pytorch/xla

Pull Requests Commits

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

Add v4 config to xla_dist (#3440)

JackCaoG committed 3 years ago

Verified cc049535

Remove shape and operand related code in Node (#3477)

wonjoo-wj committed 3 years ago

Verified 515dae01

Add optimization barrier to enable the gradient checkpointing (#3482)

JackCaoG committed 3 years ago

Verified 2c2f554a

Only include `test/cpp/build/test_ptxla` when `BUILD_CPP_TESTS=1` (#3479)

will-cromar committed 3 years ago

Verified 5d9c53d7

Removing Duplicate Ostream Definitions for at::Scalar (#3475)

Gamrix committed 3 years ago

Verified 3e3255e4

Add test_env_var_mapper.py test (#3473)

wonjoo-wj committed 3 years ago

Verified dfb04e48

Add function in ir.cpp to update upstream env vars (#3469)

wonjoo-wj committed 3 years ago

Verified 7dbb5845

Migrate to upstream torch::lazy::Value (#3454)

wonjoo-wj committed 3 years ago

Verified ed03cf92

Update README for availalbe images and wheels for GPUs (#3467)

yeounoh committed 3 years ago

Verified 063c0f62

Update docker cloud build to support A100 (#3465)

yeounoh committed 3 years ago

Verified 7bad694c

Migrate to upstream Node and Output (#3402)

wonjoo-wj committed 3 years ago

Verified 81f08457

Remove deprecated nvidia deb repo from base docker image (#3459)

yeounoh committed 3 years ago

Verified 0c7453c4

Revert "Revert "Migrate to upstream torch::lazy::BackendDevice (#3383)" (#3432)" (#3433)

wonjoo-wj committed 3 years ago

Verified 94d71650

Fix optional vector (#3334)

JackCaoG committed 3 years ago

Verified 23496c53

Pin Jinja2 for doc push (#3449)

JackCaoG committed 3 years ago

Verified 5ca6cc90

Revert "Revert "Usthe colab wheel for colab tutorials. (#3438)" (#3444)" (#3447)

yeounoh committed 3 years ago

Verified 14e9bf32

Newer Older