pytorch/xla

Pull Requests Commits

Add support to auto-generate *_out() functions.

dlibenzi committed 6 years ago

96869445

Add index_put, index_put_ to ATen XLA tensor

asuhan committed 6 years ago

f6fc28ad

Remove landed 17847 PR.

dlibenzi committed 6 years ago

55f74f0e

Fix changes in PyTorch argmax/argmin APIs.

dlibenzi committed 6 years ago

d684efe2

Make IR graph trim parameters configurable.

dlibenzi committed 6 years ago

9478f472

Fix std::string hash.

dlibenzi committed 6 years ago

17951fab

Add shape caching to more IR ops.

dlibenzi committed 6 years ago

25c4064a

Bridge in autograd PR.

dlibenzi committed 6 years ago

bf8a1770

Avoid expensive shape computations while building IR nodes.

dlibenzi committed 6 years ago

9085c8f6

Drop changes tools/build_pytorch_libs.py which landed

asuhan committed 6 years ago

ff09384e

Simplify an open coded loop.

dlibenzi committed 6 years ago

fbbf2f58

Add window factories to ATen XLA tensor

asuhan committed 6 years ago

df327ef9

Add pixel_shuffle to ATen XLA tensor

asuhan committed 6 years ago

9d376f33

Add sum_to_size to ATen XLA tensor

asuhan committed 6 years ago

a089e862

Add bilinear to ATen XLA tensor

asuhan committed 6 years ago

d0313543

Fix optimizer breakage

asuhan committed 6 years ago

3da48bf6

Add flatten to ATen XLA tensor

asuhan committed 6 years ago

6997b33d

Add diagflat to ATen XLA tensor

asuhan committed 6 years ago

de444b1e

Add pinverse to ATen XLA type

asuhan committed 6 years ago

b44eb79e

Fix diag on rank 1

asuhan committed 6 years ago

ae68a9b2

Fix the shape of the diagonal returned by SVD

asuhan committed 6 years ago

b931a6d2

Fix where implementation

asuhan committed 6 years ago

5059ebc8

Overlap XLA tensor sync with future XLA tensor work.

dlibenzi committed 6 years ago

10b8e428

Fix linear and softplus test tolerance on TPU

asuhan committed 6 years ago

2911d612

Finish up the ordering and do the XLA ATEN tensor sources as well.

dlibenzi committed 6 years ago

8097525c

Add argsort to ATen XLA tensor

asuhan committed 6 years ago

e5874ba3

Add one_hot to ATen XLA tensor

asuhan committed 6 years ago

ca7297d3

Add min, max (without indices) to ATen XLA tensor

asuhan committed 6 years ago

143d89d6

Adjust to removal of elementSizeInBytes in core PyTorch

asuhan committed 6 years ago

c0ec8392

Add {hard, soft}shrink[_backward] to ATen XLA tensor

asuhan committed 6 years ago

c3b32336

Older