saaduddinM/OXE_fractal20220817_data_embeddings

LeRobot语言表（Language Table LeRobot）— 仅嵌入版本。这是一个基于IPEC-COMMUNITY/fractal20220817_data_lerobot数据集的重编码变体，其中原始视频被替换为固定长度的图像嵌入，任务字符串增加了文本嵌入。所有索引、分割和语义与源数据集保持一致，而存储和I/O则大大减轻。为了便于上传/下载和从Hub中流式传输，还将每个小型的剧集Parquet文件合并为N个大的Parquet碎片，放在单个data/文件夹下。

Language Table (LeRobot) — Embedding-Only Release. This is a re-encoded variant of the IPEC-COMMUNITY/fractal20220817_data_lerobot dataset where raw videos are replaced by fixed-length image embeddings, and task strings are augmented with text embeddings. All indices, splits, and semantics remain consistent with the source dataset while storage and I/O are substantially lighter. For ease of upload/download and streaming from the Hub, tiny per-episode Parquet files have been consolidated into N large Parquet shards under a single data/ folder.