search

Lackel / AGLA

Code for paper "AGLA: Mitigating Object Hallucinations in Large Vision-Language Models with Assembly of Global and Local Attention"
13 stars 0 forks source link

About greedy decoding #5

Open haohaodw opened 1 month ago

haohaodw commented 1 month ago

A nice work! How can I reproduce the situation where AGLA uses greedy decoding? Which code do I need to modify?

Lackel commented 1 month ago

Hi, thanks for your interest in our work.

You can uncomment this line for greedy decoding. https://github.com/Lackel/AGLA/blob/efa126347c41631152a70d7db0a6ac0708bd9d00/eval/sample.py#L156