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.
ChemLab instrument integration and model evaluation
The latest model evaluation runs Kinex + GPT-6 Astra / high through RLE-Bench: four independent seed-0 realtime instrument tasks, one attempt each. Results, model sessions, usage, native verdicts and synchronized videos are under model-evaluations/gpt-6-astra-high-seed0-20261023/. These are authored integration tasks with no robot.
Scripted acceptance and asset coverage
The current release includes all 203 instrument catalogs, native Isaac import checks and individual screenshots, at baseline source f159fc2, with a separate final-source Ziath follow-up, plus four realtime Isaac + Kinex scripted tasks with eight original camera recordings and four fresh-process replays. That earlier scripted snapshot contains no model-policy evaluation.
Open the interactive Space. The current evidence is under realtime-v3/ and catalog-v3/. index.json pins source identity and exact coverage. Missing specialized Matterix scene helpers remain deferred; readout projections, unresolved channels and upstream-unbound controls have separate denominators.
All previous results remain byte-for-byte at their original paths. The previous index is retained under history/realtime-v2-index.json. Owner records contain evaluator goals/state and must not be supplied to evaluated agents. Simulator source/assets, credentials and host paths are not redistributed.
Attribution and terms
ChemLab assets and contracts: MBZUAI/ChemLab-Assets, revision a6c006f921c9bee62f5585dd7f769163a7f51573. Upstream code is MIT; assets, contracts, catalogs and previews are CC BY-NC 4.0. The rendered media depicts these assets and retains their attribution and noncommercial conditions. Copyright (2026): Zirui Song, Jiadong Zhang, Jiahao Zhang, Xiaosong Ma, Xiaojun Chang and Xiuying Chen. Embedded third-party data retains its upstream notices. See the pinned upstream LICENSE, LICENSE-ASSETS.md, NOTICE.md and LICENSES/.
The RoboEnv implementation is linked by commit; its source and packages are not redistributed in this dataset. The practice tasks and integration checks are RoboEnv additions.
- Downloads last month
- -