strisunshinewentingwang/pytorch

Pull Requests Commits

add hardtanh(0,6) to the set of MKLDNN fusible ops for mobilenetv2 (#56203)

Krovatkin committed 3 years ago

d6a25a58

[PyTorch] Migrate hacky wrapper removal to borrow_from_optional_tensor (#56648)

swolchok committed 3 years ago

7b7a4750

[PyTorch] Add & document borrow_from_optional_tensor (#56647)

swolchok committed 3 years ago

f2fd91cc

addmm CPU inplace implementation shouldn't resize an input tensor (#56452)

imaginary-person committed 3 years ago

02c3e6d9

Fix: Compare input against beta * threshold in softplus backwards (#56484)

jbschlosser committed 3 years ago

e5fda07e

Some simple optimizations (#51831)

cyyever committed 3 years ago

83c23703

Torchelastic: make process failure init error non-fatal (#56739)

aivanou committed 3 years ago

0a72904a

static runtime support for fb::equally_split (#56565)

Edvard Ghazaryan committed 3 years ago

a4e47ea1

moved more lowerings over (#55372)

Chillee committed 3 years ago

7c50852a

Consolidate nondeterministic error tests (#55631)

kurtamohler committed 3 years ago

1f04494c

[torch.package] is_from_package check (#56729)

Lilyjjo committed 3 years ago

88deea4e

Revert "Revert [ONNX] Redesign inplace conversion" (#56675)

BowenBao committed 3 years ago

913f1f75

CPU Convolution benchmark harness for some popular models (#56455)

bertmaher committed 3 years ago

461e887d

Move windows testers to previous image (#56626)

malfet committed 3 years ago

f84a5010

[NNC] Add unroll and flatten APIs which not require return stmt pointer (#56420)

huiguoo committed 3 years ago

29491f79

Clean up raise exception logic (#55656)

soulitzer committed 3 years ago

20788360

Enable fast gradcheck for real inputs and outputs (#55237)

soulitzer committed 3 years ago

d0130243

Disable flaky tests (#56279)

soulitzer committed 3 years ago

2ea3c24c

Print non-breaking space directly in lint.yml (#56726)

davidriazati@fb.com committed 3 years ago

5c752ead

torch: Add cpython as a dependency for torch_python_obj (#56740)

seemethere committed 3 years ago

08ce2300

Fix mp serialization for integer nn.Parameter on CUDA (#56529)

Alvant committed 3 years ago

bac4cfd5

Support factory kwargs in torch.nn modules (#54508)

jbschlosser committed 3 years ago

febff459

Create helper function for RPC profiling in _invoke_rpc and remote (#56643)

Tran Le committed 3 years ago

3a4344a7

[quant][graphmode][fx] Support preserving attributes in deepcopy of observed/quantized graphmodule (#56550)

jerryzh168 committed 3 years ago

1719cb82

Add periodic_ prefix to all jobs run by cron (#56695)

malfet committed 3 years ago

3a44d269

[sparsity] Moving the sparsity python files to OSS (#56617)

Zafar Takhirov committed 3 years ago

37568783

Remove extraneous variables in windows report stats step (#56596)

janeyx99 committed 3 years ago

31fe2bbb

Introducing JitPlugin (#56708)

janeyx99 committed 3 years ago

5b01b3e8

Fix grad_fn bindings when saved variable freed (#56499)

soulitzer committed 3 years ago

2128a84a

Re-enable fast winograd conv on IOS (#56021)

albanD committed 3 years ago

679cc7eb

Newer Older