[primTorch] Implement batch, group, and instance norm references#81191
Closed
rdspring1 wants to merge 34 commits intopytorch:masterfrom
Closed
[primTorch] Implement batch, group, and instance norm references#81191rdspring1 wants to merge 34 commits intopytorch:masterfrom
rdspring1 wants to merge 34 commits intopytorch:masterfrom
Conversation
Contributor
🔗 Helpful links
❌ 5 New FailuresAs of commit 393c6c4 (more details on the Dr. CI page): Expand to see more
🕵️ 5 new failures recognized by patternsThe following CI failures do not appear to be due to upstream breakages
|
Chillee
reviewed
Jul 12, 2022
Use prim.squeeze which supports axis list argument
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/81191
Note: Links to docs will display an error until the docs builds have been completed. ❌ 7 Failures, 1 PendingAs of commit ac0598c: The following jobs have failed:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
mruberry
reviewed
Sep 26, 2022
| return flip(a, (0,)) | ||
|
|
||
|
|
||
| def _repeat_if_defined(a: Optional[Tensor], sizes: int): |
Collaborator
There was a problem hiding this comment.
Add a comment for this (private) function
mruberry
reviewed
Sep 27, 2022
| double momentum, double eps) { | ||
|
|
||
| using accscalar_t = at::acc_type<scalar_t, false>; | ||
| TORCH_CHECK(input.dim() >= 1, |
Collaborator
There was a problem hiding this comment.
Since batch_norm_cpu_update_stats_template is not a user-facing function this should be an internal assertion -- maybe the check could occur earlier?
mruberry
reviewed
Sep 27, 2022
| auto out = input.clone(); | ||
| if (weight.defined()) out = out * weight[0]; | ||
| if (bias.defined()) out = out + bias[0]; | ||
| if (weight.defined() && weight.numel() > 0) { |
Collaborator
There was a problem hiding this comment.
This will need a comment for what the semantics are when weight and/or bias have no elements
mruberry
reviewed
Sep 27, 2022
| auto num_features = self.sizes()[1]; | ||
| auto options = self.options().dtype( | ||
| at::toAccumulateType(self.scalar_type(), /*is_cuda=*/self.is_cuda())); | ||
| auto save_mean = at::empty({num_features}, options); |
Collaborator
There was a problem hiding this comment.
Is it really OK to return tensors with random values here?
pytorchmergebot
pushed a commit
that referenced
this pull request
Oct 3, 2022
This PR adds nvFuser's implementation for batch_norm as there's no reference yet (#81191) and no in-place copy support (#84545). Pull Request resolved: #85562 Approved by: https://github.com/kevinstephano, https://github.com/ngimel
Contributor
|
/easycla As part of the transition to the PyTorch Foundation, this project now requires contributions be covered under the new CLA. See #85559 for additional details. This comment will trigger a new check of this PR. If you are already covered, you will simply see a new "EasyCLA" check that passes. If you are not covered, a bot will leave a new comment with a link to sign. |
|
|
mruberry
reviewed
Oct 14, 2022
| ((S, S, S), {'training': True, 'momentum': 0.5, 'eps': 0.6}), | ||
| ((3, 2, 4), {'training': False, 'momentum': -1.2}), | ||
| ((3, 1), {'training': True, 'momentum': 0.0}), | ||
| ((0,), {'training': True}), |
Collaborator
There was a problem hiding this comment.
Should these be added to instance and groupnorm sample inputs, too?
mruberry
reviewed
Oct 14, 2022
Collaborator
mruberry
left a comment
There was a problem hiding this comment.
Updates from offline review:
- let's separate the changes to existing operators and ensure those are tested properly
- also review #85960 and see if that should be part of the PR changing the ATen operations
- after the C++ changes to existing ATen operators, let's revisit the references
This was referenced Oct 17, 2022
jjsjann123
pushed a commit
to jjsjann123/nvfuser
that referenced
this pull request
Oct 29, 2022
This PR adds nvFuser's implementation for batch_norm as there's no reference yet (pytorch/pytorch#81191) and no in-place copy support (pytorch/pytorch#84545). Pull Request resolved: pytorch/pytorch#85562 Approved by: https://github.com/kevinstephano, https://github.com/ngimel
jjsjann123
pushed a commit
to jjsjann123/nvfuser
that referenced
this pull request
Nov 10, 2022
This PR adds nvFuser's implementation for batch_norm as there's no reference yet (pytorch/pytorch#81191) and no in-place copy support (pytorch/pytorch#84545). Pull Request resolved: pytorch/pytorch#85562 Approved by: https://github.com/kevinstephano, https://github.com/ngimel
pytorchmergebot
pushed a commit
that referenced
this pull request
Nov 11, 2022
Add group norm reference Split from #81191 Pull Request resolved: #87054 Approved by: https://github.com/mruberry
kulinseth
pushed a commit
to kulinseth/pytorch
that referenced
this pull request
Dec 10, 2022
Add group norm reference Split from pytorch#81191 Pull Request resolved: pytorch#87054 Approved by: https://github.com/mruberry
Contributor
|
Looks like this PR hasn't been updated in a while so we're going to go ahead and mark this as |
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
7 participants
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 References:
Depends on:
cc @ezyang @mruberry @ngimel @lezcano @fdrocha @peterbell10