feat(sim): add metaworld env #2088
Open
+595
−2
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
There was a problem hiding this comment.
Pull Request Overview
This PR adds support for MetaWorld environments to the lerobot package, enabling integration with the MetaWorld benchmark suite for robotic manipulation tasks.
- Implements a new
MetaworldEnvclass that wraps MetaWorld environments with configurable observation types and camera settings - Adds factory functions to create single or multiple MetaWorld environments with support for multitask evaluation
- Integrates MetaWorld configuration into the existing environment system
Reviewed Changes
Copilot reviewed 4 out of 4 changed files in this pull request and generated 5 comments.
| File | Description |
|---|---|
| src/lerobot/envs/metaworld.py | New MetaWorld environment wrapper with task management and observation formatting |
| src/lerobot/envs/factory.py | Integration of MetaWorld environment creation into the factory pattern |
| src/lerobot/envs/configs.py | Configuration class for MetaWorld environments with feature mapping |
| pyproject.toml | Dependency specification for the metaworld package |
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
Signed-off-by: Jade Choghari <[email protected]>
jadechoghari
added
enhancement
AdilZouitine
added a commit
that referenced
this pull request
Sep 29, 2025
… for normalization (#2088) Signed-off-by: AdilZouitine <[email protected]>
Co-authored-by: Copilot <[email protected]> Signed-off-by: Jade Choghari <[email protected]>
AdilZouitine
added a commit
that referenced
this pull request
Sep 29, 2025
… for normalization (#2088) (#2090) Signed-off-by: AdilZouitine <[email protected]>
AdilZouitine
reviewed
Oct 3, 2025
examples/train.sh
Outdated
| xarm = ["gym-xarm>=0.1.1"] | ||
| libero = ["lerobot[transformers-dep]", "libero @ git+https://github.com/huggingface/lerobot-libero.git@main#egg=libero"] | ||
|
|
||
| metaworld = ["metaworld>=3.0.0"] |
There was a problem hiding this comment.
put gymnasium upgrade here
src/lerobot/envs/metaworld.py
Outdated
| from huggingface_hub import hf_hub_download | ||
|
|
||
| # download the JSON | ||
| path = hf_hub_download(repo_id="jadechoghari/smolvla-metaworld-keys", filename="tasks_metaworld.json") |
There was a problem hiding this comment.
You can hard code this.
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What this does
feat(sim): add Metaworld environment
This PR add Metaworld, a standard robotic VLA simulation benchmark with 50+ manipulation tasks (e.g. pick-and-place, drawer, door, button press). It is designed to test policy generalization, multi-task learning, compositionality, and transfer across diverse manipulation skills.
Corresponding dataset: https://huggingface.co/datasets/lerobot/metaworld_mt50 ; demonstrations for all tasks, ready for training and evaluation.