Skip to content
This repository has been archived by the owner on Sep 18, 2024. It is now read-only.

Speedup supports channel pruning #2906

Merged
merged 119 commits into from
Oct 9, 2020

Conversation

chicm-ms
Copy link
Contributor

No description provided.

chicm-ms added 30 commits August 6, 2019 11:19
Filter prune algo implementation (microsoft#1655)
document the dispatcher working dir (microsoft#1866)
Parameters
----------
module_masks : ModuleMasks
The ModuleMasks instance of the batchnorm2d
Copy link
Contributor

Choose a reason for hiding this comment

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

batchnorm2d -> linear

"""
assert 'weight' in mask

nonzero_index = torch.nonzero(mask['weight'].sum(0), as_tuple=True)[0]
Copy link
Contributor

Choose a reason for hiding this comment

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

better to make it clear that this function only deals with 0th dimension, using for example asserts

@ultmaster ultmaster merged commit 8bc74a2 into microsoft:master Oct 9, 2020
@chicm-ms chicm-ms deleted the speedup-channel-pruning branch October 19, 2020 15:59
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants