main
data.jsonl
video → text
Google
num_people_gemini_pro
Tell me the total number many people are in this video, there are three categories: zero, one, many. Also tell me if Eminem the rapper is in the video: yes or no.
Respond only with the category and the binary classification.
For example example responses could be:
zero,no
one,yes
one,no
many,yes
many,no
{clip_path} Apr 2, 2026, 3:26 AM UTC
Apr 2, 2026, 6:38 AM UTC
1543 rows
749785 tokens$ 1.54
1543 rows processed, 749785 tokens used ($1.54)
completed
7 columns, 1-100 of 1543 rows