R_REDDYX.XYZ
AIC++★ 1,176⑂ 206 forksMIT

gufo-org/gufo

GUFO: INFERENCE ENGINE FOR STRIX HALO

gufo — inference engine for AMD Strix Halo in C++. Qwen Flash Next Q4_K_XL delivers 59.41 tok/s (1 user) and 157.22 tok/s (8 users). Qwen27B Q4_K_XL with DFlash2 gives 70.56 tok/s. 210 stars on GitHub.

// KEY FEATURES

  • Runs LLM on AMD Strix Halo via DFlash2
  • Qwen Flash Next Q4_K_XL: 59.41 tok/s (1 user)
  • Scales to 157.22 tok/s for 8 users
  • Qwen27B Q4_K_XL delivers 70.56 tok/s
#amd#llm#strix-halo
Open on GitHub →

New repositories every 30 minutes

REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.

Join on Telegram
← Full catalog·Full index

// SIMILAR REPOSITORIES

MCP SERVER FOR DAVINCI RESOLVE★ 1176Hand-drawn CSS annotations★ 1176SMALL ENGINEER ROBOT WITH AI ASSISTANT★ 1176DLSS 5 Feeder for Games★ 1177AWS SKILLS FOR AI AGENTS★ 1175Read Papers in Comfortable Chinese★ 1177

← FULL CATALOG

gufo-org/gufo — Gufo: Inference Engine For Strix Halo