Hi,
I’m curious, do you use LocalAI in NextCloud ? And if you use it’s, what’s is your configuration / model ? I’m using LocalAI, but I find the performance and the token conversion very very slow on my CPU.
1 Like
What are your system specs? How many users?
Hi,
currently, I test for my single user. On a Intel N100 / 16Go de RAM. “How you ?” do more than 2mn to be answered.
The issue would be that your hardware is simply not enough. N100 is a low end laptop cpu, but ai depends on cpu, ram and gpu power, which increases with each data set. And that is without running Nextcloud. I recommend asking the localai support forums or related groups for more info.
I had a friend build a dedicated machine for ai last month, using a second hand xeon with multiple gpus and 128gb ram for $2k or 3k. Let us know what you find out, and hopefully you find a setup that meets your budget.
1 Like