admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-07-26 11:16:00 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(robomme): align docs and tests with nested pixels obs layout

Browse Source 
Addresses PR #3311 review feedback:
- Docs: correct observation keys to `pixels/image` / `pixels/wrist_image`
  (mapped to `observation.images.image` / `observation.images.wrist_image`)
  and drop the now-obsolete column-rename snippet.
- Tests: assert `result["pixels"]["image"]` instead of flat `pixels/image`,
  matching the nested layout required by `preprocess_observation()`.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
Pepijn
2026-04-20 10:10:23 +02:00
parent cace58536d
commit f086caee20
2 changed files with 11 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/source/robomme.mdx
+2 -8
View File
@@ -109,15 +109,9 @@ dataset = LeRobotDataset("lerobot/robomme")
### Feature key alignment (training)
The env wrapper exposes `front_rgb` and `wrist_rgb` as observation keys. The `features_map` in `RoboMMEEnv` maps these to `obs_images.front` and `obs_images.wrist` for the policy.
The env wrapper exposes `pixels/image` and `pixels/wrist_image` as observation keys. The `features_map` in `RoboMMEEnv` maps these to `observation.images.image` and `observation.images.wrist_image` for the policy. State is exposed as `agent_pos` and maps to `observation.state`.
The dataset uses `image` and `wrist_image` as column names. When fine-tuning from the dataset, you may need to rename columns to match your policy's expected input keys:
```python
dataset = LeRobotDataset("lerobot/robomme")
# rename dataset columns if needed:
# dataset.hf_dataset = dataset.hf_dataset.rename_column("image", "front_rgb")
```
The dataset's `image` and `wrist_image` columns already align with the policy input keys, so no renaming is needed when fine-tuning.
## Action Spaces
tests/test_robomme_env.py
+9 -5
View File
@@ -118,7 +118,11 @@ def test_robomme_features_action_dim_ee_pose():
def test_convert_obs_list_format():
"""_convert_obs takes the last element from list-format obs fields and
emits the lerobot-canonical keys (pixels/image, pixels/wrist_image, agent_pos)."""
emits a nested ``pixels`` dict (image, wrist_image) plus ``agent_pos``.
The nested layout is required so ``preprocess_observation()`` in
``envs/utils.py`` maps each camera to ``observation.images.<cam>``.
"""
_install_robomme_stub()
try:
from lerobot.envs.robomme import RoboMMEGymEnv
@@ -138,8 +142,8 @@ def test_convert_obs_list_format():
}
result = env._convert_obs(obs_raw)
np.testing.assert_array_equal(result["pixels/image"], front)
np.testing.assert_array_equal(result["pixels/wrist_image"], wrist)
np.testing.assert_array_equal(result["pixels"]["image"], front)
np.testing.assert_array_equal(result["pixels"]["wrist_image"], wrist)
assert result["agent_pos"].shape == (8,)
np.testing.assert_array_almost_equal(result["agent_pos"][:7], joints)
assert result["agent_pos"][7] == gripper[0]
@@ -163,8 +167,8 @@ def test_convert_obs_array_format():
"gripper_state_list": np.zeros(2, dtype=np.float32),
}
result = env._convert_obs(obs_raw)
assert result["pixels/image"].shape == (256, 256, 3)
assert result["pixels/wrist_image"].shape == (256, 256, 3)
assert result["pixels"]["image"].shape == (256, 256, 3)
assert result["pixels"]["wrist_image"].shape == (256, 256, 3)
assert result["agent_pos"].shape == (8,)
finally:
_uninstall_robomme_stub()