Samantha Andow 53faf78143 expanded weights without fast rules (#70140) ... Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/70140 [Design Doc for Expanded Weights](https://gist.github.com/samdow/fa0a164fec7963f93ff45284989cfc55) <-- gives an overview of the design for Expanded Weights Introduces the ExpandedWeights mechanism and user-facing API without any custom implemented, faster rules. - User facing API is in `_stateless.py` (with documentation) - Testing is in test_expanded_weights - The rest is the implementation of the erroring fallback + the mechanism for being able to register faster per sample grad rules. Only linear is implemented here, but they are all implemented in #70141 Test Plan: Imported from OSS Reviewed By: mikaylagawarecki Differential Revision: D34350950 Pulled By: samdow fbshipit-source-id: 69c664b0bc3dff6951358d79d7e5d94882f7aef2 (cherry picked from commit ae1620d3b6)		2022-02-22 20:35:16 +00:00
..
_internal	expanded weights without fast rules (#70140)	2022-02-22 20:35:16 +00:00
__init__.py	refactor assert_close to be more modular (#67794)	2021-11-19 12:37:16 -08:00
_comparison.py	only compare attributes for meta tensors (#72508)	2022-02-17 02:33:08 +00:00
_core.py	use torch.testing.assert_equal in TestCase.assertEqual (#67796)	2022-01-27 08:33:55 +00:00
_creation.py	disallow requires_grad=True in make_tensor for integral inputs (#67149)	2021-10-26 16:19:28 -07:00
_deprecated.py	change sparse COO comparison strategy in assert_close (#68728)	2022-01-12 06:43:50 -08:00
_legacy.py	deprecate device getter from torch.testing namespace (#63844)	2021-09-29 02:40:52 -07:00