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/harbor/harbor.py", line 171, in _split_generators
raise DataFilesNotFoundError("No task.toml or instruction.md files found")
datasets.exceptions.DataFilesNotFoundError: No task.toml or instruction.md files found
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.
Open Science Cases
Public collection of research cases run with AIPOCH Open-Science. Cases document actual research workflows and the evidence retained in exported .science session packages.
The template contains a short README.md and a session.science file. Real case folders require the same two files and may also include an optional supplementary/ folder for additional supporting files. The repository-level manifest.json is maintained at the repository root and indexes the published packages with their paths, file sizes, and SHA-256 digests. Research goals, author or team credits, and sharing scope belong in each case's README.
Prepare and submit a case
- Choose a research task relevant to your work and run it in Open-Science using your usual research workflow. Approve or revise plans, inspect results, and request corrections as appropriate; these interactions form part of the case history.
- Export the actual session as
session.science. The package is the submitted record of the conversation, execution, and retained artifacts; its contents depend on the export selection. - Create
cases/<case-name>/README.mdfrom SUBMISSION_TEMPLATE.md. Briefly describe the research goal, identify the author or team and preferred attribution, link the package, and state the sharing scope. See cases/template/README.md for a worked example. - Submit the README and package together. This repository is public, so upload only material cleared for public distribution. Cases intended only for internal inspection should be handed to the AIPOCH team through an agreed private channel.
Submit README.md and session.science in the case folder, with an optional supplementary/ folder for additional supporting files linked from the case README. Maintainers update the root collection manifest when a package is added or replaced.
Before submission
- Confirm that the README states the research goal, author or team,
.sciencefilename, and sharing scope. - Confirm that the exported package opens in a compatible Open-Science version and contains the session and artifacts you intend to share. Briefly note any known omissions or access requirements in the README.
- Check the exported content against the agreed sharing scope, including third-party data and confidential material.
Upload a public case
Use a stable folder name with lowercase words joined by hyphens, for example nhanes-smoking-depressive-symptoms. Existing case IDs remain valid.
hf upload AIPOCH-AI/open-science-cases ./cases/<case-name> cases/<case-name> --repo-type dataset
You can also upload through the repository's Files tab, preserving the same folder paths. The destination is the dataset repository datasets/AIPOCH-AI/open-science-cases.
Inspect a case
Start with the case README, use manifest.json to identify and check the published package, then import the .science file in a compatible Open-Science version to inspect the retained history and artifacts. Any re-execution is a separate step and may require dependencies, external data, or access described in the case.
The manifest describes file identity; it does not establish scientific correctness or successful reproduction. This is a folder-based case collection rather than a standardized load_dataset() table or Dataset Viewer dataset.
Sharing and maintenance
The repository-level manifest is separate from the manifest inside each .science package. Its minimal schema and update convention are described in DIRECTORY_STRUCTURE.md.
The repository declares the Apache-2.0 license. Third-party data and other materials may have separate terms; record relevant sources and restrictions in the case README. Public availability does not expand those permissions. Permission to reuse a case description or its figures in a preprint should also be stated in the case README or confirmed with its authors.
- Downloads last month
- 385