strisunshinewentingwang/pytorch

Pull Requests Commits

Enable autograd wrt sparse CSR tensors

IvanYashchuk committed 2 years ago

bba47802

[PyTorch] Add NestedTensorCPU and NestedTensorCUDA dispatch keys (#75808)

swolchok committed 2 years ago

0a5e788a

[GH1] Add linalg related changes to merge rules

janeyx99 committed 2 years ago

a5cb0d6b

Revert "ReReland Fix public binding check for modules with `__all__`"

pytorchmergebot committed 2 years ago

347ea626

[quant][core][gpu][improvement] Suported int8 matmul for quantized linear cudnn op

dzdang committed 2 years ago

982be196

Revert "Remove `-Wno-unused-variable` from utils.cmake"

pytorchmergebot committed 2 years ago

5c56b228

Update allowlist

albanD committed 2 years ago

5635af45

Revert "Port `sort` to structured kernels."

pytorchmergebot committed 2 years ago

58782151

Remove extra conv OpInfos

ngimel committed 2 years ago

4787adb5

[easy][PTE] Remove GetMutableSizePrefixed* functions

pavithranrao committed 2 years ago

c3e67d8a

Enhance exporting torch.minimum() function to ONNX so it can handle parameters with different dtype.

fatcat-z committed 2 years ago

166568d4

[GHF] Exclude 5f37e5c2a39c3acb776756a17730b865f0953432 from sync

malfet committed 2 years ago

af13797c

[complex32] jiterator support

kshitij12345 committed 2 years ago

e23cbd63

Update symbolics policy to emit aten::ATen for Caffe2 build only

Thiago Crepaldi committed 2 years ago

eab3f428

ReReland Fix public binding check for modules with `__all__`

albanD committed 2 years ago

d2517a43

Port `sort` to structured kernels.

ysiraichi committed 2 years ago

8c37a056

[quant][fx] Move backend_config folder to torch.ao.quantization

jerryzh168 committed 2 years ago

74454bdb

Revert "Reland Fix public binding check for modules with `__all__`"

pytorchmergebot committed 2 years ago

cbb9b33c

Remove `-Wno-unused-variable` from utils.cmake

malfet committed 2 years ago

018cbe1f

add operator header

george-qi committed 2 years ago

f5517761

Reland Fix public binding check for modules with `__all__`

albanD committed 2 years ago

2e5e4be7

Reformat tools/codegen with black

ezyang committed 2 years ago

0f98fc05

Annotate some long lines with noqa: B950

ezyang committed 2 years ago

75410682

Fix complex to real casting warning in _to_copy backward

peterbell10 committed 2 years ago

cc56fac2

Fix conjugate bit discrepancy in composite compliance

peterbell10 committed 2 years ago

58fb3f01

[ROCm] enable composite compliance backward tests

jeffdaily committed 2 years ago

e587c8bc

Register torch.return_types.* as pytree nodes

zou3519 committed 2 years ago

18c74d10

torch.autograd.grad needs an extra tuple when handling single outputs and is_grads_batched=True

HarborYuan committed 2 years ago

45ac08b3

Enable windows tests for nvfuser

davidberard98 committed 2 years ago

cd0fdcca

albanD committed 2 years ago

bd7e99cb

Newer Older