strisunshinewentingwang/pytorch

Pull Requests Commits

Fix functorch test selection logic (#86944)

huydhn committed 2 years ago

d393a463

Revert "symintify nll loss fns (#86915)"

pytorchmergebot committed 2 years ago

bbd7b38d

symintify nll loss fns (#86915)

anjali411 committed 2 years ago

0ece7c86

[FSDP] Consolidate FSDP state_dict offload_to_cpu settings (#86211)

fegin committed 2 years ago

a86278b0

add super setup to test to enable disabling in test_dims.py (#86953)

clee2000 committed 2 years ago

c9a8d309

Revert "[Profiler] Move legacy profiler out of `torch/csrc/autograd` (#85512)"

pytorchmergebot committed 2 years ago

8eb579e3

[primTorch] Add a ref for `addcmul` (#86731)

nkaretnikov committed 2 years ago

4460e40d

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

pytorchmergebot committed 2 years ago

746500d5

Add optional recomputable_ops argument for the min cut partitioner (#86686)

IvanYashchuk committed 2 years ago

2cfc4cb3

Add nvFuser support for torch.Tensor.view (#84634)

IvanYashchuk committed 2 years ago

fd806847

Set up new module torch.signal.windows (#85599)

alvgaona committed 2 years ago

b48deedb

Revert "[ONNX] Update training state logic to support ScriptedModule (#86745)"

pytorchmergebot committed 2 years ago

056cfb04

[Profiler] Move legacy profiler out of `torch/csrc/autograd` (#85512)

Taylor Robie committed 2 years ago

157a3d2a

[Profiler][Minor] Separate standalone profilers from the main PyTorch profiler. (#85511)

Taylor Robie committed 2 years ago

35fb0077

[Profiler][Minor] Group and consolidate stub APIs (#85510)

Taylor Robie committed 2 years ago

b8f14b78

[FSDP] Fix load_sharded_state_dict FQN mismatches for shared parameters (#86524)

fegin committed 2 years ago

bc4ca4c2

[ONNX] Update training state logic to support ScriptedModule (#86745)

justinchuby committed 2 years ago

960b9812

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

pytorchmergebot committed 2 years ago

f451e824

Remove functorch copy of conftest.py (#86927)

huydhn committed 2 years ago

c16c4a37

Unified symbolic shape variables between AOTAutograd and Inductor (#86659)

Chillee committed 2 years ago

b3b9786f

Move TorchDynamo into PyTorch core (#86461)

jansel committed 2 years ago

c7c09722

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

Newer Older