Intel MKL ERROR: Parameter 4 was incorrect on entry to SLASCL.
Closed this issue · 3 comments
xiangly55 commented
Hi, nice work. I just run your code and I come to the case that says 'Intel MKL ERROR: Parameter 4 was incorrect on entry to SLASCL.' and the loss turns to inf. This case appears sometimes, but not always. I guess this is somewhat related to the weight orthogonalization. Have you encountered this issue?
m-tassano commented
Hi, I've never encountered this problem.
If you think it could be related to the orthogonalization, you can try to
disable it with the "no orthogonalization" flag.
Le dim. 25 oct. 2020 à 13:02, xiangly95 <notifications@github.com> a écrit :
… Hi, nice work. I just run your code and I come to the case that says
'Intel MKL ERROR: Parameter 4 was incorrect on entry to SLASCL.' and the
loss turns to inf. This case appears sometimes, but not always. I guess
this is somewhat related to the weight orthogonalization. Have you
encountered this issue?
—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
<#21>, or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AHEVKQVDIQAPAL2SS4MBPLDSMQHV5ANCNFSM4S6JBSNQ>
.
xiangly55 commented
Thanks a lot, and btw could you please provide the mkl version used in your experiment?
m-tassano commented
"conda list | grep mkl" showed me version 2019.4 was installed on the
machine I used
…On Mon, 26 Oct 2020 at 03:59, xiangly95 ***@***.***> wrote:
Thanks a lot, and btw could you please provide the mkl version used in
your experiment?
—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
<#21 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AHEVKQROPQZ5GPUV276XZZLSMTQZHANCNFSM4S6JBSNQ>
.