This repository has been archived by the owner on Sep 18, 2024. It is now read-only.
[Retiarii] Weight-sharing trainers #3137
Merged
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
QuanluZhang
reviewed
Dec 5, 2020
examples/nas/enas/micro.py
Outdated
Comment on lines
55
to
57
| sampled = self.input_choice.sampled | ||
| return out, torch.tensor([i == sampled or (isinstance(sampled, list) and i in sampled) | ||
| for i in range(len(self.input_choice))], dtype=torch.bool) |
There was a problem hiding this comment.
this logic is complicated, why?
QuanluZhang
reviewed
Dec 5, 2020
| num_epochs = args.epochs or 150 | ||
| mutator = enas.EnasMutator(model, tanh_constant=1.1, cell_exit_extra_step=True) | ||
| if args.v1: |
There was a problem hiding this comment.
suggest to directly abandon v1
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Outdated suggestions cannot be applied.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 PR migrates old trainers in NNI v1.x to Retiarii framework. Will support the following trainers on Retiarii:
The trainers will implement the core ideas in the original paper, instead of faithfully implementing all the details in paper.
Examples will be updated correspondingly. Expects no change except the part where trainers got initiated and trained (changed to
fit).TODO: