deepspeedai/DeepSpeed

Pull Requests Commits

Remove get_accelerator().amp() API and use torch.amp directly

delock committed 1 day ago

a5b04f02

Replace get_accelerator().amp().autocast() with torch.amp.autocast(device_type=...)

delock committed 1 day ago

2e856639

fix: replace 35 bare except clauses with except Exception (#7873)

haosenwang1018 committed 2 days ago

Verified ae21699a

Fix broken links and add AutoTP Training tutorial to sidebar nav (#7874)

tohtana committed 2 days ago

Verified efc0b49a

Schedule nightly full test (#7870)

tohtana committed 3 days ago

Verified 0416cf68

Fix Zero-3 static scale assertion in fp16 test (#7866)

tohtana committed 6 days ago

Verified 93524c89

Throw error when parameter is modified in GatheredParameters (#7832)

tohtana committed 6 days ago

Verified 57b10d54

Fix compilation of Evoformer (#7862)

Flamefire committed 6 days ago

Verified c9e652cf

Fix ROCm BF16 conversion intrinsics in inference v2 (#7843) (#7846)

tohtana committed 9 days ago

Verified dbc1b075

Add EXAONE 4.0 model support for Inference V2 (#7853)

Bias92 committed 10 days ago

Verified f3a9819c

Add news entry for DeepSpeed updates (#7854)

PKUWZP committed 11 days ago

Verified 8a63a2a9

Enable shm_comm support for arm (#7800)

phalani-paladugu committed 12 days ago

Verified 7f49367a

Z1/2 init: flatten params on device (#7828)

ksugama committed 15 days ago

Verified 84af8221

Merge pull request #7850 from deepspeedai/loadams/update-post-0.18.6

loadams committed 15 days ago

Verified b74df164

Replace torch.jit.script with torch.compile (#7835) (#7840)

tohtana committed 15 days ago

Verified d2ca6e75

Fix bf16 gradient norm divergence with ZeRO stage 0 (#7839)

tohtana committed 15 days ago

Verified 1752c2ab

fix: replace deprecated fractions.gcd with math.gcd (#7845)

Mr-Neutr0n committed 16 days ago

Verified 2c362837

Support new python 3.14 annotation handling (#7831)

sdvillal committed 18 days ago

Verified 72b9465c

Fix AutoTP custom patterns: respect use_default_specs (#7827)

tohtana committed 20 days ago

Verified a44fb58f

Fix gradient is ready with z2 (#7829)

sfc-gh-truwase committed 24 days ago

Verified 0ccb2bb6

Support custom partitioning patterns for AutoTP (#7806)

tohtana committed 27 days ago

Verified 6b9cab1d

Skip sequence parallel operations during eval (#7821)

jp1924 committed 28 days ago

Verified a972f7c6

Fix leaf module race condition (#7825)

tohtana committed 28 days ago

Verified 52b1d4d6

Update version.txt to 0.18.6 after latest release (#7826)

loadams committed 28 days ago

Verified 7b15a6a7

Update PyTorch to v2.9 for modal tests (#7816)

tohtana committed 29 days ago

Verified b19987c7

[Bugfix] Resolve Rank index out of range during BWD when sp_size < world_size in Ulysses (#7809)

Flink-ddd committed 30 days ago

Verified bb250a25

Raise clear error on in-place GatheredParameters edits without modifier_rank (#7817)

tohtana committed 31 days ago

Verified d9f3d40a

Fix typos in accelerator setup guide (#7818)

nathon-lee committed 31 days ago

Verified 7b763459

Ignore evoformer test (#7815)

tohtana committed 33 days ago

Verified c89e0db8

Improve engine's cleanup (#7813)

tohtana committed 33 days ago

Verified 827767a4

Older