Skip to content

MLA kv cache: fix split graph backend assignment when kv cache store on CPU #13648

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

Closed
wants to merge 1 commit into from
Closed

MLA kv cache: fix split graph backend assignment when kv cache store on CPU #13648

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
4 changes: 4 additions & 0 deletions src/llama-graph.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1156,6 +1156,10 @@ ggml_tensor * llm_graph_context::build_attn_mha(
// for MLA with the absorption optimization, we need to "decompress" from MQA back to MHA
if (v_mla) {
kqv = ggml_mul_mat(ctx0, v_mla, kqv);
// all nodes between the KV store and the attention output are run on the CPU
if (!cparams.offload_kqv) {
ggml_backend_sched_set_tensor_backend(sched, kqv, backend_cpu);
}
}

cur = ggml_permute(ctx0, kqv, 0, 2, 1, 3);
Expand Down
Loading