[bug]: Stopped working after video card replacement

[goodwin74]

Is there an existing issue for this?

• [X] I have searched the existing issues

OS

Linux

GPU

cuda

VRAM

6GB

What happened?

When installing for the first time, I had a GTX 960 4gb video card and everything worked. The other day I installed gtx 1660 6gb and now, when generating, I always get an error that there is not enough video card memory. Why does this happen?

p/s/ free gpu mem doesn't help. Moreover, 960 worked without --free_gpu_mem

Screenshots

No response

Additional context

No response

Contact Details

No response

[kriptolix]

Hi, it's been a while since I last checked this, but there was a compatibility issue with the gtx 16XX series. According to this issue https://github.com/invoke-ai/InvokeAI/issues/775 the problem seems to have been resolved, but there may have been some sort of regression.

What version of invoke are you using?

[goodwin74]

Hi, it's been a while since I last checked this, but there was a compatibility issue with the gtx 16XX series. According to this issue #775 the problem seems to have been resolved, but there may have been some sort of regression.

What version of invoke are you using?

InvokeAI 2.2.5.

Yes, apparently this is somehow connected with the 16x version of the video card, since the previous 960 gtx worked fine, although there was less memory there. Moreover, I note that the InvokeAI version for Windows works fine. Ubuntu also has other SD toolkits installed. For example "automatic" it works well on both Ubuntu and Windows. I think this is related to how InvokeAI works in Ubuntu.

[github-actions[bot]]

There has been no activity in this issue for 14 days. If this issue is still being experienced, please reply with an updated confirmation that the issue is still being experienced with the latest release.