strisunshinewentingwang/pytorch

Pull Requests Commits

allow loading model from a path in torchbench

shunting314 committed 1 year ago

e8feb41f

Add mem efficient backward (#88856)

drisspg committed 1 year ago

35e668b5

Use same retry logic as macos binary builds (#89014)

ZainRizvi committed 1 year ago

f3462833

Take input striding for conv fusion op based on eager output (#88864)

XiaobingSuper committed 1 year ago

7a37bbed

[inductor] fix could not find as_strided with config.triton.mm=triton (#88946)

jspark1105 committed 1 year ago

0544a32b

improving torch.linalg.lstsq documentation formatting (#89013)

wswartworth committed 1 year ago

92c78f37

Fix some naughty uses of reshape/flatten (#88999)

ezyang committed 1 year ago

8df64abc

Revert "support running test_mobile_profiler with buck1/buck2 and OSS (#89001)"

pytorchmergebot committed 1 year ago

c53a5ac6

[testing] fix a key in parse_namespace() (#88969)

kiszk committed 1 year ago

3c3bd55b

[Dynamo] Fix torch.is_tensor and torch.overrides.is_tensor_like (#88704)

yanboliang committed 1 year ago

911a1349

support running test_mobile_profiler with buck1/buck2 and OSS (#89001)

mikey dagitses committed 1 year ago

3b33a279

[CI] Push `latest` and hash+CUDAver tags (#88971)

malfet committed 1 year ago

074278f3

Revert "woof (#89010)"

pytorchmergebot committed 1 year ago

b2082833

ezyang committed 1 year ago

4570bd60

Remove skip (#88979)

anjali411 committed 1 year ago

f8099221

[quant][executorch] Support quant fusion for cat in quant in executorch stack (#88960)

jerryzh168 committed 1 year ago

540b42a1

Fix typos in messages under torch (#88961)

kiszk committed 1 year ago

e0c194f1

wrap_pybind_function: support member function pointers (#88932)

peterbell10 committed 1 year ago

3d79ced8

Fix long comment error on dashboard (#89002)

williamwen42 committed 1 year ago

36d87465

_get_nested_attr should return a value in the general case (#88822)

srossross committed 1 year ago

cdb798fa

[primTorch] _refs & opinfo alpha_dropout (#87989)

khushi-411 committed 1 year ago

f1a5044d

Remove cpu path from lobpcg's basis helper (#88984)

IvanYashchuk committed 1 year ago

b0c86caa

don't use prims.unsqueeze in group_norm (#88927)

ngimel committed 1 year ago

06f1b527

Run test_torchinductor_opinfo CPU tests if triton not installed (#88934)

peterbell10 committed 1 year ago

c8f3d1c1

reland "Do not use unsafe restriding for subclasses (#87610)" (#88343)

bdhirsh committed 1 year ago

ec4eadac

TorchDynamo: skip convolution fusion when convolution's padding is string (#88794)

XiaobingSuper committed 1 year ago

9943d46a

Fake Tensor For (ConvFusion) Propagation (#88414)

XiaobingSuper committed 1 year ago

15ef0660

Revert "Run test_torchinductor_opinfo CPU tests if triton not installed (#88934)"

pytorchmergebot committed 1 year ago

5e6cefd2

Run test_torchinductor_opinfo CPU tests if triton not installed (#88934)

peterbell10 committed 1 year ago

8371bb8a

TorchDynamo: Add convolution binary+unary fusion for cpu in inference mode (#88412)

XiaobingSuper committed 1 year ago

072920c2

Older