strisunshinewentingwang/pytorch

Pull Requests Commits

Replace flatten tensors with flatten loops. (#46737)

navahgar committed 3 years ago

4b6e3071

[quant][graphmode][fx] Support sigmoid/hardsigmoid/tanh in qat (#46738) (#46871)

jerryzh168 committed 3 years ago

6b50ccc4

Add single element tuple output from to_backend/to_glow (#5029)

jackm321 committed 3 years ago

60eded6c

Add a warning message that torch.sign would not support complex numbers (#43280)

xuhdev committed 3 years ago

bcbb6bac

add fburl link to error message (#46795)

linbinyu committed 3 years ago

37da6d26

Fix TripletMarginWithDistanceLoss example code (#46853)

soulitzer committed 3 years ago

9858b012

[c10] fix weak_intrusive_ptr lock() (#46007)

wanchaol committed 3 years ago

4a35280e

Remove loop_test mode (#46618)

khabinov committed 3 years ago

b3e64c86

Add Vulkan Tensor factory. (#44016)

Ashkan Aliabadi committed 3 years ago

af27da93

Add Vulkan Tensor. (#44015)

Ashkan Aliabadi committed 3 years ago

c9bf03a6

[pytorch] Improve/fix heuristics for using mkldnn vs native conv (#46675)

bertmaher committed 3 years ago

2397c8d1

[ROCm] fix bug in miopen findAlgorithm. (#46852)

jeffdaily committed 3 years ago

a602811d

Fix unused variable warning (#46838)

r-barnes committed 3 years ago

a4adc1b6

[Gradient Compression] Refactor CommHookInterface and PythonCommHook. (#46512)

Yi Wang committed 3 years ago

a6cd294c

Support RRef.backward() for local RRefs. (#46568)

pritamdamania committed 3 years ago

adafd3d4

CUDA BFloat16 gelu, hardswish, hardsigmoid (#44997)

zasdfgbnm committed 3 years ago

7731370e

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

Newer Older