CinematicT2vData/t2v_style_captions

该数据集是一个视频描述数据集，包含视频ID、视频URL、场景名称、提示文本、视频字幕、场景数量、起始帧、结束帧、起始时间码、结束时间码、起始秒数、结束秒数、视频持续时间以及与前一个片段的时间间隔等多个字段。数据集分为三个部分：caption_base_results_cinepile_checkpoint、caption_scene_text_results_cinepile_checkpoint和caption_t2v_style_results_cinepile_checkpoint，每个部分都包含2166个视频描述示例。

This dataset is a video description dataset, including fields such as video ID, video URL, scene name, prompt text, video captions, scene count, start frame, end frame, start timecode, end timecode, start seconds, end seconds, duration in seconds, and time gap from the previous segment. The dataset is divided into three parts: caption_base_results_cinepile_checkpoint, caption_scene_text_results_cinepile_checkpoint, and caption_t2v_style_results_cinepile_checkpoint, each containing 2166 video description examples.