search

google-research / tuning_playbook

A playbook for systematically maximizing the performance of deep learning models.
Other
27.02k stars 2.24k forks source link

Consider recommending scipy.stats.qmc for quasi-Monte Carlo sequence generation #2

Open shoyer opened 1 year ago

shoyer commented 1 year ago

e.g., https://docs.scipy.org/doc/scipy/reference/generated/scipy.stats.qmc.Halton.html#scipy.stats.qmc.Halton

This is likely to be better maintained than the MLCommons code.

varungodbole commented 1 year ago

@znado could you please take a look when you get a chance?

varungodbole commented 1 year ago

Thanks for the recommendation @shoyer!!

varungodbole commented 7 months ago

Hey @shoyer that's a great idea! Could you please make a pull request?

shoyer commented 7 months ago

You're welcome to act on this (or not), but I'm not planning to submit a PR.

On Sun, Mar 24, 2024 at 8:46 PM Varun Godbole @.***> wrote:

Assigned #2 https://github.com/google-research/tuning_playbook/issues/2 to @shoyer https://github.com/shoyer.

— Reply to this email directly, view it on GitHub https://github.com/google-research/tuning_playbook/issues/2#event-12229039240, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAJJFVRTMLCXDEAFHJMOTDTYZ6MYBAVCNFSM6AAAAAAUAWET2WVHI2DSMVQWIX3LMV45UABCJFZXG5LFIV3GK3TUJZXXI2LGNFRWC5DJN5XDWMJSGIZDSMBTHEZDIMA . You are receiving this because you were assigned.Message ID: @.*** com>