strisunshinewentingwang/pytorch

Pull Requests Commits

check tensor has storage before refer to tensor data ptr

Cheng Tang committed 2 years ago

75be4f9c

[PyTorch] Add native fast path for transformer encoder inference (#76333)

swolchok committed 2 years ago

e816e176

[PyTorch Edge] Add Optimized QInt8 Quantize Tensor Arm (#76245)

salilsdesai committed 2 years ago

68a9057a

Revert "[NVFuser] always use fallback if fusion fails"

pytorchmergebot committed 2 years ago

bfb39e57

[Model Averaging] Make an error message more clear in hierarchical_model_averager.py

wayi1 committed 2 years ago

25fa6235

[lint] url encode lint message

suo committed 2 years ago

20543221

Fixes black lint

janeyx99 committed 2 years ago

5b65361a

[DataPipe] clearing buffer for DataPipes during __del__

NivekT committed 2 years ago

ccd7233f

[ROCM] Enable custom tests on rocm

jpvillam-amd committed 2 years ago

6b21a337

[Dynamic RPC] Add graceful shutdown for dynamic RPC members

H-Huang committed 2 years ago

811ccde4

Pad: don't error when unused fill value is zero

peterbell10 committed 2 years ago

f02b7a9c

remove unused --ninja-global from codegen (#75869)

mikey dagitses committed 2 years ago

03e85e57

remove unused all_generator_source from generate_code (#75868)

mikey dagitses committed 2 years ago

f010b15d

move version_h to shared build structure in Buck (#75964)

mikey dagitses committed 2 years ago

8b1cf8ed

Add NVFuser Python Frontend

Kevin Stephano committed 2 years ago

b17b2b1c

Revert "[JIT] Allow registering Decompositions"

pytorchmergebot committed 2 years ago

2d72cb33

Reapply black for torchgen, this time with lint to fix!

ezyang committed 2 years ago

c2ae0b01

[quant][docs] Fix formatting for quantization.rst (#76223)

jerryzh168 committed 2 years ago

30342f6b

[JIT] Allow registering Decompositions

Elias Ellison committed 2 years ago

d9f0774f

Fix memory use after free: buffer need to be attached to the module. (#76350)

qihqi committed 2 years ago

6d0f2be3

Fix bug where __getstate__ of DDP looks for self._replicated_tensor_module

pritamdamania87 committed 2 years ago

b26df43f

E2E SymInt example narrow_copy

Krovatkin committed 2 years ago

bb60cac2

Add black linter

ezyang committed 2 years ago

d286197e

deploy: add dummy metadata for builtin packages (#76211)

d4l3k committed 2 years ago

bbc6fcd7

Fix issue in sparce_coo_tensor only supporting CUDA device.

chengjunlu committed 2 years ago

a9deda54

Update torch::lazy::BackendDevice to have a new default ordinal (#76264)

wonjoolee95 committed 2 years ago

f9d07ae6

[lint] correctly display annotations for all severities

suo committed 2 years ago

1e111895

[reland] use scatter in shard_tensor API (#75991)

wanchaol committed 2 years ago

1f3aa535

Fix forward-mode AD formula for binary_cross_entropy_with_logits

zou3519 committed 2 years ago

9cb2871f

patching 11.1 ptxas issue

jjsjann123 committed 2 years ago

e48b29b1

Newer Older