Will mm-video dataset be released?

Thank you for releasing traces for multimodal inference, it has been very helpful!

I wanted to ask whether you are also planning to release the `mm-video` or `mm-omni` datasets used in the paper. Having access to video traces (at least number of video tokens and request timestamps) would be extremely valuable for extending research on multimodal tasks.

If so, could you share whether there is a timeline or plan for making it available?

Thanks again for your great work!

No files, tests, or entry points are identified. Start by checking the paper and repository references to the `mm-video` and `mm-omni` datasets, then determine whether release details or a timeline exist. Done means the dataset availability and release status are clearly documented.

Written by the indexing model from the issue text.