Datasets:

npow
/

so101_recording_20260519_000250_augmented_v2

Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
episode_index: int64
stats: struct<observation.images.front: struct<min: list<item: list<item: list<item: double>>>, max: list<i (... 1441 chars omitted)
  child 0, observation.images.front: struct<min: list<item: list<item: list<item: double>>>, max: list<item: list<item: list<item: double (... 129 chars omitted)
      child 0, min: list<item: list<item: list<item: double>>>
          child 0, item: list<item: list<item: double>>
              child 0, item: list<item: double>
                  child 0, item: double
      child 1, max: list<item: list<item: list<item: double>>>
          child 0, item: list<item: list<item: double>>
              child 0, item: list<item: double>
                  child 0, item: double
      child 2, mean: list<item: list<item: list<item: double>>>
          child 0, item: list<item: list<item: double>>
              child 0, item: list<item: double>
                  child 0, item: double
      child 3, std: list<item: list<item: list<item: double>>>
          child 0, item: list<item: list<item: double>>
              child 0, item: list<item: double>
                  child 0, item: double
      child 4, count: list<item: int64>
          child 0, item: int64
  child 1, observation.images.side: struct<min: list<item: list<item: list<item: double>>>, max: list<item: list<item: list<item: double (... 129 chars omitted)
      child 0, min: list<item: list<item: list<item: double>>>
          child 0, item: list<item: list<item: double>>
    
...
ax: list<item: int64>, mean: list<item: double>, std: list<item: dou (... 31 chars omitted)
      child 0, min: list<item: int64>
          child 0, item: int64
      child 1, max: list<item: int64>
          child 0, item: int64
      child 2, mean: list<item: double>
          child 0, item: double
      child 3, std: list<item: double>
          child 0, item: double
      child 4, count: list<item: int64>
          child 0, item: int64
  child 7, index: struct<min: list<item: int64>, max: list<item: int64>, mean: list<item: double>, std: list<item: dou (... 31 chars omitted)
      child 0, min: list<item: int64>
          child 0, item: int64
      child 1, max: list<item: int64>
          child 0, item: int64
      child 2, mean: list<item: double>
          child 0, item: double
      child 3, std: list<item: double>
          child 0, item: double
      child 4, count: list<item: int64>
          child 0, item: int64
  child 8, task_index: struct<min: list<item: int64>, max: list<item: int64>, mean: list<item: double>, std: list<item: dou (... 31 chars omitted)
      child 0, min: list<item: int64>
          child 0, item: int64
      child 1, max: list<item: int64>
          child 0, item: int64
      child 2, mean: list<item: double>
          child 0, item: double
      child 3, std: list<item: double>
          child 0, item: double
      child 4, count: list<item: int64>
          child 0, item: int64
tasks: list<item: string>
  child 0, item: string
length: int64
to
{'episode_index': Value('int64'), 'tasks': List(Value('string')), 'length': Value('int64')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 478, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2818, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2355, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2380, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2369, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2297, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              episode_index: int64
              stats: struct<observation.images.front: struct<min: list<item: list<item: list<item: double>>>, max: list<i (... 1441 chars omitted)
                child 0, observation.images.front: struct<min: list<item: list<item: list<item: double>>>, max: list<item: list<item: list<item: double (... 129 chars omitted)
                    child 0, min: list<item: list<item: list<item: double>>>
                        child 0, item: list<item: list<item: double>>
                            child 0, item: list<item: double>
                                child 0, item: double
                    child 1, max: list<item: list<item: list<item: double>>>
                        child 0, item: list<item: list<item: double>>
                            child 0, item: list<item: double>
                                child 0, item: double
                    child 2, mean: list<item: list<item: list<item: double>>>
                        child 0, item: list<item: list<item: double>>
                            child 0, item: list<item: double>
                                child 0, item: double
                    child 3, std: list<item: list<item: list<item: double>>>
                        child 0, item: list<item: list<item: double>>
                            child 0, item: list<item: double>
                                child 0, item: double
                    child 4, count: list<item: int64>
                        child 0, item: int64
                child 1, observation.images.side: struct<min: list<item: list<item: list<item: double>>>, max: list<item: list<item: list<item: double (... 129 chars omitted)
                    child 0, min: list<item: list<item: list<item: double>>>
                        child 0, item: list<item: list<item: double>>
                  
              ...
              ax: list<item: int64>, mean: list<item: double>, std: list<item: dou (... 31 chars omitted)
                    child 0, min: list<item: int64>
                        child 0, item: int64
                    child 1, max: list<item: int64>
                        child 0, item: int64
                    child 2, mean: list<item: double>
                        child 0, item: double
                    child 3, std: list<item: double>
                        child 0, item: double
                    child 4, count: list<item: int64>
                        child 0, item: int64
                child 7, index: struct<min: list<item: int64>, max: list<item: int64>, mean: list<item: double>, std: list<item: dou (... 31 chars omitted)
                    child 0, min: list<item: int64>
                        child 0, item: int64
                    child 1, max: list<item: int64>
                        child 0, item: int64
                    child 2, mean: list<item: double>
                        child 0, item: double
                    child 3, std: list<item: double>
                        child 0, item: double
                    child 4, count: list<item: int64>
                        child 0, item: int64
                child 8, task_index: struct<min: list<item: int64>, max: list<item: int64>, mean: list<item: double>, std: list<item: dou (... 31 chars omitted)
                    child 0, min: list<item: int64>
                        child 0, item: int64
                    child 1, max: list<item: int64>
                        child 0, item: int64
                    child 2, mean: list<item: double>
                        child 0, item: double
                    child 3, std: list<item: double>
                        child 0, item: double
                    child 4, count: list<item: int64>
                        child 0, item: int64
              tasks: list<item: string>
                child 0, item: string
              length: int64
              to
              {'episode_index': Value('int64'), 'tasks': List(Value('string')), 'length': Value('int64')}
              because column names don't match

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.

so101_recording_20260519_000250 — Cosmos-augmented (v2, multi-camera)

World-model-augmented LeRobot dataset for the SO-101 arm. The camera pixels are generated by NVIDIA Cosmos rolled forward from real seed frames; the action / observation.state labels are the seed's REAL teleoperation labels, carried through unchanged. In short: pixels are world-model-generated, labels are real.

Seed dataset: ofcourseistillloveyou/so101_recording_20260519_000250 — real SO-101 teleoperation
Method: Cosmos world-model roll-forward, conditioned per-camera on each real seed frame, with real label-carry
Cameras: observation.images.front, observation.images.side — each rolled out separately (Cosmos is a single-view generator) from its own real first frame, written back under its real seed key
Augmentations (prompt-steered domain randomization): lighting, distractors
Episodes: 1 · Frames: 97 · fps: 30
action dim: [6] · observation.state dim: [6]
Task: "Teleoperate the SO-101 arm."

How it was made

Each generated trajectory conditions Cosmos on the actual initial observation frame of a seed episode and rolls the scene forward. The seed episode's per-frame action and observation.state vectors are carried onto the generated frames aligned by index, so every frame keeps a learnable, real (frame -> action) supervision signal — never zeros. The two camera views share the same per-frame labels (the robot supervision is camera-independent). Prompt-steered augmentations (lighting, distractors) vary the generated pixels while the carried labels stay invariant — the point of domain randomization.

Label fidelity (this release)

The Cosmos rollout produced 97 frames while 93 seed label frames were loaded for conditioning, so the first 93 frames carry the seed's exact per-frame labels and the final 4 frames hold the last real seed label (standard "hold last real label when the rollout outruns the seed trajectory" behavior — still a real label, never zeros/garbage). Verified: both camera keys present, videos decode to real 704x1280 frames, and labels match the seed element-wise.

Provenance: pixels world-model-generated (Cosmos); labels real (from ofcourseistillloveyou/so101_recording_20260519_000250).

Downloads last month: 43

Total file size:

1.01 MB