strisunshinewentingwang/pytorch

Pull Requests Commits

Refs and decompositions for index_{add,copy,select,fill} (#85002)

lezcano committed 2 years ago

5dd9610e

[BE] Add explicit `__all__` to torch.cuda (#85193)

malfet committed 2 years ago

45a9dcd4

Add SymInt::guard_int (#85139)

ezyang committed 2 years ago

8c9d7fab

Add nondeterministic alert for `MaxUnpool1d/2d/3d` (#84766)

kurtamohler committed 2 years ago

b0a631cd

Create Cache for Fusion Reuse in NVFuser in Python Frontend for Primtorch (#85045)

kevinstephano committed 2 years ago

b8418e02

allow changing the cuda allocator settings even after the process started (#84970)

Hector Yuen committed 2 years ago

d23ce297

Revert "Faster mul(sparse, sparse) with broadcasting in dense dims. (#83428)"

pytorchmergebot committed 2 years ago

81620c33

Add refs for sinc and sgn (#85142)

lezcano committed 2 years ago

98b8ef99

Revert "Refs and decompositions for index_{add,copy,select,fill} (#85002)"

pytorchmergebot committed 2 years ago

e33b464f

fix external codegen kernel error checking (#85029)

bdhirsh committed 2 years ago

1838957e

Don't cache model specs within PTMCoreMLCompiler (#85136)

jmdetloff committed 2 years ago

652707ab

remove symintnode bits in LTC (#85171)

Krovatkin committed 2 years ago

2dbd2673

Disable torch.library.Library with PYTORCH_DISABLE_LIBRARY (#85190)

soulitzer committed 2 years ago

02f654ab

[torchdynamo hash update] update the pinned torchdynamo hash (#85198)

pytorchmergebot committed 2 years ago

dca42ec2

[Profiler][Trivial] Make `test/profiler` folder. (#84273)

Taylor Robie committed 2 years ago

21f2d559

Make param 'option' const& to prevent unnecessary copy at call-site (#84747)

ggwg committed 2 years ago

4a5edbf0

Expose get_active_ddp_module api for torchdynamo DDP (#83333)

wconstab committed 2 years ago

32fc0b95

CoreML .mlmodel export support (#84784)

Bartek Rymkowski committed 2 years ago

0a6f3261

[NNC] add eltwise OPs: mish and elu (#80586)

chunyuan-w committed 2 years ago

ca419c33

Added additional simplifications/caching for replacements and divisibility (#84918)

Chillee committed 2 years ago

377b5d6f

[ONNX] Create decorators for symbolic function registration (#84709)

justinchuby committed 2 years ago

9d115523

Add symlink resolution in benchmark timer interface (#82734)

Mateusz Sypniewski committed 2 years ago

05ff3f89

Refs and decompositions for index_{add,copy,select,fill} (#85002)

lezcano committed 2 years ago

2f0b3de4

[ONNX] Update ONNX documentation to include unsupported operators (#84496)

justinchuby committed 2 years ago

d6c2080e

[ONNX] Update error messages (#85179)

justinchuby committed 2 years ago

46843be1

Retry installing lintrunner if download fails (#85189)

ZainRizvi committed 2 years ago

a4c7cadc

Revert "[FSDP] Option to keep grads in lower prec (#85134)"

pytorchmergebot committed 2 years ago

14b3bdc0

Remove assertEqualIgnoreType from test_pooling (#85112)

mshuaibii committed 2 years ago

4382da5d

[ONNX] New symbolic function registry (#84382)

justinchuby committed 2 years ago

cd7e6d4a

update torch.narrow doc (#85180)

kshitij12345 committed 2 years ago

73515435

Newer Older