ggml-cpu: FA split across kv for faster TG (#19209)

* ggml-cpu: split across kv for faster TG * simplify sinks application * add ref impl
2026-02-03 01:19:55 +08:00
parent a3fa035822
commit 9f682fb640
6 changed files with 220 additions and 69 deletions
@@ -24,6 +24,9 @@ struct ggml_compute_params {
    void * wdata;

    struct ggml_threadpool * threadpool;
+
+    // use reference implementation
+    bool use_ref;
 };