mirror of
https://github.com/huggingface/lerobot.git
synced 2026-08-08 17:39:44 +00:00
CarolinePascal
5a8a86ec70
Documents every remaining public class/function across the 21 files in lerobot.datasets: LeRobotDataset/LeRobotDatasetMetadata/MultiLeRobotDataset/ StreamingLeRobotDataset, the EpisodeAwareSampler, the dataset_tools.py editing API (add/remove/modify features, split/merge/delete episodes, reencode, recompute stats), aggregate.py's dataset aggregation pipeline, video encode/decode utilities, depth quantization, image writing, and misc dataset I/O helpers. Reformats every function's existing Args: block to the machine-checked Args:/Returns:/Raises: standard (verified against each function's real signature), fixing several stale/misattributed descriptions found along the way (a "_description_" placeholder, an orphaned continuation misattributed to the wrong parameter, an "ep_idx" arg documented that no longer exists in the signature). Also replaces a bare `object()` sentinel default in VideoDecoderCache with a small class carrying a fixed `__repr__`, since the sentinel's memory-address repr made a stable "defaults to X" docstring clause structurally impossible — the change preserves identical `is`-based sentinel semantics. Expands docs/source/api/datasets.mdx with the dataset-editing and aggregation API surface. Adds lerobot.datasets to check_docstrings.py's MODULES_TO_CHECK ratchet and removes the module's ruff D-ignore.
58 lines
1.7 KiB
Plaintext
58 lines
1.7 KiB
Plaintext
# Datasets
|
|
|
|
[`LeRobotDataset`] is the format every LeRobot script reads and writes. It is episode-aware, decodes video
|
|
observations on the fly, and round-trips to the Hugging Face Hub.
|
|
|
|
See [Using LeRobotDataset](../lerobot-dataset-v3) for the format and the common operations,
|
|
[Porting Large Datasets](../porting_datasets_v3) for migration, and [Tools](../tools) for the CLI.
|
|
|
|
## LeRobotDataset
|
|
|
|
[[autodoc]] lerobot.datasets.LeRobotDataset
|
|
|
|
## LeRobotDatasetMetadata
|
|
|
|
[[autodoc]] lerobot.datasets.LeRobotDatasetMetadata
|
|
|
|
## MultiLeRobotDataset
|
|
|
|
[[autodoc]] lerobot.datasets.MultiLeRobotDataset
|
|
|
|
## StreamingLeRobotDataset
|
|
|
|
[[autodoc]] lerobot.datasets.StreamingLeRobotDataset
|
|
|
|
## EpisodeAwareSampler
|
|
|
|
[[autodoc]] lerobot.datasets.sampler.EpisodeAwareSampler
|
|
|
|
## Editing a dataset
|
|
|
|
Functions in `lerobot.datasets.dataset_tools` for editing an existing `LeRobotDataset` on disk: adding,
|
|
removing, or modifying features; splitting, merging, or deleting episodes; re-encoding video; and
|
|
recomputing statistics. Each returns a new dataset rather than mutating the source in place.
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.add_features
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.remove_feature
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.modify_features
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.modify_tasks
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.delete_episodes
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.split_dataset
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.merge_datasets
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.recompute_stats
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.reencode_dataset
|
|
|
|
[[autodoc]] lerobot.datasets.dataset_tools.convert_image_to_video_dataset
|
|
|
|
## Aggregating datasets
|
|
|
|
[[autodoc]] lerobot.datasets.aggregate.aggregate_datasets
|