models : support qwen3.5 series (#19468)

* support qwen3.5 series * remove deepstack for now, and some code clean * code clean * add FULL_ATTENTION_INTERVAL metadata * code clean * reorder v heads for linear attention to avoid expensive interleaved repeat
2026-02-10 18:00:26 +02:00
parent 9a96352729
commit fc0fe40049
17 changed files with 2096 additions and 10 deletions
@@ -54,6 +54,7 @@ enum llama_vocab_pre_type {
    LLAMA_VOCAB_PRE_TYPE_SOLAR_OPEN      = 43,
    LLAMA_VOCAB_PRE_TYPE_YOUTU           = 44,
    LLAMA_VOCAB_PRE_TYPE_EXAONE_MOE      = 45,
+    LLAMA_VOCAB_PRE_TYPE_QWEN35          = 46,
 };

 struct LLM_KV;