The dataset viewer is not available for this split.
Error code: StreamingRowsError
Exception: ArrowInvalid
Message: JSON parse error: Invalid value. in row 0
Traceback: Traceback (most recent call last):
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 324, in _generate_tables
df = pandas_read_json(f)
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 38, in pandas_read_json
return pd.read_json(path_or_buf, **kwargs)
~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/pandas/io/json/_json.py", line 791, in read_json
json_reader = JsonReader(
path_or_buf,
...<16 lines>...
engine=engine,
)
File "/usr/local/lib/python3.14/site-packages/pandas/io/json/_json.py", line 905, in __init__
self.data = self._preprocess_data(data)
~~~~~~~~~~~~~~~~~~~~~^^^^^^
File "/usr/local/lib/python3.14/site-packages/pandas/io/json/_json.py", line 917, in _preprocess_data
data = data.read()
File "/usr/local/lib/python3.14/site-packages/datasets/utils/file_utils.py", line 839, in read_with_retries
out = read(*args, **kwargs)
File "<frozen codecs>", line 325, in decode
UnicodeDecodeError: 'utf-8' codec can't decode byte 0x80 in position 128: invalid start byte
During handling of the above exception, another exception occurred:
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 2951, in __iter__
for key, example in ex_iterable:
^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2461, in __iter__
for key, pa_table in self._iter_arrow():
~~~~~~~~~~~~~~~~^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2486, 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 547, in _iter_arrow
for key, pa_table in iterator:
^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 430, 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 327, in _generate_tables
raise e
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 290, in _generate_tables
pa_table = paj.read_json(
io.BytesIO(batch), read_options=paj.ReadOptions(block_size=block_size)
)
File "pyarrow/_json.pyx", line 364, in pyarrow._json.read_json
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowInvalid: JSON parse error: Invalid value. in row 0Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
EgoRecover reproducibility data
Code and protocols: https://github.com/sxh-kk/EgoRecover Models: https://huggingface.co/sxhkk/EgoRecover
Public migration snapshot requested by the project owner. Contains 10,424 files (47.686 GiB logical size), restored at their original project-relative paths by the release utility:
- The eight locally used processed EE4D-Motion files (35.531 GiB), including train/val motion, DINOv2 features, evaluation GT, normalization, take metadata and splits.
- Project-created mismatch data, frozen train/dev/test manifests, 4,815 paired episodes with verified markers, original and expanded P caches, final merged predicted histories and startup caches.
- Final selected-checkpoint evaluation predictions and official E7 reconstruction/forecast/generation evidence. These include fitted motion GT; they are data artifacts, not model weights.
Source processed dataset: https://huggingface.co/datasets/chaitanya100100/uniegomotion and https://github.com/chaitanya100100/UniEgoMotion/blob/main/DATASET.md . Original EgoExo4D data and terms: https://docs.ego-exo4d-data.org/getting-started/ . Preserve source attribution and follow the applicable source terms; this mirror does not relicense the data or imply unrestricted reuse. SMPL-X model files are excluded and require separate authorized download.
All selected source files were hashed without changing frozen identities or file contents. release-index.json lists bytes/SHA256/roles/presets. Restore with python tools/restore_release.py --preset data; use g-evaluation for the 222 project-test episode subset and required models, or p-evaluation for P weights and fixed dev caches. GitHub docs/release/artifacts.json pins exact revisions. --verify-only checks local contents after transfer. Shared directories are restored as relative symlinks, avoiding duplicate copies of the same episode cache.
The benchmark was used during prior project analysis; the E7 initializer's upstream validation also used official val. Historical manifests and identity metadata retain their original paths. This release supports evaluating stored models and restarting new runs from recorded inputs; exact cross-machine continuation remains subject to path/config/RNG and environment checks.
Physical storage and verification
The original 35,067,467,786-byte processed ZIP is copied server-side from official revision 2686aa6a5042c6003c4de8f6fb17602b2b171cb2, with exact archive SHA256 verified. All eight local source members matched the fixed source directory sizes/CRCs; restored members additionally verify SHA256. Other binary data is grouped into 17 lossless ZIP bundles; file hashes were rechecked while packing. Dataset payload is about 40.983 GiB; restoration produces 47.686 GiB. Small JSON/NPZ metadata keeps its original path. A minimum g-smoke preset is available for one test episode and R0 seed62.
Large prepared archives and checkpoints are stored as ordered, content-addressed parts of at most 64 MiB. Use the GitHub restore utility to assemble exact original bytes and verify both part and whole-file SHA256. Original large .pt/.zip paths are logical restoration paths; see parts in release-index.json for physical Hub objects.
Publication complete. All selected transport objects are committed; exact sizes and SHA256 are verified against remote metadata. A clean public-Hub g-smoke restoration and CPU inference passed. GitHub docs/release/artifacts.json pins the final revision.
- Downloads last month
- 276