This repository has been archived by the owner on Sep 18, 2024. It is now read-only.
Replace layer choice with selected module after applied fixed architecture #2420
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
liuzhe-lz
reviewed
May 13, 2020
| """ | ||
| if module is None: | ||
| module = self.model | ||
| for name, mutable in module.named_children(): |
There was a problem hiding this comment.
Why not use the util function to traverse here?
Seems because setting module by full name is not trival.
liuzhe-lz
approved these changes
May 13, 2020
QuanluZhang
reviewed
May 13, 2020
| global_name = (prefix + "." if prefix else "") + name | ||
| if isinstance(mutable, LayerChoice): | ||
| chosen = self._fixed_arc[mutable.key] | ||
| if sum(chosen) == 1 and max(chosen) == 1 and not mutable.return_mask: |
There was a problem hiding this comment.
what is the meaning of not mutable.return_mask?
There was a problem hiding this comment.
If true, layer choice will return the choice vector. If we replace the layer choice module with the selected module, choice vector will be no longer available, and users’ code will crash.
|
please update doc at proper place. |
QuanluZhang
reviewed
May 18, 2020
docs/en_US/NAS/NasGuide.md
Outdated
| @@ -176,7 +176,7 @@ For example, | |||
| } | |||
| ``` | |||
|
|
|||
| After applying, the model is then fixed and ready for final training. The model works as a single model, although it might contain more parameters than expected. This comes with pros and cons. The good side is, you can directly load the checkpoint dumped from supernet during the search phase and start retraining from there. However, this is also a model with redundant parameters and this may cause problems when trying to count the number of parameters in the model. For deeper reasons and possible workarounds, see [Trainers](./NasReference.md). | |||
| After applying, the model is then fixed and ready for final training. The model works as a single model, and unused parameters and modules will be pruned. | |||
QuanluZhang
approved these changes
May 18, 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.
This should solve the problem #2401 is trying to solve.