Dataset Viewer
Duplicate
The dataset viewer is not available for this split.
Cannot load the dataset split (in streaming mode) to extract the first rows.
Error code:   StreamingRowsError
Exception:    CastError
Message:      Couldn't cast
absolute_path_hits: list<item: null>
  child 0, item: null
artifact_count: int64
created_at: timestamp[s]
existing_registry_prefix_byte_stable: bool
failures: list<item: null>
  child 0, item: null
new_registry_rows: int64
new_registry_unique_paths: int64
pdb_files_in_staging: int64
status: string
token_hits: list<item: null>
  child 0, item: null
bundled_task_count: int64
registry_counts: struct<envs: int64, records: int64, tasks: int64>
  child 0, envs: int64
  child 1, records: int64
  child 2, tasks: int64
new_release_counts: struct<envs: int64, records: int64, result_rows: int64, tasks: int64, trajectory_rows: int64>
  child 0, envs: int64
  child 1, records: int64
  child 2, result_rows: int64
  child 3, tasks: int64
  child 4, trajectory_rows: int64
builds: struct<linux: struct<status: string, unity_build_job: string, wsl_headless_smoke: string>, windows:  (... 48 chars omitted)
  child 0, linux: struct<status: string, unity_build_job: string, wsl_headless_smoke: string>
      child 0, status: string
      child 1, unity_build_job: string
      child 2, wsl_headless_smoke: string
  child 1, windows: struct<status: string, unity_build_job: string>
      child 0, status: string
      child 1, unity_build_job: string
created_at_utc: string
compatibility_verdict_sha256: string
release_id: string
base_registry_preserved: bool
env_artifacts: list<item: struct<bundled_task_count: int64, linux_executable_mode: string, path: string, platform:  (... 80 chars omitted)
  child 0, item: struct<bundled_task_count: int64, linux_executable_mode: string, path: string, platform: string, req (... 68 chars omitted)
      child 0, bundled_task_count: int64
      child 1, linux_executable_mode: string
      child 2, path: string
      child 3, platform: string
      child 4, required_managed_dlls_present: bool
      child 5, sha256: string
      child 6, size_bytes: int64
passed: bool
schema: string
failure_count: int64
repaired_task_count: int64
manifest_sha256_before_report: string
to
{'base_registry_preserved': Value('bool'), 'builds': {'linux': {'status': Value('string'), 'unity_build_job': Value('string'), 'wsl_headless_smoke': Value('string')}, 'windows': {'status': Value('string'), 'unity_build_job': Value('string')}}, 'bundled_task_count': Value('int64'), 'compatibility_verdict_sha256': Value('string'), 'created_at_utc': Value('string'), 'env_artifacts': List({'bundled_task_count': Value('int64'), 'linux_executable_mode': Value('string'), 'path': Value('string'), 'platform': Value('string'), 'required_managed_dlls_present': Value('bool'), 'sha256': Value('string'), 'size_bytes': Value('int64')}), 'failure_count': Value('int64'), 'failures': List(Value('null')), 'manifest_sha256_before_report': Value('string'), 'new_release_counts': {'envs': Value('int64'), 'records': Value('int64'), 'result_rows': Value('int64'), 'tasks': Value('int64'), 'trajectory_rows': Value('int64')}, 'passed': Value('bool'), 'registry_counts': {'envs': Value('int64'), 'records': Value('int64'), 'tasks': Value('int64')}, 'release_id': Value('string'), 'repaired_task_count': Value('int64'), 'schema': Value('string')}
because column names don't match
Traceback:    Traceback (most recent call last):
                File "/src/services/worker/src/worker/utils.py", line 147, in get_rows_or_raise
                  return get_rows(
                      dataset=dataset,
                  ...<4 lines>...
                      column_names=column_names,
                  )
                File "/src/libs/libcommon/src/libcommon/utils.py", line 272, in decorator
                  return func(*args, **kwargs)
                File "/src/services/worker/src/worker/utils.py", line 127, in get_rows
                  rows_plus_one = list(itertools.islice(safe_iter(ds, dataset=dataset), rows_max_number + 1))
                File "/src/services/worker/src/worker/utils.py", line 483, in safe_iter
                  yield from ds.decode(False) if ds.features else ds
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2840, in __iter__
                  for key, example in ex_iterable:
                                      ^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2373, in __iter__
                  for key, pa_table in self._iter_arrow():
                                       ~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 2398, in _iter_arrow
                  for key, pa_table in self.ex_iterable._iter_arrow():
                                       ~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 536, in _iter_arrow
                  for key, pa_table in iterator:
                                       ^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/iterable_dataset.py", line 419, in _iter_arrow
                  for key, pa_table in self.generate_tables_fn(**gen_kwags):
                                       ~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 343, in _generate_tables
                  self._cast_table(pa_table, json_field_paths=json_field_paths),
                  ~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
                File "/usr/local/lib/python3.14/site-packages/datasets/packaged_modules/json/json.py", line 132, in _cast_table
                  pa_table = table_cast(pa_table, self.info.features.arrow_schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2378, in table_cast
                  return cast_table_to_schema(table, schema)
                File "/usr/local/lib/python3.14/site-packages/datasets/table.py", line 2306, in cast_table_to_schema
                  raise CastError(
                  ...<3 lines>...
                  )
              datasets.table.CastError: Couldn't cast
              absolute_path_hits: list<item: null>
                child 0, item: null
              artifact_count: int64
              created_at: timestamp[s]
              existing_registry_prefix_byte_stable: bool
              failures: list<item: null>
                child 0, item: null
              new_registry_rows: int64
              new_registry_unique_paths: int64
              pdb_files_in_staging: int64
              status: string
              token_hits: list<item: null>
                child 0, item: null
              bundled_task_count: int64
              registry_counts: struct<envs: int64, records: int64, tasks: int64>
                child 0, envs: int64
                child 1, records: int64
                child 2, tasks: int64
              new_release_counts: struct<envs: int64, records: int64, result_rows: int64, tasks: int64, trajectory_rows: int64>
                child 0, envs: int64
                child 1, records: int64
                child 2, result_rows: int64
                child 3, tasks: int64
                child 4, trajectory_rows: int64
              builds: struct<linux: struct<status: string, unity_build_job: string, wsl_headless_smoke: string>, windows:  (... 48 chars omitted)
                child 0, linux: struct<status: string, unity_build_job: string, wsl_headless_smoke: string>
                    child 0, status: string
                    child 1, unity_build_job: string
                    child 2, wsl_headless_smoke: string
                child 1, windows: struct<status: string, unity_build_job: string>
                    child 0, status: string
                    child 1, unity_build_job: string
              created_at_utc: string
              compatibility_verdict_sha256: string
              release_id: string
              base_registry_preserved: bool
              env_artifacts: list<item: struct<bundled_task_count: int64, linux_executable_mode: string, path: string, platform:  (... 80 chars omitted)
                child 0, item: struct<bundled_task_count: int64, linux_executable_mode: string, path: string, platform: string, req (... 68 chars omitted)
                    child 0, bundled_task_count: int64
                    child 1, linux_executable_mode: string
                    child 2, path: string
                    child 3, platform: string
                    child 4, required_managed_dlls_present: bool
                    child 5, sha256: string
                    child 6, size_bytes: int64
              passed: bool
              schema: string
              failure_count: int64
              repaired_task_count: int64
              manifest_sha256_before_report: string
              to
              {'base_registry_preserved': Value('bool'), 'builds': {'linux': {'status': Value('string'), 'unity_build_job': Value('string'), 'wsl_headless_smoke': Value('string')}, 'windows': {'status': Value('string'), 'unity_build_job': Value('string')}}, 'bundled_task_count': Value('int64'), 'compatibility_verdict_sha256': Value('string'), 'created_at_utc': Value('string'), 'env_artifacts': List({'bundled_task_count': Value('int64'), 'linux_executable_mode': Value('string'), 'path': Value('string'), 'platform': Value('string'), 'required_managed_dlls_present': Value('bool'), 'sha256': Value('string'), 'size_bytes': Value('int64')}), 'failure_count': Value('int64'), 'failures': List(Value('null')), 'manifest_sha256_before_report': Value('string'), 'new_release_counts': {'envs': Value('int64'), 'records': Value('int64'), 'result_rows': Value('int64'), 'tasks': Value('int64'), 'trajectory_rows': Value('int64')}, 'passed': Value('bool'), 'registry_counts': {'envs': Value('int64'), 'records': Value('int64'), 'tasks': Value('int64')}, 'release_id': Value('string'), 'repaired_task_count': Value('int64'), 'schema': Value('string')}
              because column names don't match

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.

AgentArk Assets

Versioned Unity runtimes, task Mods, and multimodal replay/evaluation records.

Environment compatibility / 环境兼容性

A task's labeled env version is its minimum supported environment version. That version and all later versions are compatible, with no upper version bound.

任务标注的 env 版本是最低环境版本;该版本及之后的所有版本均兼容,没有版本上限。

Task artifact label / 任务标注 Compatible environments / 可用环境
env-1.0.3 >=1.0.3: 1.0.3, 1.0.4, 1.0.5, and all later versions
env-1.0.4 >=1.0.4: 1.0.4 and all later versions; not 1.0.3

An env-1.0.3 directory is not an exact-version restriction. Reuse its task ZIP on later environments; no re-export is required solely because the environment version increased. Match the operating-system platform and honor any separately declared Python wrapper/capability requirements.

In registry/tasks.jsonl, min_env_version and env_version_spec express this rule. compatible_env_versions enumerates known published compatible versions; it does not exclude future versions and is not a test matrix. Actual validation and recorded runtime versions remain separately documented. 兼容声明不等于已在所有后续版本逐一实测;回放版本也不会因此改写。

Read COMPATIBILITY.md for field definitions. Use compatibility.py, which needs only Python's standard library, to select the latest compatible task build for each task/platform:

python compatibility.py --registry registry/tasks.jsonl --env-version 1.0.4 --platform windows
python compatibility.py --registry registry/tasks.jsonl --env-version 1.0.4 --platform linux --task-id 156 --task-id 165

Download the registry and script first. The selector prints JSONL metadata and does not install anything. It uses numeric version comparison and falls back to an older compatible task version when necessary. Archive paths/hashes are stable.

Available environments / 可用环境

  • Default release: env-1.0.3.
  • Latest published environment: env-1.0.4, available explicitly; the default has not been switched.
  • env-1.0.4 downloads: Windows starter, Linux starter.
  • env-1.0.4 fixes recoverable Func argument-binding errors. Task156 v2 and Task165 v3 additionally require a Python wrapper containing commit 157620c and the host capability log_error_v1; replacing a task Mod alone does not update the host.
  • Linux export/artifact checks do not imply Linux Player execution. Consult each release's validation evidence.

Release task lists are selection snapshots, not compatibility ceilings. For any chosen environment version, use the complete task registry and the minimum-version rule. Release record lists describe their recorded runs, not all compatible tasks.

Catalog

Consumers should read manifest.json and registry/*.jsonl instead of scanning artifact directories. At this metadata update:

  • Environment artifacts: 10
  • Task artifacts: 580
  • Record artifacts: 402

The manifest contains the authoritative current counts.

License

This dataset and the AgentArk runtime, task, and record artifacts in this repository are released under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0), unless otherwise noted.

The materials may be used, copied, modified, and redistributed for non-commercial academic research and educational purposes with appropriate attribution.

Commercial use, including use in commercial products, services, hosted platforms, paid evaluations, or proprietary training/evaluation pipelines, is not permitted without prior written authorization from the AgentArk authors.

Third-party components, including Unity runtime components and any bundled third-party assets or libraries, remain subject to their respective upstream licenses.

Downloads last month
490