Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Server error while post-processing the rows. Please report the issue.
Error code:   RowsPostProcessingError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

Filtering 统计

Filtering 文件每行格式为 Category VideoID ClipID。Video ID 数量按第二列去重统计,不是 Clip ID 数量。

文件 筛选条件 Clip 数量 Video ID 数量
PanoData/filtering_v2.txt 全部 30,120 1,186
PanoData/filtering_3840HR_v2.txt W >= 3840 19,096 836
PanoData/filtering_5120HR_v2.txt W >= 5120 11,090 203

完整下载

先安装 Hugging Face CLI:

python3 -m pip install -U "huggingface_hub>=0.32"

若数据集需要权限,先登录:

hf auth login

在希望保存数据的目录中执行:

mkdir -p Pano-World
cd Pano-World
hf download "cszhilu1998/data" --repo-type dataset --local-dir .

该命令会下载数据集中的全部文件。如果当前目录已有 archives/ 中的部分或全部文件,重新执行同一命令即可继续下载并补齐其他文件。

校验分卷

下载完成后,在 Pano-World 目录中执行:

cd archives
sha256sum -c SHA256SUMS
cd ..

所有分卷都显示 OK 后再解压。

解压到 ./PanoData

tar 分卷内已包含 PanoData/<directory> 路径。在 Pano-World 目录中执行以下命令,不要再使用 -C PanoData

for name in intrinsics pose video video_1024_512; do
  cat "archives/${name}.tar.part-"* | tar -xf -
done

解压完成后的结构如下:

PanoData/
├── intrinsics/
├── pose/
├── video/
├── video_1024_512/
├── filtering_v2.txt
├── filtering_3840HR_v2.txt
├── filtering_5120HR_v2.txt
├── video_resolution.jsonl
└── wan22_pano_prompts.jsonl
Downloads last month
46