Update CUDA graph on scale change plus clear nodes/params (#9550)
* Avoid using saved CUDA graph if scale changes and reset nodes/params on update Fixes https://github.com/ggerganov/llama.cpp/issues/9451 * clear before resize
A
agray3 committed
41f477879fd5ccc31211634292e8e293ec700e85
Parent: e948a7d
Committed by GitHub <noreply@github.com>
on 9/21/2024, 12:41:07 AM