From c52ed2a0b0b865e82eb1b393106c48df1c39cb32 Mon Sep 17 00:00:00 2001 From: danielhanchen Date: Wed, 26 Aug 2026 17:37:20 +0000 Subject: [PATCH] llama: give qwen4exp a large-graph node budget (cherry picked from commit 37c8c194e6a30e4c46ac29bee3fb264f091596ef) --- src/llama-context.cpp | 1 + 1 file changed, 1 insertion(+) diff --git a/src/llama-context.cpp b/src/llama-context.cpp index 0402044da6..97fe39b7b1 100644 --- a/src/llama-context.cpp +++ b/src/llama-context.cpp @@ -2301,6 +2301,7 @@ uint32_t llama_context::graph_max_nodes(uint32_t n_tokens) const { model.arch == LLM_ARCH_BAILINGMOE3 || model.arch == LLM_ARCH_QWEN35 || model.arch == LLM_ARCH_QWEN35MOE || + model.arch == LLM_ARCH_QWEN4EXP || model.arch == LLM_ARCH_DEEPSEEK4 || (model.arch == LLM_ARCH_DFLASH && model.hparams.dsv4_hc_mult > 0) || model.arch == LLM_ARCH_NANBEIGE ||