strisunshinewentingwang/pytorch

Pull Requests Commits

[package] remove save_source_file API (#57340)

suo committed 3 years ago

44b021d2

[package] remove PackageExporter.file_structure (#57339)

suo committed 3 years ago

a3cba770

[package] use digraph to back dependency visualization (#57338)

suo committed 3 years ago

f326f7dd

[package] add simple graph data structure (#57337)

suo committed 3 years ago

53c21172

[package] make extern a dict (#57336)

suo committed 3 years ago

a39c685a

[package] factor out `PackageExporter._get_dependencies` (#57335)

suo committed 3 years ago

dedf9fbe

hardswish reland (#57652)

Elias Ellison committed 3 years ago

7627dd56

[NNC] ported over sum and softmax to new scheme (#56775)

Chillee committed 3 years ago

56211524

Add missing return statement from 57057 (#57669)

albanD committed 3 years ago

0b51ee31

[PyTorch] Autoformat c10, round 2 (#57645)

swolchok committed 3 years ago

cd22bdf2

Share VS Code settings/extensions nicely (#57671)

samestep committed 3 years ago

e5179e96

Expand Kineto platform support (ci-all) (#56323)

Ilia Cherniavskii committed 3 years ago

65fad0eb

[iOS GPU] Add Metal API availability check (#57663)

xta0 committed 3 years ago

30c96c94

[Flaky tests] Fix flaky rpc profiling tests (#57517)

rohan-varma committed 3 years ago

69e64b26

NNAPI: flex size support for upsample_nearest2d op (#57563)

Akshit Khurana committed 3 years ago

c4bb6a57

NNAPI: Add qadd flexible size support (#57562)

Akshit Khurana committed 3 years ago

4c609a97

NNAPI: add flexible size support for conv2d (#57561)

Akshit Khurana committed 3 years ago

28cd04ea

Make torch.linalg.eigvalsh differentiable (#57189)

IvanYashchuk committed 3 years ago

049152fa

Make torch.linalg.svdvals differentiable (#57188)

IvanYashchuk committed 3 years ago

babae61f

add a standalone extra file loader for pytorch model (#57591)

linbinyu committed 3 years ago

534c457d

Revert "Make grad mode error just a warning (#56401)" (#57640)

albanD committed 3 years ago

15c092b8

Clang format ProcessGroupNCCL.cpp (#56840)

rohan-varma committed 3 years ago

7115a4b8

[c10d] Profiler support for nccl p2p collectives (#56427)

rohan-varma committed 3 years ago

a948e279

Speedup render_junit (#57641)

malfet committed 3 years ago

17035f6a

[PyTorch][Edge] Add api to get bytecode model version (#56801)

cccclai committed 3 years ago

fb9a32b7

Reland: [TensorExpr] Add methods for inspecting generated code in `TensorExprKernel`. (#57560)

Mikhail Zolotukhin committed 3 years ago

dedaf4fa

Reland: [StaticRuntime] Use NNC's call_raw API to reduce call overheads. (#57553)

Mikhail Zolotukhin committed 3 years ago

9e7814d5

Reland: [TensorExpr] Add `TensorExprKernel::runFast` method. (#57552)

Mikhail Zolotukhin committed 3 years ago

e686c66f

Reland: [TensorExpr] Add `CodeGen::call_raw` method. (#57551)

Mikhail Zolotukhin committed 3 years ago

0bf69278

Relax tp_new so that it is OK to call (#57544)

ezyang committed 3 years ago

da8cc355

Newer Older