From 57bb668674d9fb0d382885e5b04911c6437f8e83 Mon Sep 17 00:00:00 2001 From: drluoto Date: Sat, 29 Aug 2026 08:58:37 +0200 Subject: [PATCH] convert: keep the model-level hc head mixer in --mtp exports The mtp_only whitelist kept only embed/norm/lm_head, but llama_model_qwen4exp requires the model-level output_hc_norm/_down/_up at load, so a detached sidecar exported by --mtp failed with 'tensor output_hc_norm.weight not found'. Keep model.hyper_connection_mixer.* as well (34 -> 37 tensors); verified loading and generating on gfx1151/ROCm with both a self-converted sidecar and the pre-existing community one. Co-Authored-By: Claude Fable 5 --- conversion/qwen.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/conversion/qwen.py b/conversion/qwen.py index c5297418c99b..edb0209e05fc 100644 --- a/conversion/qwen.py +++ b/conversion/qwen.py @@ -341,7 +341,7 @@ def filter_tensors(cls, item): keep = name in ( "model.embed_tokens.weight", "model.norm.weight", "lm_head.weight", "embed_tokens.weight", "norm.weight", - ) + ) or name.startswith("model.hyper_connection_mixer.") if not keep: return None return name, gen