-
Notifications
You must be signed in to change notification settings - Fork 52
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
Add a tutorial for the new Warm-start feature #1029
base: develop
Are you sure you want to change the base?
Conversation
ec69241
to
cec5d5a
Compare
8c46f4b
to
403c191
Compare
storage=current_experiment_storage, | ||
executor=SingleExecutor(), | ||
max_trials=current_experiment_n_runs, | ||
# NOTE: This n_initial_points is changed slightly, since it also counts the trials from |
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.
It's interesting to make this fair comparison. In practice however would not need to force random points first at the previous results already provide a good basis to start bayesian optimization.
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.
Ok, should I keep this then? or not?
Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Use more samples for training of Profet model Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Warm-start example is a bit better Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Add more comments in the notebook Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Use Pytorch mnist/fashionmnist instead, bad Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> [dirty] Trying to find good settings Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Update example a bit, seems to work much better! Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Try to make the plot appear Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Add the pickleddb files for the databases Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Rename things, image still does show :( Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Add a thumbnail (image shows up now!) Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Removed the torch multiprocessing context setting Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Make path relative so CI will work too Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Add tutorial to index, add text to describe plot Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Move databases to avoid any possible abs paths Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Attempt to fix KnowledgeBase config merge issue Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Update the pickle files for the example Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Seed all the randomness, update pickles Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Fix issue with portability of PickledDB objects Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com> Update the pickleddb files for the example Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com>
44d1eb3
to
1084f01
Compare
Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com>
Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com>
Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com>
Signed-off-by: Fabrice Normandin <fabrice.normandin@gmail.com>
Adds an example of how to use the new warm-start feature.
Currently uses the Profet tasks, which require quite a bit of compute to run the first time (GPUS also help, but aren't required).
TODOS:
Current results: