strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D29132956: Add some TORCH_API annotations to RPC

Mike Ruberry committed 4 years ago

3c3bb911

Update on "[DDP] Support for multiple backwards"

rohan-varma committed 4 years ago

a26f20bc

Update base for Update on "[DDP] Support for multiple backwards"

rohan-varma committed 4 years ago

6d3ab327

Revert D28875276: Move RPC agents to libtorch

Mike Ruberry committed 4 years ago

f233274f

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

Mike Ruberry committed 4 years ago

e5c99d99

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

thomasjpfan committed 4 years ago

a0ad4c24

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

xta0 committed 4 years ago

9e79a8a5

nonzero: Default to transposed output strides (#59370)

peterbell10 committed 4 years ago

0e7b5ea6

[quant] Equalization Observer modifications (#59953)

angelayi committed 4 years ago

c0b7c59e

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

angelayi committed 4 years ago

45c31cab

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

angelayi committed 4 years ago

7ce74f33

Refactor ZeroRedundancyOptimizer Assuming SPSD (#59834)

Andrew Gu committed 4 years ago

c6cdb4f1

[TensorExpr] More python binding cleanups (#60058)

jansel committed 4 years ago

85517a2b

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

Meghan Lele committed 4 years ago

c01939a9

Update on "[DDP] Support for multiple backwards"

rohan-varma committed 4 years ago

b309e453

Update base for Update on "[DDP] Support for multiple backwards"

rohan-varma committed 4 years ago

02298ede

Fix version comparison for defining CUDA11OrLater (#60010)

IvanYashchuk committed 4 years ago

d99a8a31

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

bdhirsh committed 4 years ago

c458bb98

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

kshitij12345 committed 4 years ago

3288c9d3

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

jfix71 committed 4 years ago

f6579350

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

cyyever committed 4 years ago

5f017e91

[testing] OpInfoHelper tool (#58698)

kshitij12345 committed 4 years ago

64aec8d2

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

ezyang committed 4 years ago

0bf12607

Remove dead code in SavedVariable (#59838)

Varal7 committed 4 years ago

dab1e596

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

Varal7 committed 4 years ago

1efa8638

removed gelu from autocast fp32 list (#59639)

Fuzzkatt committed 4 years ago

5948e6f6

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

jerryzh168 committed 4 years ago

a95207da

Revert D29154971: Training resnext with msuru_suru_union and ig_msuru_suru_union datasets

Mike Ruberry committed 4 years ago

5686fe58

Some fixes to vec256_bfloat16.h (#59957)

r-barnes committed 4 years ago

4c8c61f2

[torch deploy] add register_module_source (#58290)

zdevito committed 4 years ago

8ce6d0c4

Newer Older