vllm/model_executor/models/aya_vision.py · 56d4aefa33f3f8ffaf74d02a8d7eef9523651864 · OpenDAS / vllm_cscc · GitLab

Find file Blame History Permalink

[VLM] Avoid unnecessary dummy multimodal data during processing (#16416) · 56d4aefa
Cyrus Leung authored Apr 11, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
56d4aefa

aya_vision.py 18.8 KB