Bot-AGI-1 – a robotics benchmark for VLMs
bot-agi.orgGreat idea. Interesting to see qwen's textual state evaluation for each frame.
Must say though that the keyboard controls didn't work as advertised for me at all, it seemed very borked.
Great idea. Interesting to see qwen's textual state evaluation for each frame.
Must say though that the keyboard controls didn't work as advertised for me at all, it seemed very borked.