-
Notifications
You must be signed in to change notification settings - Fork 6.9k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Conda] pillow 9.5.0 incompatible with any torchvision version due to switch from jpeg to libjpeg-turbo #7660
Comments
There are no pillow-9.5.0 on Anaconda https://anaconda.org/anaconda/pillow is there? |
The default channels are currently ok. But the conda-forge channel has already been updated. Since the conda-forge channel is more widely adopted (29M downloads) compare to anaconda (1.8M downloads), the impact is there. |
We're seeing this impact our conda environment where we cannot upgrade dependencies b/c we have deps from I'm wondering if we could unblock this scenario by removing |
Quick update on this: for conda builds, we're going to try to make torchvision depend on |
@NicolasHug Thanks for moving this along! Do you have a timeline for when a new build will be pushed to conda? |
I believe the next release is planned for early October |
Hi all, this should be fixed now with the 0.16 release as we now ship with libjpeg-turbo. I'll close the issue, but LMK if there are still problems. Thanks |
🐛 Describe the bug
Pillow no longer compatible with
torchvision
starting from version 9.5.0 due to issues discussed in conda-forge/libjpeg-turbo-feedstock#28The core issue was that the
jpeg
conda package was "deprecated" by the conda-forge community and they want all users ofjpeg
to migrate tolibjpeg-turbo
.Versions
all torchvision conda package
The text was updated successfully, but these errors were encountered: