4 Replies
because in runpod its not supported, but if you want to put some suggestions feel free to put it in #🧐｜feedback
usually they assign it, higher type of gpu = higher cpu model
@Jason Thank you Jason
The last question is, is it possible to specify the worker's RAM?
not directly modifiable, i guess higher type of gpu has more ram
and so if you use multiple gpu in the endpoint it will have multiples of the ram too
the hardware specs are varying between data-centers on GPU deployments
Check your workers tab and see which region it runs, remove that insufficient specs region from Edit>Advanced tab and see what happens. it will assign another worker from another region check it's specs then