File size: 3,412 Bytes
f67ee80 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 | ---
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
- so100
- tutorial
- robotics
- tsfile
- time-series
- modality:timeseries
pretty_name: SO-100 Sorting (TsFile)
size_categories:
- 10K<n<100K
configs:
- config_name: default
data_files:
- split: train
path: data/so100_sorting.tsfile
---
# SO-100 Sorting (TsFile)
This dataset is an Apache TsFile conversion of the Hugging Face dataset
[`dragon-95/so100_sorting`](https://huggingface.co/datasets/dragon-95/so100_sorting).
The source dataset was created using [LeRobot](https://github.com/huggingface/lerobot).
Modalities: Time-series. The original repository also contains synchronized video
streams; videos are not included in this converted repository.
## Source Dataset
- Original dataset: [`dragon-95/so100_sorting`](https://huggingface.co/datasets/dragon-95/so100_sorting)
- License: `apache-2.0`
- LeRobot codebase version: `v2.0`
- Robot type: `so100`
- Task: `Put the object in box A into box B`
- Split: `train` (`0:61`)
- Source scale from `meta/info.json`: `61` episodes, `95,346` frames, `1` task
- Source video count: `122`
- Sampling rate: `50` fps
- Source data layout: `data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet`
- Source video layout: `videos/chunk-{episode_chunk:03d}/{video_key}/episode_{episode_index:06d}.mp4`
## Converted Files
- TsFile: `data/so100_sorting.tsfile`
- Converted rows: `95,346`
- TsFile table: `so100_sorting`
- Time precision: milliseconds
- TAG columns: `episode_index`, `task_index`
## Schema
`Time` is synthesized as `round(timestamp * 1000)` in milliseconds. The source
`timestamp` column is dropped because it is redundant with `Time / 1000` seconds.
At 50 fps, consecutive frames are spaced by about 20 ms.
TAG columns:
- `episode_index`
- `task_index`
FIELD columns:
- `frame_index`
- `sample_index` (renamed from source `index`)
- `action_0` to `action_5`
- `observation_state_0` to `observation_state_5`
Vector features are flattened by preserving the source feature name and replacing
`.` with `_`. For example, `observation.state` becomes
`observation_state_0` to `observation_state_5`. The 6-element `action` and
`observation.state` vectors use the source joint order:
`main_shoulder_pan`, `main_shoulder_lift`, `main_elbow_flex`, `main_wrist_flex`,
`main_wrist_roll`, and `main_gripper`.
## Video Policy
The following source video features are not converted into TsFile and are not
uploaded here:
- `observation.images.laptop`
- `observation.images.phone`
Use the original dataset for videos:
[`dragon-95/so100_sorting/videos`](https://huggingface.co/datasets/dragon-95/so100_sorting/tree/main/videos).
## Metadata
The source `meta/` files are mirrored in this repository. `meta/info.json` is
updated so `data_path` points to `data/so100_sorting.tsfile` and includes a
`tsfile_conversion` object documenting the Time mapping, TAG columns, flattened
features, dropped fields, and video policy.
## Validation
The converted TsFile was validated with the project pipeline and read back using
the TsFile Python SDK:
- staged Parquet rows: `95,346`
- TsFile metadata rows: `95,346`
- TsFile query rows: `95,346`
- TsFile size: `1,682,521` bytes
## Usage
```python
from tsfile import TsFileReader
path = "data/so100_sorting.tsfile"
with TsFileReader(path) as reader:
schemas = reader.get_all_table_schemas()
print(schemas.keys())
```
|