strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D29147009: [pytorch][PR] refine disabled test

Mike Ruberry committed 3 years ago

e5c99d99

MAINT Migrates rrelu_with_noise from THC to ATen on Cuda (#57864)

thomasjpfan committed 3 years ago

a0ad4c24

[iOS GPU][MaskRCNN] Force the temporaryImage to become static when doing synchronization (#60155)

xta0 committed 3 years ago

9e79a8a5

nonzero: Default to transposed output strides (#59370)

peterbell10 committed 3 years ago

0e7b5ea6

[quant] Equalization Observer modifications (#59953)

angelayi committed 3 years ago

c0b7c59e

[quant] Input Weight Equalization - prepare modifications (#59747)

angelayi committed 3 years ago

45c31cab

[quant] EqualizationQConfig to distinguish input/output activations (#59739)

angelayi committed 3 years ago

7ce74f33

Refactor ZeroRedundancyOptimizer Assuming SPSD (#59834)

awgu committed 3 years ago

c6cdb4f1

[TensorExpr] More python binding cleanups (#60058)

jansel committed 3 years ago

85517a2b

[JIT] Handle modules that already have __constants__ (#60003)

Meghan Lele committed 3 years ago

c01939a9

Fix version comparison for defining CUDA11OrLater (#60010)

IvanYashchuk committed 3 years ago

d99a8a31

make it easier to grep for unary/binary op kernels (#60128)

bdhirsh committed 3 years ago

c458bb98

[numpy] mvlgamma: int -> float promotion (#59934)

kshitij12345 committed 3 years ago

3288c9d3

[fx][Transformer] Add override for call_function (#60057)

jfix71 committed 3 years ago

f6579350

don't use moved field in the second lambda (#59914)

cyyever committed 3 years ago

5f017e91

[testing] OpInfoHelper tool (#58698)

kshitij12345 committed 3 years ago

64aec8d2

Fix Python 3.8 expecttest machinery again, this time for good. (#60044)

ezyang committed 3 years ago

0bf12607

Remove dead code in SavedVariable (#59838)

Varal7 committed 3 years ago

dab1e596

Avoid un-necessary unwrapping of Tensor in SavedVariable (#59837)

Varal7 committed 3 years ago

1efa8638

removed gelu from autocast fp32 list (#59639)

Fuzzkatt committed 3 years ago

5948e6f6

[quant] Add a quantize_per_tensor overload that takes Tensor quantization parameters (#59773)

jerryzh168 committed 3 years ago

a95207da

Revert D29154971: Training resnext with msuru_suru_union and ig_msuru_suru_union datasets

Mike Ruberry committed 3 years ago

5686fe58

Some fixes to vec256_bfloat16.h (#59957)

r-barnes committed 3 years ago

4c8c61f2

[torch deploy] add register_module_source (#58290)

zdevito committed 3 years ago

8ce6d0c4

[Profiler] Fix timestamp discrepancy in profiler_kineto.cpp (#60070)

gdankel committed 3 years ago

fd1e9253

Use GitHub's diff directly in clang-tidy (#60048)

driazati committed 3 years ago

9d776464

support parsing of PR stats in run_test.py (#60026)

Rong Rong (AI Infra) committed 3 years ago

b2fc6de2

Fix compile failure on CUDA92 (#60017)

ejguan committed 3 years ago

691183bb

[torch][segment_reduce] Add missing cuda kernel launch check (#60114)

Serhat Yilmaz committed 3 years ago

15dbc566

Add c10d filestore functionality to the current c10d_rendezvous_backend (#59719)

SirLegolot committed 3 years ago

2c5db9a4

Newer Older