[BugFix] Initialize model_config for Qwen3-VL MoE (#44863)

Signed-off-by: wenpengw-nv <wenpengw@nvidia.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
This commit is contained in:
wenpengw-nv
2026-07-13 13:43:53 -07:00
committed by GitHub
co-authored by Roger Wang
parent 8ac8375270
commit 9a21f0d1a3
@@ -216,6 +216,7 @@ class Qwen3VLMoeForConditionalGeneration(
multimodal_config = vllm_config.model_config.multimodal_config
self.config = config
self.model_config = vllm_config.model_config
self._tokenizer = cached_tokenizer_from_config(vllm_config.model_config)
self.multimodal_config = multimodal_config
self.use_data_parallel = multimodal_config.mm_encoder_tp_mode == "data"