strisunshinewentingwang/pytorch

Pull Requests Commits

Strenghten preconditions of linalg.cross (#83798)

lezcano committed 2 years ago

7f0198e7

Enable NCCL_DESYNC_DEBUG when TORCH_DISTRIBUTED_DEBUG=DETAIL (#83881)

kwen2501 committed 2 years ago

9beddde1

Allow None arguments for elementwise type promotion wrapper and fix clamp with None arguments (#83586)

IvanYashchuk committed 2 years ago

cb488e6d

Move ATenNVRTC.h include from `jit_utils.h` to `jit_utils.cpp` (#83886)

malfet committed 2 years ago

8793cd2f

reinplace pass: special handling for view_scatter ops (#83846)

bdhirsh committed 2 years ago

8db04c11

reinplace pass: bugfix for output node replacement (#83845)

bdhirsh committed 2 years ago

75ec7b75

Improve DistanceKernel.cu (#83811)

chengscott committed 2 years ago

01434c2d

[functorch] add linalg cross batch rule (#83759)

samdow committed 2 years ago

df048414

[PyTorch] Remove unused sstream/string includes from c10/macros/Macros.h (#83353)

swolchok committed 2 years ago

e4af53c1

Update retry action to latest version (#83911)

janeyx99 committed 2 years ago

7e386845

[ROCm] restore MIOpen benchmark flag default to true (#82656)

jeffdaily committed 2 years ago

a315a2c7

Fix stride issue with faketensors (#83822)

Chillee committed 2 years ago

0270a707

Refactored ops on size to be dispatcher ops (#83719)

Chillee committed 2 years ago

7ebdb4c7

Remove DBR quantization from the codebase (#83642)

vkuzo committed 2 years ago

58170fb8

Normalize DLPack stride to 1 where shape < 2 (#83158)

mattip committed 2 years ago

4dfa6d28

[ROCm] More Sparse UTs enablement and more hipification mappings. (#78939)

jpvillam-amd committed 2 years ago

247468ba

[xla hash update] update the pinned xla hash (#83899)

pytorchmergebot committed 2 years ago

ed949e22

[optim] rprop: handle complex params as independent real params (#83858)

kshitij12345 committed 2 years ago

7c20ad3d

[nn] split rnn_utils test from test_nn.py (#83675)

kshitij12345 committed 2 years ago

dd67d52b

[quant][fx] Add support for quantized matmul (#83885)

jerryzh168 committed 2 years ago

a419e483

[ONNX] Use `errors.SymbolicValueError` for more context (#83332)

justinchuby committed 2 years ago

3dfb8dfc

Optimize transpose copy on CPU using fbgemm transpose (#83327)

CaoE committed 2 years ago

04d8da88

[BE] Revert distributed change in https://github.com/pytorch/pytorch/pull/68779 (#83181)

rohan-varma committed 2 years ago

b29a0748

[FSDP] Remove unneeded checks (#83150)

rohan-varma committed 2 years ago

4e90526a

hash update - bug fix for branches (#83865)

clee2000 committed 2 years ago

8e074f45

[MPS] Move mps_linear to mps dispatch key (#80068)

drisspg committed 2 years ago

7cfc8b78

[cmake] Change COLORIZE_OUTPUT option to USE_COLORIZE_OUTPUT (#83716)

xwang233 committed 2 years ago

b18f9843

Lazy deprecation import function in torch.nn (#83834)

z-a-f committed 2 years ago

b7afee8a

fix upsample bf16 issue for channels last path by using high pricsion to compute index (#83847)

XiaobingSuper committed 2 years ago

658f958b

[ONNX] Add quantization support to more single output ops (#83008)

justinchuby committed 2 years ago

80cfafc3

Newer Older