strisunshinewentingwang/pytorch

Pull Requests Commits

Pin NumPy version on MacOS testers to 1.18.5 (#42409)

malfet committed 4 years ago

115d2264

Limits cpu scalar error message to where it's appropriate (#42360)

Mike Ruberry committed 4 years ago

29123906

Improve `torch.norm` functionality, errors, and tests (#41956)

kurtamohler committed 4 years ago

206db5c1

Convert ProcessGroupNCCLTest.cpp to gtest unittest (#42365)

saidmoglu committed 4 years ago

44b018dd

[NNC] Bounds Inference: make inferred bounds respect gaps (#42185)

nickgg committed 4 years ago

f47e00bd

[TensorExpr] Make tensorOrConstant non-templatized function. (#42202)

Mikhail Zolotukhin committed 4 years ago

dcc4d11f

[TensorExpr] Implement shape inference for TE. (#41451)

Mikhail Zolotukhin committed 4 years ago

2decccea

[TensorExpr] Explicitly cast to bool results of comparison ops in kernel.cpp. (#42201)

Mikhail Zolotukhin committed 4 years ago

f41bb1f9

[TensorExpr] Add debug dumps to kernel.cpp. (#42196)

Mikhail Zolotukhin committed 4 years ago

f8c5800b

Implement Enum sugared value and Enum constant support (#42085)

gmagogsfm committed 4 years ago

655f3764

Changes to match Fused Op: Dequantize->Swish->Quantize (#42255)

venkatacrc committed 4 years ago

ff91b169

Change C++ frontend to take optional<Tensor> arguments (#41947)

smessmer committed 4 years ago

1542c41a

Make operators with optional Tensor? arguments c10-full (#41610)

smessmer committed 4 years ago

3a19af24

[JIT] Make create autodiff subgraphs do in place updates to aliasDb (#42141)

Elias Ellison committed 4 years ago

f502290e

refactor canonical ordering to also be able to do isAfter checks (#42140)

Elias Ellison committed 4 years ago

2285a2fc

[Dper3] Implementation of squeezed input to DC++

Jing Ma committed 4 years ago

4fc525e7

[pytorch] include all overloads for OSS custom build

ljk53 committed 4 years ago

a01e91e6

[quant] Use PlaceholderObserver instead of Fp16Observer and NoopObserver (#42348)

supriyar committed 4 years ago

38bf5be2

[quant][graph] Add support for FP16 dynamic quant (#42222)

supriyar committed 4 years ago

6bd46b58

[quant] Add FP16Observer for fp16 quant support (#42221)

supriyar committed 4 years ago

8c5bf102

[quant] Add saturate_to_fp16 op for FP16 quant support (#42147)

supriyar committed 4 years ago

a9eebaf6

Support RowWiseSparseAdam on GPU (#35404)

Yan Xie committed 4 years ago

bdd9ef19

[jit] make clone works for interface type (#42121)

wanchaol committed 4 years ago

a9e7e787

Revert D22812445: Update TensorPipe submodule

ezyang committed 4 years ago

352e15f1

Fix missing attribute when loading model from older version (#42242) (#42290)

sharvil committed 4 years ago

832b1659

[gloo] change ProcessGroupGlooAsyncTest to use gtest (#42313)

Brandon Lin committed 4 years ago

4c6878c9

Make resize_ use normal device dispatch (#42240)

Ailing Zhang committed 4 years ago

0adb5843

Warns when TensorIterator would resize its output (#42079)

Mike Ruberry committed 4 years ago

2f840b16

Enables torch.full bool and integer type inference (#41912)

Mike Ruberry committed 4 years ago

e54f268a

torch.where : Scalar Support (#40336)

kshitij12345 committed 4 years ago

31d41f98

Newer Older