pytorch/xla

Pull Requests Commits

fix cpp test build

JackCaoG committed 2 years ago

87e6739b

manually add gcs to our dep

JackCaoG committed 2 years ago

5eee2ed1

Update pin one more time to pass a gs file system bug

JackCaoG committed 2 years ago

d82e5403

Update XLA pin to 2023/08/08

JackCaoG committed 2 years ago

9ebea48d

Cache sharded placeholder for dynamo execution (#5446)

JackCaoG committed 2 years ago

Verified ee6ef4d6

Update to GCC 10 (#5451)

will-cromar committed 2 years ago

Verified ca2b4aac

Remove base image override from TPU CI build (#5453)

will-cromar committed 2 years ago

Verified 63febd9e

parallelize SPMD inputhandler and GetDataShards (#5447)

JackCaoG committed 2 years ago

Verified 481d0b04

Revert "Install gcc-10 (#5450)" (#5452)

will-cromar committed 2 years ago

Verified 493bb734

Install gcc-10 (#5450)

qihqi committed 2 years ago

Verified 65b7639c

Revert "Update gcc to 10. (#5445)" (#5449)

lsy323 committed 2 years ago

Verified 55249edd

Update the random seed for every dynamo execution (#5444)

JackCaoG committed 2 years ago

Verified 13f9382c

Update gcc to 10. (#5445)

qihqi committed 2 years ago

Verified 454e916e

Move the C++ test build to CI build job instead of test job (#5442)

JackCaoG committed 2 years ago

Verified 4ef4dd34

Bugfix + add more test for llama (#5439)

qihqi committed 2 years ago

Verified a2a4ff4c

[PJRT] Support `torchrun` with `pjrt://` `init_method` (#5438)

will-cromar committed 2 years ago

Verified 9db8fac4

Update fix_includes.sh (#5441)

aazzolini committed 2 years ago

Verified 5b88b5f8

Remove TPU loading hack (#5435)

will-cromar committed 2 years ago

Verified 709c75a0

Add repo for libcudnn8=8.7.0.84 and CUDA 11.8 (#5425)

mateuszlewko committed 2 years ago

Verified f06c31f5

[PJRT|TPU] Update `test_xla_devices_single_process_all_chips` for expected device number (#5421)

ManfeiBai committed 2 years ago

Verified 2d97d3a1

[SPMD] named partition spec support (#5415)

yeounoh committed 2 years ago

Verified 85453d27

Fix the incorect clone arg condition in dynamo bridge (#5414)

JackCaoG committed 2 years ago

Verified 1835771c

SPMD cross slice-replication using partial_replication sharding (#5411)

yeounoh committed 2 years ago

Verified 6852df1d

Revert "Partially replicate lower-rank tensors (#5409)" (#5412)

yeounoh committed 2 years ago

Verified 628ba698

Partially replicate lower-rank tensors (#5409)

jonb377 committed 2 years ago

Verified 56a6a02a

Fix TPU collective ops test for multi-host TPUs (#5408)

will-cromar committed 2 years ago

Verified 2658c882

Refactor to share code between export_torch_model and save_as_stablehlo (#5388)

qihqi committed 2 years ago

Verified 2971efca

[SPMD] Multi-host batch sharded data loading (#5331)

khatwanimohit committed 2 years ago

Verified f2f877f8

Update the dynamo backend name to `openxla` (#5402)

JackCaoG committed 2 years ago

Verified b199b4dc

Fix ReplicateShardedData for int type (#5404)

JackCaoG committed 2 years ago

Verified 17570e20

Older