Datasets:

azithromycin
/

LSY-lab_libero_plus_goal_camera_all_kinova3_failures

Exception:    SplitsNotFoundError
Message:      The split names could not be parsed from the dataset config.
Traceback:    Traceback (most recent call last):
                File "tsfile/tsfile_py_cpp.pyx", line 567, in tsfile.tsfile_py_cpp.tsfile_reader_new_c
              tsfile.exceptions.FileOpenError: 28: 
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 286, in get_dataset_config_info
                  for split_generator in builder._split_generators(
                                         ~~~~~~~~~~~~~~~~~~~~~~~~~^
                      StreamingDownloadManager(base_path=builder.base_path, download_config=download_config)
                      ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/tsfile/tsfile.py", line 271, in _split_generators
                  scan = self._scan_metadata(all_files)
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/tsfile/tsfile.py", line 318, in _scan_metadata
                  with self._open_reader(file) as reader:
                       ~~~~~~~~~~~~~~~~~^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/tsfile/tsfile.py", line 742, in _open_reader
                  return TsFileReader(file)
                File "tsfile/tsfile_reader.pyx", line 323, in tsfile.tsfile_reader.TsFileReaderPy.__init__
              SystemError: <class '_weakrefset.WeakSet'> returned a result with an exception set
              
              The above exception was the direct cause of the following exception:
              
              Traceback (most recent call last):
                File "/src/services/worker/src/worker/job_runners/config/split_names.py", line 71, in compute_split_names_from_streaming_response
                  for split in get_dataset_split_names(
                               ~~~~~~~~~~~~~~~~~~~~~~~^
                      path=dataset,
                      ^^^^^^^^^^^^^
                      config_name=config,
                      ^^^^^^^^^^^^^^^^^^^
                      token=hf_token,
                      ^^^^^^^^^^^^^^^
                  )
                  ^
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 340, in get_dataset_split_names
                  info = get_dataset_config_info(
                      path,
                  ...<6 lines>...
                      **config_kwargs,
                  )
                File "/usr/local/lib/python3.14/site-packages/datasets/inspect.py", line 291, in get_dataset_config_info
                  raise SplitsNotFoundError("The split names could not be parsed from the dataset config.") from err
              datasets.inspect.SplitsNotFoundError: The split names could not be parsed from the dataset config.

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

LIBERO Plus Goal Camera All Kinova3 Failures (TsFile)

This repository is an Apache TsFile conversion of LSY-lab/libero_plus_goal_camera_all_kinova3_failures, a LeRobot v3.0 dataset for Kinova3 manipulation trajectories. The source repository is published by the Learning Systems Lab (LSY-lab) and the source upload is attributed to Maximilian Christof (max-chr). The source card declares the Apache-2.0 license and does not provide a dataset-specific paper citation; the tasks are part of the LIBERO-Plus project.

Source dataset and videos

Source dataset: https://huggingface.co/datasets/LSY-lab/libero_plus_goal_camera_all_kinova3_failures
Source revision inspected: 23c505b64fcdeedb558f478776c79aea4c9de4a9
Robot: kinova3; LeRobot codebase: v3.0; sampling rate: 20 fps
Source frame layout: data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet
Source video layout: videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4
Camera streams: observation.images.image, observation.images.wrist_image, observation.images.frontview_image, and observation.images.sideview_image (256×256 RGB, AV1, 20 fps).

The TsFile repository intentionally does not include videos/. Retrieve videos from the original Hugging Face video tree. Align a video with numeric rows using episode_index and frame_index; meta/episodes/chunk-000/file-000.parquet contains the per-stream chunk/file and timestamp ranges.

Converted artifact

TsFile: data/lsy_lab_libero_plus_goal_camera_all_kinova3_failures.tsfile
Table: lsy_lab_libero_plus_goal_camera_all_kinova3_failures
Split: train
Source shards: 1 Parquet file; merged output: 1 TsFile
Rows: 56,874; episodes: 258; tasks: 10
Source Parquet size: 10,202,862 bytes; TsFile size: 7,033,403 bytes (smaller after encoding/compression)

Schema

Time is an INT64 millisecond timestamp computed as round(timestamp * 1000). The source timestamp is then dropped because it is redundant with Time / 1000. Time remains monotonic within each (episode_index, task_index) device segment; the source episode_index and task_index values are preserved as TsFile TAG columns.

Role	Columns
TIME	`Time` (INT64, ms)
TAG/device	`episode_index` (INT64), `task_index` (INT64)
FIELD metadata	`frame_index`, `sample_index` (renamed from `index`)
FIELD vectors	`observation_state_0..7`, `observation_states_ee_state_0..5`, `observation_states_joint_state_0..6`, `observation_states_gripper_state_0..1`, `action_0..6` (FLOAT32)

Vector columns are flattened row-major to scalar FLOAT fields. Source dots are replaced with underscores: for example, observation.states.ee_state becomes observation_states_ee_state_0 … _5. No numeric rows or state/action dimensions are dropped; only timestamp is dropped and index is renamed.

The source contains some IEEE-754 NaN values in state/action dimensions (preserved in the TsFile); these are data values rather than dropped rows.

TsFile encoding and compression

The table writer uses the requested compact policy:

FLOAT/DOUBLE → GORILLA + LZ4
INT32/INT64 → TS_2DIFF + LZ4
TIME → TS_2DIFF + LZ4
BOOLEAN (if present in a future source revision) → RLE + LZ4
TAG values → TsFile table-model device/TAG segments

The staged Parquet and final TsFile were read back with the TsFile Python SDK. Metadata and query readback both report 56,874 rows, and the converted schema contains 1 TIME, 2 TAG, and 32 FIELD columns.

Task index mapping

put the wine bottle on the rack
open the middle layer of the drawer
put the wine bottle on the top of the drawer
put the bowl on the plate
put the bowl on the top of the drawer
open the top layer of the drawer and put the bowl inside
push the plate to the front of the stove
put the bowl on the stove
put the cream cheese on the bowl
turn on the stove

Minimal read example

from tsfile import TsFileReader

path = "data/lsy_lab_libero_plus_goal_camera_all_kinova3_failures.tsfile"
reader = TsFileReader(path)
table_name = "lsy_lab_libero_plus_goal_camera_all_kinova3_failures"
schema = reader.get_all_table_schemas()[table_name]
columns = [c.get_column_name() for c in schema.get_columns()
           if c.get_column_name() != "Time"]

with reader.query_table(table_name, columns, batch_size=65536) as result:
    batch = result.read_arrow_batch()

The dataset-specific YAML configuration and converter are kept locally under D:/code/config/ and D:/code/scripts/; validation reports are local conversion artifacts and are not part of the TsFile upload set.

Downloads last month: 27

Total file size:

7.7 MB