[Feature]: 4D Attention Mask

[littletomatodonkey]

🚀 The feature, motivation and pitch

I am workning on 4D attention mask input and LLM generateion process. Huggingface provides an interface for the 4D attention mask. Does vllm have any plan? huggingface/transformers#27539

Alternatives

No response

Additional context

No response

[iiLaurens]

AFAIK custom attention masks are not supported in optimized implementations of attention (FlashInfer, FlashAttention), so this wouldn't be possible without also incurring a dramatic drop in performance.

[littletomatodonkey]

AFAIK custom attention masks are not supported in optimized implementations of attention (FlashInfer, FlashAttention), so this wouldn't be possible without also incurring a dramatic drop in performance.

Thanks for your reply, i might have to realize the function myself.

[github-actions]

This issue has been automatically marked as stale because it has not had any activity within 90 days. It will be automatically closed if no further activity occurs within 30 days. Leave a comment if you feel this issue should remain open. Thank you!

[github-actions]

This issue has been automatically closed due to inactivity. Please feel free to reopen if you feel it is still relevant. Thank you!