Vertex AI tunning chat-bison exceeded quota exception

Google Cloud Build with AI Custom ML & MLOps
1

I’ve been trying to run tune jobs in both us-central1 and europe-west-4, but I keep getting exception about quotas. I submitted an increase of quota request for us-central1 for the metrics:

“Restricted image training Nvidia A100 80GB GPUs per region”

and

“Restricted image training CPUs for A2 CPU types per region”

and they were increased from 0 to 10 but I still get exception that the quota for “Restricted image training CPUs for A2 CPU types per region” is exceeded. Should I request a further increase or I am doing something wrong?

2

Hi @vesofielv ,

Welcome and thank you for reaching out to our community for clarifications.

I understand that you are having concerns with your quotas and limits, requesting to increase them is one sure way to go about it but please check your quota page first, make sure that you are indeed at your limit prior requesting more.

You might also want to visit your logs and try to understand what’s depleting your Vertex AI resources.

Adding some related articles that I found online, it may be helpful.

Hope this helps.