model: support Qwen3.5 (#18489)

Co-authored-by: 瑀澈 <yuche.lz@alibaba-inc.com>
This commit is contained in:
Zheng Li
2026-02-10 00:27:59 +08:00
committed by GitHub
co-authored by 瑀澈
parent 0b4d4f2838
commit 27c447653d
17 changed files with 1923 additions and 9 deletions
@@ -66,6 +66,7 @@ def get_model_config(
"Qwen3MoeForCausalLM",
"Qwen3NextForCausalLM",
"Qwen3VLMoeForConditionalGeneration",
"Qwen3_5MoeForConditionalGeneration",
]:
E = config.num_experts // ep_size
topk = config.num_experts_per_tok