pytorch/xla

Pull Requests Commits

JackCaoG committed 2 years ago

989bb0eb

revert unneeded change

JackCaoG committed 2 years ago

1e444486

Check the actual device instead of query env var for virtual device

JackCaoG committed 2 years ago

2fca2ce9

add fs linker flag (#5347)

lsy323 committed 2 years ago

Verified 5673ac4a

Make all-reduce a no-op when world size is 1 (#5342)

will-cromar committed 2 years ago

Verified 61816b0b

Also dump output sharding on HLO file (#5339)

JackCaoG committed 2 years ago

Verified efb1e6fd

Check out the release branch instead of origin/master in ansible (#5344)

vanbasten23 committed 2 years ago

Verified 68d7fe80

OpenXLA pin update 07052023 (#5304)

lsy323 committed 2 years ago

Verified 8ffbd1a4

Update Python device API for SPMD (#5129)

JackCaoG committed 2 years ago

Verified f996701a

Sharding should be per output of IR Node, instead of per IR Node (#5330)

JackCaoG committed 2 years ago

Verified 901d154a

handle single tensor for method send_to_device_single (#5317)

JackCaoG committed 2 years ago

Verified f5edcb20

Use `_sharded_cpu_state_dict` functionality to Write Items for SPMD Save Planner (#5315)

shahyash10 committed 2 years ago

Verified 74d115ec

Enable Some input output aliasing under SPMD (#5320)

JackCaoG committed 2 years ago

Verified ff8f7a49

Disable coverage for now (#5321)

JackCaoG committed 2 years ago

Verified e45cf404

Add tf independent serialization (#5308)

qihqi committed 2 years ago

Verified 0b5b5197

quick refactor on _get_group_assignment (#5318)

JackCaoG committed 2 years ago

Verified a8f723b6

Fix mismatched GPU docker image in the doc. (#5319)

vanbasten23 committed 2 years ago

Verified 42ae0bf9

Support unordered sharding spec for partial replication (#5316)

JackCaoG committed 2 years ago

Verified 78c6840c

Update gcc version in CI (#5297)

lsy323 committed 2 years ago

Verified e7e18996

Add docs for AMP (#5313)

cowanmeg committed 2 years ago

Verified 1cb5a524

Disable test failing on TPU CI (#5314)

JackCaoG committed 2 years ago

Verified cf491f9e

Fix race condition when use > 1 threads to transfer data in parallel loader (#5267)

aws-tianquaw committed 2 years ago

Verified 1dc5af55

Supoort unordered sharding spec correctly (#5305)

JackCaoG committed 2 years ago

Verified 67705065

Add `_sharded_cpu_state_dict` for distributed checkpointing (#5288)

shahyash10 committed 2 years ago

Verified 46a0117b

[SPMD] Allow dumping post optimizations hlo (#5302)

alanwaketan committed 2 years ago

Verified 15ee2191

[SPMD] Support mark_sharding on IRs (#5301)

alanwaketan committed 2 years ago

Verified 6fe5cb98

[SPMD] Avoid recompilations in xs.mark_sharding() (#5300)

alanwaketan committed 2 years ago

Verified caf51687

[SPMD] Add FSDP sharding for test_train_spmd_linear_model.py (#5299)

alanwaketan committed 2 years ago

Verified d01880c5

[SPMD] Use xs.Mesh in test_2d_tensor_3d_mesh (#5295)

khatwanimohit committed 2 years ago

Verified df680da7

Make README more actionable (#5262)

will-cromar committed 2 years ago

Verified c27a493e

Older