vllm/model_executor/models/deepseek_vl2.py · 27d7638b9476062931a6770ed90714792e77cc83 · OpenDAS / vllm_cscc

[Bugfix] Merge MM embeddings by index instead of token IDs (#16229) · 27d7638b

Cyrus Leung authored Sep 27, 2025


Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: NickLucche <nlucches@redhat.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: NickLucche <nlucches@redhat.com>
Co-authored-by: Roger Wang <hey@rogerw.io>

27d7638b

deepseek_vl2.py 25 KB

Replace deepseek_vl2.py