Heitor Schueroff de Souza ffc3da35f4 Don't materialize output grads (#41821) ... Summary: Added a new option in AutogradContext to tell autograd to not materialize output grad tensors, that is, don't expand undefined/None tensors into tensors full of zeros before passing them as input to the backward function. This PR is the second part that closes https://github.com/pytorch/pytorch/issues/41359. The first PR is https://github.com/pytorch/pytorch/pull/41490. Pull Request resolved: https://github.com/pytorch/pytorch/pull/41821 Reviewed By: albanD Differential Revision: D22693163 Pulled By: heitorschueroff fbshipit-source-id: a8d060405a17ab1280a8506a06a2bbd85cb86461		2020-08-11 04:27:07 -07:00
..
api	Don't materialize output grads (#41821)	2020-08-11 04:27:07 -07:00
common
dist_autograd	Release DistAutogradContainer context for each dist_autograd test case (#38711)	2020-06-18 20:58:55 -07:00
jit	Fix lite trainer unit test submodule registration (#42714)	2020-08-07 18:26:56 -07:00
rpc	test_cpp_rpc: Build test_e2e_process_group.cpp only if USE_GLOO is true (#42836)	2020-08-10 16:54:26 -07:00
tensorexpr	[NNC] Remove VarBinding and go back to Let stmts (#42634)	2020-08-07 10:50:38 -07:00
__init__.py	remediation of S205607	2020-07-17 17:19:47 -07:00