Replies: 4 comments
-
Just build it yourself. Here is part of the Dockerfile I'm using to build it for amd64 and arm64 only +cpu but I guess it won't be hard to adapt to use cu111.
|
Beta Was this translation helpful? Give feedback.
-
You can try my repository for building packages and PyPI simple index and see if it works for you: |
Beta Was this translation helpful? Give feedback.
-
@ssoooong do we have any solution for this, or still we need to wait for it; |
Beta Was this translation helpful? Give feedback.
-
Why was this never answered, python 3.10 is over 3 years old at this point? |
Beta Was this translation helpful? Give feedback.
-
Hello.
I have a question about detectron2 version.
I understand that detectron2 supports up to torch 1.10.
Is there any way to get detectron2 from python 3.10?
(We estimate python 3.10 supports from torch 1.11)
(My Environment: CUDA version 11.1, python 3.10, GPU)
After searching, I found it difficult to use detectron2 in python 3.10, but I am asking if there is a way.
(case 1) Run well when python 3.8
pip 설치 torch==1.9.0+cu111 torchvison==0.10.0+cu111 -f https://download.pytorch.org/whl/torch_stable.html
pip install detectron2 -f https://dl.fbaipublicfiles.com/detectron2/wheels/cu111/torch1.9/index.html
(case 2) Python 3.10
Not executed at execution of the code above, tried running with a different version, but not executed
(Error when torch version is set to 1.11)
Thank you!
Beta Was this translation helpful? Give feedback.
All reactions