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/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.
Human-Last Digital Agent Taskbenchmark
v2 Executable Tasks (8) — use these
executable_v2.tar.gz contains 8 tasks that follow the DeepSeek-V4.1 §5.1
pipeline: each has a real Dockerfile, pinned immutable data, a reference
solution that has been run end-to-end, and a real verifier (not a
file-exists check).
| Task | Domain | Reference MAE / result | Verifier checks |
|---|---|---|---|
| realtime-gdp-nowcast | macro-finance | AR(4) expanding window, MAE 0.56 | 16 (incl. anti-look-ahead static sniff) |
| heart-logrank | clinical survival | KM + log-rank chi2=3.14, p=0.076 | 17 |
| china-pooled-or | epidemiology | MH OR=2.17, I2=0% | 17 |
| elnino-trend-changepoint | climate | 0.135 C/decade, Pettitt 1976 | 15 |
| nile-ar-forecast | hydrology | AR(1) MAE 101 | 18 |
| tiny-statsutils | SWE bugfix | 3 fail-to-pass points | 6 (anti-hardcode) |
| randhie-demand | health econ | lncoins coef -0.169, t=-10.9 | 16 |
| statecrime-ols | public policy | R2=0.68 | 16 |
Each task contains:
instruction.md(agent-facing, no hints)environment/Dockerfile(python:3.11-slim, pinned deps)environment/init.sh,network_policy.json(egress locked to PyPI)environment/data/(immutable pinned dataset)solution/(reference solution — stripped by cleanup.sh before shipping)tests/verify.py(real checks: compile, run, determinism, schema, numeric bands, anti-look-ahead)cleanup.sh(strip solution + git history before RL rollout)
Legacy task spec cards (537)
human-last-digital-agent-taskbenchmark.tar.gz contains 537 task spec cards.
These are not yet executable environments — they are problem statements
only. They are being upgraded to the v2 pipeline incrementally.
- Downloads last month
- 35