--- license: mit task_categories: - video-text-to-text --- # VideoChat2 TextVR archive An unofficial, unmodified copy of `textvr.zip` from [shenxq/VideoChat2](https://huggingface.co/datasets/shenxq/VideoChat2/tree/e57b64db316632942279351a025e220940e75dc6). Only the requested archive is included, alongside this card and a checksum manifest. - Source revision: `e57b64db316632942279351a025e220940e75dc6` - Size: 5,409,479,797 bytes (5.41 GB) - SHA256: `ae86d8ea109a8d9404070f82ad0e821be0c07de088e91fb4ed9eb70abbee96bd` - Original TextVR project: https://github.com/callsys/textvr ## Download ```bash hf download axiong/VideoChat2-textvr textvr.zip --repo-type dataset --local-dir ./VideoChat2-textvr ``` ## Source and terms The [source card](https://huggingface.co/datasets/shenxq/VideoChat2/blob/e57b64db316632942279351a025e220940e75dc6/README.md) describes video training data for [LongVU](https://github.com/Vision-CAIR/LongVU) downloaded from [OpenGVLab/VideoChat2-IT](https://huggingface.co/datasets/OpenGVLab/VideoChat2-IT). The MIT metadata above is inherited from that card. Original media retains its applicable terms; this mirror grants no additional rights. No data transformation or extraction has been performed.