Closed SXKJames closed 1 year ago
Hi SXKJames,
Good idea! We did not try Zen + NTK. It might worth a try. Please be aware that the NTK also suffers from the BN normalization problem, so you should calibrate the gradient norm too.
Please let me know if it works!
这是来自QQ邮箱的假期自动回复邮件。 邮件我已经成功接收,我会尽快处理并答复,谢谢!
Hello MingLin,
I found your NAS approach to be very interesting.
Did you ever try to combine the ZEN-Score with the NTK-score like the TE-NAS paper did? The TE-NAS paper suggests that combining the NTK-Score with a expressivity measure would improve the overall performance.
Thanks, SXK