strisunshinewentingwang/pytorch

Pull Requests Commits

Add native masked_softmax (#69268)

zrphercule2 committed 2 years ago

aab67c6d

[SR] Wrap check_for_memory_leak with DCHECK (#69588)

Hao Lu committed 2 years ago

a5996a68

Make c10d tests -Werror clean (#69703)

malfet committed 2 years ago

3bb20ae4

Do not use `std::labs` (#69704)

malfet committed 2 years ago

be757add

[ONNX] shapeValueMap: Represent symbolic shape as value (#68203) (#69545)

BowenBao committed 2 years ago

3f02ad09

Back out "[wip][quant][graphmode] produce reference pattern for binary ops and then rewrite to quantized op" (#69713)

Ha-nyung Chung committed 2 years ago

3d32a0c1

[nnc][quant] Fix quantized concat (#69596)

IvanKobzarev committed 2 years ago

7dba88df

Remove WindowsTorchApiMacro.h in favor of Export.h (#69585)

peterbell10 committed 2 years ago

b2e79ed5

[SR] assignStorageToManagedTensors returns a vector (#69568)

Mike Iovine committed 2 years ago

f87f1d08

[Static Runtime] [Code Cleanup] Encapsulate function objects within ProcessedFunction (#69595)

d1jang committed 2 years ago

9aa1b3e3

Introduce isTensorSubclassLike; add special cases to backwards formulas (#69534)

zou3519 committed 2 years ago

41e1ab07

[pytorch][PR] Add ability for a mobile::Module to save as flatbuffer (#69306)

qihqi committed 2 years ago

d3649309

.github: Add fbsync to push triggers (#69718)

seemethere committed 2 years ago

193e3c48

[SR] Update memory planner docs (#69559)

Mike Iovine committed 2 years ago

3e20a74b

Extend explanation of `torch.cholesky_inverse` to consider batched inputs. (#69069)

nonconvexopt committed 2 years ago

e963b436

Upgrade oneDNN to v2.3.3 and package oneDNN Graph API together (#63748)

chunyuan-w committed 2 years ago

9ad05f2c

Revert D32942007: OpInfo: Convert more sample_input_funcs to generators

suo committed 2 years ago

17641fed

Fix inference_mode decorator (#68617)

milesial committed 2 years ago

0ccb1dcd

use irange for loops 10 (#69394)

r-barnes committed 2 years ago

afb74238

Added ScriptFunction pkl exception for issue #61210 #61381 (#67076)

BashkirovN committed 2 years ago

2d5b3101

OpInfo: Convert more sample_input_funcs to generators (#69257)

peterbell10 committed 2 years ago

d21646c4

OpInfo: Allow sample_inputs_func to be any iterable (#69256)

peterbell10 committed 2 years ago

6de9f0fc

Fix errors in `common_utils.py` (#69578)

zasdfgbnm committed 2 years ago

d2917f70

[sparsity] Convert function for sparse kernels without a context manager (#66778)

z-a-f committed 2 years ago

07932e27

Replace issue templates with new issue forms - v2 (#69361)

NicolasHug committed 2 years ago

b957b82d

[sparsity] Add ability to keep sparsity parameters in modules (#66777)

z-a-f committed 2 years ago

e948856c

[Operator Versioning][Edge] Implement register function for upgrader (#67730)

cccclai committed 2 years ago

13faaff5

[AO] Clear the contents of the torch/ao/__init__.py (#69415)

z-a-f committed 2 years ago

4f5806de

[AutoAccept][Codemod][FBSourceClangFormatLinter] Daily `arc lint --take CLANGFORMAT`

CodemodService FBSourceClangFormatLinterBot committed 2 years ago

015e481a

Fixes mem_get_info when querying on a device other than the current device (#69640)

Mike Ruberry committed 2 years ago

dc87cf5f

Newer Older