strisunshinewentingwang/pytorch

Pull Requests Commits

[quant][graphmode] Add support for functional linear (#40331)

jerryzh168 committed 4 years ago

64f925eb

qat eager: remove unneeded modules (#40396)

vkuzo committed 4 years ago

b02c932f

Add state dict for LSTM and RNNCell and helper functions for accessing weights and bias (#40333)

raghuramank100 committed 4 years ago

d7d75e37

[RELAND2] Change AccumulateGrad to yield `.grad`s that match weights' memory layout (#40358)

definitelynotmcarilli committed 4 years ago

8066fba2

[android][gradle] packaging headers in aars for publishing (#40392)

IvanKobzarev committed 4 years ago

9e5d6258

[DDP Note] Remove refs to RoundRobin PG until we officially support it (#40380)

rohan-varma committed 4 years ago

ae2f1f03

Revert D22102408: DNNL: enable conv3d

zhaojuanmao committed 4 years ago

016cf7d6

Revert D22102407: DNNL: enable batchnorm3d

zhaojuanmao committed 4 years ago

17fe0e2b

[JIT] Fork/Join inline docs (#39952)

Elias Ellison committed 4 years ago

0d060853

Revert D22102406: DNNL: enable max_pool3d and avg_pool3d

zhaojuanmao committed 4 years ago

13a8ec3c

Revert D22138737: DNNL: enable dilation conv

zhaojuanmao committed 4 years ago

9498e24c

Add view_as_real, view_as_complex for complex tensors (#39099)

anjali411 committed 4 years ago

8ec2ae9a

Migrate `var` & `std` to ATen (#39967)

ShawnZhong committed 4 years ago

7a3c223b

Build docker for CUDA11 (#40231)

zasdfgbnm committed 4 years ago

c4fc278f

add TypeError to c10 and fix segfault in error checking in Tensor constructor (#40106)

albanD committed 4 years ago

02ae9a15

Added a link to Contribution guide in Readme (#40353)

postmalloc committed 4 years ago

a8ab78c8

DNNL: enable dilation conv (#40220)

XiaobingSuper committed 4 years ago

dbcc5b75

Port addmm, addbmm, addr to ATen (CUDA) (#38421)

Baranowski committed 4 years ago

43331609

DNNL: enable max_pool3d and avg_pool3d (#35664)

XiaobingSuper committed 4 years ago

c8738957

DNNL: enable batchnorm3d (#35663)

XiaobingSuper committed 4 years ago

8df35fd7

DNNL: enable conv3d (#35662)

XiaobingSuper committed 4 years ago

6ba807cb

Utilise the vector version for sinh and cosh (UnaryOpsKernel) (#36396)

kshitij12345 committed 4 years ago

03af4dcb

jit: Conv3d + BatchNorm3d fusion (#40082)

XiaobingSuper committed 4 years ago

87c5f02f

Add prefix of remote events for RPC profiling (#40066)

rohan-varma committed 4 years ago

14f7e95c

Relax cudnn conditions for channels-last convolutions (#38904)

xwang233 committed 4 years ago

17d3f74e

Add addmv for complex dtypes (#40238)

anjali411 committed 4 years ago

c72ab194

Reenable memory format test for some unary functions (#39102)

xuhdev committed 4 years ago

3894de56

[quant][refactor] Tests for torch.jit.quantized (#40330)

jerryzh168 committed 4 years ago

9f9e7c1d

[ONNX] Support large attribute and subgraph for large model (#38793)

BowenBao committed 4 years ago

eaa91071

reference Swish implementation (#40150)

Hector Yuen committed 4 years ago

49887d1f

Newer Older