Persistent error after migration to ZeroGPU: requested GPU duration (300s) is larger than maximum allowed

#11
by TastyFruitSalad - opened

Good day.

After this space has been migrated to a ZeroGPU, I keep getting same error on any attempt to generate samples:
error_message

It is reproducing on any settings given, even if they are not extreme or edge cases. For example, it can be reproduced on "Custom" tab by clicking random example button and trying to generate from it. No other adjustments to input needed to get the error message, so model used is default one: "acestep-v15-turbo".
Before migration of this space to ZeroGPU the error was not present.

Please, take a look at possible causes of this issue - hopefully it is some kind of minor problem. If any additional information on the error is needed I'll provide it.

I have that issue too.

ACE-Step org

image
I cannot reproduce your issue

Sign up or log in to comment