Dataset Viewer
The dataset viewer is not available for this subset.
Cannot get the split names for the config 'default' of the dataset.
Exception: SplitsNotFoundError
Message: The split names could not be parsed from the dataset config.
Traceback: Traceback (most recent call last):
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 342, 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: Column(/aesthetic_quality/[]) changed from number to array in row 0
During handling of the above exception, another exception occurred:
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/json/json.py", line 101, in _split_generators
pa_table = next(iter(self._generate_tables(**splits[0].gen_kwargs, allow_full_read=False)))[1]
~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 304, in _generate_tables
batch = json_encode_fields_in_json_lines(original_batch, json_field_paths)
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 111, in json_encode_fields_in_json_lines
examples = [ujson_loads(line) for line in original_batch.splitlines()]
~~~~~~~~~~~^^^^^^
File "/usr/local/lib/python3.14/site-packages/datasets/utils/json.py", line 20, in ujson_loads
return pd.io.json.ujson_loads(*args, **kwargs)
~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^
ValueError: Expected object or value
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 68, 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.
Clean Forcing — videos and per-video measurements
Generated videos and measurements behind Clean Forcing: Drift-Resistant Autoregressive Video Diffusion with a Frozen Base.
- Project page and paper: https://clean-forcing.github.io/
- Code: https://github.com/wnqw/clean_forcing · Checkpoints: https://huggingface.co/illustro1/clean-forcing
Protocol
Un-seeded text-to-video, 50 s per video (832×480, 16 fps), on the 128 held-out MovieGen prompts in
prompts_finals128.txt. pNNN is the prompt index and the seed; every configuration uses the same
prompt–seed pairs. Rollouts use block-causal attention over 3-latent chunks, a rolling 21-latent KV
window and a 20-step solver.
Layout
<row>/videos/pNNN.mp4 50 s video
<row>/musiq_4hz/pNNN.npy per-frame MUSIQ sampled at 4 Hz (201 values, t = 0..50 s)
<row>/musiq_every_frame/pNNN.npy every-frame MUSIQ (rows where it was computed)
causal_forcing/<row>/... same layout for the Causal-Forcing base rows, plus scores.npz
wan14b/<row>/videos/pNNN.mp4 Wan2.1-T2V-14B, 32 prompts (every 4th index); wan14b/ood_scores.json
measurements/vbench/ official VBench custom-input results (long_horizon_50s, short_horizon_5s, wan14b)
measurements/per_config/ metric_suite.json (drift suite: ΔQualityDrift, color shift, survival,
temporal LPIPS, flicker), ci.json (MUSIQ / Δ-drift mean ± SEM),
*_posthoc.npz (anchoring, lag-2 s identity, cuts), *_fullrate.npz, *_dino_latesim.npy
measurements/drift_suite/ survival curves, temporal and anchored-consistency evaluations (in-domain subset)
Rows (Table 1 of the paper)
| Folder | Paper row | Internal tag (inside measurement files) |
|---|---|---|
adapted_base |
Adapted base | abase |
context_noise |
Context noise (σ = 0.2) | adfs |
history_guidance |
History guidance (w = 1.2) | ahg |
pathwise_ttc_500_250 |
Pathwise TTC ({500, 250}) | attc |
| — | Pathwise TTC ({750, 500}) — videos and per-prompt curves were not retained; VBench results only | attc750 |
clean_forcing_zero_real |
Clean Forcing (zero real videos) | av2s |
clean_forcing_real |
Clean Forcing (+40 real clips) | av2 |
clean_forcing_zero_real_one_step |
one-step corrector, zero real (ablation) | av1s |
self_forcing |
Self Forcing | sfd |
skyreels_v2_df |
SkyReels-V2-DF | skyr |
causal_forcing/causal_forcing_base |
Causal-Forcing base | cfb |
causal_forcing/causal_forcing_clean_forcing |
+ Clean Forcing (zero real videos) | cfc |
wan14b/wan14b_adapted_base, wan14b/wan14b_clean_forcing |
Wan2.1-14B appendix | abase14, av2_14 |
The Self Forcing and SkyReels-V2 videos were generated with the authors' publicly released models (Self-Forcing, SkyReels-V2) under our prompts and protocol; please also credit those works if you use them.
Citation
@article{wang2026cleanforcing,
title={Clean Forcing: Drift-Resistant Autoregressive Video Diffusion with a Frozen Base},
author={Wang, Wenqing and Shin, Joonghyuk and Tremblay, Jonathan and Song, Chan Hee and Fu, Yun},
journal={arXiv preprint},
year={2026}
}
- Downloads last month
- 345