This repository has been archived by the owner on Sep 18, 2024. It is now read-only.
Bug-fixed for _compute_hessian in Pytorch NAS Darts #3058
Merged
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Outdated suggestions cannot be applied.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.
In Issues#3044,
I report a bug about Pytorch darts in train.py, which is located at
”nni\algorithms\nas\pytorch\darts\trainer.py“
In the darts trainer.py, which function is _compute_hessian, shows :
"hessian = (dalpha { L_trn(w+, alpha) } - dalpha { L_trn(w-, alpha) }) / (2*eps)"
but the below is written by this :
hessian = [(p - n) / 2. * eps for p, n in zip(dalpha_pos, dalpha_neg)]
I change it to
hessian = [(p - n) / (2. * eps) for p, n in zip(dalpha_pos, dalpha_neg)]
Good luck!