microsoft/DeepSpeed

Pull Requests Commits

addressing Conglong's feedback

HeyangQin committed 1 year ago

Verified da61c095

HeyangQin committed 1 year ago

8c6e4cab

chinese version of long sequence blog

HeyangQin committed 1 year ago

5a7a50eb

awan-10 committed 1 year ago

b908f80f

Update README.md

samadejacobs committed 1 year ago

Verified f9a52ccd

Merge branch 'master' into deepspeed-ulysses

samadejacobs committed 1 year ago

Verified 4da253fd

Update README.md

samadejacobs committed 1 year ago

Verified 8931c3ac

Some edit and clean up

samadejacobs committed 1 year ago

70ba7497

Remove deprecated figs

samadejacobs committed 1 year ago

48d4f502

samadejacobs committed 1 year ago

1a3c01a8

Replace DS Sequence with DS-Ulysses in all figs

samadejacobs committed 1 year ago

1661cf6e

Fix ZeRO parameter initialization for tensors with `requires_grad=True` (#4138)

XuehaiPan committed 1 year ago

Verified 426810a2

Add link to tutorial

samadejacobs committed 1 year ago

dcc503cf

samadejacobs committed 1 year ago

a377599f

More formatting, resize figures, add links

samadejacobs committed 1 year ago

5aa90d2e

Formatting and figures centering

samadejacobs committed 1 year ago

659b6a62

Add unit test to check HF low_cpu_mem_usage_flag (#4184)

loadams committed 1 year ago

Verified 9723a879

Title formatting

samadejacobs committed 1 year ago

56269a5c

Minor cleanup of text

Microsoft Inc committed 1 year ago

bb23350d

Add blog text and related media

Microsoft Inc committed 1 year ago

3be23d72

Fix path (#4193)

kytimmylai committed 1 year ago

Verified 6c684e12

Merge branch 'master' of github.com:microsoft/DeepSpeed

jeffra committed 1 year ago

956ed2fa

Fix nv-nightly workflow (#4163)

mrwyattii committed 1 year ago

Verified d9a889d5

enable autoTP for mpt in huggingface model hub without trust_remote_code (#4062)

sywangyi committed 1 year ago

Verified 5e16eb2c

Merge branch 'master' of github.com:microsoft/DeepSpeed

jeffra committed 1 year ago

f0afcf3b

Merge branch 'master' of github.com:microsoft/DeepSpeed

jeffra committed 1 year ago

a746aca8

Treat empty environment variables as unset in (#4185)

loadams committed 1 year ago

Verified 8fb111c0

do allgather only in shared optimizer states groups (#4167)

inkcherry committed 1 year ago

Verified 7f3e82fe

Merge branch 'master' of github.com:microsoft/DeepSpeed

jeffra committed 1 year ago

d9b16722

MP ZeRO++ (#3954)

HeyangQin committed 1 year ago

Verified 7711bdbb

Older