This repository has been archived by the owner on May 1, 2023. It is now read-only.
Refine pruning logic #222
Merged
Refine pruning logic #222
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add finer control over the pruning logic, to accomodate more pruning use-cases. The full description of the new logic will be available as documentation. In this commit: - Added a new callback to the CompressionScheduler: ```compression_scheduler.before_parameter_optimization``` which is invoked after the gradients are are computed, but before the weights are updated by the optimizer. - We skip the first mini-batch of the first epoch (global_mini_batch_id == 0) because of PyTorch's SGD implementation (details later). - We register to the parameter backward hook in order to mask the gradients. This gives us finer control over the parameter updates.
DropFilter is a method to regularize networks, and it can also be used to prepare a network for permanent filter pruning.
'pages' was deprecated and replaced by 'nav'. Also removed unused earlyexit.md file.
A new version of mkdocs moved files around - creating and deleting files.
michaelbeale-IL
pushed a commit
that referenced
this pull request
Apr 24, 2023
Add finer control over the pruning logic, to accommodate more pruning use-cases. The full description of the new logic is available in the updated [documentation of the CompressionScheduler](https://nervanasystems.github.io/distiller/schedule.html#pruning-fine-control), which is also part of this PR. In this PR: * Added a new callback to the CompressionScheduler: compression_scheduler.before_parameter_optimization which is invoked after the gradients are are computed, but before the weights are updated by the optimizer. * We provide an option to mask the gradients, before the weights are updated by the optimizer. We register to the parameter backward hook in order to mask the gradients. This gives us finer control over the parameter updates. * Added several DropFilter schedules. DropFilter is a method to regularize networks, and it can also be used to "prepare" a network for permanent filter pruning. *Added documentation of pruning fine-control
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add finer control over the pruning logic, to accomodate more pruning
use-cases.
The full description of the new logic will be available as documentation.
In this commit:
Added a new callback to the CompressionScheduler:
compression_scheduler.before_parameter_optimizationwhich is invokedafter the gradients are are computed, but before the weights are updated
by the optimizer.
We skip the first mini-batch of the first epoch (global_mini_batch_id == 0)
because of PyTorch's SGD implementation (details later).
We register to the parameter backward hook in order to mask the gradients.
This gives us finer control over the parameter updates.
Add several DropFilter schedules
DropFilter is a method to regularize networks, and it can also be
used to prepare a network for permanent filter pruning.
Add documentation of pruning fine-control