Scott Wolchok e88d1c4f10 [PyTorch] Add tuple inline storage (#64066) ... Summary: Pull Request resolved: https://github.com/pytorch/pytorch/pull/64066 I noticed a bunch of time being spent heap-allocating Tuples in the unpickler. 1-, 2-, and 3-element Tuples are apparently common enough that they get their own bytecode instructions, so I decided to try also giving them their own representation. We store up to 3 IValues inline in `Tuple` rather than doing a second heap allocation for a `std::vector<IValue>`. ghstack-source-id: 140695395 Test Plan: Added automated tests for TupleElements. Pixel 3 before: https://www.internalfb.com/intern/aibench/details/761596366576284 Pixel 3 after: https://www.internalfb.com/intern/aibench/details/591414145082422 We went from 347 ms to 302 ms. Reviewed By: dhruvbird Differential Revision: D30592622 fbshipit-source-id: 93625c54c9dca5f765ef6d5c191944179cb281a8		2021-10-15 12:16:51 -07:00
..
compilation_unit.h
function_impl.cpp	irange-ify 2 (#62113)	2021-07-26 12:00:52 -07:00
function_impl.h	Make pytorch clang-tidy clean (#60649)	2021-07-01 12:21:07 -07:00
method.h	Make name() part of IMethod interface (#63995)	2021-08-30 13:31:55 -07:00
module_save.cpp
module.cpp	[PyTorch] Add tuple inline storage (#64066)	2021-10-15 12:16:51 -07:00
module.h	[jit] Reduce refcounting of Types (#65345)	2021-10-08 09:03:04 -07:00
object.cpp
object.h	[jit] Reduce refcounting of Types (#65345)	2021-10-08 09:03:04 -07:00