The dataset viewer is not available for this subset.
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/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/webdataset/webdataset.py", line 80, in _split_generators
raise ValueError(
...<2 lines>...
)
ValueError: The TAR archives of the dataset should be in WebDataset format, but the files in the archive don't share the same prefix or the same types.
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.
FrameDiff H200, 1,408 residues per assembly
160,012 preprocessed PDB assemblies: 75,865 monomers and 84,147 multimers. 159,674 pass the training filters. The fixed 10% subset has 15,967 assemblies. Both sets include per-chain ellipsoid annotations. No separate symmetry-extension set is included. Biological assemblies may themselves have natural symmetry. All 136,180 old A100 dense examples are retained; 23,832 larger assemblies are added.
Download an exact commit revision using the FrameDiff cluster.dataset fetch
command. It verifies SHA-256 and extracts all shards into one directory. CSV
structure paths are relative to the metadata root. training_data*.yaml and
annotation source paths record the original build location; the cluster launcher
relocates runtime paths. Annotation matching uses structure content hashes.
The metadata archive includes filtering, source hashes, and build provenance.
Use release_manifest.json for archive sizes, checksums and counts. The full and
10p sets share structure shards. PDB source provenance and third-party rights
remain with the original depositors/providers; MIT applies to project code.
Do not load pickle files from untrusted or modified releases.
- Downloads last month
- 28