strisunshinewentingwang/pytorch

Pull Requests Commits

Fix sign-compare violations in cpp tests

malfet committed 2 years ago

81d765ef

Fix sign-compare in `c10d/Utils.hpp`

malfet committed 2 years ago

ef56497e

Add "pytorch/metamates" to merge_rules.json (#82)

bigfootjon committed 2 years ago

ff404cec

quantization: autogenerate quantization backend configs for documentation (#75126)

vkuzo committed 2 years ago

74b23b20

[JIT] nvfuser CI fixes

davidberard98 committed 2 years ago

83400e83

[Static Runtime] Fix handling relu in quantized linear relu dynamic op

navahgar committed 2 years ago

60bda4d0

Fix Team Tagging in CODEOWNERS

osalpekar committed 2 years ago

211c6bc0

Extend CSR constructor to support batched indices and values

IvanYashchuk committed 2 years ago

eead5990

Revert "Support masked sum on CSR tensors [CPU, CUDA]"

pytorchmergebot committed 2 years ago

f6b9a1d4

Revert "WIP Jiterator reduction"

pytorchmergebot committed 2 years ago

7d2f36b8

[Model Averaging] Code simplification for _find_process_group function (#75007)

wayi1 committed 2 years ago

4fb7fa08

[PyTorch] Existing MHA: fuse the attn_mask addition (#73219)

swolchok committed 2 years ago

87f40ee6

Add BFloat16 support for logsigmoid, hardsigmoid, hardshrink, softshrink, hardswish and softplus on CPU (#63134)

CaoE committed 2 years ago

77c7a50d

Go through codebase and consolidate checkout steps

janeyx99 committed 2 years ago

63189698

Fix parameterlist dir func error (#74404)

纪少敏 committed 2 years ago

29de7924

[DataPipe] apply dill serialization for _Demux and add cache to traverse

ejguan committed 2 years ago

841a7f51

Support masked sum on CSR tensors [CPU, CUDA]

pearu committed 2 years ago

cda3f586

[GHA] Small updates to syncbranches.yml

malfet committed 2 years ago

caa40308

changed documentation from cosine distance to cosine similarity; fixe…

hahahannes committed 2 years ago

6a586512

Typo in Dockerfile

mimoralea committed 2 years ago

0765a804

make functionalization work better with subclasses

bdhirsh committed 2 years ago

9429dbb4

Use the same checks in all `grid_sampler` functions

Nikita Karetnikov committed 2 years ago

936a6505

Dedupe no parsing __torch_function__ handler

ezyang committed 2 years ago

e3848d75

Introduce SafePyObject, make TorchDispatchTypeObject use it

ezyang committed 2 years ago

de6353ba

Split PyInterpreter into its own file.

ezyang committed 2 years ago

1faf1cdf

[Quant][fx] Define native backend_config_dict for linear and conv (#74636)

andrewor14 committed 2 years ago

ee9335a6

Revert D35352705: fix: buck2 build fbcode//caffe2:torch_types_gen

ezyang committed 2 years ago

7b506e88

fix: buck2 build fbcode//caffe2:torch_types_gen (#75176)

luciang committed 2 years ago

152489a8

Add BFloat16 support for smooth_l1_loss on CPU (#62558)

CaoE committed 2 years ago

c5872e6d

Deprecate bytecode v3 and bump kMinSupportedBytecodeVersion to 4 (#75149)

cccclai committed 2 years ago

96050ee0

Newer Older