nepfaff/steerable-scene-generation-combined

这是一个包含场景标签、子数据集索引和语言注释的数据集。数据集被划分为训练集，共有约4479万示例，总大小为约1.3 PB。数据集支持默认配置，训练集数据文件以data/train-*的模式存储。This dataset includes scene labels, subdataset indices, and language annotations. The dataset is split into a training set with approximately 44,790,942 examples and a total size of about 1.3 PB. The dataset supports a default configuration, with training set data files stored in the pattern data/train-*.