pytorch/xla

Pull Requests Commits

resolve merge conflicts and switched to 1e-6

zcain117 committed 5 years ago

9a91ffda

Added more meaningful device spec print for multi-processing.

dlibenzi committed 5 years ago

68f9db8b

Added MP API documentation.

dlibenzi committed 5 years ago

45b9e4a2

add issue templates (#1036)

ailzhang committed 5 years ago

Verified 7d0abbc7

Use static for auto-unregistration object.

dlibenzi committed 5 years ago

429c89ad

draft contribution guide for discussion (#1033)

ailzhang committed 5 years ago

Verified fb6b0788

Fixed the XLA _out() code generation and added a few more overrides.

dlibenzi committed 5 years ago

040166bc

Ailing Zhang committed 5 years ago

a7e47149

Ailing Zhang committed 5 years ago

e672f3ec

Allow reduce operations in every point of the graph.

dlibenzi committed 5 years ago

acff0b9a

allow update any env (#1028)

taylanbil committed 5 years ago

Verified 8bc314b2

Added XLA multi-processing support and API.

dlibenzi committed 5 years ago

1f45f0a3

log optimizers learning rate

zcain117 committed 5 years ago

6db60c95

git checkout origin/master third_party/tensorflow

zcain117 committed 5 years ago

47c7ecb6

pass scheduler args; not FLAGS

zcain117 committed 5 years ago

585fa91e

zcain117 committed 5 years ago

c5d5384b

fetch third_party/tensorflow

zcain117 committed 5 years ago

10622f23

git rm -r --cached third_party/tensorflow

zcain117 committed 5 years ago

d31ec4d0

fix num devices

zcain117 committed 5 years ago

e7ef064c

zcain117 committed 5 years ago

092c60f3

remove unused constant

zcain117 committed 5 years ago

824bc6bf

model args and naming

zcain117 committed 5 years ago

c302fddf

Merge branch 'master' of https://github.com/pytorch/xla into add-imagenet-learning-rate-scheduler

zcain117 committed 5 years ago

fd6ae7e4

match pytorch api

zcain117 committed 5 years ago

7b4b0728

Make add_scatter_combiner a std::function

asuhan committed 5 years ago

271c6fcf

allow copy.copy() to return xla tensor (#1018)

ailzhang committed 5 years ago

Verified 2e67d89a

Fix conv backward hash population and ToString formatting.

dlibenzi committed 5 years ago

2ba4ec4c

Fix convolution ToString() formatting.

dlibenzi committed 5 years ago

856b864a

Merge branch 'master' of https://github.com/pytorch/xla into add-imagenet-learning-rate-scheduler

zcain117 committed 5 years ago

fa5d2333

Override is_floating_point() and is_signed().

dlibenzi committed 5 years ago

da89b09c

Older