strisunshinewentingwang/pytorch

Pull Requests Commits

C10D extension to enable per-thread PG (#86348)

Rodrigo Kumpera committed 2 years ago

97abc21f

Improve complex lerp performance (#84844)

peterbell10 committed 2 years ago

66979fbf

disable failing circleci test jobs (#86940)

clee2000 committed 2 years ago

ae45dab5

Add BFloat16 dtype support for oneDNN Graph JIT fuser (#85591)

sanchitintel committed 2 years ago

974ad8fa

[fsdp] Fix test for 2d parallel integration to trigger the load hooks. (#86272)

Rodrigo Kumpera committed 2 years ago

14dd5db2

[quant][be] Rename node_name_to_target_dtype to node_name_to_target_dtype_info (#86860)

jerryzh168 committed 2 years ago

18f58e2d

add _freeze for embedding op (#86769)

inisis committed 2 years ago

158a0710

set the correct size of aten tensor in presence of mkldnn padding (#86767)

Krovatkin committed 2 years ago

e737f2d8

[ONNX] Fix FindCommonAncestor in function_extraction (#86650)

BowenBao committed 2 years ago

860ad049

[ONNX] Fix triu/tril export with diagonal input (#86843)

BowenBao committed 2 years ago

af1dcef7

Skip test_nvfuser_extremal_values for native_batch_norm (#86897)

IvanYashchuk committed 2 years ago

dbdfb8dd

[ONNX] Fix scalar_type_analysis metadata for copied constant (#86716)

BowenBao committed 2 years ago

2ce6150d

Fix incorrect tensor storage check (#86845)

smk2007 committed 2 years ago

4839f73f

Fix path to nested_tensor in example (#86891)

frankier committed 2 years ago

afc99638

[nn] module: full_backward_pre_hook (#86700)

kshitij12345 committed 2 years ago

54ee95c8

Fix LinearLR scheduler start_factor (#86695)

mikael10j committed 2 years ago

7dcfbedc

[functorch] Add shard to run functorch tests with asan (#82164)

samdow committed 2 years ago

6ee94b57

[cuDNN] Enable cuDNN Frontend v8 API by Default (#84948)

eqy committed 2 years ago

427e0a6b

[ONNX] Clarify phrasing of skipScriptTest/skipTraceTest decorators (#86216)

BowenBao committed 2 years ago

b0d80f43

[ONNX] Renable assert diagnostic test (#85999)

BowenBao committed 2 years ago

0ee09996

Enable max.unary_out (#86855)

tugsbayasgalan committed 2 years ago

cff333bd

[FSDP] restricts meta model check to non ignored modules in FSDP (#86766)

colin2328 committed 2 years ago

25811663

Add nested squeeze.dim and unsqueeze (#86813)

mikaylagawarecki committed 2 years ago

ab695506

[ao] fixing public v private for fx.backend_config_utils.py (#86037)

HDCharles committed 2 years ago

e531cf7b

Revert "Use CUTLASS GEMM for NT bmm [OSS-only] (#85894)"

pytorchmergebot committed 2 years ago

d169f950

Change legacy wrap_dim to work with symint == (#86842)

wconstab committed 2 years ago

b97ae59e

[functorch] Add more details to the functorch install page (#86823)

zou3519 committed 2 years ago

3d9fd060

OpInfo: Sample input cleanup (2/n) (#86379)

peterbell10 committed 2 years ago

cbc01c43

OpInfo: Sample input cleanup (1/n) (#86231)

peterbell10 committed 2 years ago

2efc56d9

[ONNX] Partially re-enable RoiAlign and RoiPool unit tests (#86169)

BowenBao committed 2 years ago

45274c56

Newer Older