strisunshinewentingwang/pytorch

Pull Requests Commits

[jit] bind backward in TorchScript

wanchaol committed 5 years ago

e596a9e3

Enable OSS quantization tests (#23858)

James Reed committed 5 years ago

40f0b1c8

Add flag to temporarily disable MKL-DNN conv (#23837)

James Reed committed 5 years ago

6ba60ec9

weight_names bug fix (#23848)

amylittleyang committed 5 years ago

9588cd92

format init.cpp (#23840)

Elias Ellison committed 5 years ago

d413f2d3

Updating submodules

svcscm committed 5 years ago

43c4bcba

Docs: Delete placeholder to use top-level file (#23869)

hugovk committed 5 years ago

52be1448

Fix SliceGradientOp to handle properly empty batches (#23784)

kennyhorror committed 5 years ago

d58059bc

Delete Travis CI config (#23788)

ezyang committed 5 years ago

c002ede1

Define toIValue conversion for dtype (#23708)

James Reed committed 5 years ago

489cc466

Rename previously THNN conv kernels to have naive_ prefix.

ezyang committed 5 years ago

e8cf9b68

Remove nightly suffix from nightlies; upload to pytorch-nightly. (#23752)

ezyang committed 5 years ago

60a4ef30

Properly mangle `nn.Module.__construct` (#23779)

suo committed 5 years ago

e2f5bc5c

fix some compiler warnings

suo committed 5 years ago

8fc349f7

fix torch.frac documentation (#23830)

Ailing Zhang committed 5 years ago

51d59a43

make nn.LSTM accept PackedSequence instead of Tuples

wanchaol committed 5 years ago

8fb0d198

Negate halves on GPU using __hneg() when possible, instead of using float conversion.

xuhdev committed 5 years ago

a1584555

Recommend `~` and `bitwise_not()` when user tries to apply neg (`-`) on a bool tensor.

xuhdev committed 5 years ago

f90afff3

Std opset export (#22310)

neginraoof committed 5 years ago

f278aee7

Add CUDA 10.1 to CI. (#23791)

ezyang committed 5 years ago

a710f816

galv committed 5 years ago

0015b188

Remove unnecessary fetch and reset on builder checkout. (#23792)

ezyang committed 5 years ago

44ba092e

Revert D16627326: [pytorch][PR] [ROCm] Improve hip-clang support in build_amd.py

ezyang committed 5 years ago

4050de5b

Adam implementation minor fix (#23737)

farhadrgh committed 5 years ago

ab15d384

Revert D15996322: Open up AliasAnalysisKind for any ops

suo committed 5 years ago

8e9fef61

Improve hip-clang support in build_amd.py (#23699)

yxsamliu committed 5 years ago

f0a58180

Open up AliasAnalysisKind for any ops (#23810)

smessmer committed 5 years ago

3ad9dbf9

Document bool tensors for bitwise_not. (#23800)

xuhdev committed 5 years ago

1aa4afde

Channels last stored in tensor (#23391)

VitalyFedyunin committed 5 years ago

6e4a83ab

Revert D16452539: support Gather different indices for different examples in one batch

suo committed 5 years ago

a3c165f9

Older