Skip to content

[Fix][Builtin] Fix "GetQueryPosition" of PagedKVCache #16746

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Mar 20, 2024
Merged

[Fix][Builtin] Fix "GetQueryPosition" of PagedKVCache #16746

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion src/runtime/relax_vm/kv_state.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -159,7 +159,7 @@ class AttentionKVCacheObj : public KVStateObj {
* This function is supposed to be invoked after calling BeginForward.
* \return The in-sequence query positions, in shape `(total_length,)`.
*/
virtual NDArray GetQueryPositions() const = 0;
virtual NDArray GetQueryPositions() = 0;

/************** Debug Helpers **************/

Expand Down
9 changes: 5 additions & 4 deletions src/runtime/relax_vm/paged_kv_cache.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -838,10 +838,11 @@ class PagedAttentionKVCacheObj : public AttentionKVCacheObj {
}
}

NDArray GetQueryPositions() const final {
CHECK(!dirty_aux_data_device_)
<< "The auxiliary arrays are not synchronized to device. Please call "
"`BeginForward` to synchronize before calling `GetQueryPositions`.";
NDArray GetQueryPositions() final {
// Sync the copy stream and the compute stream.
ComputeStreamWaitForCopyStream();
// The auxiliary data structure on device must have been synchronized.
ICHECK(!dirty_aux_data_device_);
return q_rope_position_map_view_;
};

Expand Down