-
Notifications
You must be signed in to change notification settings - Fork 46
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
Henry/wallclock #488
Henry/wallclock #488
Conversation
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.
Looks good. Various comments but happy for you to merge once you've addressed them.
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.
not sure I can contribute much technically to this PR, but should we consider reflecting this functionality in a notebook somehow?
Mentioned it in the logging notebook |
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.
Looks much better. Just a couple more comments (sorry).
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.
That was lightning fast, thanks Henry! Only nitpicking at this point
trieste/bayesian_optimizer.py
Outdated
with timeit() as total_wallclock_timer: | ||
if step == 0 and fit_initial_model: | ||
with timeit() as initial_model_fitting_timer: | ||
for tag, model in models.items(): | ||
dataset = datasets[tag] | ||
model.update(dataset) | ||
model.optimize(dataset) |
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.
That looks like way too many levels of nesting, but i don't have a good alternative. @uri-granta any ideas?
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.
I think it's fine (though I used to write LISP code so maybe I'm not scared of nesting as much as I should be)
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.
this method overall looks pretty scary to me. half-way through it's hard to figure out at which level you are
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.
LGTM
As requested by @apaleyes and @vpicheny , I have added in some wallclock timing to our Tensorboard logging.
Its very self explanatory.