-
Notifications
You must be signed in to change notification settings - Fork 260
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
Fix module import #698
Fix module import #698
Conversation
These enhancements provide additional flexibility and options for implementing and experimenting with different recurrence methods in the Mamba and Jamba models, potentially improving performance and accuracy for various tasks.
- fixed functions redundant definitions - fixed Incorrect Module Import in layers.py
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thx!
Hi @markblee, Both ruomingp and jiarui-lu2 have approved the changes to the PR. Could you please take a final look when you get a chance? Thanks again for your time! Best, |
Hi @vishesh9131, I think PR check was on hold due to a required approval in CI. I've approved it and CI should be starting now |
@jiarui-lu2 Do I Need to modify
|
I don't think upgrading aqt is recommended. Minimum python version supported by |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
See comment about aqt versions.
Hi @jiarui-lu2 , I understand your concerns regarding the aqt package and its compatibility with Best regards, |
The main concern I have is not of Circle CI. axlearn repo formally supports python3.9 as stated here. Depending on a package
|
Hi @jiarui-lu2 , I understand your concern about maintaining Python 3.9 support for the axlearn repo. Given that aqt>=0.5.0 requires Python 3.10, I agree that we should close this PR until either aqt addresses this issue or axlearn is ready to upgrade its minimum Python version requirements. Best, |
According to the discussions, closing PR due to python minimum version support. We should revisit if axlearn increases its minimum python version support to > =3.10 |
Hi @jiarui-lu2, Since Axlearn now supports Python 3.10, could you reopen this closed PR? According to the previous discussions, the closure was due to the minimum Python version support issue. Now that the support has been updated, it seems like a good time to revisit. Best regards, |
Pull Request: Fix Incorrect Module Import in
layers.py
Summary
This pull request addresses an issue with an incorrect module import in the
axlearn/common/quantized_dot_general/layers.py
file. The import and usage of theContext
class from theaqt_dot_general
module have been corrected to use the appropriateContext
class from theaqt_utils
module.Changes
1. Import Correction:
aqt.jax.v2.aqt_dot_general
toaqt.jax.v2.utils
and aliased it asaqt_utils
.2. Context Class Usage:
Context
class to useaqt_utils.Context
instead ofaqt_dot_general.Context
.Code Changes
1. Importing aqt_utils
2. Taken Reference from the official README of aqt
to
Testing
AttributeError
and the code runs successfully without any issues.