File size: 4,925 Bytes
2246112
 
 
 
 
 
 
 
 
d808a8f
2246112
 
 
 
 
 
 
 
291c3d9
 
 
2246112
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
---
license: apache-2.0
task_categories:
- robotics
tags:
- LeRobot
- robotics
- tsfile
- timeseries
- format:tsfile
pretty_name: SAM Frames5 (TsFile)
size_categories:
- 10K<n<100K
configs:
- config_name: default
  data_files:
  - split: train
    path: data/sam_frames5_train.tsfile
modality:
- tabular
- timeseries
---

# SAM Frames5 (TsFile)

This dataset is an Apache TsFile conversion of the Hugging Face dataset
[`1g0rrr/sam_frames5`](https://huggingface.co/datasets/1g0rrr/sam_frames5).
The source was created with
[LeRobot](https://github.com/huggingface/lerobot) and contains `sam_double`
robot demonstrations for peeling the protective layer from adhesive tape.

Modalities: Time-series. The source repository also contains three synchronized
camera streams; videos are not included in this converted repository.

## Source Dataset

- Original dataset: [`1g0rrr/sam_frames5`](https://huggingface.co/datasets/1g0rrr/sam_frames5)
- License: `apache-2.0`
- LeRobot codebase version: `v2.1`
- Robot type: `sam_double`
- Task: `Peel off the protective layer from the adhesive tape.`
- Split: `train` (`0:51`)
- Scale: `51` episodes, `32,369` frames, `1` task
- Sampling rate: `30` fps
- Source frame files: `51` Parquet files
- Source videos: `153` MP4 files across `3` camera streams
- 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`

The camera streams are `observation.images.laptop`,
`observation.images.phone`, and `observation.images.side`. Each stream contains
51 videos at 30 fps. Source video frames are 480 x 640 RGB, encoded with AV1,
and contain no audio.

## Converted File

- TsFile: `data/sam_frames5_train.tsfile`
- TsFile table: `sam_frames5_train`
- Converted rows: `32,369`
- Episodes: `51`
- Time precision: milliseconds
- TAG columns: `episode_index`, `task_index`
- File size: `537,440` bytes
- SHA-256: `7415819c4cfe7786f3fe18ff489fe58e8c22abb1c7128c51ebc324c47c4bc788`

All train episodes are merged into one TsFile. The source `episode_index` and
`task_index` columns are retained as TAG columns, so queries can select an
episode without synthetic aliases.

## Schema

`Time` is computed as `Time = round(timestamp * 1000)` in milliseconds and
restarts in each episode. At 30 fps, consecutive frames are approximately
33.333 ms apart. The source `timestamp` column is not retained because it is
redundant with `Time / 1000` seconds. No source rows are dropped.

TAG columns:

- `episode_index`
- `task_index`

FIELD columns:

- `frame_index`
- `sample_index` (renamed from source `index`)
- `action_0` through `action_6` (FLOAT)
- `observation_state_0` through `observation_state_6` (FLOAT)

The seven vector elements use this source order for both `action` and
`observation.state`:

1. `main_shoulder_pan`
2. `main_shoulder_lift`
3. `main_elbow_flex`
4. `main_wrist_flex`
5. `main_wrist_side`
6. `main_wrist_roll`
7. `main_gripper`

Vector column names preserve the complete source feature name: `.` is replaced
with `_`, then the zero-based element index is appended.

## Video Policy

The three source video features are not converted or uploaded. Use the original
dataset for synchronized videos:
[`1g0rrr/sam_frames5/videos`](https://huggingface.co/datasets/1g0rrr/sam_frames5/tree/main/videos).

Each numeric row retains `episode_index`, `frame_index`, `task_index`, and
`sample_index`, preserving alignment with the original per-episode videos.

## Metadata

The source `meta/` files are mirrored in this repository. `meta/info.json` is
updated so `data_path` points to `data/sam_frames5_train.tsfile`. Its
`tsfile_conversion` object records the source and converted file counts, table
name, Time formula, TAG columns, row count, feature mappings, and frame/video
alignment. The converted `total_videos` is `0`; the source count of `153` is
preserved as `tsfile_conversion.source_video_count`.

## Validation

The converted file was compared with all 51 source Parquet files and read back
with the TsFile Python SDK:

- source and staged rows: `32,369`
- duplicate `(episode_index, task_index, Time)` rows: `0`
- maximum action-vector difference: `0`
- maximum observation-state-vector difference: `0`
- source `index` to converted `sample_index` mismatches: `0`

## Usage

```python
from tsfile import TsFileReader

path = "data/sam_frames5_train.tsfile"
with TsFileReader(path) as reader:
    schemas = reader.get_all_table_schemas()
    table = schemas["sam_frames5_train"]
    print([(column.get_column_name(), column.get_category())
           for column in table.get_columns()])
```

## Source & License

The source dataset is maintained at
[`1g0rrr/sam_frames5`](https://huggingface.co/datasets/1g0rrr/sam_frames5) and
is distributed under the Apache License 2.0. The source card does not provide a
paper, author list, or citation entry.