THULab/ReubenLim_so101_tape_in_square

SO101 Tape in Square TsFile

This is an Apache TsFile conversion of ReubenLim/so101_tape_in_square, a LeRobot v2.1 SO101 dataset for pushing a circle block to the blue square in the center.

Source and provenance

• Original dataset: ReubenLim/so101_tape_in_square
• Source repository snapshot: 44444e34637363c81246c1cb93ca915a5dfc9fb8
• Repository owner, uploader, and commit author: Reuben Lim Yaw Hui (ReubenLim)
• License: Apache-2.0
• Paper/citation: not provided by the source repository
• Robot: so101; LeRobot codebase: v2.1
• Split: train; sampling frequency: 30 fps
• Task 0: Push a circle block to the blue square in the center.
• Scale: 62 episodes, 51,284 rows, one task, 62 source Parquet shards
• Source data path: data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet

The source README still embeds an obsolete 3-episode info.json example. The counts above come from the repository meta/info.json, all source Parquet footers, and the complete repository tree.

TsFile dataset

• TsFile: data/reubenlim_so101_tape_in_square.tsfile
• Table: reubenlim_so101_tape_in_square
• Rows/devices: 51,284 / 62
• Time precision: milliseconds
• Source Parquet size: 2,426,794 bytes
• TsFile size: 711,154 bytes
• TsFile/source size ratio: 0.293043
• meta/ is preserved, with meta/info.json updated for this TsFile artifact.

Schema

Column or group	TsFile role	Type	Notes
Time	TIME	INT64	round(timestamp * 1000) ms; restarts at zero per episode
episode_index	TAG	STRING	Original integer value stored as a TsFile device/TAG segment
task_index	TAG	STRING	Original integer value stored as a TsFile device/TAG segment
frame_index	FIELD	INT64	Original per-episode frame index
sample_index	FIELD	INT64	Renamed from source index
action_0 ... action_5	FIELD	FLOAT	Flattened source action[6]
observation_state_0 ... observation_state_5	FIELD	FLOAT	Flattened source observation.state[6]

Vector prefixes preserve their source names, with . replaced by _.

Conversion details

• Time: TS_2DIFF + LZ4
• FLOAT/DOUBLE: GORILLA + LZ4
• INT32/INT64: TS_2DIFF + LZ4
• BOOLEAN: RLE + LZ4 when present; this source has no BOOLEAN field
• TAG values use the TsFile table-model device/tag mechanism.
• All train episodes are merged into one table and sorted by TAG columns and Time.
• Source timestamp is dropped because it is represented by Time / 1000 seconds.
• No numeric rows, action/state dimensions, frame indexes, episode indexes, or task indexes are dropped.
• Camera pixels are video assets rather than Parquet columns and are not embedded in TsFile.

Videos

Videos are not included in this TsFile dataset. The original repository stores 124 frame-aligned AV1 MP4 files: 62 episodes for each of the birdEye and thirdPerson streams under videos/chunk-000/. The exact template is videos/chunk-{episode_chunk:03d}/{video_key}/episode_{episode_index:06d}.mp4. The source videos total 1,157,458,986 bytes and are 640x480 at 30 fps without audio. Stream sizes are 472,915,377 bytes (birdEye) and 684,543,609 bytes (thirdPerson). Use episode_index and frame_index to align TsFile rows with both views.

Minimal read example

from tsfile import TsFileReader

reader = TsFileReader("data/reubenlim_so101_tape_in_square.tsfile")
columns = [
    "episode_index",
    "task_index",
    "frame_index",
    "sample_index",
    "action_0",
    "observation_state_0",
]
with reader.query_table("reubenlim_so101_tape_in_square", columns, batch_size=65536) as result:
    batch = result.read_arrow_batch()
    print(batch.to_pandas().head())
reader.close()

Citation

The source repository provides no formal citation. Cite the original Hugging Face dataset and its repository author Reuben Lim Yaw Hui (ReubenLim) when using this conversion.