ny_insert_connector_viktor_crop

This dataset is an Apache TsFile conversion of the Hugging Face dataset 1g0rrr/ny_insert_connector_viktor_crop. The source dataset was created with LeRobot.

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

Source Dataset

• Original dataset: 1g0rrr/ny_insert_connector_viktor_crop
• Source license: apache-2.0
• LeRobot codebase version: v2.1
• Robot type: sam_evt2
• Task name: ny_insert_connector
• Task: Insert connector
• Split: train (0:26)
• Source scale from meta/info.json: 26 episodes, 14,948 frames, 1 task
• Source video count from meta/info.json: 104
• Sampling rate: 60 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

Episode lengths range from 449 to 718 frames, with 14,948 frames in total.

Converted Files

• TsFile: data/ny_insert_connector_viktor_crop.tsfile
• Converted rows: 14,948
• Episodes: 26
• TsFile table: ny_insert_connector_viktor_crop
• Time precision: milliseconds
• TAG columns: episode_index, task_index
• TsFile size: 709,196 bytes

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 60 fps, consecutive frames are spaced by about 17 ms.

TAG columns:

• episode_index
• task_index

FIELD columns:

• frame_index
• sample_index (renamed from source index)
• action_0 to action_13
• observation_state_0 to observation_state_13

Vector features are flattened by preserving the source feature name and replacing . with _. For example, observation.state becomes observation_state_0 to observation_state_13.

The 14-element action and observation.state vectors use the source joint order: left_shoulder_pan.pos, left_shoulder_lift.pos, left_elbow_flex.pos, left_wrist_flex.pos, left_wrist_side.pos, left_wrist_roll.pos, right_shoulder_pan.pos, right_shoulder_lift.pos, right_elbow_flex.pos, right_wrist_flex.pos, right_wrist_side.pos, right_wrist_roll.pos, left_gripper.pos, and right_gripper.pos.

Video Policy

The source video features are not converted into TsFile and are not uploaded here:

• observation.images.wrist_left
• observation.images.wrist_right
• observation.images.front
• observation.images.side

Use the original dataset for videos: 1g0rrr/ny_insert_connector_viktor_crop/videos.

Metadata

The source meta/ files are mirrored in this repository. meta/info.json is updated so data_path points to data/ny_insert_connector_viktor_crop.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: 14,948
• TsFile metadata rows: 14,948
• TsFile query rows: 14,948
• TsFile size: 709,196 bytes

Usage

from tsfile import TsFileReader

path = "data/ny_insert_connector_viktor_crop.tsfile"
with TsFileReader(path) as reader:
    schemas = reader.get_all_table_schemas()
    print(schemas.keys())