strisunshinewentingwang/pytorch

Pull Requests Commits

Bound shape inferencer: fix int8fc scale and bias

tracelogfb committed 4 years ago

cdd7db1f

List[index]::toOptionalStringRef (#42263)

smessmer committed 4 years ago

b44a10c1

All Gather and gather APIs for Python Objects (#42189)

rohan-varma committed 4 years ago

f22aa601

suppress all Autograd keys in AutoNonVariableTypeMode (#42610)

Basil Hosmer committed 4 years ago

1f689b6e

Skips spectral tests to prevent ROCm build from timing out (#42667)

Mike Ruberry committed 4 years ago

85a00c4c

Delete dead is_named_tensor_only (#42672)

ezyang committed 4 years ago

40b6dacb

Add benchmark for calculate_qparams (#42138)

durumu committed 4 years ago

5ca08b88

Remove VS2017 workaround for autocasting (#42352)

zasdfgbnm committed 4 years ago

79de9c02

Turn on non ASCII string literals serialization (#40719)

bzinodev committed 4 years ago

e28a98a9

[JIT] Clone runOptimizations and similar functions for profiling executor. (#42656)

Mikhail Zolotukhin committed 4 years ago

57854e7f

Add documentation for PYTORCH_JIT_TYPE_VERBOSITY (#42241)

bertmaher committed 4 years ago

a4dbc648

Add fastrnns benchmark to CI and upload data to scribe (#42030)

wconstab committed 4 years ago

65066d77

NVMified NE Eval

ehsanardestani committed 4 years ago

a5af2434

pin numpy version to 1.18.5 (#42670)

jeffdaily committed 4 years ago

049c1b97

And type annotations for cpp_extension, utils.data, signal_handling (#42647)

rgommers committed 4 years ago

bcab2d68

Fix cudnn version on build_environment of Windows CI (#42615)

zasdfgbnm committed 4 years ago

608f99e4

Bump up NCCL to 2.7.6 (#42645)

zasdfgbnm committed 4 years ago

576aab50

Enable C++ RPC tests (#42636)

mrshenli committed 4 years ago

0642d17e

Update TensorPipe submodule (#42522)

lw committed 4 years ago

c30bc6d4

Don't reference TensorPipe headers in our headers (#42521)

lw committed 4 years ago

bd458b7d

Remove ProfiledType (#42570)

Ilia Cherniavskii committed 4 years ago

a53fdaa2

Adds fft namespace (#41911)

Mike Ruberry committed 4 years ago

ccfce9d4

find rccl properly (#42072)

acxz committed 4 years ago

644d787c

Create CuBLAS PointerModeGuard (#42639)

kurtamohler committed 4 years ago

23607441

Implement `gpu_kernel_multiple_outputs` (#37969)

crcrpar committed 4 years ago

eb9ae7c0

[NNC] Add loop unroll transformation (#42465)

asuhan committed 4 years ago

1848b43c

Add __torch_function__ for methods (#37091)

hameerabbasi committed 4 years ago

3d46e02e

Enforce counter value to double type in rowwise_counter

Rui Liu committed 4 years ago

92b7347f

Update docs about CUDA stream priority (#41364)

zasdfgbnm committed 4 years ago

c14fbc36

Fix method stub used for fixing mypy issue to work with pylint (#42356)

mberr committed 4 years ago

ddb8849f

Newer Older