Skip to content

[shardformer]test GPT-2 to check flash attention 2 in conjunction with pipeline parallelism.. #4390

@flybird11111

Description

@flybird11111
No description provided.

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

Status

✅ Done

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions