strisunshinewentingwang/pytorch

Pull Requests Commits

Update ONNX Export for Gather and Scatter for Opset 11

lara-hdr committed 5 years ago

c79d116a

Fix Exporting RNN/LSTM's Initial State (h0/c0) to ONNX

lara-hdr committed 5 years ago

3569a1c6

quantized torch.topk (#26486)

James Reed committed 5 years ago

cb9fd0ce

Allow batch size of 0 in Conv

James Reed committed 5 years ago

64d58c2f

add CondValue to unify refinements and code emission (#26145)

zdevito committed 5 years ago

fcd13549

Fix the flaky test_qlinear test caused by hypothesis deadline (#26663)

jianyuh committed 5 years ago

cbdbdd3c

"fixing" gcc bug introduced with cuda 10.1 (#26445)

Mingbo Wan committed 5 years ago

21314cfd

Serialization for per channel qtensor (#26339)

Dmytro Dzhulgakov committed 5 years ago

ebc2365f

NHWC specialization for quantized::cat (#26524)

James Reed committed 5 years ago

c0aa6a01

nightly prefix for android nightly jobs (#26652)

IvanKobzarev committed 5 years ago

69631c3e

BlobReference __getattr__ can only throw AttributeError (#26654)

Aapo Kyrola committed 5 years ago

aeb6532e

Import torch.quantization when one imports torch

Dmytro Dzhulgakov committed 5 years ago

8fc86525

Fix ellipsis behavior for `Tensor.align_to` to glob all missing dims (#26648)

zou3519 committed 5 years ago

567a1981

Revert D17450502: [pytorch][PR] [WIP] Enabled bfloat16 dtype on CUDA

ezyang committed 5 years ago

fdf2bdef

Add namedtensor build & tests to default sets (#26633)

zou3519 committed 5 years ago

e0f86f7a

Makes test_indexing.py device generic (#26634)

Mike Ruberry committed 5 years ago

a9a9d362

Sync docker images

IvanKobzarev committed 5 years ago

2a574e49

Add testing script for iOS x86 build (#26632)

xta0 committed 5 years ago

781f8618

fix operator level benchmark to have NHWC layout (#26577)

dskhudia committed 5 years ago

fc926d92

Simplify observers declaration with functools.partial (#26492)

Dmytro Dzhulgakov committed 5 years ago

a79b3685

Enabled bfloat16 dtype on CUDA

izdeby committed 5 years ago

76697a3b

prevent generating caffe2::mkldnn for multiple times (#25257)

jasjuang committed 5 years ago

e4821012

ATen port of lgamma (cuda) (#26600)

pearu committed 5 years ago

786d2259

Remove deprecated torch.gels (#26480)

vishwakftw committed 5 years ago

15b50606

Fixing the calling parameters of write_gif function of the moviepy.

danielmatte committed 5 years ago

557246b7

Revert D17521607: Name inference for min(Tensor, dim?) / max(Tensor, dim?)

zou3519 committed 5 years ago

808f4a4d

Renames `tensor.renamed -> rename`, `tensor.names_ -> rename_` (#26548)

zou3519 committed 5 years ago

4fada962

Name inference for min(Tensor, dim?) / max(Tensor, dim?) (#25582)

zou3519 committed 5 years ago

d3e90bc4

C++ API parity: at::Tensor::detach

pbelevich committed 5 years ago

6b255624

Whenever possible, use function pointers rather than std::function to represent Operation's. (#26560)

Owen Anderson committed 5 years ago

bdf10380

Newer Older