strisunshinewentingwang/pytorch

Pull Requests Commits

Corrected pre- and post-division comments

Olga Andreeva committed 2 years ago

b6f82b3b

Fix nvFuser's where(tensor, python_scalar, tensor) type promotion (#80347)

IvanYashchuk committed 2 years ago

12dc410f

FSDP communication hook interface for NO_SHARD strategy (#79833)

Olga Andreeva committed 2 years ago

5fc209ed

Register nested tensor linear kernel (#80397)

drisspg committed 2 years ago

2a09e951

Use AutogradNestedTensor key for nested tensors (#80396)

drisspg committed 2 years ago

2afbc998

Fix FSDP when not all outputs get gradient in backward (#80245)

rohan-varma committed 2 years ago

e88dadd5

[ci] disable docs build (#80313)

suo committed 2 years ago

fd8a15bd

[ao][sparsity] Fixing failing internal pruner tests (#80111)

zaf committed 2 years ago

cb5ef130

Made Proxy Tensor Mode also trace overloads (#80403)

Chillee committed 2 years ago

615dd250

[Quant][fx] Hide equalization_config from prepare APIs (#80164)

andrewor14 committed 2 years ago

8aedd8fb

[5] move XNNPACK to shared BUCK build (#80209)

linbinyu committed 2 years ago

8a45ef23

Mark underlying type as C10_UNUSED (#80415)

malfet committed 2 years ago

17ee5de0

[ao] Implemented report generation for InputWeightEqualization Detector (#80191)

vspenubarthi committed 2 years ago

d4f8a6d0

[ci] fix viable/strict promotion (#80399)

suo committed 2 years ago

08c32d7f

Revert "torch.special.gamma (#78904)"

pytorchmergebot committed 2 years ago

602c38ff

[Resubmission] fix mul_out CUDA config for COO tensors (#80254)

cpuhrsch committed 2 years ago

5da776dd

`__launch_bounds__` for `torch.mode` with CUDA 11.7 (#79710)

crcrpar committed 2 years ago

7549f901

Update README.md (#80190)

skmonga committed 2 years ago

ed7464f9

[GHA] remove newlines from COMMIT_MESSAGES to avoid batch annoyance (#80385)

janeyx99 committed 2 years ago

4d80b2dc

[quant] Modify APoT global methods to align with uniform API (#80364)

asl3 committed 2 years ago

2727d885

[ao] Implemented InputWeightEqualization Detector observer insertion (#79962)

vspenubarthi committed 2 years ago

4f94c8e0

[mergebot] Create land time check options (#77943)

zengk95 committed 2 years ago

2acd2317

[c10d] Make send/recv as custom ops (#79779)

Jiewen Tan committed 2 years ago

79c2dfcd

Fix distributed store to use add for the counter of DL shared seed (#80348)

ejguan committed 2 years ago

3ec9d34f

Raise warning for unpickable local function (#547) (#80232)

ejguan committed 2 years ago

3d218e1c

Add __all__ to torch.nn.quantized, fx.passes, ao.nn and amp submodules (#80376)

anjali411 committed 2 years ago

f68f7761

Add a note on CUDA 11.6 (#80363)

Svetlana Karslioglu committed 2 years ago

7394de4e

More stable computation of KL between two Bernoulli distributions (#79944)

nicola-decao committed 2 years ago

27fc9fcd

StringUtils: Avoid unnecessary allocation in ReplaceAll (#79915)

pytorchmergebot committed 2 years ago

be71f663

Add __all__ to various submodules in torch.fx, distributions, distributed, package (#80367)

anjali411 committed 2 years ago

3bcc19b2

Older