search

whwu95 / FreeVA

FreeVA: Offline MLLM as Training-Free Video Assistant
Apache License 2.0
48 stars 0 forks source link

Failed to run "single_video_inference.py" #1

Closed DemoGit4LIANG closed 5 months ago

DemoGit4LIANG commented 5 months ago

Thanks for your interesting work~

I try to run single_video_inference.py but the program returns nothing. Besides, I noticed the D2 aggregation codes in "def temporal_aggregation()" have been canceled. So could you give some guidance to run your program? Thanks.

whwu95 commented 5 months ago

Thank you for your interest in my work!

The code defaults to using simplest D1. Please change the "num_frames" from 8 to 4. If you need to use D2 aggregation, comment D1 and uncomment D2. Then, the "num_frames" can be increased to 8.