pytorch/xla

Pull Requests Commits

Rotate the nvidia repo key in the base docker image (#3538)

yeounoh committed 3 years ago

Verified f684fd43

Move GetXlaTensorOrCreateForWrappedNumber (#3534)

JackCaoG committed 3 years ago

Verified facc7126

Revert "Start XRT server in a separate process in CircleCI CPU test (#3519)" (#3536)

JackCaoG committed 3 years ago

Verified 093d443b

Move device lock before the execution instead of tensor gathering (#3457)

JackCaoG committed 3 years ago

Verified 5d1c4210

Start XRT server in a separate process in CircleCI CPU test (#3519)

JackCaoG committed 3 years ago

Verified 935b6024

Remove TPU_HOST_BOUNDS setting for TPU single host (#3525)

JackCaoG committed 3 years ago

Verified 9f9db76a

Lower Gelu and GeluBackward (#3531)

wonjoo-wj committed 3 years ago

Verified 89c5dad4

Rename torch_xla::ir Node to XlaNode and Value to XlaValue (#3526)

wonjoo-wj committed 3 years ago

Verified d9dc45fc

Add lazy::shape and lazy::shapefn to XlaNode constructor with some clean up (#3528)

JackCaoG committed 3 years ago

Verified cf7e123c

lowering roll op (#3505)

cicirori committed 3 years ago

Verified 4f6b2833

Add Checkpoint api to use optimization barrier (#3524)

JackCaoG committed 3 years ago

Verified 16498c34

Remove torch_xla::Device (#3495)

wonjoo-wj committed 3 years ago

Verified dbc67604

Track tools.codegen rename to torchgen (#3509)

ezyang committed 3 years ago

Verified 7f9310c6

LTC phase 2 clean up (#3521)

JackCaoG committed 3 years ago

Verified 1f0fae92

Revert "Update tf submodule to 04/07/2022 (#3507)" (#3522)

JackCaoG committed 3 years ago

Verified da2d9887

Revert "Revert "Node hash new (#3514)" (#3517)" (#3518)

JackCaoG committed 3 years ago

Verified f4b64885

Update tf submodule to 04/07/2022 (#3507)

JackCaoG committed 3 years ago

Verified 1d0a6e75

Revert "Node hash new (#3514)" (#3517)

JackCaoG committed 3 years ago

Verified 7c572100

Node hash new (#3514)

JackCaoG committed 3 years ago

Verified e8353135

Fix stable docker container tag in README (#3512)

d4l3k committed 3 years ago

Verified 706c2039

Make layout pining optional for cross core communication (#3511)

JackCaoG committed 3 years ago

Verified 5ece4ca7

Lower hardswish (#3496)

lucasc896 committed 3 years ago

Verified 734f6103

Make Optimization Barrier in place (#3493)

JackCaoG committed 3 years ago

Verified 92d6ac26

Remove test_autocast_torch_fp16 (#3499)

wonjoo-wj committed 3 years ago

Verified 6b0efa6e

Add performance profiling tool to enable programmatic/interactive/automatic PT-XLA profile capturing (#3498)

thisisalbertliang committed 3 years ago

Verified db0c7e80

Use upstream torch::lazy::Node in xla NodePtr (#3481)

wonjoo-wj committed 3 years ago

Verified 4037ca7f

Fix gather by using xla sparse gather (#3450)

ymwangg committed 3 years ago

Verified cfb6e690

Update disabled test name for test_upsamplingBicubic2d_xla (#3489)

wonjoo-wj committed 3 years ago

Verified c496278c

Remove reduce_scaterr layout pining (#3484)

JackCaoG committed 3 years ago

Verified 5af279ee

CUDA image stable version tag fix (#3480)

yeounoh committed 3 years ago

Verified 770c98d1

Newer Older