strisunshinewentingwang/pytorch

Pull Requests Commits

[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

Restore TestTorchFunctionOverride

peterbell10 committed 2 years ago

bf165526

torch.tensor: add tests for list of numpy arrays case

kshitij12345 committed 2 years ago

05090224

Update forward AD not supported error message

soulitzer committed 2 years ago

631f0351

WIP Jiterator reduction

ngimel committed 2 years ago

ad028e5e

[quant][fx] Remove is_output_quantized from QuantizeHandler (#74843)

jerryzh168 committed 2 years ago

bd032cd8

[WIP] Add support to `Tensor[]` for structured kernel codegen.

ezyang committed 2 years ago

dca42dc9

Skip TorchScript backend for OVRSource as well (#75138)

malfet committed 2 years ago

8b7e2bf7

Save disable_torch_function in ThreadLocalState

ezyang committed 2 years ago

3f108a5c

Rewrite upgrader bytecode version from 3 to 4 (content unchanged) (#75120)

cccclai committed 2 years ago

6efc5c1a

Extend jit::load to work on flatbuffer file (#75022)

pavithranrao committed 2 years ago

d9d34922

Newer Older