Dataset Viewer
Auto-converted to Parquet Duplicate
Search is not available for this dataset
action
list
timestamp
float32
0
10.5
frame_index
int64
0
314
episode_index
int64
0
148
index
int64
0
31.6k
task_index
int64
0
2
observation.state
list
[ 0, 0, 0, 0 ]
0
0
0
0
0
[ 80, 0 ]
[ 0, 0, 0, 0 ]
0.033333
1
0
1
0
[ 80, 0 ]
[ 0, 0, 0, 0 ]
0.066667
2
0
2
0
[ 80, 0 ]
[ 0, 0, 0, 0 ]
0.1
3
0
3
0
[ 80, 0 ]
[ 0, 0, 0, 0 ]
0.133333
4
0
4
0
[ 80, 0 ]
[ 0, 0, 0, 0 ]
0.166667
5
0
5
0
[ 80, 0 ]
[ 0, 0, 0, 0 ]
0.2
6
0
6
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.233333
7
0
7
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.266667
8
0
8
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.3
9
0
9
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.333333
10
0
10
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.366667
11
0
11
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.4
12
0
12
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.433333
13
0
13
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.466667
14
0
14
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.5
15
0
15
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.533333
16
0
16
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.566667
17
0
17
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.6
18
0
18
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.633333
19
0
19
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.666667
20
0
20
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.7
21
0
21
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.733333
22
0
22
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.766667
23
0
23
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
0.8
24
0
24
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
0.833333
25
0
25
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
0.866667
26
0
26
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
0.9
27
0
27
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
0.933333
28
0
28
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
0.966667
29
0
29
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1
30
0
30
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.033333
31
0
31
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.066667
32
0
32
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.1
33
0
33
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.133333
34
0
34
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.166667
35
0
35
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.2
36
0
36
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.233333
37
0
37
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.266667
38
0
38
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.3
39
0
39
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.333333
40
0
40
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.366667
41
0
41
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.4
42
0
42
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.433333
43
0
43
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.466667
44
0
44
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.5
45
0
45
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.533333
46
0
46
0
[ 80, 0 ]
[ 1, 0, -1, -1 ]
1.566667
47
0
47
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.6
48
0
48
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.633333
49
0
49
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.666667
50
0
50
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.7
51
0
51
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.733333
52
0
52
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.766667
53
0
53
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.8
54
0
54
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
1.833333
55
0
55
0
[ 70, -300 ]
[ 1, 0, 0, -1 ]
1.866667
56
0
56
0
[ 70, 0 ]
[ 1, 0, 0, -1 ]
1.9
57
0
57
0
[ 70, 0 ]
[ 1, 0, 0, -1 ]
1.933333
58
0
58
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
1.966667
59
0
59
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2
60
0
60
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.033333
61
0
61
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.066667
62
0
62
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.1
63
0
63
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.133333
64
0
64
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.166667
65
0
65
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.2
66
0
66
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.233333
67
0
67
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.266667
68
0
68
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.3
69
0
69
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.333333
70
0
70
0
[ 70, 0 ]
[ 1, 0, 0, 0 ]
2.366667
71
0
71
0
[ 80, 300 ]
[ 1, 0, 0, 0 ]
2.4
72
0
72
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.433333
73
0
73
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.466667
74
0
74
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.5
75
0
75
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
2.533333
76
0
76
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
2.566667
77
0
77
0
[ 80, 0 ]
[ 1, 0, 0, -1 ]
2.6
78
0
78
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.633333
79
0
79
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.666667
80
0
80
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.7
81
0
81
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.733333
82
0
82
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.766667
83
0
83
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.8
84
0
84
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.833333
85
0
85
0
[ 80, 0 ]
[ 1, 0, 0, 1 ]
2.866667
86
0
86
0
[ 80, 0 ]
[ 1, 0, 0, 1 ]
2.9
87
0
87
0
[ 80, 0 ]
[ 1, 0, 0, 1 ]
2.933333
88
0
88
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
2.966667
89
0
89
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
3
90
0
90
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
3.033333
91
0
91
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
3.066667
92
0
92
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
3.1
93
0
93
0
[ 80, 0 ]
[ 1, 0, -1, 0 ]
3.133333
94
0
94
0
[ 80, 0 ]
[ 1, 0, -1, 0 ]
3.166667
95
0
95
0
[ 80, 0 ]
[ 1, 0, -1, 0 ]
3.2
96
0
96
0
[ 80, 0 ]
[ 1, 0, -1, 0 ]
3.233333
97
0
97
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
3.266667
98
0
98
0
[ 80, 0 ]
[ 1, 0, 0, 0 ]
3.3
99
0
99
0
[ 80, 0 ]
End of preview. Expand in Data Studio

This dataset was created using LeRobot.

Dataset Description

Repaired copy of Grigorij/Tello_multifruit_sum (Tello drone, 149 episodes, 31624 frames, 30 fps). Video, actions, tasks, timestamps and all indices are byte-for-byte identical to the original; only observation.state changed.

Why: in the original, observation.state was declared with shape=[1], which LeRobot deliberately stores as a bare scalar Arrow column (see get_hf_features_from_features). Policies that load several observation timesteps then receive a 2-D tensor [B, T] with no feature axis, and mistake the time axis for the feature axis. With vla_jepa (num_video_frames=8) this crashes in the action head: RuntimeError: mat1 and mat2 shapes cannot be multiplied (64x8 and 1x1024).

Fix: observation.state is now a genuine 2-channel vector, shape=[2], stored as fixed_size_list<float>[2], so it always carries a real feature axis:

channel name units definition
0 height cm the original altitude, copied verbatim and unchanged
1 height_velocity cm/s backward finite difference of channel 0 within each episode: v[0] = 0, v[t] = (h[t] - h[t-1]) * 30

Channel 0 is bitwise identical to the original observation.state for all 31624 frames, so the original signal is fully recoverable with state[..., 0].

Caveat on channel 1: the Tello reports altitude quantised to 10 cm, so the finite difference is 0 for 97.8% of frames and exactly +/-300 cm/s otherwise. It marks vertical transitions rather than providing a smooth velocity estimate.

Built with lerobot.datasets.dataset_tools (remove_feature, then add_features, then recompute_stats); meta/stats.json and the per-episode stats in meta/episodes/ were both regenerated for the new 2-channel shape.

  • Homepage: [More Information Needed]
  • Paper: [More Information Needed]
  • License: apache-2.0

Dataset Structure

meta/info.json:

{
    "codebase_version": "v3.0",
    "fps": 30,
    "features": {
        "observation.images.camera_front": {
            "dtype": "video",
            "shape": [
                480,
                640,
                3
            ],
            "names": [
                "height",
                "width",
                "channels"
            ],
            "info": {
                "video.height": 480,
                "video.width": 640,
                "video.codec": "av1",
                "video.pix_fmt": "yuv420p",
                "video.is_depth_map": false,
                "video.fps": 30,
                "video.channels": 3,
                "has_audio": false
            }
        },
        "action": {
            "dtype": "float32",
            "shape": [
                4
            ],
            "names": [
                "forward",
                "right",
                "up",
                "yaw"
            ]
        },
        "timestamp": {
            "dtype": "float32",
            "shape": [
                1
            ],
            "names": null
        },
        "frame_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "episode_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "task_index": {
            "dtype": "int64",
            "shape": [
                1
            ],
            "names": null
        },
        "observation.state": {
            "dtype": "float32",
            "shape": [
                2
            ],
            "names": [
                "height",
                "height_velocity"
            ]
        }
    },
    "total_episodes": 149,
    "total_frames": 31624,
    "total_tasks": 3,
    "chunks_size": 1000,
    "data_files_size_in_mb": 100,
    "video_files_size_in_mb": 200,
    "data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
    "video_path": "videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4",
    "robot_type": null,
    "splits": {
        "train": "0:149"
    }
}

Citation

BibTeX:

[More Information Needed]
Downloads last month
79

Models trained or fine-tuned on Grigorij/Tello_multifruit_sum_two_state_vars