Skip to content

[Feature]: Enable MLA for V1 on AMD [Triton MLA] #46

Open
@tjtanaa

Description

@tjtanaa

🚀 The feature, motivation and pitch

As a quick enablement of DeepSeekV3 for V1 engine, we will open up the code path that uses existing mla class.

Blocker

The lm_eval result show low accuracy.

Alternatives

No response

Additional context

No response

Before submitting a new issue...

  • Make sure you already searched for relevant issues, and asked the chatbot living at the bottom right corner of the documentation page, which can answer lots of frequently asked questions.

Metadata

Metadata

Assignees

Labels

enhancementNew feature or request

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions