deepspeedai/DeepSpeed

Pull Requests Commits

delock committed 127 days ago

8ffc055e

remove memory data

delock committed 127 days ago

c855b035

delock committed 127 days ago

77183c38

expand memory analysis

delock committed 127 days ago

9a818424

delock committed 127 days ago

db3e2f0a

add contributor list

delock committed 128 days ago

1e06bf35

Muon optimizer blog draft

delock committed 128 days ago

ef4b6235

Use pytorch utils to detect ninja (#7687)

Emrys-Merlin committed 129 days ago

Verified 4cb26cab

modal ci: fix group concurrency (#7691)

stas00 committed 129 days ago

Verified dff81cb6

stage 1-2: don't pin memory if not configured (#7689)

stas00 committed 129 days ago

Verified c502d3ac

make debug utils more resilient (#7690)

stas00 committed 129 days ago

Verified 7cb37efc

see_mem_usage: make always work (#7688)

stas00 committed 129 days ago

Verified c2c090b7

allow seperate learning rate "muon_lr" and "adam_lr" for muon optimizer (#7658)

delock committed 130 days ago

Verified df59f203

disable nv-lightning-v100.yml cI (#7681)

stas00 committed 133 days ago

Verified 283f6f5f

leaf modules: explain better (#7674)

stas00 committed 134 days ago

Verified d25fd486

[modal ci] fixes (#7676)

stas00 committed 135 days ago

Verified b073a557

Update version.txt after release (#7675)

loadams committed 136 days ago

Verified c4b8c3f3

README refresh (#7668)

sfc-gh-truwase committed 137 days ago

Verified e993fea3

Fix typo in pytorch-profiler.md documentation (#7652)

kunheek committed 138 days ago

Verified 4691bebd

resolved a 0-dim tensor slicing bug from _get_state_without_padding (#7659)

therealnaveenkamal committed 138 days ago

Verified 6fe3c7ab

UlyssesSP: TiledMLP doc - recomputes forward twice (#7664)

stas00 committed 138 days ago

Verified 76a40757

[bug]: fixed comm_dtype in extra_large_param_to_reduce (#7660)

therealnaveenkamal committed 141 days ago

Verified 3e64f495

Fix misplaced overflow handling return in fused_optimizer.py (#7645)

rraminen committed 142 days ago

Verified d56e847b

ALST/UlyssesSP: more intuitive API wrt variable seqlen (#7656)

stas00 committed 144 days ago

Verified 02da3732

ulysses mpu: additional api (#7649)

stas00 committed 145 days ago

Verified 433e3c72

Deduplicate fp32 weights under torch autocast and ZeRO3 (#7651)

eternalNight committed 145 days ago

Verified 706f6e82

Update version after 0.18.1 release (#7647)

loadams committed 149 days ago

Verified 9a012d21

Update version.txt in advance of next release

loadams committed 149 days ago

3631712b

Add DataStates-LLM: Asynchronous Checkpointing Engine Support (#7166)

mauryaavinash95 committed 149 days ago

Verified d1e62ff2

Ulysses HF Accelerate integration (#7638)

stas00 committed 150 days ago

Verified 64c0052f

Older