From 56505fdad50d043f199a4a2a1bfe478198b7fa7b Mon Sep 17 00:00:00 2001 From: Junmo Kim Date: Mon, 10 Aug 2026 18:54:24 +0900 Subject: [PATCH] model : fix SWA not being enabled for EXAONE 4.5 load_arch_hparams tests `hparams.n_layer() == 64` before LLM_KV_NEXTN_PREDICT_LAYERS has been read. n_layer() returns n_layer_all - n_layer_nextn and n_layer_nextn defaults to 0, so a GGUF carrying the MTP head (block_count=65, nextn=1) evaluates to 65 and the whole SWA block is skipped. The model type switch further down in the same function reads 64, because by then the key has been loaded. n_swa is still filled in by the unconditional get_key below the block, so llama_model_n_swa() reports 4096 and the logs look correct while only swa_type stays LLAMA_SWA_TYPE_NONE. This affects the official LGAI-EXAONE GGUF release as well. EXAONE 4.0 has no MTP head, so block_count is 64 there and the check matches. --- src/models/exaone4.cpp | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/src/models/exaone4.cpp b/src/models/exaone4.cpp index 863268abce..a06819a67c 100644 --- a/src/models/exaone4.cpp +++ b/src/models/exaone4.cpp @@ -1,6 +1,9 @@ #include "models.h" void llama_model_exaone4::load_arch_hparams(llama_model_loader & ml) { + ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.n_layer_nextn, false); + GGML_ASSERT(hparams.n_layer_nextn < hparams.n_layer_all && "n_layer_nextn must be < n_layer"); + if (hparams.n_layer() == 64) { // 32B hparams.swa_type = LLAMA_SWA_TYPE_STANDARD; hparams.n_swa = 4096; @@ -15,9 +18,6 @@ void llama_model_exaone4::load_arch_hparams(llama_model_loader & ml) { ml.get_key(LLM_KV_ATTENTION_SLIDING_WINDOW, hparams.n_swa, false); ml.get_key(LLM_KV_ATTENTION_LAYERNORM_RMS_EPS, hparams.f_norm_rms_eps); - ml.get_key(LLM_KV_NEXTN_PREDICT_LAYERS, hparams.n_layer_nextn, false); - - GGML_ASSERT(hparams.n_layer_nextn < hparams.n_layer_all && "n_layer_nextn must be < n_layer"); switch (hparams.n_layer()) { case 30: type = LLM_TYPE_1_2B; break;