strisunshinewentingwang/pytorch

Pull Requests Commits

[quant] QNNPACK deconv kernel and tests (#36790)

Zafar Takhirov committed 4 years ago

172f3117

[jit] avoid pre-convert tensor to cpu in pickling (#38898)

wanchaol committed 4 years ago

6c56671f

[quant] QNNPACK deconvolution packing (#37405)

Zafar Takhirov committed 4 years ago

1db4a31d

Fix smoke test jobs (#39638)

peterjc123 committed 4 years ago

ee2bc13f

remove double registered ops (#39609)

linbinyu committed 4 years ago

b06b792b

remove duplicated op schema for aten::pow (#39606)

linbinyu committed 4 years ago

81776373

Optimize GroupNorm on CUDA (#28204)

xiaomengy committed 4 years ago

614dd032

instancenorm: static quant graph mode support (#39096)

vkuzo committed 4 years ago

ebdff07d

groupnorm: graph mode static quant support (#39095)

vkuzo committed 4 years ago

b443ca26

layernorm: eager mode qat support (#39094)

vkuzo committed 4 years ago

952deba8

instancenorm: eager mode QAT support (#39093)

vkuzo committed 4 years ago

b530176d

groupnorm: eager mode QAT support (#39092)

vkuzo committed 4 years ago

202625ba

instancenorm: eager static quant support (#39091)

vkuzo committed 4 years ago

21408742

groupnorm: eager static quant support (#39090)

vkuzo committed 4 years ago

f9b675f7

quant: clean up normalization channels_last handling (#37802)

vkuzo committed 4 years ago

26bc2727

[quant][graphmode] Dynamic quantInsert observers for module output (#39458)

supriyar committed 4 years ago

8a4597b8

[quant][graphmode] Dynamic Quant Do not depend on input shapes (#39412)

supriyar committed 4 years ago

67115b22

[quant][graphmode] Support conv*d_relu in traced models (#39490)

jerryzh168 committed 4 years ago

6d13b583

Move bernoulli_() to DistributionTemplates (#38558)

Kiyosora committed 4 years ago

faf0a3bd

Fix __STDC_FORMAT_MACROS redefinition issue for TypeDerived (#39608)

Kaikai Wang committed 4 years ago

a25b1b91

[pytorch] remove tracing logic from gen_variable_factories.py (#39514)

ljk53 committed 4 years ago

183b04da

[pytorch] add operator name to callBoxed() error message (#39562)

ljk53 committed 4 years ago

9db27a50

[quant][graphmode] Fix add_relu patterns for scripting and tracing (#39455)

jerryzh168 committed 4 years ago

e4627e5d

[Resubmit] Fix argmin/max bug (#39576)

ShawnZhong committed 4 years ago

2da54442

add overload name for aten::as_tensor (#39610)

linbinyu committed 4 years ago

644d6a09

add overload name for str cmp (#39607)

linbinyu committed 4 years ago

b28422d4

Improve DistributedSampler docs and add seed option (#39628)

ssnl committed 4 years ago

479b04e2

Fix circleci postnightly jobs (#39627)

peterjc123 committed 4 years ago

f2af07d7

Update TensorPipe submodule (#39598)

lw committed 4 years ago

53c19423

Implement Quantized Version of Threshold Function (#39352)

Paul Shao committed 4 years ago

6a75f650

Newer Older