Skip to content

Conversation

nitnelave
Copy link
Contributor

With these additions, it becomes possible to use the solver to handle the learning rate while setting the inputs/diffs in python.

self.assertTrue((data - -0.01 * np.ones(data.shape)).max() < 1e-7)



Copy link
Contributor

Choose a reason for hiding this comment

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

Why so much space?

@longjon
Copy link
Contributor

longjon commented Jul 12, 2016

Seems like a bit of an odd case, but I guess it makes sense; one might also want to change the iter for debugging or other purposes.

@nitnelave
Copy link
Contributor Author

@longjon updated as required

return test_nets_;
}
int iter() { return iter_; }
void SetIter(int iter) { iter_ = iter; }
Copy link
Contributor

Choose a reason for hiding this comment

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

Ah, I noticed this before but neglected to comment (sorry!): this should be set_iter; trivial functions use underscore_case (like the getter above).

@nitnelave
Copy link
Contributor Author

@longjon fixed function name.

@shelhamer
Copy link
Member

Thanks for this @nitnelave, but a fuller interface has been exposed with the merge of #6238. Note that it includes c5b5b55 for the update test, with your authorship of course.

@shelhamer shelhamer closed this Jun 8, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants