This repository was archived by the owner on Oct 20, 2025. It is now read-only.
This repository was archived by the owner on Oct 20, 2025. It is now read-only.
Do vtensor need 64K/128K physical memory policy? #24
vAttention said that: if use 2M pageSize, 128M physical memory can be wasted per-request in the worst-case in Llama-3-8B (TP-1), but if use 64KB, 128M would be only 4M
Do vtensor have the same problem？ Will vtensor integrate 64K/128K pageSize in the future?