strisunshinewentingwang/pytorch

Pull Requests Commits

[inductor][triton] if device is a torch.device, then make cuda_properties index it correctly

soumith committed 3 years ago

b6432340

Revert "[BE] Get rid of deprecation warnings in workflows (#87152)"

pytorchmergebot committed 3 years ago

11915b31

[triton] allow cuda properties to be queried from workers (#87101)

zdevito committed 3 years ago

d36c284d

[BE] Get rid of deprecation warnings in workflows (#87152)

malfet committed 3 years ago

9da032ec

Revert "[BE] Get rid of deprecation warnings in workflows (#87152)"

pytorchmergebot committed 3 years ago

66658e1d

[Inductor] Lift the maximum depth of the Python interpreter stack to adapt large/deep models (#87130)

yanboliang committed 3 years ago

8ca7820e

[BE] Get rid of deprecation warnings in workflows (#87152)

malfet committed 3 years ago

acaf484f

Enable sdp_forward for NestedTensors (#86720)

drisspg committed 3 years ago

5fb68718

Use conda-forge in mac mps test (#87155)

huydhn committed 3 years ago

74138a8d

[vulkan] Use 2D texture types for convolution weights and biases (#86972)

ssjia committed 3 years ago

9d1a8edc

[vulkan] Enable 2D texture types (#86971)

ssjia committed 3 years ago

5b588036

Check all CUDA API calls in aten/src/ATen/test for errors (#74919) (#83556)

r-barnes committed 3 years ago

a7ed398c

Install blas from conda-forge (#87150)

huydhn committed 3 years ago

f02f0e3a

Small update to Module note (#87142)

albanD committed 3 years ago

9db7270e

Enable UBSAN mode for test_jit (#85735)

mehtanirav committed 3 years ago

fb614b18

[ci] put more logs in a folded group (#86138)

Catherine Lee committed 3 years ago

18cc00d3

remove dynamo hash updates (#87092)

Catherine Lee committed 3 years ago

e3b84f6c

Don't only apply DDP optimizer on forward frames (#87097)

davidberard98 committed 3 years ago

4fd98dfe

Add venv to gitignore (#86702)

justinchuby committed 3 years ago

09d72091

[DataPipe] Fixing interface generation in setup.py (#87081)

NivekT committed 3 years ago

0cb273b5

[ci] fix bot comment (#87127)

suo committed 3 years ago

f5ee2d88

[Docs] Remove outdated comment for sparse all-reduce (#87018)

Andrew Gu committed 3 years ago

f552eee4

handle libomp update on circleci (#86979)

Catherine Lee committed 3 years ago

d023e839

Use 12xlarge for nightly cpp doc generation job (#86859)

huydhn committed 3 years ago

5acf6e0e

[dynamo] Introduce `get_real_value` API to TensorVariable (#87091)

suo committed 3 years ago

48142707

[docs] Fix ScalarTensor __repr__ in Extending PyTorch example (#86330)

jmargeta committed 3 years ago

e85dbcc9

[Dynamo] More robust pyop support, module properties as args (#87020)

voznesenskym committed 3 years ago

b8007742

[ONNX] Ignore print(Tensor) during tracing (#86223)

Thiago Crepaldi committed 3 years ago

1167949b

Workarounds for cudnn_batch_norm with TorchRefsNvfuserCapabilityMode (#86796)

IvanYashchuk committed 3 years ago

31931515

add XLA backend into tensor type strings (#86881)

holimion committed 3 years ago

33343def

Older