strisunshinewentingwang/pytorch

Pull Requests Commits

Fix incorrect runtime error in mul_() when the tensor layout is Mkldnn (#51758)

kazutoshi-ak committed 3 years ago

0b939740

Revert D26644079: [pytorch][PR] Adding functional way of stacking DataPipes

H-Huang committed 3 years ago

da732c76

[vulkan] Add nonVarTypeModeGuard to vulkan tests and speed_benchmark_torch (#52535)

SS-JIA committed 3 years ago

c2558b4b

Use touch() in pathlib for better compatibility on Windows (#52729)

skyline75489 committed 3 years ago

e94940b1

quant: fix conv transpose with qconfig == None (#52844)

vkuzo committed 3 years ago

19a8ada8

Added torch.no_grad() to update_bn (#52654)

MiHarsh committed 3 years ago

c871abec

Upgrade onednn to v.1.8.1 (#51184)

XiaobingSuper committed 3 years ago

0e86f14e

Adding functional way of stacking DataPipes (#52507)

VitalyFedyunin committed 3 years ago

7972036b

Expose Store's timeout and TCPStore's host and port in Python API (#52784)

cbalioglu committed 3 years ago

a11b6011

Test for distributed RL with RPC (#52393)

jbschlosser committed 3 years ago

f974cf46

Fix TensorPipe agent trying to double-set error (#52837)

lw committed 3 years ago

163a91be

Update TensorPipe submodule (#52677)

lw committed 3 years ago

3ff6c917

Add scatter_add to amp promote list (#52133)

pbialecki committed 3 years ago

39fa0b5d

fix(docs): remove redundant hardsigmoid() in docstring to show up `inplace` parameter (#52559)

Jeff Yang committed 3 years ago

316eabe9

ns for fx: update graph matching to handle dicts and tuples in node args (#52681)

vkuzo committed 3 years ago

1618dc2a

ns for fx: update graph matching to not match nodes with equal types (#52402)

vkuzo committed 3 years ago

608f44b2

ns for fx: support linear_relu for weight matching (#52395)

vkuzo committed 3 years ago

4483c48e

ns for fx: allow graph matching of parents of cat (#52368)

vkuzo committed 3 years ago

64b4e37c

[Pytorch, sparsity] Bug fix to update requantization and zp parameters of input (#52797)

kimishpatel committed 3 years ago

13121598

Lower ReLu6 to aten (#52723)

bmon581 committed 3 years ago

99a428ab

Update backwards compatibility check to ignore reverted op (#52841)

lw committed 3 years ago

fa7575ea

Fix typos in tools/test_history.py helpstring (#52840)

samestep committed 3 years ago

91412690

Fix RPC get_worker_info for rank=0 (#52804)

mrshenli committed 3 years ago

1ac59d9d

Store test filename in test report path (#52791)

janeyx99 committed 3 years ago

f71d9e28

Revert D26375734: Implemented torch.linalg.multi_dot

lw committed 3 years ago

92a4ee1c

remove index_fill side-effect for scalar tensors (#52209)

nikitaved committed 3 years ago

0048d97e

[TensorExpr] Add `Placeholder::handle` method to get the corresponding `BufHandle`. (#52793)

Mikhail Zolotukhin committed 3 years ago

57947c5d

[TensorExpr] Add an unmasked `Load` constructor. (#52790)

Mikhail Zolotukhin committed 3 years ago

d3b427a0

Introduce `mlc` device (ML Compute device) to PyTorch's device list (#50634)

bmon581 committed 3 years ago

30cb6ac5

Drop unused variables (#52643)

r-barnes committed 3 years ago

2bdf6305

Newer Older