The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: TypeError
Message: Couldn't cast array of type
struct<canvas_ref: string, joint_pos: string, pen_down: string, ts_ns: string>
to
{'joint_pos': Value('string'), 'pen_down': Value('string'), 'ts_ns': Value('string')}
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 483, 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 2840, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, 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 2378, in table_cast
return cast_table_to_schema(table, schema)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2312, in cast_table_to_schema
cast_array_to_feature(
~~~~~~~~~~~~~~~~~~~~~^
table[name] if name in table_column_names else pa.array([None] * len(table), type=schema.field(name).type),
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
feature,
^^^^^^^^
)
^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1861, in wrapper
return pa.chunked_array([func(chunk, *args, **kwargs) for chunk in array.chunks])
~~~~^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2068, in cast_array_to_feature
_c(array.field(name) if name in array_fields else null_array, subfeature)
~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1863, in wrapper
return func(array, *args, **kwargs)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2068, in cast_array_to_feature
_c(array.field(name) if name in array_fields else null_array, subfeature)
~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 1863, in wrapper
return func(array, *args, **kwargs)
File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2158, in cast_array_to_feature
raise TypeError(f"Couldn't cast array of type\n{_short_str(array.type)}\nto\n{_short_str(feature)}")
TypeError: Couldn't cast array of type
struct<canvas_ref: string, joint_pos: string, pen_down: string, ts_ns: string>
to
{'joint_pos': Value('string'), 'pen_down': Value('string'), 'ts_ns': Value('string')}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.
muretai network episodes (RLDS)
Provenance-verified RLDS-compatible episodes recorded on the muretai agent network. This is a corpus of signed records, not a leaderboard and not a claim that any recorded behaviour was correct.
The public shelf is this dataset. There is no catalog on muretai.com.
The screenshots are the local world-model lab (console.py) over the
same harvest.
What the files are
| path | what it is |
|---|---|
<dataset>/episodes/<id>.json |
canonical sealed episode (jsonl-canonical/0.1). Re-hash it. This is the artifact. |
<dataset>/sts/<id>.jsonl |
Hugging Face STS-Format rendering of the same episode. The first line is a session header whose extra fields include episode_hash. |
<dataset>/manifest.jsonl |
harvest index (verified is hash or chain) |
<dataset>/blobs/ |
content-addressed sha256:<hex> payloads (canvas frames, artifacts) |
verify_episode.py |
stdlib-only checker — no muretai install |
The STS file is generated from the sealed JSON at pack time. If they
disagree, trust the JSON. A test in this tooling repo asserts
round-trip agreement (episode_hash on the session header equals the
SHA-256 of the sealed bytes) for every packed episode.
Guarantee (read this before citing)
The protocol proves provenance: who asked, what bytes were sealed, who signed the completion (when envelopes are present). It does not prove the task was solved well, that a review verdict was true, or that a robot motion was safe.
Remote harvests are verified: hash — the file matches its published
digest and schema. verified: chain needs the robot's message log and
is a stronger label; this pack says which one it is, per row.
A pseudonymized public edition (salted requester DID) deliberately breaks chain step 3. This pack keeps the DIDs the recorder wrote. If a later edition salts them, the card must say so and must not show a green verification tick for step 3.
Schemas
rlds-muretai/0.2— embodied robot episodes (T62). Spec: https://github.com/muretai/muretai-core/blob/main/docs/RLDS_MURETAI.mdrlds-muretai/0.2-rooms— multi-party room episodes (T64). Empty in this pack until a cross-owner room records with opt-in on.rlds-muretai/0.2-web— open-door walks (T-WEBRLDS). Empty here for the same reason (recording default off).
This pack
- generated_at: 2026-08-29T07:17:22Z
- datasets: robot_painter_live: 27 episodes, 27 STS files, 7220 blobs
- harvest verification: hash=27
- requester DIDs (distinct): 15
- of which recorded as muretai-operated (
own-dids): 0 - remaining (upper bound on external requesters): 15
Self-generated or operator-owned runs are labelled, not hidden.
control_mode stays auto or teleop as recorded.
Verify
python3 verify_episode.py <dataset>/episodes/<id>.json
# optional pin:
python3 verify_episode.py <dataset>/episodes/<id>.json --expect sha256:<hex>
That is hash-level. To walk the signature chain you need the request and completion envelopes and the core verifier:
python3 shared/episode.py verify episode.json request.json completion.json
Consent
Episodes carry a machine-readable consent field (typically
training-ok/0.1) written at recording time. Contact
muretaicom@gmail.com for licensing of larger corpora.
Citation
Sealed bytes are content-addressed. Cite the episode_hash, not a row
number.
- Downloads last month
- 54