b-speedhack-gx10-dgx-spark Public

   Qwen3.6 27B × DFlash — 30-35 tok/s on NVIDIA DGX Spark (GB10) - LLama.Cpp

   C++ 17 3

3. llama.cpp-gx10-dgx-sparks-deepseekv4 llama.cpp-gx10-dgx-sparks-deepseekv4 Public

   Forked from ggml-org/llama.cpp

   llama.cpp optimized for DeepSeek V4 Flash on NVIDIA DGX Spark / ASUS GX10 (128GB unified, 273GB/s). Fixes graph buffer overflow ▎ for 8K+ context. Achieves 6-7 tok/s on 284B IQ2_XXS.

   C++ 7

4. roocode-bridge roocode-bridge Public

   Transform Your AI Coding Experience: Claude/Antigravity as the Brains, Roo Code as the Brawn, and Telegram as a Bonus Remote Control!

   JavaScript 2

5. llama-ready llama-ready Public

   A smart Layer-7 Auto-Scaler & Lifecycle Proxy for llama.cpp. Save 100% VRAM when idle, auto-spawn multiple instances on demand, and prevent AI Agent bottlenecks with zero-config.

   Python 1

6. nemotron-solver-fleet nemotron-solver-fleet Public

   Deterministic solver fleet that cracked 9,500/9,500 NVIDIA Nemotron reasoning puzzles (Private 0.86, Silver). Solvers + CoT generators + quality gates.

   Python