strisunshinewentingwang/pytorch

Pull Requests Commits

[CUDA graphs] Fixes errors in RNG seed (#84967)

Aidyn-A committed 2 years ago

5271494e

[numpy] add `torch.concatenate`, alias of torch.cat (#82946)

khushi-411 committed 2 years ago

270e5e51

Revert "Create Cache for Fusion Reuse in NVFuser in Python Frontend for Primtorch (#83267)"

pytorchmergebot committed 2 years ago

94b67f4c

[MPS] Fix mps to cpu casting from a smaller dtype to a bigger dtype (#84928)

DenisVieriu97 committed 2 years ago

4247cc98

Test tracing consecutive comms on the same input tensor (#84980)

mrshenli committed 2 years ago

1a81ab3b

Remove eager mode support form CommTensor (#84978)

mrshenli committed 2 years ago

0f300592

[ROCM] test_batchnorm_cudnn_nhwc (#84603)

micmelesse committed 2 years ago

b6d6a78c

Revert "Python Dispatcher integration with C++ dispatcher (#84826)"

pytorchmergebot committed 2 years ago

706b9903

[2/N] [Dispatchable Collectives] Extract ProcessGroup::Work into a separate class and update references (#83680)

H-Huang committed 2 years ago

74ead619

Upgrade to CUDNN version for cuda 11.7 (#84964)

atalman committed 2 years ago

54c46e4f

Skip validate_view_consistency for nvFuser tests (#84858)

IvanYashchuk committed 2 years ago

6750946b

Python Dispatcher integration with C++ dispatcher (#84826)

voznesenskym committed 2 years ago

35f6a691

[quant][docs] Add example for the error message for fixed qparam ops (#84666)

jerryzh168 committed 2 years ago

44c30c5d

Remove enable_recursive_torch_dispatch (#84945)

ezyang committed 2 years ago

55ca297d

Removes unnecessary namespace of functions used only in einsum (#84955)

janeyx99 committed 2 years ago

922560b8

[vision hash update] update the pinned vision hash (#84975)

pytorchmergebot committed 2 years ago

d26e9cd9

[torchdynamo hash update] update the pinned torchdynamo hash (#84912)

pytorchmergebot committed 2 years ago

b28d82cb

Fix `expectedFailureMeta` to avoid skipping tests (#84875)

kurtamohler committed 2 years ago

d00cabae

Avoid nested CommTensor wrapping (#84963)

mrshenli committed 2 years ago

8cbbd3a2

Revert "Don't detach when making views; force caller to detach (#84893)"

pytorchmergebot committed 2 years ago

8ca057eb

[ROCm] re-enable tensorexpr and test_openmp (#81367)

arindamroy-eng committed 2 years ago

a185dc2e

Updated library level maintainers for torchtext (#84950)

Nayef211 committed 2 years ago

cb9ef466

[CMake] Add functorch target (#83464)

malfet committed 2 years ago

d05a1133

`ReflectionPad` supports `BFloat16` (#84949)

crcrpar committed 2 years ago

26b59862

[Profiler] Make `LibKinetoClient::stop()` directly call `ProfilerStateBase::pop` (#83965)

Taylor Robie committed 2 years ago

fdd36654

Don't detach when making views; force caller to detach (#84893)

ezyang committed 2 years ago

3bb8d6a9

Create Cache for Fusion Reuse in NVFuser in Python Frontend for Primtorch (#83267)

kevinstephano committed 2 years ago

ec916bf6

Adds mruberry as superuser (#84869)

mruberry committed 2 years ago

59bb5c93

disable onednn gelu for empty input (#84926)

XiaobingSuper committed 2 years ago

c61e8954

Updating cudnn_frontend to 0.7.1 (#84943)

atalman committed 2 years ago

25d91e0a

Newer Older