strisunshinewentingwang/pytorch

Pull Requests Commits

[CUDA Pinned Memory] [Retry] Alternative implementation of pinned memory allocator focusing on multi-threaded scalability (#69299)

Andrew Tulloch committed 2 years ago

00997969

[warnings][caffe2] Fix -Wstring-conversion warnings (#71874)

Nolan O'Brien committed 2 years ago

ebeeee7b

Propagate full autocast state to CheckpointFunction's forward-inside-backward (#71169)

mcarilli committed 2 years ago

cf3ef237

[ONNX] Update opset_version restriction for local function

BowenBao committed 2 years ago

804f1328

[bc-breaking][quant][be] Refactor fuser_method to include `is_qat` argument (#70009)

jerryzh168 committed 2 years ago

ef501e8f

fixing of usage of rel_tol for test adadelta (#71880)

puririshi98 committed 2 years ago

b0669311

Revert "Disable XLA config"

pytorchmergebot committed 2 years ago

c224f82e

Revert "Fix lint"

pytorchmergebot committed 2 years ago

bbe6144b

.github: Exclude rocm from ciflow/all,ciflow/trunk

seemethere committed 2 years ago

7beb030e

[GHF] Add revert workflow

malfet committed 2 years ago

5bd33247

.github: Enable windows binary builds (#71484)

seemethere committed 2 years ago

5ee629e5

Refactor flatbuffer loader to allow overriding how IValues are parsed. (#71661)

qihqi committed 2 years ago

9f4bdf78

Update _create_c10d_store to check port value (#71863)

H-Huang committed 2 years ago

666ff0ae

Fixes pr-labels workflow trigger (#71871)

atalman committed 2 years ago

d7e5870b

[GHF] Small cleanups

malfet committed 2 years ago

d73dc9b7

[Opt Overlap] Clean up code in _OptimizerHookState (#71620)

rohan-varma committed 2 years ago

bdcdf94b

[Opt Overlap] Support optimizing partial set of parameters (#71608)

rohan-varma committed 2 years ago

1c8fcc44

Implement labelling for release notes and topics check (#71726)

atalman committed 2 years ago

c44d0ac1

[Profiler] Split observer implementations based on ProfilerState (#71135)

Taylor Robie committed 2 years ago

46817895

[numpy] add decimals argument to round (#66195)

kshitij12345 committed 2 years ago

d3bbb281

[SR] Reverse iteration order in resetMemory (#71705)

Mike Iovine committed 2 years ago

7e6312a5

Skip compiledWithCuDNN() call for mobile to avoid segfault (#71775)

jbschlosser committed 2 years ago

e04ade92

Revert D33768645: Set correct device id on efficientzerotensors

Mike Ruberry committed 2 years ago

0891c908

Revert D33778917: Disable some forward mode AD tests

Mike Ruberry committed 2 years ago

adcf34f6

Add forward AD formulas for some losses (#71026)

soulitzer committed 2 years ago

25e84fa4

extract out //c10/core:base library (#70857)

Michael Dagitses committed 2 years ago

c6d885e4

extract final two libraries out of //c10/util (#70856)

Michael Dagitses committed 2 years ago

130ca586

extract //c10/core:ScalarType to its own library (#70855)

Michael Dagitses committed 2 years ago

bfc481cf

Implements allreduce_coalesced for ProcessGroupNCCL (#62140)

mcarilli committed 2 years ago

f37d2046

Remove state_dict from AveragedModel and use buffers instead (#71763)

prabhat00155 committed 2 years ago

942a084c

Newer Older