This repository has been archived by the owner on Sep 18, 2024. It is now read-only.
update lottery ticket pruner based on refactored compression code #1989
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
Contributor
QuanluZhang
commented
Feb 3, 2020
•
QuanluZhang
commented
- update lottery ticket pruner based on refactored compression code
- fix bug: allow users to specify device of the model when exporting mask and model, used to place dummy input tensor for exporting onnx model.
added 2 commits
February 1, 2020 14:44
Cjkkkk
reviewed
Feb 4, 2020
| self._unwrap_model() | ||
| self.bound_model.load_state_dict(model_state) | ||
| self._wrap_model() | ||
| else: |
There was a problem hiding this comment.
when will is_wrapped be false ? Before calling compress method?
There was a problem hiding this comment.
yes.
it is possible that users call load_model_state before compress, we need an approach to check whether the model has been wrapped or not. any suggestion?
Cjkkkk
reviewed
Feb 4, 2020
| @@ -291,7 +294,10 @@ def _wrap_modules(self, layer, config): | |||
| the configuration for generating the mask | |||
| """ | |||
| _logger.info("compressing module %s.", layer.name) | |||
| return PrunerModuleWrapper(layer.module, layer.name, layer.type, config, self) | |||
| wrapper = PrunerModuleWrapper(layer.module, layer.name, layer.type, config, self) | |||
| assert hasattr(layer.module, 'weight') | |||
There was a problem hiding this comment.
why not using check_weight method?
There was a problem hiding this comment.
where is this method defined, and how to use it?
There was a problem hiding this comment.
Cjkkkk
reviewed
Feb 4, 2020
| return PrunerModuleWrapper(layer.module, layer.name, layer.type, config, self) | ||
| wrapper = PrunerModuleWrapper(layer.module, layer.name, layer.type, config, self) | ||
| assert hasattr(layer.module, 'weight') | ||
| wrapper.to(layer.module.weight.device) |
There was a problem hiding this comment.
if the module is placed on GPU, buffers in the wrapped module are put on CPU, thus, I need to move the whole wrapper to the device that the module is placed
chicm-ms
approved these changes
Feb 5, 2020
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.