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 92, in _split_generators
inferred_arrow_schema = pa.concat_tables(pa_tables, promote_options="default").schema
~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "pyarrow/table.pxi", line 6268, in pyarrow.lib.concat_tables
File "pyarrow/error.pxi", line 155, in pyarrow.lib.pyarrow_internal_check_status
return check_status(status)
File "pyarrow/error.pxi", line 92, in pyarrow.lib.check_status
raise convert_status(status)
pyarrow.lib.ArrowTypeError: Unable to merge: Field json has incompatible types: struct<schemaVersion: int64, mediaType: string, manifests: list<item: struct<mediaType: string, digest: string, size: int64, annotations: struct<io.containerd.image.name: string, org.opencontainers.image.ref.name: string>>>> vs list<item: struct<Config: string, RepoTags: list<item: string>, Layers: list<item: string>>>
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.
ALE-Daily-VM
Prebuilt Docker images for the 23 self-built ALE-Daily apps (openbank, opencall, openticket, ...): the servers, their dependencies, build outputs, and the Android clients running in an emulator.
| File | Image | Compressed | Contents |
|---|---|---|---|
ale-daily-apps-linux-fr7.tar.zst.part-00 |
ale-daily-apps:fr7 |
3.4 GB (14 GB loaded) | Servers for all 23 apps, plus the ale-app launcher |
ale-daily-apps-android-fr7.tar.zst.part-00 |
ale-daily-apps-android:fr7 |
20 GB (74 GB loaded) | Everything in the Linux image, plus JDK, Android SDK 36, an x86_64 Android 16 emulator, and 19 APKs |
The Android image already contains everything in the Linux image, so you only need the Linux image if you don't need the Android clients.
Both images were built from ALE-Daily-Apps main at d2c4cdc8. The apps tree is b49fdd3e. The ALE_CODE_REF env var inside the image is e1d80854, which is the commit's hash before the author rewrite. The tooling (ale-app, ale-android) is baked in at /opt/ale/ale, so you don't need to clone the repo.
Download and load
pip install -U huggingface_hub
hf download Cloudriver/ALE-Daily-VM --repo-type dataset --local-dir ale-vm
cd ale-vm
sha256sum -c SHA256SUMS
cat ale-daily-apps-linux-fr7.tar.zst.part-* | zstd -d | docker load # ale-daily-apps:fr7
cat ale-daily-apps-android-fr7.tar.zst.part-* | zstd -d | docker load # ale-daily-apps-android:fr7
To download only one image, add --include "ale-daily-apps-linux-*" --include SHA256SUMS to the hf download command. If you do, sha256sum -c will report the other file as missing; that is expected. Loading the images needs about 74 GB plus 14 GB of free Docker storage.
Linux: run the app servers
docker run -d --name ale ale-daily-apps:fr7 sleep infinity
docker exec ale ale-app list # 23 apps and their ports
docker exec ale ale-app start openbank # start one app
docker exec ale ale-app probe all --log /out # cold start → mutate → reset → reset==cold → reproducible
docker exec ale ale-app test all --log /out # each app's bundled headless tests
The default command (docker run -d ale-daily-apps:fr7 with no arguments) starts all apps in the foreground.
Android: emulator plus clients
This needs an x86_64 Linux host with KVM (/dev/kvm). The emulator uses 4 cores and 4 GB of RAM.
docker run -d --name ale-android --device /dev/kvm ale-daily-apps-android:fr7
docker exec ale-android python3 /opt/ale/ale/android/ale-android check all --log /out
docker exec ale-android python3 /opt/ale/ale/android/ale-android connect --log /out
The check command boots the emulator, then for each app it starts the server, installs and launches the APK, and checks that the app did not crash, that there are no FATAL or ANR lines in logcat, that the client actually reached its server, and that the WebView console shows no JS errors. Inside the container, 10.0.2.2 is the container's own loopback.
The connect command clicks Connect on the clients that stop at a connection screen (opendrive, openinsure). In check, those two clients report reached_server as FAIL by design.
Reference results (this build)
| Check | Result |
|---|---|
ale-app test all |
23/23 rc=0 |
ale-app probe all (23 × 7) |
161 PASS / 0 FAIL |
ale-android check all (3 runs) |
93 PASS / 2 FAIL each; the 2 are reached_server for opendrive and openinsure, which is by design |
ale-android connect |
2/2 PASS |
The logs are in the repo under images/evidence/fr7-*.
What these images do not provide
- No isolation. All apps and any agent you put in the container run as the same user, so state files, tokens, and
/proc/<pid>/environare readable and writable. A score read from inside the container can be forged (see the repo's AGENTS.md §3.4 and §3.8). - No browser (Playwright Chromium) is installed.
- The emulator is not a real device.
- The probe token
ale-image-probe-token-…is a test-only dummy value, not a secret.
- Downloads last month
- 46