AIPython★ 238⑂ 38 forksMIT
Yamz-Labs/kyojin
AMD Strix Halo Inference Engine
Kyojin runs ExLlamaV3 on AMD Strix Halo (ROCm gfx1151) to execute 300B‑parameter Mixture‑of‑Experts models on a single 128 GB mini PC.
// KEY FEATURES
- Optimized ExLlamaV3 kernels for AMD ROCm gfx1151 architecture.
- Supports 300B‑class Mixture‑of‑Experts models within 128 GB RAM.
- Enables low‑latency inference on a compact mini‑PC workstation.
New repositories every 30 minutes
REDDYX AI scans GitHub 24/7 and ships the best AI/ML/Web3 projects to Telegram.
Join on Telegram