pytorch/xla

Pull Requests Commits

Remove one-use, one liner, tensor API.

dlibenzi committed 6 years ago

867ca179

Added aten::svd operation.

dlibenzi committed 6 years ago

b06ebd48

Introduce and use XLA SVD.

dlibenzi committed 6 years ago

18a92743

Let frobenius_norm dispatch through AtenXlaType

asuhan committed 6 years ago

e94b4e1d

Add smooth_l1_loss_backward to ATen XLA tensor

asuhan committed 6 years ago

ac430133

Add smooth_l1_loss to ATen XLA tensor

asuhan committed 6 years ago

774a946b

Use GenericOp to implement eye()

asuhan committed 6 years ago

3d498efd

Add trace to ATen XLA tensor

asuhan committed 6 years ago

439b15b6

Add eye to XLATensor

asuhan committed 6 years ago

c6f932b7

Added aten::norm operations.

dlibenzi committed 6 years ago

63de0b54

Remove PR patch that landed.

dlibenzi committed 6 years ago

6db4fff3

Add at::frac op

vishwakftw committed 6 years ago

168b4e9f

Added aten::scatter API set.

dlibenzi committed 6 years ago

93ab70c9

Add trunc to ATen XLA tensor

asuhan committed 6 years ago

3375ed13

Add missing relu_ test

asuhan committed 6 years ago

96f85b1b

Add leaky_relu_ to ATen XLA tensor

asuhan committed 6 years ago

d17a91c9

Add clamp_{max, min}, clamp_{max_, min_} to ATen XLA tensor

asuhan committed 6 years ago

88d8efa0

Add view_as to ATen XLA tensor

asuhan committed 6 years ago

91639e26

Merge pull request #257 from pytorch/colab

pkanwar23 committed 6 years ago

Verified 2c7566b6

Add fill_ to ATen XLA tensor

asuhan committed 6 years ago

ea187006

Extend XLA tensor view to support partial/sliced updates.

dlibenzi committed 6 years ago

f7a48a90

Changed the Half::from_bits PR to use constexpr function.

dlibenzi committed 6 years ago

c9b13131

Added Nx TPU replication test.

dlibenzi committed 6 years ago

6bface05

Add API to list all available devices.

dlibenzi committed 6 years ago

159cf6c9

Use XLA build versioning only when needed, to avoid spamming the Conda env folder.

dlibenzi committed 6 years ago

45855192

Add inplace version for ne, eq

vishwakftw committed 6 years ago

1909c002

Use at::rand for GetTestTensor, remove GetTestTensor

vishwakftw committed 6 years ago

5bf2fe7d

Remove redundant self_tensor instantiation

asuhan committed 6 years ago

e56ea1b1

Add more arithmetic with scalar to ATen XLA tensor

asuhan committed 6 years ago

d23cfaff

Fix formatting of aten_xla_type.cpp

asuhan committed 6 years ago

ef1accea

Older