strisunshinewentingwang/pytorch

Pull Requests Commits

CUDA BFloat16 signal windows (#45155)

zasdfgbnm committed 3 years ago

99cf3b1c

Enable complex backward for torch.take() and tensor.fill_() (#46860)

anjali411 committed 3 years ago

13a5be57

vmap fallback: gracefully error out when vmap over dim of size 0 (#46846)

zou3519 committed 3 years ago

02dc52f2

Add NCCL_ASYNC_ERROR_HANDLING to docs (#46856)

osalpekar committed 3 years ago

5e2f17d7

[docs] Changing the wording on quantization versioning and support (#46858)

z-a-f committed 3 years ago

57bf0b59

Added context manager enabling all futures returned by rpc_async and custom build rpc functions to be automatically waited on (#41807)

osandoval-fb committed 3 years ago

58ed60c2

Revert D24486972: [quant][graphmode][fx] Support sigmoid/hardsigmoid/tanh in qat

albanD committed 3 years ago

25db74bf

Update TensorPipe submodule (#46842)

mrshenli committed 3 years ago

0c74b43a

[NVFuser]Benchmark minor update (#46778)

shmsong committed 3 years ago

56a3831b

[quant][graphmode][fx] Support sigmoid/hardsigmoid/tanh in qat (#46738)

jerryzh168 committed 3 years ago

e927b62e

[uhm][0/n] add cuda Mod Op (#46732)

olittle committed 3 years ago

b5662ba0

Add error messages and workaround for RET failure of containers with a torch class type (#46543)

jwpark1985 committed 3 years ago

5a2b537b

Upgrading lcov install to install v1.15 to be compatible with GCC9 (#46847)

janeyx99 committed 3 years ago

3e606da0

Fix LAPACK functionality detection from static OpenBLAS (#46710)

malfet committed 3 years ago

83d358da

Enable dtype arg for torch.linalg.norm with order 'fro' and 'nuc' (#46637)

kurtamohler committed 3 years ago

b61671cc

Update backward formulas (Re #44444) (#46275)

anjali411 committed 3 years ago

d94bd998

Fix hash type (#46769)

r-barnes committed 3 years ago

edbc84aa

Perform explicit cast (#46771)

r-barnes committed 3 years ago

fa8cd06a

Fix compiler warning

r-barnes committed 3 years ago

9cbdd84e

Support doc_string for TorchBind custom classes (#46576)

gmagogsfm committed 3 years ago

f9b94301

Fix type warning (#46770)

r-barnes committed 3 years ago

7d4c1a5a

[quant][eagermode] Add additional_fuser_method_mapping to config (#46355)

jerryzh168 committed 3 years ago

37dbc611

Support hashing of various data types by implementing generic hashing for IValues (#46441)

gmagogsfm committed 3 years ago

13b7855f

Annotate torch.nn.cpp (#46490)

guilhermeleobas committed 3 years ago

789e9353

[pytorch][tensorexpr] Promote integer arguments to sin/cos/tan to float (#46776)

bertmaher committed 3 years ago

c4892c8e

[quant][graphmode][fx] Add support for additional_{fusion/quant}_pattern (#46346)

jerryzh168 committed 3 years ago

343260a1

Use new_zeros in evenly_distribute_backward (#46674)

zou3519 committed 3 years ago

74d81080

Support undefined grads in vmap fallback (#46671)

zou3519 committed 3 years ago

aa828bf0

fix minor bug, message variable does not exist (#46777)

janeyx99 committed 3 years ago

85954164

Enable XNNPACK on Windows & Update XNNPACK (#45830)

peterjc123 committed 3 years ago

89f368be

Newer Older