From 3a915484e26ad2e56202047a4c07de3459a9f967 Mon Sep 17 00:00:00 2001
From: Iwan Kawrakow <iwan.kawrakow@gmail.com>
Date: Tue, 2 Dec 2025 08:44:46 +0000
Subject: [PATCH] Is this better for multi-GPU and split mode "graph"?

---
 src/llama-build-context.cpp | 6 ------
 1 file changed, 6 deletions(-)

diff --git a/src/llama-build-context.cpp b/src/llama-build-context.cpp
index 123fc185b..f51e8587b 100644
--- a/src/llama-build-context.cpp
+++ b/src/llama-build-context.cpp
@@ -688,12 +688,6 @@ ggml_tensor * llm_build_context::llm_build_ffn(
             cur = ggml_add(ctx, cur, ffn[id]);
             cb(cur, "combine_ffn", il);
         }
-        if (ffn.size() > 2) {
-            cur->op_params[0] = 0xff;
-        }
-        //if (cur->type != GGML_TYPE_F32) {
-        //    cur = ggml_cast(ctx, cur, GGML_TYPE_F32);
-        //}
 
         return cur;
     }