Decouple/use older CUDA models with 0.13? #9552
Unanswered
Gwindalmir
asked this question in
Question
Replies: 1 comment
-
So, through experimentation, I found one tag, |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Is there a way to decouple the tensorrt models from frigate, or use the older versions with the new 0.13?
My distribution (OpenMediaVault 6, based on Debian Bullseye) is stuck at nVidia driver version 470. I checked the next distro version, bookworm, and it only goes to 525.
I upgraded my docker without thinking, and my DB upgraded and now I can't go back, and I can't go forward.
Frigate is basically dead now, at least if I want to keep using object detection. :-(
EDIT: As an aside, is there a 0.12 tag? Github doesn't seem to allow filtering or sorting of the image tags, so all I see are a few
0.13
s and 30+ pages ofdev-<hash>
?I'm using
ghcr.io/blakeblackshear/frigate:stable-tensorrt
for lack of a more precise version. I like to specify major/minor version explicitly so I can avoid unfortunate upgrade issues like this until I can spend time to work on them.Beta Was this translation helpful? Give feedback.
All reactions