projects / platform / upstream / pytorch.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 7a37e06) | patch
Allow registering an operator schema without a kernel (#18551)
authorSebastian Messmer <messmer@fb.com>
Tue, 2 Apr 2019 19:23:13 +0000 (12:23 -0700)
committerFacebook Github Bot <facebook-github-bot@users.noreply.github.com>
Tue, 2 Apr 2019 19:33:30 +0000 (12:33 -0700)
commit58f5954252ba8970aae965d4c5b010f28ad15c06
tree416aefe84a1f58edd93f0c9ca2d96dcc879cfdcdtree | snapshot
parent7a37e066e6d67ce6a79c9f80e00ec44fc90e2114commit | diff
Allow registering an operator schema without a kernel (#18551)

Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/18551

This is helpful for defining a set of operators as an interface but not adding concrete kernels just yet.
The registration logic will ensure that any other libraries that add kernels for these schemas exactly match the schema defined here.

Reviewed By: dzhulgakov

Differential Revision: D14660208

fbshipit-source-id: 7adb5a4876cff5a0ad21d92d8c450cb889f00cc3
aten/src/ATen/core/op_registration/op_registration.cpp diff | blob | history
aten/src/ATen/core/op_registration/op_registration_test.cpp diff | blob | history
Domain: Machine Learning / ML Framework;