microsoft/DeepSpeed

Pull Requests Commits

Merge branch 'master' into loadams/fix-docs-rendering

loadams committed 1 year ago

53402a93

Update version.txt after 0.16.0 release (#6786)

loadams committed 1 year ago

Verified f743feca

Revert release workflow (#6785)

loadams committed 1 year ago

Verified e5570b10

Update version.txt before release (#6784)

loadams committed 1 year ago

Verified 03845dbc

Domino Blog (#6776)

GuanhuaWang committed 1 year ago

Verified ec6cc490

ReadTheDocs/Sphinx autodoc mock config updates to get docs to render properly

loadams committed 1 year ago

0ee2caae

erge branch 'master' into loadams/fix-docs-rendering

loadams committed 1 year ago

c74e8af1

Cleanup code docs warnings (#6783)

loadams committed 1 year ago

Verified fabcf407

Fix explicit markup needing to end with a blank line

loadams committed 1 year ago

c9188ee5

Revert "Fix imports prefixed with deepspeed"

loadams committed 1 year ago

dedcf41b

Fix imports prefixed with deepspeed

loadams committed 1 year ago

5710c5d3

Fix warnings on lengths of underlines

loadams committed 1 year ago

9a0d258b

Fix Doc Error: ZeRO Stage 2 gradient partitioning (#6775)

yewentao256 committed 1 year ago

Verified d6410f90

docs: fix HF links (#6780)

imba-tjd committed 1 year ago

Verified 5e16f255

Unpin with latest transformers fixes (#6763)

loadams committed 1 year ago

Verified f57b1ef1

Fix potential memory issues when use deepspeed Z3 (#6726)

wenbinc-Bin committed 1 year ago

Verified cd20a3bb

Removes unnecessary cloning (#6761)

swigls committed 1 year ago

Verified f515104e

Enable torch compile on _allgather_params (#6769)

deepcharm committed 1 year ago

Verified b5709cce

Use `json_schema_extra` instead of extra keyword in `Field` (#6764)

qgallouedec committed 1 year ago

Verified 83e4364f

Fix setup.py bash cmd generation to correctly extract git info (#6762)

nelyahu committed 1 year ago

Verified 065398d5

Add explicit parameters for torch.load (#6751)

loadams committed 1 year ago

Verified 2e0c39b5

make xpu ops compatible with oneapi 2025.0 (#6760)

baodii committed 1 year ago

Verified 1fdad1fa

Pin transformers version to work around latest torch requirements (#6759)

loadams committed 1 year ago

Verified 8488beea

A faster and more memory-efficient implementation of `zero_to_fp32` (#6658)

xu-song committed 1 year ago

Verified dd402694

Disable failing python tests (#6758)

loadams committed 1 year ago

Verified f594dbe3

Gaudi2 Nightly job for daily check (#6753)

raza-sikander committed 1 year ago

Verified e3b5a4b6

Add no_sync context manager (#6675)

tjruwase committed 1 year ago

Verified fc4e7337

Adding the governance doc (#6748)

minjiazhang committed 1 year ago

Verified d702eb5f

Sanitize inputs to eval() (#6745)

loadams committed 1 year ago

Verified 9a2c209c

Update path for BingBertSquad from DeepSpeedExamples (#6746)

loadams committed 1 year ago

Verified 877aa0db

Older