[Model] Introduce Kimi Linear to vLLM (#27809)

Signed-off-by: lizhiyuan <lizhiyuan@moonshot.cn> Signed-off-by: Zhiyuan Li <uniartisan2017@gmail.com>
2025-10-30 21:02:27 +08:00
parent 1994de99ea
commit 4e68cc9b6a
15 changed files with 1325 additions and 48 deletions
--- a/vllm/model_executor/layers/fla/ops/kda.py
+++ b/vllm/model_executor/layers/fla/ops/kda.py
@@ -1304,7 +1304,7 @@ def kda_gate_fwd_kernel(
    tl.store(y_ptr, b_y.to(y.dtype.element_ty), boundary_check=(0, 1))


-def kda_gate_fwd(
+def fused_kda_gate(
    g: torch.Tensor,
    A: torch.Tensor,
    head_k_dim: int,