strisunshinewentingwang/pytorch

Pull Requests Commits

Fix JIT test config (#46230)

malfet committed 3 years ago

4ad47156

Fix incorrect CUDA `torch.nn.Embedding` result when max_norm is not None and indices are not sorted (#45248)

kurtamohler committed 3 years ago

66505b64

[fx] use a linked list for nodes (#45708)

zdevito committed 3 years ago

88dcb95e

Adding information how to control randomness with DataLoader (#45749)

VitalyFedyunin committed 3 years ago

31ee5d8d

[pytorch][PR][Gradient Compression] Reduce the peak memory of fp16 compression provided by ddp comm hook (#46078)

Yi Wang committed 3 years ago

ee3d3e6d

[pt][quant] Support either min or max in qclamp (#45937)

dskhudia committed 3 years ago

87a4baf6

Implement ravel (#46098)

ejguan committed 3 years ago

bed3b405

fix test_serialization not working with Windows. (#46120)

Rong Rong committed 3 years ago

b98e3594

[NNC] Fix two more bugs in Cuda Half support (#46129)

nickgg committed 3 years ago

f3db6877

adding complex support for distributed functions and . fix #45760 (#45879)

bdhirsh committed 3 years ago

c02efdef

clean up dataclasses installation to only <3.7 (#46182)

Rong Rong committed 3 years ago

8de9aa19

including tensorexpr tests in CI for all configs (#46188)

janeyx99 committed 3 years ago

ba78eb80

[caffe2] add PlanExecutorTest ErrorPlanWithCancellableStuckNet (#46110)

dahsh committed 3 years ago

85c3ba55

Made exception message for torch.LongTensor() legacy constructor more readable (#46147)

partypyro committed 3 years ago

8d5256e6

Improve error checking of Storage._writeFile. (#46036)

gchanan committed 3 years ago

2070834b

Fix error in Binomial to retain lazy logit initialization (#46055)

neerajprad committed 3 years ago

9202c443

Fix typing errors in the torch.distributions module (#45689)

xuzhao9 committed 3 years ago

146721f1

[quant][test] Add mul_scalar test (#46106)

jerryzh168 committed 3 years ago

6a001dec

Only use hacky_wrapper_for_legacy_signatures if an op needs it (#45742)

smessmer committed 3 years ago

6ba6ecb0

Fix mkldnn build on legacy x64 arch (#46082)

pinzhenx committed 3 years ago

e1f74b18

[fix] torch.kthvalue : handle non-contiguous CUDA tensor (#45802)

kshitij12345 committed 3 years ago

a8142316

TensorInferenceFunction checks

John Lundell committed 3 years ago

3883cdb8

[caffe2] Fix preprocessor checks for FMA

marksantaniello committed 3 years ago

1a99689d

Automated submodule update: FBGEMM (#46151)

facebook-github-bot committed 3 years ago

bbb3f093

Fix mistakes and increase clarity of norm documentation (#42696)

kurtamohler committed 3 years ago

a0a8bc88

[TensorExpr] Disable and/or fix some failing tests. (#46146)

Mikhail Zolotukhin committed 3 years ago

496d72d7

[Caffe2] use the real new fbgemm sparse adagrad interface (#46132)

jspark1105 committed 3 years ago

4c87d337

a few more comments on dispatch key computation methods (#46128)

Basil Hosmer committed 3 years ago

9f743015

[pytorch][te] Add compilation time benchmark (#46124)

bertmaher committed 3 years ago

b7261de0

[JIT] Add dynamic shape benchmark for NV Fuser (#46107)

shmsong committed 3 years ago

43fe45ab

Newer Older