strisunshinewentingwang/pytorch

Pull Requests Commits

Use std::mutex instead of std::call_once in Function when we initialize GraphExecutor. (#26571)

Mikhail Zolotukhin committed 4 years ago

c034f979

Remove 'recurse' parameter from Inline. (#26487)

Mikhail Zolotukhin committed 4 years ago

76e2ffc8

Enable registering stackbased kernels with lambdas (#26658)

smessmer committed 4 years ago

a65db650

Revert D17495679: [pytorch][PR] A few hub improvements

Karl Ostmo committed 4 years ago

839e636f

Updates and extends TestNNDeviceType (#26638)

Mike Ruberry committed 4 years ago

98bbb778

Allow per-channel QTensor accept any floating type for scales (#26676)

Dmytro Dzhulgakov committed 4 years ago

ade60f8a

Per-channel quantized tensor to have only a single axis (#26675)

Dmytro Dzhulgakov committed 4 years ago

b93823cb

Fix _empty_per_channel_affine_quantized to be less hacky (#26243)

Dmytro Dzhulgakov committed 4 years ago

9aad4d7b

adding OpProfile proto into ProfDAGProtos to support storing operation cost (#26677)

wenqicaofb committed 4 years ago

fbc3c148

Quantized Interpolate Kernel(upsample_nearest2d) (#26617)

lly-zero-one committed 4 years ago

ba8002ec

_per_channel_affine_qtensor -> _make_per_channel_quantized_tensor (#26679)

jerryzh168 committed 4 years ago

aa95c795

Skip observing bias across function call hierarchy (#26642)

jerryzh168 committed 4 years ago

8a919f4f

Whitelist ATen/core sources and headers for Caffe2 (#26609)

dreiss committed 4 years ago

af96e0cb

_per_tensor_affine_qtensor -> _make_per_tensor_quantized_tensor (#26678)

jerryzh168 committed 4 years ago

d63143dc

Add ObserveHelper and remove some common function parameters (#26641)

jerryzh168 committed 4 years ago

7d612066

fix github actions for forked PRs (#26562)

suo committed 4 years ago

7f89464b

Update qengine flag in python to string (#26620)

supriyar committed 4 years ago

45391cce

remove unneeded code (#26640)

jerryzh168 committed 4 years ago

5d82cefa

A few hub improvements (#25980)

Ailing Zhang committed 4 years ago

1eaaf8b6

Update ONNX Export for Gather and Scatter for Opset 11

lara-hdr committed 4 years ago

c79d116a

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

lara-hdr committed 4 years ago

3569a1c6

quantized torch.topk (#26486)

James Reed committed 4 years ago

cb9fd0ce

Allow batch size of 0 in Conv

James Reed committed 4 years ago

64d58c2f

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

zdevito committed 4 years ago

fcd13549

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

jianyuh committed 4 years ago

cbdbdd3c

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

Mingbo Wan committed 4 years ago

21314cfd

Serialization for per channel qtensor (#26339)

Dmytro Dzhulgakov committed 4 years ago

ebc2365f

NHWC specialization for quantized::cat (#26524)

James Reed committed 4 years ago

c0aa6a01

nightly prefix for android nightly jobs (#26652)

IvanKobzarev committed 4 years ago

69631c3e

BlobReference __getattr__ can only throw AttributeError (#26654)

Aapo Kyrola committed 4 years ago

aeb6532e

Newer Older