pytorch/xla

Pull Requests Commits

JackCaoG committed 1 year ago

48e250d9

update troubleshooting

JackCaoG committed 1 year ago

02e956e7

JackCaoG committed 1 year ago

11cd90c4

JackCaoG committed 1 year ago

794263a3

Update memory usage to GB instead of MB

JackCaoG committed 1 year ago

350bc7f3

add PT_XLA_DEBUG_LEVEL

JackCaoG committed 1 year ago

64c93be5

add test for onehot to make sure no fallback (#7127)

JackCaoG committed 1 year ago

Verified ad73e070

Add Python 3.11 version for release build (#7143)

wonjoo-wj committed 1 year ago

Verified bd70d3f0

Revert "`upsample_bilinear`: fix output data-type." (#7142)

JackCaoG committed 1 year ago

Verified f1141980

[Pallas] Support tgmm (#7137)

alanwaketan committed 1 year ago

Verified 50d81b3e

Update contribute to include bdist_wheel (#7135)

JackCaoG committed 1 year ago

Verified c367b66f

Facebook->Meta in README (#7141)

kit1980 committed 1 year ago

Verified 0eded8da

Add function for retrieving fallback operations. (#7116)

ysiraichi committed 1 year ago

Verified 6d271232

`upsample_bilinear`: fix output data-type. (#7111)

ysiraichi committed 1 year ago

Verified 468a5c93

update example dir's README (#7136)

JackCaoG committed 1 year ago

Verified c7bbdfbb

Update XLA pin as of 20240528 (#7131)

wonjoo-wj committed 1 year ago

Verified 6f406b77

[FSDPv2] Shard on the maximal dim of weights (#7134)

alanwaketan committed 1 year ago

Verified 15fc0f1c

[Pallas] Make gmm support bf16 (#7133)

alanwaketan committed 1 year ago

Verified fb373129

[MoE] Test sorting lhs for gmm (#7121)

alanwaketan committed 1 year ago

Verified 8531d1c5

remove torch_xla2 jax dependency in native install (#7126)

zpcore committed 1 year ago

Verified ed90be1e

Use remote cache for `push` events too (#7124)

will-cromar committed 1 year ago

Verified 7770a494

`index`: fix index of 0-element tensor by 0-element tensor. (#7113)

ysiraichi committed 1 year ago

Verified be3b08e6

[Pallas] Set a better tiling for gmm (#7119)

alanwaketan committed 1 year ago

Verified fd4900ce

[Pallas] Make gmm output a tensor (#7120)

alanwaketan committed 1 year ago

Verified 65b5ace8

Only use remote cache for main repository, not forks (#7112)

will-cromar committed 1 year ago

Verified 1a8c2fe2

[Pallas] Make gmm functional (#7117)

alanwaketan committed 1 year ago

Verified a9b4fadf

Update XLA pin to 2024/05/24 and fix Hermetic Python integration (#7110)

vam-google committed 1 year ago

Verified 7d31f7de

[Pallas] Support _histogram (#7115)

alanwaketan committed 1 year ago

Verified cb805837

implement Repeat with fixed output shape (#7114)

JackCaoG committed 1 year ago

Verified 3369bf7d

[Pallas] Introduce _make_group_metadata (#7107)

alanwaketan committed 1 year ago

Verified 22e912ea

Older