pytorch/xla

Pull Requests Commits

Update test_spmd_debugging.py

ManfeiBai committed 1 year ago

Verified 7486969f

Update test_spmd_debugging.py

ManfeiBai committed 1 year ago

Verified 3029e3b9

Update test_spmd_debugging.py

ManfeiBai committed 1 year ago

Verified eea642cd

Use SGD for list of models from PyTorch. (#6324)

ysiraichi committed 1 year ago

Verified a7918a72

Allow xla bloat16 torch test results with some variances (#6338)

yeounoh committed 1 year ago

Verified ab28b23f

Update test_spmd_debugging.py to avoid code test code self (#6263)

ManfeiBai committed 1 year ago

Verified df097d75

Update pytorch_test_base.py (#6337)

wonjoo-wj committed 1 year ago

Verified ec23c917

Clean-up CUDA cache. (#6325)

ysiraichi committed 1 year ago

Verified 423bb0b2

Raise an error in `xmp.spawn` if runtime is initialized (#6291)

will-cromar committed 1 year ago

Verified e93597de

Update metric check for resnet18 dynamo unit test (#6307)

wonjoo-wj committed 1 year ago

Verified 8c5e2897

Add option to register installed plugins automatically (#6328)

will-cromar committed 1 year ago

Verified d7ddecdb

Revert "Set all_reduce_token to None when exiting" (#6321)

vanbasten23 committed 1 year ago

Verified f33634e4

Increase absolute tolerance for f16 aten_addmm test (#6332)

LukeBoyer committed 1 year ago

Verified 151dc1d8

Update contribution guide to use python setup.py develop. (#6327)

changm committed 1 year ago

Verified 5f6d2b36

Propagates requires_grad over to AllReduce output (#6326)

yeounoh committed 1 year ago

Verified 40727e4d

Update doc on new modules. (#6318)

golechwierowicz committed 1 year ago

Verified 68515da9

OpenXLA pin update (#6313)

yeounoh committed 1 year ago

Verified 0bd8b8d4

Use spawn as the fork method for the profiler test. (#6302)

ysiraichi committed 1 year ago

Verified 52ef8efc

Add doc images for auto-sharding (#6323)

yeounoh committed 1 year ago

Verified e632897b

Re-land: Fix model initialization. (#6296)

ysiraichi committed 1 year ago

Verified a8b27eb1

[HOT FIX] Disable GPU CI Coverage in master (#6317)

yeounoh committed 1 year ago

Verified ec2db46c

Handle dynamo counter change after fake tensor cache (#6308)

JackCaoG committed 1 year ago

Verified d6dc1a01

Address GPU CI coverage timeout (#6306)

yeounoh committed 1 year ago

Verified 7fae2ab1

update release build to use rc8 tag (#6310)

zpcore committed 1 year ago

Verified 1cfd0cf5

Support mixed precision for torch.where (#6303)

qihqi committed 1 year ago

Verified d49fc70d

Fix global_runtime_device_count() for spmd case (#6266)

vanbasten23 committed 1 year ago

Verified 27ae6dc9

fix gpu spmd core dump (#6275)

mars1248 committed 1 year ago

Verified de27258c

Fix some more aten core ops (#6301)

wonjoo-wj committed 1 year ago

Verified 5d0de94b

Correct order of constant vs args (#6300)

qihqi committed 1 year ago

Verified 55177798

Enable Tensor index check in both dimension (#6299)

ManfeiBai committed 1 year ago

Verified 28502884

Older