Add VideoChat3/Seeker-173K format SFT jsonl

Update SFT json: +578 holmes samples (17100 total)

Add holmes_videos.tar sha256

Add holmes videos tar (217 videos, 1.7GB)

Add VideoChat3 training dataset entry config

Add VideoChat3/Seeker-173K format SFT jsonl (16522 samples)

Add original nested-format SFT json (16522 samples)

Upload part-03 (19GB)

Upload part-02 (40GB)

Upload part-01 (40GB)

Upload part-00 (40GB)

Add sha256 checksums

Add README

Upload video tool call SFT dataset (16522 samples, 8670 videos, 139GB)

initial commit