The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: CastError
Message: Couldn't cast
generated_at: timestamp[s]
gate: struct<min_face_valid: double, min_valid_s: double, valid_minutes_definition: string, prescreen: str (... 87 chars omitted)
child 0, min_face_valid: double
child 1, min_valid_s: double
child 2, valid_minutes_definition: string
child 3, prescreen: string
child 4, channel_cap: string
child 5, speaker_cap: double
child 6, speaker_cap_note: string
child 7, audio: string
totals: struct<kept: int64, kept_hours: double, valid_hours: double, raw_valid_min: double, capped_valid_min (... 659 chars omitted)
child 0, kept: int64
child 1, kept_hours: double
child 2, valid_hours: double
child 3, raw_valid_min: double
child 4, capped_valid_min: double
child 5, pushed: int64
child 6, items_by_state: struct<captured: struct<n: int64, hours: double>, capturing: struct<n: int64, hours: double>, droppe (... 76 chars omitted)
child 0, captured: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 1, capturing: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 2, dropped: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 3, queued: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 7, speakers: int64
child 8, hours_by_language: struct<en: double, en-US: double>
child 0, en: double
child 1, en-US: double
ch
...
(... 107 chars omitted)
child 0, api: string
child 1, basis: string
child 2, agency: string
child 3, channel_url: string
child 4, channel_id: string
child 5, statement: string
child 6, license: string
child 7, channel: string
child 8, uploader_id: string
child 9, upload_date: string
child 10, webpage_url: string
n_src_frames: int64
source_file_url: string
sign_convention: string
harvest: struct<item_id: string, source_kind: string, channel_key: string, channel_name: string, speaker: str (... 277 chars omitted)
child 0, item_id: string
child 1, source_kind: string
child 2, channel_key: string
child 3, channel_name: string
child 4, speaker: string
child 5, series: string
child 6, language: string
child 7, lang_evidence: string
child 8, chunk: int64
child 9, chunk_start_s: double
child 10, chunk_len_s: double
child 11, prescreen: struct<ok: bool, n: int64, frac_one_face: double, frac_multi: double, frac_none: double, median_face (... 11 chars omitted)
child 0, ok: bool
child 1, n: int64
child 2, frac_one_face: double
child 3, frac_multi: double
child 4, frac_none: double
child 5, median_face_w: double
child 12, category: string
child 13, worker: string
arm: string
subject: string
source_url: string
audio: struct<file: string, codec: string, sr: int64, restore: string>
child 0, file: string
child 1, codec: string
child 2, sr: int64
child 3, restore: string
schema: string
src_fps: double
to
{'source': Value('string'), 'source_path': Value('string'), 'source_url': Value('string'), 'source_file_url': Value('string'), 'title': Value('string'), 'license': Value('string'), 'license_evidence': {'api': Value('string'), 'basis': Value('string'), 'agency': Value('string'), 'channel_url': Value('string'), 'channel_id': Value('string'), 'statement': Value('string'), 'license': Value('string'), 'channel': Value('string'), 'uploader_id': Value('string'), 'upload_date': Value('string'), 'webpage_url': Value('string')}, 'artist': Value('string'), 'rate_hz': Value('float64'), 'subject': Value('string'), 'arm': Value('string'), 'neutral': {'head_angles_deg': List(Value('float64')), 'head_trans_mm': List(Value('float64')), 'n_frames': Value('int64'), 'seconds': Value('float64'), 'torso_deg': List(Value('float64')), 'face_raw': {'gaze_yaw': Value('float64'), 'gaze_pitch': Value('float64'), 'brow_l_signed': Value('float64'), 'brow_r_signed': Value('float64'), 'eye_open_l': Value('float64'), 'eye_open_r': Value('float64'), 'mouth_open': Value('float64'), 'smile': Value('float64')}}, 'tool_versions': {'animacy': Value('string'), 'python': Value('string'), 'mediapipe': Value('string'), 'cv2': Value('string'), 'numpy': Value('string'), 'scipy': Value('string'), 'torch': Value('string')}, 'models': {'face': Value('string'), 'pose': Value('string'), 'models_dir': Value('string')}, 'vad': Value('string'), 'audio_backend': Value('string'), 'src_fps': Value('float64'), 'src_size': List(Valu
...
s_visible_frac': Value('float64'), 'face_crop_frac': Value('float64'), 'head_yaw_std': Value('float64'), 'head_yaw_p05_p95': List(Value('float64')), 'head_pitch_std': Value('float64'), 'head_pitch_p05_p95': List(Value('float64')), 'head_roll_std': Value('float64'), 'head_roll_p05_p95': List(Value('float64')), 'head_x_std': Value('float64'), 'head_x_p05_p95': List(Value('float64')), 'head_y_std': Value('float64'), 'head_y_p05_p95': List(Value('float64')), 'head_z_std': Value('float64'), 'head_z_p05_p95': List(Value('float64')), 'gaze_yaw_std': Value('float64'), 'gaze_yaw_p05_p95': List(Value('float64')), 'mouth_open_std': Value('float64'), 'mouth_open_p05_p95': List(Value('float64')), 'speaking_frac': Value('float64')}, 'captured_at': Value('timestamp[s]'), 'schema': Value('string'), 'harvest': {'item_id': Value('string'), 'source_kind': Value('string'), 'channel_key': Value('string'), 'channel_name': Value('string'), 'speaker': Value('string'), 'series': Value('string'), 'language': Value('string'), 'lang_evidence': Value('string'), 'chunk': Value('int64'), 'chunk_start_s': Value('float64'), 'chunk_len_s': Value('float64'), 'prescreen': {'ok': Value('bool'), 'n': Value('int64'), 'frac_one_face': Value('float64'), 'frac_multi': Value('float64'), 'frac_none': Value('float64'), 'median_face_w': Value('float64')}, 'category': Value('string'), 'worker': Value('string')}, 'audio': {'file': Value('string'), 'codec': Value('string'), 'sr': Value('int64'), 'restore': Value('string')}}
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 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 2306, in cast_table_to_schema
raise CastError(
...<3 lines>...
)
datasets.table.CastError: Couldn't cast
generated_at: timestamp[s]
gate: struct<min_face_valid: double, min_valid_s: double, valid_minutes_definition: string, prescreen: str (... 87 chars omitted)
child 0, min_face_valid: double
child 1, min_valid_s: double
child 2, valid_minutes_definition: string
child 3, prescreen: string
child 4, channel_cap: string
child 5, speaker_cap: double
child 6, speaker_cap_note: string
child 7, audio: string
totals: struct<kept: int64, kept_hours: double, valid_hours: double, raw_valid_min: double, capped_valid_min (... 659 chars omitted)
child 0, kept: int64
child 1, kept_hours: double
child 2, valid_hours: double
child 3, raw_valid_min: double
child 4, capped_valid_min: double
child 5, pushed: int64
child 6, items_by_state: struct<captured: struct<n: int64, hours: double>, capturing: struct<n: int64, hours: double>, droppe (... 76 chars omitted)
child 0, captured: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 1, capturing: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 2, dropped: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 3, queued: struct<n: int64, hours: double>
child 0, n: int64
child 1, hours: double
child 7, speakers: int64
child 8, hours_by_language: struct<en: double, en-US: double>
child 0, en: double
child 1, en-US: double
ch
...
(... 107 chars omitted)
child 0, api: string
child 1, basis: string
child 2, agency: string
child 3, channel_url: string
child 4, channel_id: string
child 5, statement: string
child 6, license: string
child 7, channel: string
child 8, uploader_id: string
child 9, upload_date: string
child 10, webpage_url: string
n_src_frames: int64
source_file_url: string
sign_convention: string
harvest: struct<item_id: string, source_kind: string, channel_key: string, channel_name: string, speaker: str (... 277 chars omitted)
child 0, item_id: string
child 1, source_kind: string
child 2, channel_key: string
child 3, channel_name: string
child 4, speaker: string
child 5, series: string
child 6, language: string
child 7, lang_evidence: string
child 8, chunk: int64
child 9, chunk_start_s: double
child 10, chunk_len_s: double
child 11, prescreen: struct<ok: bool, n: int64, frac_one_face: double, frac_multi: double, frac_none: double, median_face (... 11 chars omitted)
child 0, ok: bool
child 1, n: int64
child 2, frac_one_face: double
child 3, frac_multi: double
child 4, frac_none: double
child 5, median_face_w: double
child 12, category: string
child 13, worker: string
arm: string
subject: string
source_url: string
audio: struct<file: string, codec: string, sr: int64, restore: string>
child 0, file: string
child 1, codec: string
child 2, sr: int64
child 3, restore: string
schema: string
src_fps: double
to
{'source': Value('string'), 'source_path': Value('string'), 'source_url': Value('string'), 'source_file_url': Value('string'), 'title': Value('string'), 'license': Value('string'), 'license_evidence': {'api': Value('string'), 'basis': Value('string'), 'agency': Value('string'), 'channel_url': Value('string'), 'channel_id': Value('string'), 'statement': Value('string'), 'license': Value('string'), 'channel': Value('string'), 'uploader_id': Value('string'), 'upload_date': Value('string'), 'webpage_url': Value('string')}, 'artist': Value('string'), 'rate_hz': Value('float64'), 'subject': Value('string'), 'arm': Value('string'), 'neutral': {'head_angles_deg': List(Value('float64')), 'head_trans_mm': List(Value('float64')), 'n_frames': Value('int64'), 'seconds': Value('float64'), 'torso_deg': List(Value('float64')), 'face_raw': {'gaze_yaw': Value('float64'), 'gaze_pitch': Value('float64'), 'brow_l_signed': Value('float64'), 'brow_r_signed': Value('float64'), 'eye_open_l': Value('float64'), 'eye_open_r': Value('float64'), 'mouth_open': Value('float64'), 'smile': Value('float64')}}, 'tool_versions': {'animacy': Value('string'), 'python': Value('string'), 'mediapipe': Value('string'), 'cv2': Value('string'), 'numpy': Value('string'), 'scipy': Value('string'), 'torch': Value('string')}, 'models': {'face': Value('string'), 'pose': Value('string'), 'models_dir': Value('string')}, 'vad': Value('string'), 'audio_backend': Value('string'), 'src_fps': Value('float64'), 'src_size': List(Valu
...
s_visible_frac': Value('float64'), 'face_crop_frac': Value('float64'), 'head_yaw_std': Value('float64'), 'head_yaw_p05_p95': List(Value('float64')), 'head_pitch_std': Value('float64'), 'head_pitch_p05_p95': List(Value('float64')), 'head_roll_std': Value('float64'), 'head_roll_p05_p95': List(Value('float64')), 'head_x_std': Value('float64'), 'head_x_p05_p95': List(Value('float64')), 'head_y_std': Value('float64'), 'head_y_p05_p95': List(Value('float64')), 'head_z_std': Value('float64'), 'head_z_p05_p95': List(Value('float64')), 'gaze_yaw_std': Value('float64'), 'gaze_yaw_p05_p95': List(Value('float64')), 'mouth_open_std': Value('float64'), 'mouth_open_p05_p95': List(Value('float64')), 'speaking_frac': Value('float64')}, 'captured_at': Value('timestamp[s]'), 'schema': Value('string'), 'harvest': {'item_id': Value('string'), 'source_kind': Value('string'), 'channel_key': Value('string'), 'channel_name': Value('string'), 'speaker': Value('string'), 'series': Value('string'), 'language': Value('string'), 'lang_evidence': Value('string'), 'chunk': Value('int64'), 'chunk_start_s': Value('float64'), 'chunk_len_s': Value('float64'), 'prescreen': {'ok': Value('bool'), 'n': Value('int64'), 'frac_one_face': Value('float64'), 'frac_multi': Value('float64'), 'frac_none': Value('float64'), 'median_face_w': Value('float64')}, 'category': Value('string'), 'worker': Value('string')}, 'audio': {'file': Value('string'), 'codec': Value('string'), 'sr': Value('int64'), 'restore': Value('string')}}
because column names don't matchNeed help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
squaredcuber/animacy-human-motion-large
Canonical human conversational motion (animacy.human.v1, 30 Hz) captured at scale with
animacy — the open interaction layer for expressive robots.
Every clip is a directory clips/<shard>/<name>/ holding
motion.parquet— 28 channels: head 6-DoF, gaze, brows, eyes, mouth, torso, a puppet arm chain,speaking/ validity flags (schema:docs/CANONICAL.mdin the repo);audio.opus— the clip's audio, mono, Opus 32 kb/s from the 16 kHz capture track, on the same clock asmotion.t(ffmpeg -i audio.opus -ar 16000 -ac 1 audio.wavrestores the 16 kHz wav thatanimacy.schema.HumanClip.loadreads);meta.json— source url, license and the machine-readable license evidence, capture settings, neutral pose, validity stats, and aharvestblock (source family, channel, speaker key, language guess, chunk offset, prescreen result).
index.json is one row per clip (all shards). manifests/<shard>.json is the same for one commit.
26 clips, 2.9 hours (2.7 face-valid hours), 2 speaker keys, 2 languages.
License policy (enforced in code, not remembered per file)
Only material whose machine-readable license metadata says public domain / CC0 / CC-BY is
captured; anything ND, NC, SA, or with missing license metadata is refused
(scripts/fetch_sources.py::classify_license, applied by scripts/harvest/fetch.py). Evidence per source:
| source | evidence recorded in meta.json.license_evidence |
|---|---|
| YouTube (CC search) | info-json license == "Creative Commons Attribution license (reuse allowed)" (CC-BY 3.0), channel id, upload date |
| U.S. Government channels | allowlisted official agency channel + channel id match; basis 17 U.S.C. § 105 (works of the U.S. Government). VOA clips carry a caveat about third-party newswire material |
| Wikimedia Commons | videoinfo.extmetadata LicenseShortName / License / LicenseUrl / Copyrighted=False |
| archive.org | item metadata licenseurl / rights |
Attribution: every clip's meta.json has source_url, artist/channel, and the license label;
CC-BY clips must be credited to that channel/author when redistributed. This dataset is a derived
work (motion parameters + downsampled audio), not the source video.
Quality gate
Chunks of <= 600 s; prescreen (>= 60% of 16 sampled frames show exactly one face);
kept if face_valid >= 60% and face_valid * duration >= 60 s. No channel contributes
more than 5% of the kept seconds (fetch-time cap). Speaker keys over-merge (a channel = one
speaker), so a per-speaker cap at training time is conservative.
Counts
| source | hours |
|---|---|
| usgov | 2.9 |
| license | hours |
|---|---|
| Public Domain | 2.9 |
| language (guess) | hours |
|---|---|
| en | 2.5 |
| en-US | 0.5 |
| series | hours |
|---|---|
| briefing | 1.9 |
| usgov_other | 0.7 |
| address | 0.3 |
Provenance and reproducibility
Harvested by scripts/harvest/ (crawl -> fetch -> workers -> push) on a CPU-only workstation; the
harvest block in each meta.json names the worker and chunk. Full pipeline description:
docs/HARVEST.md. Nothing here was captured from a source whose license metadata was missing.
- Downloads last month
- 26