Loading video...
Video Failed to Load
Ask-Anything, tool for chatting about video with chatGPT, miniGPT4 and StableLM github: Gradio demo:
143,643 views • 3 years ago •via X (Twitter)
10 Comments

@Gradio That link is http and an up with an exposed gradio port is like straight out of the 90s

@Gradio I just cannot catchup with the rate of progress in AI lately.

@Gradio @DotCSV mira esto!

@Gradio Congrats to our team for making this happen, excellent work!

@Gradio Can this be connected to AutoGPT?

@Gradio Amazing tool for chatting! 🚀 Try our new ChatGPT-O-Matic for endless fun! But wait, there's more, did you hear about the #kangaroo who learned to hop on a unicycle? 😉 #ChatGPTOMatic

@Scobleizer @Gradio Looks great but these questions seems answerable with current stills from the video. Are there intertemporal questions? Of which multiple previous frames/scenes are required to know the answer.

@Gradio Does it work on static images too like previous methods, and can we expect better performance in non-temporal task? I suspect that temporal insight increases the robustness on non-temporal interpretation with crude causal reasoning, and grounding may improve the LM itself as well

@pro_raze @Gradio Crazy..

@Gradio @_akhaliq is MiniGPT-4 really that good?

