Lean-QwenFlash: Qwen3.8-Flash-Next (125B) running at ~20 tok/s on 16gb vram and 8gb ram via direct nvme SSD offloading

🌐 链接: https://www.reddit.com/r/LocalLLaMA/comments/1xev6bo/leanqwenflash_qwen38flashnext_125b_running_at_20/
🏷️ 分组: Reddit｜AI·自托管·VPS
🕒 时间: 2026-10-25 14:41:28
Reddit
From the LocalLLaMA community on Reddit: Lean-QwenFlash: Qwen3.8-Flash-Next (125B) running at ~20 tok/s on 16gb vram and 8gb ram…
From the LocalLLaMA community on Reddit: Lean-QwenFlash: Qwen3.8-Flash-Next (125B) running at ~20 tok/s on 16gb vram and 8gb ram…
Explore this post and more from the LocalLLaMA community