Add LazyConvXd and LazyConvTransposeXd #47350
Conversation
|
Hi @asi1024! Thank you for your pull request and welcome to our community. We require contributors to sign our Contributor License Agreement, and we don't seem to have you on file. In order for us to review and merge your code, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA. If you have received this in error or have any questions, please contact us at cla@fb.com. Thanks! |
💊 CI failures summary and remediationsAs of commit 87f3823 (more details on the Dr. CI page): 💚 💚 Looks good so far! There are no failures yet. 💚 💚 This comment was automatically generated by Dr. CI (expand for details).Follow this link to opt-out of these comments for your Pull Requests.Please report bugs/suggestions on the GitHub issue tracker or post in the (internal) Dr. CI Users group. This comment has been revised 25 times. |
ejguan
added
module: nn
|
Can you please rebase this PR against |
ejguan
added
the
triaged
|
Rebased! |
Codecov Report
@@ Coverage Diff @@
## master #47350 +/- ##
=======================================
Coverage 80.95% 80.95%
=======================================
Files 1853 1853
Lines 199835 199879 +44
=======================================
+ Hits 161771 161813 +42
- Misses 38064 38066 +2 |
|
Thanks! |
There was a problem hiding this comment.
@albanD has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.
Summary: Some minor improvement for lazy modules introduced in #44538, #47350 and #51548. This PR mainly turn the bias to `UninitializedParameter` and instead of creating empty tensors like ```python self.bias = Parameter(torch.Tensor(0)) self.bias = UninitializedParameter() ``` I think it would be better to ```python self.register_parameter('bias', None) self.bias = UninitializedParameter() ``` In addition, I change the constructor of the `LazyBatchNorm` from ```python self.running_mean = UninitializedBuffer() ``` to ```python self.register_buffer('running_mean', UninitializedBuffer()) ``` as the original one would not change the underlying `self._buffers`. Thank you for your time on reviewing this PR :). Gently ping albanD, mruberry Pull Request resolved: #52212 Reviewed By: jbschlosser Differential Revision: D26504508 Pulled By: albanD fbshipit-source-id: 7094d0bb4fa9e2a40a07b79d350ea12a6ebfd080
…h#52212) Summary: Some minor improvement for lazy modules introduced in pytorch#44538, pytorch#47350 and pytorch#51548. This PR mainly turn the bias to `UninitializedParameter` and instead of creating empty tensors like ```python self.bias = Parameter(torch.Tensor(0)) self.bias = UninitializedParameter() ``` I think it would be better to ```python self.register_parameter('bias', None) self.bias = UninitializedParameter() ``` In addition, I change the constructor of the `LazyBatchNorm` from ```python self.running_mean = UninitializedBuffer() ``` to ```python self.register_buffer('running_mean', UninitializedBuffer()) ``` as the original one would not change the underlying `self._buffers`. Thank you for your time on reviewing this PR :). Gently ping albanD, mruberry Pull Request resolved: pytorch#52212 Reviewed By: jbschlosser Differential Revision: D26504508 Pulled By: albanD fbshipit-source-id: 7094d0bb4fa9e2a40a07b79d350ea12a6ebfd080
6 participants
This PR implements LazyConvXd and LazyConvTransposeXd based on #44538. (cc. @emcastillo and @albanD)