hexuan21/for_test

该数据集包含多个特征，如视频ID、视觉质量、时间一致性、动态程度、文本到视频对齐、事实一致性、文本、视频链接和文本到视频模型。数据集的特征数据类型包括字符串和整型。数据集配置中有一个名为pika的配置，包含一个训练集的数据文件路径。

The dataset includes multiple features such as video ID, visual quality, temporal consistency, dynamic degree, text-to-video alignment, fact consistency, text, video link, and text-to-video model. The data types of the features include string and integer. The dataset configuration includes a configuration named pika, which contains a data file path for the training set.

原始信息汇总

数据集概述

数据集配置

• config_name: all

  • split: train
    • path: ans_detailed_all.json

• config_name: pika

  • split: train
    • path: ans_detailed_pika.json

• config_name: t2vz

  • split: train
    • path: ans_detailed_t2vz.json

• config_name: vc2

  • split: train
    • path: ans_detailed_vc2.json

• config_name: ms

  • split: train
    • path: ans_detailed_ms.json

• config_name: lavie_base

  • split: train
    • path: ans_detailed_lavie_base.json

• config_name: anidiff

  • split: train
    • path: ans_detailed_anidiff.json

• config_name: lvdm

  • split: train
    • path: ans_detailed_lvdm.json

• config_name: hotshot

  • split: train
    • path: ans_detailed_hotshot.json

• config_name: zs_576w

  • split: train
    • path: ans_detailed_zs_576w.json

• config_name: didemo_high_res

  • split: train
    • path: ans_detailed_didemo_high_res.json

许可证

• license: apache-2.0