Maya-XS runs DeepSeek-V4.1-Flash in 120.6 GB, with all 384 experts resident in GPU memory on a 128 GB Ryzen AI Max.

🌐 链接: https://www.reddit.com/r/LocalLLaMA/comments/1xdbtvq/mayaxs_runs_deepseekv41flash_in_1206_gb_with_all/
🏷️ 分组: Reddit｜AI·自托管·VPS
🕒 时间: 2026-10-23 19:49:10
Reddit
From the LocalLLaMA community on Reddit: Maya-XS runs DeepSeek-V4.1-Flash in 120.6 GB, with all 384 experts resident in GPU memory…
From the LocalLLaMA community on Reddit: Maya-XS runs DeepSeek-V4.1-Flash in 120.6 GB, with all 384 experts resident in GPU memory…
Explore this post and more from the LocalLLaMA community