Skip to content

Selecting MTile (128 or 64) for calling grouped_infer_pagedkv_mask_bi…

04c4ff7
Select commit
Loading
Failed to load commit list.
Merged

Add attention pagedkv prefill pipeline(for improving large seq_len_q performance) #68

Selecting MTile (128 or 64) for calling grouped_infer_pagedkv_mask_bi…
04c4ff7
Select commit
Loading
Failed to load commit list.