The Keye-VL-8B-Preview model supports understanding and interpreting videos, but currently, this node only allows for understanding in the form of images. Is it possible to add video input? I tested a video with 81 frames in mp4 format, and the prompts were also explained in video form, but ultimately, it resulted in 81 individual images, each accompanied by an explanation. Can we enable video input since the model itself supports understanding videos, but it seems like the node does not?
The Keye-VL-8B-Preview model supports understanding and interpreting videos, but currently, this node only allows for understanding in the form of images. Is it possible to add video input? I tested a video with 81 frames in mp4 format, and the prompts were also explained in video form, but ultimately, it resulted in 81 individual images, each accompanied by an explanation. Can we enable video input since the model itself supports understanding videos, but it seems like the node does not?