LocalAI icon indicating copy to clipboard operation
LocalAI copied to clipboard

Published 20 hours ago • verified publisher icon mudler

llama-cpp CPU 1500%，Very slow...What to do?

Open fairyqb opened this issue 4 months ago • 2 comments

llama-cpp CPU 1500%，Very slow... 企业微信截图_20240418225637 企业微信截图_20240418225603

my server:centos,20 core, 32GB memory

fairyqb avatar Apr 18 '24 14:04 fairyqb

Hello fairyqb,

Thank you for bringing this to our attention. It seems like you are experiencing high CPU usage with llama-cpp on your server. We understand this can be a concern and would like to help. Please provide us with more information regarding your LocalAI configuration and any commands used to run llama-cpp. It would also be helpful to get the output of the server load and the top

localai-bot avatar Apr 19 '24 08:04 localai-bot

Do you have a gpu? If so, I’m seeing the same issue, on docker, at least. https://github.com/mudler/LocalAI/discussions/2191

atljoseph avatar Apr 30 '24 13:04 atljoseph