From: Sigbjørn Skjæret <redacted>
Date: Mon, 8 Dec 2025 20:29:59 +0000 (+0100)
Subject: graph : use fill instead of scale_bias in grouped expert selection (#17867)
X-Git-Tag: upstream/0.0.7446~119
X-Git-Url: https://git.djapps.eu/?a=commitdiff_plain;h=c8554b66e0ed397f7457ed5f3ce3b466dd508d5c;p=pkg%2Fggml%2Fsources%2Fllama.cpp

graph : use fill instead of scale_bias in grouped expert selection (#17867)

* use fill instead of scale_bias in grouped expert selection

* do not explicitly use _inplace
---

diff --git a/src/llama-graph.cpp b/src/llama-graph.cpp
index 42ccb5b7..43620df7 100644
--- a/src/llama-graph.cpp
+++ b/src/llama-graph.cpp
@@ -973,7 +973,7 @@ ggml_tensor * llm_graph_context::build_moe_ffn(
 
         // mask out the other groups
         selection_probs = ggml_get_rows(ctx0, selection_groups, expert_groups); // [n_exp_per_group, n_group_used, n_tokens]
-        selection_probs = ggml_set_rows(ctx0, ggml_scale_bias(ctx0, selection_groups, 0.0f, -INFINITY), selection_probs, expert_groups); // [n_exp_per_group, n_expert_groups, n_tokens]
+        selection_probs = ggml_set_rows(ctx0, ggml_fill(ctx0, selection_groups, -INFINITY), selection_probs, expert_groups); // [n_exp_per_group, n_expert_groups, n_tokens]
         selection_probs = ggml_reshape_2d(ctx0, selection_probs, n_expert, n_tokens); // [n_expert, n_tokens]
         cb(selection_probs, "ffn_moe_probs_masked", il);
     }