strisunshinewentingwang/pytorch

Pull Requests Commits

Emit warning from deprecated torch function signatures (#31514)

peterbell10 committed 4 years ago

0e5a6700

Ensure autograd callbacks are called only once for reentrant backward. (#31909)

pritamdamania committed 4 years ago

5cc62f29

Support PyTorch ROCm CI on Ubuntu18.04 (#31886)

iotamudelta committed 4 years ago

4ee9c562

Raise ValueError if CUDA device is specified without specifying the : (#29087)

v0dro committed 4 years ago

2f5eefe5

Don't unconditionally compile runJITCPPTests (#31236)

ezyang committed 4 years ago

3c7db5cc

Enable personalized FC weight_init and sparse_emb weight_init (#31707)

Fei Tian committed 4 years ago

809ee9d0

Use TORCH_CHECK instead of AT_ASSERT in torch::cuda::gather() (#27456)

andreaskoepf committed 4 years ago

22044c6f

Add stub for transformer.py and MultiheadAttention Class. (#28396)

yyb1995 committed 4 years ago

20c5dd59

Update all instances of 1.4.0 -> 1.5.0 (#31785)

seemethere committed 4 years ago

346a3491

Enable BFloat16 support for Convolutions on ROCm (#30948)

rohithkrn committed 4 years ago

985fd970

minor doc tweak to use mp.spawn in example (#30381)

rohan-varma committed 4 years ago

a561a844

Don't handle bias inside cudnn_convolution* (#31524)

zasdfgbnm committed 4 years ago

34561dad

no_grad, enable_grad: support for decorating generator functions (#31792)

peterbell10 committed 4 years ago

5d80f634

Add missing TORCH_CUDA_API annotation to throw_nccl_error (#31157)

ezyang committed 4 years ago

58cffbff

Remove dead CAFFE2_LIBS variable (#31155)

ezyang committed 4 years ago

4ef9daf7

Remove LibIRC logic from cmake. (#31152)

ezyang committed 4 years ago

a9dae70b

Fix index put (#31552)

neginraoof committed 4 years ago

112196fd

Enable constant folding for Reshape (#31054)

neginraoof committed 4 years ago

78cba90a

Fix androidTest - exclude host tests from it

IvanKobzarev committed 4 years ago

492ca46e

Add a check method for custom type tensor (#31290)

Yinghai Lu committed 4 years ago

c65305e9

Refactor tests in pytorch's test/dist_autograd_test.py file (#31803)

fkhan1337 committed 4 years ago

1f2b6d63

fixed scale_factor calculation for uint8 tensor (#31778)

anjali411 committed 4 years ago

ddff014b

C++ added 3rd arg of false to BatchNorm/InstanceNorm register_parameter … (#31873)

meganset committed 4 years ago

1ba1799a

Revert D18643137: Implement backend-agnostic rpc._wait_all_workers() utility

mrshenli committed 4 years ago

33430cf0

Provide async mode for local autograd engine. (#31230)

pritamdamania committed 4 years ago

fde94e75

corrected keyword argument name in docs for Tensor.scatter (#31617)

ShnitzelKiller committed 4 years ago

3f0b3307

Updating submodules

svcscm committed 4 years ago

9020d30f

Implement backend-agnostic rpc._wait_all_workers() utility (#30710)

xush6528 committed 4 years ago

502533cf

Move prim ops from JIT registration to C10 (#30612)

Martin Yuan committed 4 years ago

f362cd51

Enable foldbn tests (#29220)

jerryzh168 committed 4 years ago

55796115

Newer Older