pytorch/xla

Pull Requests Commits

Typo fix in docstr. (#2634)

taylanbil committed 5 years ago

Verified f5d679a2

Remove backward of layer_norm and group_norm and add double backward … (#2631)

ailzhang committed 5 years ago

Verified 966a5381

Fix `global_device_map` sorting (#2630)

jysohn23 committed 5 years ago

Verified c8a725c9

Remove native_layer_norm since it's upstreamed. (#2626)

ailzhang committed 5 years ago

Verified 1f11dd73

Revert "revert empty() signature to TensorOptions version" (#2571)

ailzhang committed 5 years ago

Verified 58ccf11c

Add dynamic shape support for nms op (#2619)

JackCaoG committed 5 years ago

Verified 40020fcc

Disable test_index_put_accumulate_large_tensor_xla on GPU due to memory issue (#2625)

JackCaoG committed 5 years ago

Verified aa0b6602

Disable test_kthvalue_xla_* to unblock the master build (#2624)

JackCaoG committed 5 years ago

Verified b1594dad

Bring in new tf (#2621)

JackCaoG committed 5 years ago

Verified ae4dd3c1

Fix issue with any/all when all reduce dimensions of input have size 1 (#2590)

JackCaoG committed 5 years ago

Verified a917acb6

Modify the cpp Enisum test to check different xla counter (#2600)

JackCaoG committed 5 years ago

Verified 5290dcf6

Disable test_copysign_xla_bfloat16 for precision issue (#2618)

JackCaoG committed 5 years ago

Verified 29e2969c

Switch Colab notebooks to 1.7 pytorch/xla wheels. (#2596)

zcain117 committed 5 years ago

Verified 02e27000

Add support for UpsampleNearest2D/backward with scale (#2597)

JackCaoG committed 5 years ago

Verified 1da19b07

Revert "Revert "Accept int32 as index for index_add op"" (#2595)

JackCaoG committed 5 years ago

Verified 2f846fd2

Revert "Accept int32 as index for index_add op (#2580)" (#2594)

JackCaoG committed 5 years ago

Verified 2cd4f072

Skip test_muldiv_scalar_xla_bfloat16. (#2593)

ailzhang committed 5 years ago

Verified ae01130a

Minor cleanups (#2581)

jysohn23 committed 5 years ago

Verified 9917e41f

Accept int32 as index for index_add op (#2580)

JackCaoG committed 5 years ago

Verified a61eb490

Fix max_pool2d padding in ceil_mode under some corner cases (#2583)

JackCaoG committed 5 years ago

Verified 12d6d969

Update README with 1.7 version. (#2586)

zcain117 committed 5 years ago

Verified 016c8700

Add DCGAN colab to Readme. (#2579)

taylanbil committed 5 years ago

Verified 12b4a285

Disable test_EmbeddingBag_empty_per_sample_weights_and_offsets_xla (#2572)

JackCaoG committed 5 years ago

Verified 82d5850a

Revert `TF_CPP_MIN_LOG_LEVEL` change and document its usage (#2568)

jysohn23 committed 5 years ago

Verified 87cb0714

Handle Duplicated tensors in sync list and interop tensors (#2559)

JackCaoG committed 5 years ago

Verified be999bdb

Fix exponential samples 0 (#2563)

JackCaoG committed 5 years ago

Verified f572937c

Handle the nan genreated by inf/-intf in nll_loss (#2555)

JackCaoG committed 5 years ago

Verified 5cb46329

Revert "Revert "Rename math to default to align with RegistrationDeclarations.h."" (#2554)

ailzhang committed 5 years ago

Verified 53eb3fc5

Delete .torch_pin (#2551)

JackCaoG committed 5 years ago

Verified 1f0e917d

Skip tensors which already have data in XLATensor::GatherTensorsXlaData (#2544)

asuhan committed 5 years ago

Verified 44268fa4

Newer Older