Closed ecm200 closed 1 year ago
Hi @ecm200, we have confirmed the bug and so sorry for the inconvenience. It seems the problem on the latest wheels. Please temporarily use the old version or install from source pip install abess==0.4.6 --no-binary=abess
.
Thank you for the feedback! We will fix the problem soon.
No worries at all, no need to apologize. Glad I could help pointing something out.
As I did above, I confirmed that v0.4.5
is indeed working as expected.
Hi @ecm200, we have confirmed the bug and so sorry for the inconvenience. It seems the problem on the latest wheels. Please temporarily use the old version or install from source
pip install abess==0.4.6 --no-binary=abess
.Thank you for the feedback! We will fix the problem soon.
Thanks for the workaround, I can confirm that I've built the package from source, and it is now running successfully using v0.4.6
.
It is a weird bug... I have checked the wheels built by our PyPI workflow and the problem exists even for:
cibuildwheel
, e.g. v2.12.0
;However I haven't reproduced it in a simpler demo package. One solution is to revert the version of cibuildwheel
to v2.3.1
, which used to build ABESS v0.4.5.
@Mamba413 Do u think we should revert the version and publish a workable version first?
@oooo26 ,thanks! Just revert the version of cibuildwheel
. Also, shall we report this bug to the team for cibuildwheel
?
@oooo26 ,thanks! Just revert the version of
cibuildwheel
. Also, shall we report this bug to the team forcibuildwheel
?
Sure, I will revert it. And actually I tend to try making a simpler demo before reporting.
Describe the bug
When installing
v0.4.6
any calls to theabess
library causes a segmentation fault.The same simple script with
v0.4.5
results in a normal execution.Code for Reproduction
Using
v0.4.6
results in the following error:Expected behavior
I expected the predictions to be printed without segmentation fault.
Example run with
v0.4.5
:Desktop (please complete the following information):