Skip to content

Conversation

lcsfrey
Copy link
Contributor

@lcsfrey lcsfrey commented Feb 27, 2020

The current CyclicalLearningRate subclasses (TriangularCyclicalLearningRate, Triangular2CyclicalLearningRate, and ExponentialCyclicalLearningRate) are responsible for generating their own scale_fn. So they should not have scale_fn in their function arguments.

@seanpmorgan seanpmorgan added the backport r0.8 Will backport any PR merged with this label to the r0.8 branch label Feb 27, 2020
Copy link
Member

@seanpmorgan seanpmorgan left a comment

Choose a reason for hiding this comment

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

Thanks! For each of these 3 could you remove the arguments in the docstring that are not valid. For example:

scale_fn: A function. Scheduling function applied in cycle

@lcsfrey
Copy link
Contributor Author

lcsfrey commented Feb 28, 2020

Woops, my bad. It's fixed!

Copy link
Member

@seanpmorgan seanpmorgan left a comment

Choose a reason for hiding this comment

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

Thank you very much for your contribution to Addons! Failing test is unrelated.

@seanpmorgan seanpmorgan merged commit 78b0e6d into tensorflow:master Feb 28, 2020
jrruijli pushed a commit to jrruijli/addons that referenced this pull request Dec 23, 2020
* Removed erroneous scale_fn from CLR subclasses
* Updated CLR docstrings
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

backport r0.8 Will backport any PR merged with this label to the r0.8 branch cla: yes optimizers

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants