strisunshinewentingwang/pytorch

Pull Requests Commits

Revert D24924736: [pytorch][PR] Hipify revamp

malfet committed 3 years ago

8af9f2cc

Add `torch.swapdims` and `torch.swapaxes` (#46041)

kshitij12345 committed 3 years ago

68a3a3f3

[JIT] Pass TypePtr by reference in Argument::type() and Type::isSubtypeOfExt(). (#48061)

Mikhail Zolotukhin committed 3 years ago

d256e388

Document that `remainder` does not support complex inputs (#48024)

kurtamohler committed 3 years ago

df88cc3f

Fix default value of `num_replicas` in DistributedSampler docstring (#48135)

mx-iao committed 3 years ago

0387f2a6

Disable distributed collectives profiling tests (#48129)

rohan-varma committed 3 years ago

140e946f

Small documentation changes for RRef and Dist Autograd (#48123)

H-Huang committed 3 years ago

a6898cb5

Enable complex tests that depend on batched matmul on CUDA (#47910)

IvanYashchuk committed 3 years ago

81b1673a

Install magma on CUDA 11.1 (#48164)

zasdfgbnm committed 3 years ago

3ca4c656

Hipify revamp (#45451)

jithunnair-amd committed 3 years ago

10b490a3

Make numpy optional dependency for torch.cuda.amp (#48154)

malfet committed 3 years ago

1454cbf0

Enable the faster combined weight branch in MHA when query/key/value is same object with nan (#48126)

Guanheng Zhang committed 3 years ago

e2b4c63d

Add max supported SM for nvrtc-11.0 (#48151)

malfet committed 3 years ago

9ead5588

[ROCm] remove sccache wrappers post build (#47947)

jeffdaily committed 3 years ago

21c82397

There should be a newline between BUILD WITH CUDA and NVTX (#48048)

mszhanyi committed 3 years ago

98722ab8

Move kthvalue scalar test to separate method for XLA (#48042)

heitorschueroff committed 3 years ago

2ff748a6

Add type annotations for a few torch.nn.modules (#46013)

guilhermeleobas committed 3 years ago

ca8b9437

Fix inconsistent environment variable naming for setting NVTOOLEXT_HOME in TorchConfig.cmake (#48012)

samuelaldana committed 3 years ago

8c00221f

[TensorPipe] Avoid using deprecated alias for error (#48168)

lw committed 3 years ago

2832e325

[static runtime] Add out_ variant for aten::stack and aten::nan_to_num (#48150)

Hao Lu committed 3 years ago

df0ae244

[quant][graphmode][fx] Keep linear op unchanged when qconfig is not supported (#48067)

jerryzh168 committed 3 years ago

6049653c

Fix test_inverse_singular for cublas path; fix cusolver inverse multi-stream issue (#47026)

xwang233 committed 3 years ago

a1f494cb

[c10d][jit] initial torchbind bindings for ProcessGroupNCCL (#42944)

wanchaol committed 3 years ago

bc484cfe

Revert D24862372: [PyTorch Mobile] Fix for messenger: avoid error with [-Werror,-Wglobal-constructors]

malfet committed 3 years ago

cc611280

Avoid direct reference to at::native::tensor from TensorDataContainer (#47567)

iseeyuan committed 3 years ago

4883d39c

[JIT] Fix function schema subtype checking (#47965)

Meghan Lele committed 3 years ago

c6c6a53b

Added foreach_frac API (#47384)

Iurii Zdebskyi committed 3 years ago

94cd048b

Adding bunch of unary foreach APIs (#47875)

Iurii Zdebskyi committed 3 years ago

134bce7c

fix calculate_extra_mem_bytes_needed_for (#48102)

scottxu0730 committed 3 years ago

0adace37

[PyTorch Mobile] Fix for messenger: avoid error with [-Werror,-Wglobal-constructors]

iseeyuan committed 3 years ago

9392137d

Newer Older