---
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
- tsfile
- timeseries
- format:tsfile
modality:
- tabular
- timeseries
configs:
- config_name: default
  data_files:
  - split: train
    path: data/yellow_block_grasp_vary_train.tsfile
size_categories:
- 10K<n<100K
---
# Yellow Block Grasp Vary (TsFile)

Converted from [`yangfengzzz/yellow_block_grasp_vary`](https://huggingface.co/datasets/yangfengzzz/yellow_block_grasp_vary) at pinned revision [`806ac92c4163d3b15dc5a3b3a58088d053a79cf8`](https://huggingface.co/datasets/yangfengzzz/yellow_block_grasp_vary/tree/806ac92c4163d3b15dc5a3b3a58088d053a79cf8). **Modalities: Time-series.**

## Dataset description

This LeRobot v2.1 SO-101 dataset records demonstrations of grasping a yellow block and putting it in a box.

The pinned task instruction is: “Grasp a yellow block and put it in the box”.

- **Source repository owner/publisher:** yangfengzzz
- **License:** apache-2.0 (declared by the source card).
- **Paper/homepage/citation:** no completed paper, homepage, or citation is documented in the pinned source card unless linked above.

## Dataset Scale

| Split | Episodes | Tasks | Source trajectory Parquets | TsFile rows | Sampling rate | TsFile files/shards |
|---|---:|---:|---:|---:|---:|---:|
| `train` | 100 | 1 | 100 | 28,373 | 30 Hz | 1 |

The staged Parquet has 28,373 rows and 17 columns including `Time`; TsFile chunk metadata independently reports the same 28,373 rows.

## TsFile schema

| Column | Role | TsFile type | Observed/source range |
|---|---|---|---|
| `Time` | TIME | INT64 | 0–16,500 ms; restarts per episode |
| `episode_index` | TAG | STRING | source integer 0–99 |
| `task_index` | TAG | STRING | source integer 0–0 |
| `frame_index` | FIELD | INT64 | 0–495 within an episode |
| `sample_index` | FIELD | INT64 | 0–28,372 globally |

Exact remaining FIELD names/ranges and imported types:

- `action_0`–`action_5` (FLOAT): shoulder pan/lift, elbow flex, wrist flex/roll, and gripper action
- `observation_state_0`–`observation_state_5` (FLOAT): the corresponding SO-101 state

## Conversion

- All source episodes in the train split are merged into `data/yellow_block_grasp_vary_train.tsfile`; `episode_index` and `task_index` are TAG dimensions.
- `Time = round(timestamp * 1000)` in milliseconds. The source `timestamp` column is omitted because it is redundant with `Time / 1000` seconds.
- `frame_index` is retained. Source `index` is retained as `sample_index`.
- Every vector is fully flattened: the complete source name is kept, `.` becomes `_`, and element indices are appended. Float vectors are imported as single-precision FLOAT fields.
- Other scalar source columns shown above are retained; no trajectory rows are intentionally dropped.
- Source metadata is mirrored for publication, with copied `meta/info.json` rewritten to describe the converted data path and conversion semantics.

## Video policy

The pinned metadata declares 200 source videos across top and wrist camera streams. MP4 files were not downloaded or uploaded; they remain in [the pinned source `videos/` tree](https://huggingface.co/datasets/yangfengzzz/yellow_block_grasp_vary/tree/806ac92c4163d3b15dc5a3b3a58088d053a79cf8/videos).

## Minimal read example

```python
from tsfile import TsFileReader

reader = TsFileReader("data/yellow_block_grasp_vary_train.tsfile")
print(reader.get_all_table_schemas().keys())
reader.close()
```

## Source and provenance

- Source dataset: [`yangfengzzz/yellow_block_grasp_vary`](https://huggingface.co/datasets/yangfengzzz/yellow_block_grasp_vary)
- Pinned source revision: [`806ac92c4163d3b15dc5a3b3a58088d053a79cf8`](https://huggingface.co/datasets/yangfengzzz/yellow_block_grasp_vary/tree/806ac92c4163d3b15dc5a3b3a58088d053a79cf8)
- Converted artifact: `data/yellow_block_grasp_vary_train.tsfile`