Skip to content

Conversation

@motus
Copy link
Member

@motus motus commented Aug 6, 2024

Experimental implementation of quantization of tunable parameters

Closes #803

@motus motus added ready for review Ready for review mlos-bench labels Aug 6, 2024
@motus motus self-assigned this Aug 6, 2024
@motus motus requested a review from a team as a code owner August 6, 2024 00:02
"type": "int",
"default": 2000000,
"range": [0, 1000000000],
"quantization": 1000000,
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Probably need tests to check this now, right?

Copy link
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

absolutely. I'll add a ton more unit tests to this PR shortly

else:
raise TypeError(f"Invalid Parameter Type: {tunable.type}")

if tunable.quantization:
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
if tunable.quantization:
if tunable.quantization:
# Temporary workaround to dropped quantization support in ConfigSpace 1.0
# See Also: https://github.com/automl/ConfigSpace/issues/390

Copy link
Contributor

@bpkroth bpkroth left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Waiting on tests and then let's merge this.

@motus
Copy link
Member Author

motus commented Aug 8, 2024

superseded by #835

@motus motus closed this Aug 8, 2024
@motus motus deleted the sergiym/tunable/quantization branch August 15, 2024 22:07
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Implement quantization for ConfigSpace >= 1.0

2 participants