-
Notifications
You must be signed in to change notification settings - Fork 5.7k
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
[Serve] Only install dataclasses on Python 3.6 #10936
[Serve] Only install dataclasses on Python 3.6 #10936
Conversation
dup image hash check build ray-ml uninstall dataclasses in the last docker image build whl remove GPU tag and dataclasses uninstall patch ray-project#10909 ray-project#10936
dup image hash check build ray-ml uninstall dataclasses in the last docker image build whl remove GPU tag and dataclasses uninstall patch ray-project#10909 ray-project#10936
Would say this is urgent - currently the landing page tune example will not work because of this (on a distributed cluster). |
I am still encountering this issue as of dacite 1.6.0, with python 3.10 and using the mamba package manager. The issue occurs with regular conda as well. Abridged output of the install command:
|
@jhayes-py It sounds like your issue is unrelated to Ray. The command |
Why are these changes needed?
Before this PR, running the following script
on the product (a linux machine running python 3.7) would cause the following error:
A couple sources (HumanCellAtlas/metadata-api#95, konradhalas/dacite#48) suggest this error comes from having the
dataclasses
package (which is a backport for python 3.6) installed alongside python 3.7 (or presumably 3.8). Indeed, runningpip uninstall dataclasses
fixes the error, and reinstallingdataclasses
reintroduces the error.The solution is to only include
dataclasses
as a dependency if the python version is less than 3.7.Related issue number
Checks
scripts/format.sh
to lint the changes in this PR.