strisunshinewentingwang/pytorch

Pull Requests Commits

Expose certain methods in OpaqueTensorImpl. (#41060)

pritamdamania committed 4 years ago

93778f3b

Decouple DataParallel/DistributedDataParallel from CUDA (#38454)

chengjunlu committed 4 years ago

8d570bc7

Doc warnings (#41068)

mattip committed 4 years ago

75155df8

.circleci: Output binary sizes, store binaries (#41074)

seemethere committed 4 years ago

ff3ba25b

Insert parentheses around kernel name argument to hipLaunchKernelGGL (#41022)

jithunnair-amd committed 4 years ago

0e6b7502

Splitting embedding_bag to embedding_bag_forward_only and embedding_bag (#40557)

gmagogsfm committed 4 years ago

630e7ed9

Fix link to PyTorch organization (from Governance) (#40984)

mathemage committed 4 years ago

00ee54d2

Grammatically updated the tech docs (#41031)

lento committed 4 years ago

452d5e19

If ninja is being used, force build_ext to run. (#40837)

ezyang committed 4 years ago

22c7d183

Fix several quantization documentation typos (#40567)

Edward Leardi committed 4 years ago

733b8c23

Add option to warn if elements in a Compare table are suspect (#41011)

Taylor Robie committed 4 years ago

2d98f817

Revert D22396896: [pytorch][PR] run single-threaded gradgradcheck in test_nn

nairbv committed 4 years ago

a04af4dc

type annotations for dataloader, dataset, sampler (#39392)

Baranowski committed 4 years ago

0e09511a

Make `torch_cpu` compileable when `USE_TENSORPIPE` is not set. (#40846)

malfet committed 4 years ago

a6b703cc

Remove unused Logger in get_matching_activations (#41023)

Haixin Liu committed 4 years ago

12b5bdc6

Fix unordered-map-over-enum for GCC 5.4 (#41063)

malfet committed 4 years ago

4aa543ed

Fix CUDA jit codegen compilation with gcc-5.4 (#41055)

malfet committed 4 years ago

50df0975

ONNX: support view_as operator (#40496)

yaeldMS committed 4 years ago

56396ad0

[ONNX]Fix export of full_like (#40063)

KsenijaS committed 4 years ago

b2cc8a26

Improve error message for Pad operator (#39651)

yaeldMS committed 4 years ago

6e4f501f

Fix HTTP links in documentation to HTTPS (#40878)

Edward Leardi committed 4 years ago

6b50874c

[ATen] Add `native_cuda_h` list to CMakeLists.txt (#41038)

malfet committed 4 years ago

63ef7069

Enable `in_dims` for vmap frontend api (#40717)

zou3519 committed 4 years ago

5d1d8a58

run single-threaded gradgradcheck in test_nn (#40999)

ngimel committed 4 years ago

dac63a13

fix grad thrashing of shape analysis (#40939)

Elias Ellison committed 4 years ago

37a572f3

shape analysis fix for default dtype' (#40938)

Elias Ellison committed 4 years ago

4af84243

Back out "[2/n][Compute Meta] support analysis for null flag features"

Will Feng (FAIAR) committed 4 years ago

078669f6

Port `equal` from THC to ATen (CUDA) (#36483)

Baranowski committed 4 years ago

a7802447

s/torch::jit::class_/torch::class_/ (#40795)

James Reed committed 4 years ago

c0f9bf9b

Mish Activation Function (#40856)

YesterdayDX committed 4 years ago

cbe52d76

Newer Older