strisunshinewentingwang/pytorch

Pull Requests Commits

reland Add offsets-based reduction to segment_reduce (CPU, CUDA)

mikaylagawarecki committed 2 years ago

7360b53f

Use cuda instead of cudatoolkit for cuda 11.6 (#77164)

atalman committed 2 years ago

64f3742b

[fix] composite compliance: isclose (#79700)

kshitij12345 committed 2 years ago

bcc4dba4

Fix typo in ChainedScheduler docstring (#79775)

bmsknight committed 2 years ago

9acbaaaf

[fix] composite compliance: eig, eigh, symeig (#79698)

kshitij12345 committed 2 years ago

04b98df8

Reference implementations for rsqrt and native_layer_norm (#79413)

IvanYashchuk committed 2 years ago

bc1fef96

[BE] Use CamelCase for enum class members (#79772)

malfet committed 2 years ago

4a4890cf

Revert "Support dropout(nested tensor) (#79318)"

pytorchmergebot committed 2 years ago

8a7a5def

Add pyproject.toml for black configuration (#79399)

KanchiShimono committed 2 years ago

e85dfb62

Capitalized first letters in the contents of readme table (#77953)

Sadeedpv committed 2 years ago

9aad40e8

numpy_T: typo fixed in warning message (#75859)

SaltedFishLZ committed 2 years ago

1c4fe2a4

[JIT] Nested fix (#79480)

Elias Ellison committed 2 years ago

43aff84c

[BE] Test trymerge against hardcoded set of rules (#79764)

malfet committed 2 years ago

ac89ff56

Adjust wording for consistency (#79758)

orionr committed 2 years ago

4df76d1d

Ensure tensors are contiguous in functional all_gather.

pritamdamania87 committed 2 years ago

500fb247

[BE][TryMerge] Do not re-import trymerge multiple times (#79760)

malfet committed 2 years ago

4661c75f

[PyTorch] Don't create a new Storage in FreeMemory unnecessarily

swolchok committed 2 years ago

6e90572b

Support dropout(nested tensor) (#79318)

YifanShenSZ committed 2 years ago

1211ab67

fx quant: unbreak Mac OS tests when MKLDNN is available

vkuzo committed 2 years ago

91c5fc32

add utility to compute self time of events

davidchencsl committed 2 years ago

9955fff4

Functionality/pickling for commhooks (#79334)

Olga Andreeva committed 2 years ago

8a6d8307

Revert "Enable `dim=None` for `torch.sum` (#75845)"

pytorchmergebot committed 2 years ago

ee6ebfc0

Add nested tensor support to autograd (#79446)

drisspg committed 2 years ago

f9656817

CI Disable xla for infra issues (#79726)

zengk95 committed 2 years ago

4b342b30

[Static runtime] Pass parent graph metadata to forked subgraphs (#79578)

akshayParashar1995 committed 2 years ago

ca7ab170

[MPS]: Add fix for squeezed input axes handling in BCE loss (#79676)

kulinseth committed 2 years ago

4615f6aa

Enable `dim=None` for `torch.sum` (#75845)

kurtamohler committed 2 years ago

e79a51f7

[BE][FSDP] Enhance summon full params test (#79642)

rohan-varma committed 2 years ago

c8b073c5

true up c10::FunctionSchema::operator<< to print native_functions.yaml syntax

bdhirsh committed 2 years ago

9d419e81

Revert "Reland PySymInt (#79617)"

pytorchmergebot committed 2 years ago

44436947

Newer Older