feat(dataset-viz): render generic custom scalar columns

Salvage of huggingface/lerobot#3918 by @nathon-lee — rebased to main.

lerobot-dataset-viz only logged known scalars; custom float/bool fields
now appear in the Rerun blueprint and frame stream when shape is scalar.

refactor(viz): add new non-default colums viz

Co-authored-by: nathon-lee <leejianwoo@gmail.com>

.github/dependabot.yml

@@ -0,0 +1,11 @@
+version: 2
+updates:
+  - package-ecosystem: "github-actions"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+    cooldown:
+      default-days: 7
+    groups:
+      actions:
+        patterns: ["*"]

AGENTS.md

@@ -51,6 +51,7 @@ pre-commit run --all-files                           # Lint + format (ruff, typo
 ## Notes
 
 - **Mypy is gradual**: strict only for `lerobot.envs`, `lerobot.configs`, `lerobot.optim`, `lerobot.model`, `lerobot.cameras`, `lerobot.motors`, `lerobot.transport`. Add type annotations when modifying these modules.
-- **Optional dependencies**: many policies, envs, and robots are behind extras (e.g., `lerobot[aloha]`). New imports for optional packages must be guarded or lazy. See `pyproject.toml [project.optional-dependencies]`.
+- **Imports**: prefer top-level imports; relative (`from .sibling import X`) across sibling files within a module, absolute (`from lerobot.module import X`) across modules.
+- **Optional dependencies**: many policies, envs, and robots are behind extras (e.g., `lerobot[aloha]`, see `pyproject.toml`). Guard optional imports with `TYPE_CHECKING or _foo_available` at module top + a `require_package(...)` check at use time. Reuse the `_foo_available` flags in `utils/import_utils.py`; don't call `is_package_available`.
 - **Video decoding**: datasets can store observations as video files. `LeRobotDataset` handles frame extraction, but tests need ffmpeg installed.
 - **Prioritize use of `uv run`** to execute Python commands (not raw `python` or `pip`).

AGENT_GUIDE.md

@@ -61,16 +61,20 @@ Full details in [`docs/source/so101.mdx`](./docs/source/so101.mdx) and [`docs/so
 **4.1 Install**
 
 ```bash
-pip install 'lerobot[feetech]'              # SO-100/SO-101 motor stack
-# pip install 'lerobot[all]'                # everything
-# pip install 'lerobot[aloha,pusht]'        # specific features
-# pip install 'lerobot[smolvla]'            # add SmolVLA deps
+# uv (recommended — see AGENTS.md and CLAUDE.md)
+uv sync --locked --extra feetech          # SO-100/SO-101 motor stack
+# uv sync --locked --extra all            # everything
+# uv sync --locked --extra smolvla        # add SmolVLA deps
+
+# pip (alternative, e.g. when not working from source)
+# pip install 'lerobot[feetech]'
+# pip install 'lerobot[all]'
+# pip install 'lerobot[smolvla]'
+
 git lfs install && git lfs pull
 hf auth login                             # required to push datasets/policies
 ```
 
-Contributors can alternatively use `uv sync --locked --extra feetech` (see `AGENTS.md`).
-
 **4.2 Find USB ports** — run once per arm, unplug when prompted.
 
 ```bash

README.md

@@ -83,7 +83,7 @@ episode_index=0
 print(f"{dataset[episode_index]['action'].shape=}\n")
 ```
 
-Learn more about it in the [LeRobotDataset Documentation](https://huggingface.co/docs/lerobot/lerobot-dataset-v3)
+Learn more about it in the [LeRobotDataset Documentation](https://huggingface.co/docs/lerobot/lerobot-dataset-v3).
 
 ## SoTA Models
 
@@ -109,7 +109,7 @@ lerobot-train \
 | **World Models**           | [VLA-JEPA](./docs/source/vla_jepa.mdx), [LingBot-VA](./docs/source/lingbot_va.mdx), [FastWAM](./docs/source/fastwam.mdx)                                                                                                                                                                                                                                                                   |
 | **Reward Models**          | [SARM](./docs/source/sarm.mdx), [TOPReward](./docs/source/topreward.mdx), [Robometer](./docs/source/robometer.mdx)                                                                                                                                                                                                                                                                         |
 
-Similarly to the hardware, you can easily implement your own policy & leverage LeRobot's data collection, training, and visualization tools, and share your model to the HF Hub
+Similarly to the hardware, you can easily implement your own policy & leverage LeRobot's data collection, training, and visualization tools, and share your model to the HF Hub.
 
 For detailed policy setup guides, see the [Policy Documentation](https://huggingface.co/docs/lerobot/bring_your_own_policies). For GPU/RAM requirements and expected training time per policy, see the [Compute Hardware Guide](https://huggingface.co/docs/lerobot/hardware_guide).
 
@@ -126,7 +126,7 @@ lerobot-eval \
   --eval.n_episodes=10
 ```
 
-Learn how to implement your own simulation environment or benchmark and distribute it from the HF Hub by following the [EnvHub Documentation](https://huggingface.co/docs/lerobot/envhub)
+Learn how to implement your own simulation environment or benchmark and distribute it from the HF Hub by following the [EnvHub Documentation](https://huggingface.co/docs/lerobot/envhub).
 
 ## Resources
 

docker/Dockerfile.internal

@@ -68,17 +68,16 @@ ENV HOME=/home/user_lerobot \
 # issues with MuJoCo and OpenGL drivers.
 RUN uv venv --python python${PYTHON_VERSION}
 
-# Install Python dependencies for caching
+# Install third-party dependencies separately for layer caching
 COPY --chown=user_lerobot:user_lerobot setup.py pyproject.toml uv.lock README.md MANIFEST.in ./
-COPY --chown=user_lerobot:user_lerobot src/ src/
-
-RUN uv sync --locked --extra all --no-cache
+RUN uv sync --locked --extra all --no-install-project --no-cache
 
 RUN chmod +x /lerobot/.venv/lib/python${PYTHON_VERSION}/site-packages/triton/backends/nvidia/bin/ptxas
 
-# Copy the rest of the application source code
+# Copy the application source code and install the local project
 # Make sure to have the git-LFS files for testing
 COPY --chown=user_lerobot:user_lerobot . .
+RUN uv sync --locked --extra all --no-cache
 
 # Set the default command
 CMD ["/bin/bash"]

docker/Dockerfile.user

@@ -60,15 +60,14 @@ ENV HOME=/home/user_lerobot \
 # run other Python projects in the same container without dependency conflicts.
 RUN uv venv
 
-# Install Python dependencies for caching
+# Install third-party dependencies separately for layer caching
 COPY --chown=user_lerobot:user_lerobot setup.py pyproject.toml uv.lock README.md MANIFEST.in ./
-COPY --chown=user_lerobot:user_lerobot src/ src/
+RUN uv sync --locked --extra all --no-install-project --no-cache
 
-RUN uv sync --locked --extra all --no-cache
-
-# Copy the rest of the application code
+# Copy the application code and install the local project
 # Make sure to have the git-LFS files for testing
 COPY --chown=user_lerobot:user_lerobot . .
+RUN uv sync --locked --extra all --no-cache
 
 # Set the default command
 CMD ["/bin/bash"]

docs/source/annotation_pipeline.mdx

@@ -89,8 +89,8 @@ subtask.
 
 The resulting spans are then stitched into a gap-free, full-episode
 cover, so **every frame has exactly one active subtask**. See
-[`run_hf_job.py`](https://github.com/huggingface/lerobot/blob/main/examples/annotations/run_hf_job.py)
-for the production settings (single camera, timestamped contact sheets,
+[Running on Hugging Face Jobs](#running-on-hugging-face-jobs) for the
+production settings (single camera, timestamped contact sheets,
 auto-windowed subtask generation).
 
 ### Tools
@@ -110,28 +110,67 @@ not-yet-implemented.
 
 ## Running on Hugging Face Jobs
 
-Annotation runs on [Hugging Face Jobs](https://huggingface.co/docs/hub/en/jobs).
-The repo ships a launcher script you copy and tweak for your dataset:
+Annotating a real dataset needs a GPU big enough to serve the VLM, so
+`lerobot-annotate` can dispatch itself to
+[Hugging Face Jobs](https://huggingface.co/docs/hub/en/jobs) — same as
+`lerobot-train`. Add `--job.target=<flavor>` to the exact command you'd
+run locally and it runs on that hardware instead:
 
 ```bash
-HF_TOKEN=hf_... uv run python examples/annotations/run_hf_job.py
+hf auth login   # once
+
+uv run lerobot-annotate \
+    --repo_id=user/my_dataset \
+    --new_repo_id=user/my_dataset_annotated \
+    --push_to_hub=true \
+    --vlm.model_id=Qwen/Qwen3.6-27B \
+    --vlm.num_gpus=1 \
+    --vlm.serve_command="vllm serve Qwen/Qwen3.6-27B --tensor-parallel-size 1 \
+        --max-model-len 32768 --gpu-memory-utilization 0.8 \
+        --uvicorn-log-level warning --port {port}" \
+    --vlm.serve_ready_timeout_s=1800 \
+    --vlm.chat_template_kwargs='{"enable_thinking": false}' \
+    --job.target=h200
 ```
 
-[`run_hf_job.py`](https://github.com/huggingface/lerobot/blob/main/examples/annotations/run_hf_job.py)
-starts a single-GPU `h200` job (bump it to `h200x4` for big datasets)
-that:
+That submits a single-GPU `h200` job that:
 
-1. installs `lerobot` (from `main`) plus the annotation extras,
-2. boots one vLLM server per GPU (using the `vllm/vllm-openai` image) and
-   drives it over the OpenAI-compatible API,
-3. runs the `plan` / `interjections` / `vqa` modules across the dataset
-   with `lerobot-annotate`,
+1. starts from the `vllm/vllm-openai` image and installs `lerobot` on top,
+2. boots one vLLM server per GPU and drives it over the OpenAI-compatible API,
+3. runs the `plan` / `interjections` / `vqa` modules across the dataset,
 4. with `--push_to_hub=true`, uploads the result to `--new_repo_id` (or
    back to `--repo_id` in place if you leave that unset).
 
-To use a different dataset, model, or hub repo, edit the `CMD` block in
-the script. Every flag there maps directly to a `lerobot-annotate` flag
-(run `lerobot-annotate --help` for the full list).
+The command streams the job's logs; `Ctrl-C` detaches without cancelling
+it. List the available flavors and their pricing with `hf jobs hardware`.
+
+<Tip warning={true}>
+
+Qwen3.6 ships with thinking enabled, which eats the token budget the
+annotator needs for its JSON answer — `--vlm.chat_template_kwargs='{"enable_thinking": false}'`
+turns it off. Without `--push_to_hub=true` the annotated dataset is
+discarded when the pod exits.
+
+</Tip>
+
+### Job options
+
+| Flag                | Default                   | What it does                                                                    |
+| ------------------- | ------------------------- | ------------------------------------------------------------------------------- |
+| `--job.target`      | `local`                   | HF Jobs flavor to run on (e.g. `h200`, `h200x4`). Omitted/`local` runs here.    |
+| `--job.image`       | `vllm/vllm-openai:latest` | Runtime image for the pod.                                                      |
+| `--job.timeout`     | `2h`                      | Wall-clock cap. Raise it for large datasets.                                    |
+| `--job.detach`      | `false`                   | Submit and exit instead of streaming logs.                                      |
+| `--job.lerobot_ref` | `main`                    | Git ref of lerobot installed on the pod — point it at a branch to test changes. |
+| `--job.tags`        | `[]`                      | Extra tags on the job and on any dataset it pushes (`lerobot` is always added). |
+
+For a bigger dataset, scale to `h200x4` and raise
+`--vlm.parallel_servers` / `--vlm.num_gpus` to match, and give the job
+more headroom with e.g. `--job.timeout=8h`.
+
+Remote runs need `--repo_id` (the pod pulls the dataset from the Hub;
+`--root` names a directory only your machine has). A dataset that exists
+only in your local cache is pushed to a **private** repo first.
 
 ## Key options
 

docs/source/backwardcomp.mdx

@@ -58,7 +58,7 @@ final_action = postprocessor(action)
 
 ## Hardware API redesign
 
-PR [#777](https://github.com/huggingface/lerobot/pull/777) improves the LeRobot calibration but is **not backward-compatible**. Below is a overview of what changed and how you can continue to work with datasets created before this pull request.
+PR [#777](https://github.com/huggingface/lerobot/pull/777) improves the LeRobot calibration but is **not backward-compatible**. Below is an overview of what changed and how you can continue to work with datasets created before this pull request.
 
 ### What changed?
 
@@ -129,8 +129,8 @@ python examples/backward_compatibility/replay.py \
 
 Policies output actions in the same format as the datasets (`torch.Tensors`). Therefore, the same transformations should be applied.
 
-To find these transformations, we recommend to first try and and replay an episode of the dataset your policy was trained on using the section above.
-Then, add these same transformations on your inference script (shown here in the `record.py` script):
+To find these transformations, we recommend first replaying an episode of the dataset your policy was trained on using the section above.
+Then, add these same transformations to your inference script (shown here in the `record.py` script):
 
 ```diff
 action_values = predict_action(

docs/source/bring_your_own_policies.mdx

@@ -165,6 +165,8 @@ Batches are flat dictionaries keyed by the constants in [`lerobot.utils.constant
 
 LeRobot uses `PolicyProcessorPipeline`s to normalize inputs and de-normalize outputs around your policy. For a concrete reference, see [`processor_act.py`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/act/processor_act.py) or [`processor_diffusion.py`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/diffusion/processor_diffusion.py).
 
+Pay close attention here: processors are the most common reproducibility pain point. A mismatch in normalization mode (`IDENTITY` vs `MEAN_STD` vs `MIN_MAX` vs `QUANTILES`/`QUANTILE10`) or in which features get normalized will train and eval without erroring, yet silently wreck results. Make sure the modes match how the checkpoint was trained, that the required stats exist (e.g. `QUANTILES` needs `q01`/`q99`), and that the pre- and post-processors stay consistent.
+
 ```python
 # processor_my_policy.py
 from typing import Any
@@ -304,7 +306,9 @@ Mirror an existing policy that's structurally similar to yours; the diff is smal
 
 ### Heavy / optional dependencies
 
-Most policies need a heavy backbone (transformers, diffusers, a specific VLM SDK). The convention is **two-step gating**: a `TYPE_CHECKING`-guarded import at module top, and a `require_package` runtime check in the constructor. [`modeling_diffusion.py`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/diffusion/modeling_diffusion.py) is the canonical reference:
+Most policies need a heavy backbone (transformers, diffusers, a specific VLM SDK). Wherever one exists, prefer loading it e.g from `transformers` or `diffusers` rather than re-implementing the architecture in-tree.
+
+The convention is **two-step gating**: a `TYPE_CHECKING`-guarded import at module top, and a `require_package` runtime check in the constructor. [`modeling_diffusion.py`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/diffusion/modeling_diffusion.py) is the canonical reference:
 
 ```python
 from typing import TYPE_CHECKING
@@ -374,6 +378,7 @@ The general expectations are in [`CONTRIBUTING.md`](https://github.com/huggingfa
 - [ ] Optional deps live behind a `[project.optional-dependencies]` extra and the `TYPE_CHECKING + require_package` guard.
 - [ ] `tests/policies/` updated; backward-compat artifact committed & policy-specific tests.
 - [ ] `src/lerobot/policies/<name>/README.md` symlinked into `docs/source/policy_<name>_README.md`; user-facing `docs/source/<name>.mdx` written and added to `_toctree.yml`.
+- [ ] `lerobot-train --policy.type my_policy ...` runs end-to-end for at least a few steps + save a checkpoint that can be loaded and run by `lerobot-eval` or `lerobot-rollout`.
 - [ ] `templates/lerobot_modelcard_template.md` has a description entry and a `policy_docs` link for your policy.
 - [ ] The models table in the root `README.md` lists your policy in the right category, linking to your doc page.
 - [ ] At least one reproducible benchmark eval in the policy MDX with a published checkpoint (sim benchmark, or real-robot dataset + checkpoint).

docs/source/cameras.mdx

@@ -136,6 +136,10 @@ config = RealSenseCameraConfig(
     height=480,
     color_mode=ColorMode.RGB,
     use_depth=True,
+    # Optional fixed color controls. Omit them to leave the current sensor settings unchanged.
+    exposure=120,
+    gain=64,
+    white_balance=4600,
     rotation=Cv2Rotation.NO_ROTATION
 )
 
@@ -154,6 +158,15 @@ finally:
 ```
 <!-- prettier-ignore-end -->
 
+Manual color controls disable the corresponding automatic exposure or white-balance mode. Their
+supported ranges vary by camera model; an invalid value raises an error at connection time that
+includes the range reported by the sensor. Requesting an unsupported control also raises an error.
+Omitted controls leave the sensor's existing automatic or manual setting unchanged. These options
+require `use_rgb=True`.
+
+On the RealSense D405, the color stream is provided by the Stereo Module, so changing manual
+exposure or gain also affects the depth stream.
+
 </hfoption>
 </hfoptions>
 

docs/source/env_processor.mdx

@@ -88,20 +88,6 @@ policy_preprocessor = NormalizerProcessorStep(stats=dataset_stats)
 
 The same policy can work with different environment processors, and the same environment processor can work with different policies:
 
-````python
-# Use SmolVLA policy with LIBERO environment
-# Use SmolVLA policy with LIBERO environment
-libero_preprocessor, libero_postprocessor = make_env_pre_post_processors(
-    env_cfg=libero_cfg,
-    policy_cfg=smolvla_cfg,
-)
-smolvla_preprocessor, smolvla_postprocessor = make_pre_post_processors(smolvla_cfg)
-# Or use ACT policy with the same LIBERO environment
-libero_preprocessor, libero_postprocessor = make_env_pre_post_processors(
-    env_cfg=libero_cfg,
-    policy_cfg=act_cfg,
-)
-act_preprocessor, act_postprocessor = make_pre_post_processors(act_cfg)
 ```python
 # Use SmolVLA policy with LIBERO environment
 libero_preprocessor, libero_postprocessor = make_env_pre_post_processors(
@@ -116,6 +102,7 @@ libero_preprocessor, libero_postprocessor = make_env_pre_post_processors(
     policy_cfg=act_cfg,
 )
 act_preprocessor, act_postprocessor = make_pre_post_processors(act_cfg)
+```
 
 ### 3. **Easier Experimentation**
 
@@ -145,7 +132,7 @@ class LiberoVelocityProcessorStep(ObservationProcessorStep):
         state = torch.cat([eef_pos, eef_axisangle, eef_vel,
                           gripper_pos, gripper_vel], dim=-1)  # 14D
         return state
-````
+```
 
 ### 4. **Cleaner Environment Code**
 

docs/source/feetech.mdx

@@ -40,10 +40,10 @@ This tutorial guides you through updating the firmware of Feetech motors using t
 For each motor you want to update:
 
 1. **Select the motor** from the list by clicking on it
-2. **Click on Upgrade tab**:
-3. **Click on Online button**:
-   - If an potential firmware update is found, it will be displayed in the box
-4. **Click on Upgrade button**:
+2. **Click the Upgrade tab**:
+3. **Click the Online button**:
+   - If a potential firmware update is found, it will be displayed in the box
+4. **Click the Upgrade button**:
    - The update progress will be displayed
 
 ## Step 6: Verify Update

docs/source/hope_jr.mdx

@@ -211,7 +211,7 @@ Record, Replay and Train with Hope-JR is still experimental.
 
 ### Record
 
-This step records the dataset, which can be seen as an example [here](https://huggingface.co/datasets/nepyope/hand_record_test_with_video_data/settings).
+This step records the dataset, which can be seen as an example [here](https://huggingface.co/datasets/nepyope/hand_record_test_with_video_data).
 
 ```bash
 lerobot-record \

docs/source/integrate_hardware.mdx

@@ -18,7 +18,7 @@ If you're using Feetech or Dynamixel motors, LeRobot provides built-in bus inter
 - [`DynamixelMotorsBus`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/motors/dynamixel/dynamixel.py) – for controlling Dynamixel servos
 
 Please refer to the [`MotorsBus`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/motors/motors_bus.py) abstract class to learn about its API.
-For a good example of how it can be used, you can have a look at our own [SO101 follower implementation](https://github.com/huggingface/lerobot/blob/main/src/lerobot/robots/so_follower/so101_follower/so101_follower.py)
+For a good example of how it can be used, you can have a look at our own [SO101 follower implementation](https://github.com/huggingface/lerobot/blob/main/src/lerobot/robots/so_follower/so_follower.py)
 
 Use these if compatible. Otherwise, you'll need to find or write a Python interface (not covered in this tutorial):
 

docs/source/lekiwi.mdx

@@ -51,7 +51,7 @@ In addition to these instructions, you need to install the Feetech SDK & ZeroMQ
 pip install -e ".[lekiwi]"
 ```
 
-Great :hugs:! You are now done installing LeRobot, and we can begin assembling the SO100/SO101 arms and the mobile base :robot:.
+Great 🤗! You are now done installing LeRobot, and we can begin assembling the SO100/SO101 arms and the mobile base 🤖.
 Every time you now want to use LeRobot, you can go to the `~/lerobot` folder where we installed LeRobot and run one of the commands.
 
 # Step-by-Step Assembly Instructions

docs/source/omx.mdx

@@ -1,3 +1,11 @@
+# OMX
+
+<img
+  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/omx_mainimage.png"
+  alt="OMX"
+  width=600
+/>
+
 ## Order and Assemble the parts
 
 First, assemble the OMX hardware following the official assembly guide.

docs/source/pi0fast.mdx

@@ -174,7 +174,7 @@ The model takes images, text instructions, and robot state as input, and outputs
 
 ## Reproducing π₀Fast results
 
-We reproduce the results of π₀Fast on the LIBERO benchmark using the LeRobot implementation. We take the LeRobot PiFast base model [lerobot/pi0fast-base](https://huggingface.co/lerobot/pi0fast-base) and finetune for an additional 40kk steps in bfloat16, with batch size of 256 on 8 H100 GPUs using the [HuggingFace LIBERO dataset](https://huggingface.co/datasets/HuggingFaceVLA/libero).
+We reproduce the results of π₀Fast on the LIBERO benchmark using the LeRobot implementation. We take the LeRobot PiFast base model [lerobot/pi0fast-base](https://huggingface.co/lerobot/pi0fast-base) and finetune for an additional 40k steps in bfloat16, with batch size of 256 on 8 H100 GPUs using the [HuggingFace LIBERO dataset](https://huggingface.co/datasets/HuggingFaceVLA/libero).
 
 The finetuned model can be found here:
 

docs/source/processors_robots_teleop.mdx

@@ -22,7 +22,7 @@ With processors, you choose the learning features you want to use for your polic
 ## Three pipelines
 
 We often compose three pipelines. Depending on your setup, some can be empty if action and observation spaces already match.
-Each of these pipelines handle different conversions between different action and observation spaces. Below is a quick explanation of each pipeline.
+Each of these pipelines handles different conversions between different action and observation spaces. Below is a quick explanation of each pipeline.
 
 1. Pipeline 1: Teleop action space → dataset action space (phone pose → EE targets)
 2. Pipeline 2: Dataset action space → robot command space (EE targets → joints)
@@ -74,7 +74,7 @@ In the phone to SO-100 follower examples we use the following adapters:
 - `robot_action_to_transition`: transforms the teleop action dict to a pipeline transition.
 - `transition_to_robot_action`: transforms the pipeline transition to a robot action dict.
 - `observation_to_transition`: transforms the robot observation dict to a pipeline transition.
-- `transition_to_observation`: transforms the pipeline transition to a observation dict.
+- `transition_to_observation`: transforms the pipeline transition to an observation dict.
 
 Check out [src/lerobot/processor/converters.py](https://github.com/huggingface/lerobot/blob/main/src/lerobot/processor/converters.py) for more details.
 
@@ -82,7 +82,7 @@ Checkout [src/lerobot/processor/converters.py](https://github.com/huggingface/le
 
 Dataset features are determined by the keys saved in the dataset. Each step can declare what features it modifies in a contract called `transform_features(...)`. Once you build a processor, the processor can then aggregate all of these features with `aggregate_pipeline_dataset_features()` and merge multiple feature dicts with `combine_feature_dicts(...)`.
 
-Below is and example of how we declare features with the `transform_features` method in the phone to SO-100 follower examples:
+Below is an example of how we declare features with the `transform_features` method in the phone to SO-100 follower examples:
 
 ```python
     def transform_features(

docs/source/rtc.mdx

@@ -57,7 +57,7 @@ policy_cfg.rtc_config = RTCConfig(
 policy = PI0Policy.from_pretrained("lerobot/pi0_base", policy_cfg=policy_cfg, device="cuda")
 
 # Now use predict_action_chunk with RTC parameters
-inference_delay = 4  # How many steps of inference latency, this values should be calculated based on the inference latency of the policy
+inference_delay = 4  # How many steps of inference latency, this value should be calculated based on the inference latency of the policy
 
 # Initialize the action queue
 action_queue = ActionQueue(policy_cfg.rtc_config)
@@ -100,7 +100,7 @@ Typical values: 8-12 steps
 RTCConfig(execution_horizon=10)
 ```
 
-**`max_guidance_weight`**: How strongly to enforce consistency with the previous chunk. This is a hyperparameter that can be tuned to balance the smoothness of the transitions and the reactivity of the policy. For 10 steps flow matching (SmolVLA, Pi0, Pi0.5), a value of 10.0 is a optimal value.
+**`max_guidance_weight`**: How strongly to enforce consistency with the previous chunk. This is a hyperparameter that can be tuned to balance the smoothness of the transitions and the reactivity of the policy. For 10 steps flow matching (SmolVLA, Pi0, Pi0.5), a value of 10.0 is an optimal value.
 
 **`prefix_attention_schedule`**: How to weight consistency across the overlap region.
 

docs/source/smolvla.mdx

@@ -93,7 +93,7 @@ lerobot-train --help
 
 ## Evaluate the finetuned model and run it in real-time
 
-Similarly for when recording an episode, it is recommended that you are logged in to the HuggingFace Hub. You can follow the corresponding steps: [Record a dataset](./il_robots).
+Similarly for when recording an episode, it is recommended that you are logged in to the HuggingFace Hub. You can follow the corresponding steps: [Record a dataset](./il_robots#record-a-dataset).
 Once you are logged in, you can run inference in your setup by doing:
 
 ```bash

docs/source/using_dataset_tools.mdx

@@ -50,11 +50,11 @@ lerobot-edit-dataset \
 Divide a dataset into multiple subsets.
 
 ```bash
-# Split by fractions (e.g. 80% train, 20% test, 20% val)
+# Split by fractions (e.g. 60% train, 20% val, 20% test)
 lerobot-edit-dataset \
     --repo_id lerobot/pusht \
     --operation.type split \
-    --operation.splits '{"train": 0.8, "test": 0.2, "val": 0.2}'
+    --operation.splits '{"train": 0.6, "val": 0.2, "test": 0.2}'
 
 # Split by specific episode indices
 lerobot-edit-dataset \
@@ -252,6 +252,10 @@ lerobot-dataset-viz \
     --episode-index 0
 ```
 
+For a private or gated dataset, authenticate first with `hf auth login`, or set the
+`HF_TOKEN` environment variable. The Hub client then discovers the credential
+automatically; no token argument is needed.
+
 **From a local folder:**
 Add the `--root` option and set `--mode local`. For example, to search in `./my_local_data_dir/lerobot/pusht`:
 

examples/annotations/run_hf_job.py

@@ -1,80 +0,0 @@
-#!/usr/bin/env python
-
-# Copyright 2026 The HuggingFace Inc. team. All rights reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-"""Launch ``lerobot-annotate`` on a Hugging Face job (vllm + Qwen3.6-27B VLM).
-
-Spawns one single-GPU ``h200`` job that:
-
-  1. installs ``lerobot`` from ``main`` plus the annotation extras,
-  2. boots one vllm server with Qwen3.6-27B (dense VLM),
-  3. runs the plan / interjections / vqa modules across the dataset
-     in free-form mode (each episode generates its own subtasks +
-     memory),
-  4. uploads the annotated dataset to ``--new_repo_id`` (when set)
-     or back to ``--repo_id``.
-
-Usage:
-
-    HF_TOKEN=hf_... uv run python examples/annotations/run_hf_job.py
-
-Adjust ``CMD`` (dataset, model, hub repo) and ``flavor`` below for your
-run. For larger datasets, scale to ``h200x4`` and raise
-``--vlm.parallel_servers`` / ``--vlm.num_gpus`` to match.
-"""
-
-import os
-
-from huggingface_hub import get_token, run_job
-
-token = os.environ.get("HF_TOKEN") or get_token()
-if not token:
-    raise RuntimeError("No HF token. Run `huggingface-cli login` or `export HF_TOKEN=hf_...`")
-
-CMD = (
-    "apt-get update -qq && apt-get install -y -qq git ffmpeg && "
-    "pip install --no-deps "
-    "'lerobot @ git+https://github.com/huggingface/lerobot.git@main' && "
-    # Pins mirror pyproject.toml — unpinned installs pull av 18 / datasets 5 /
-    # draccus 0.11, which break lerobot at import time.
-    "pip install --upgrade-strategy only-if-needed "
-    "'datasets>=4.7.0,<5.0.0' 'pyarrow>=21.0.0,<30.0.0' 'av>=15.0.0,<16.0.0' 'draccus==0.10.0' "
-    "'pandas>=2.0.0,<3.0.0' jsonlines gymnasium torchcodec mergedeep pyyaml-include toml typing-inspect "
-    "openai && "
-    "export VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0 && "
-    "export VLLM_VIDEO_BACKEND=pyav && "
-    "lerobot-annotate "
-    "--repo_id=pepijn223/robocasa_pretrain_human300_v4 "
-    "--new_repo_id=pepijn223/robocasa_pretrain_human300_v4_annotated "
-    "--push_to_hub=true "
-    "--vlm.backend=openai "
-    "--vlm.model_id=Qwen/Qwen3.6-27B "
-    "--vlm.num_gpus=1 "
-    '--vlm.serve_command="vllm serve Qwen/Qwen3.6-27B '
-    "--tensor-parallel-size 1 --max-model-len 32768 "
-    '--gpu-memory-utilization 0.8 --uvicorn-log-level warning --port {port}" '
-    "--vlm.serve_ready_timeout_s=1800 "
-    # Qwen3.6 ships with thinking on; annotation wants plain JSON answers.
-    "--vlm.chat_template_kwargs='{\"enable_thinking\": false}'"
-)
-
-job = run_job(
-    image="vllm/vllm-openai:latest",
-    command=["bash", "-c", CMD],
-    flavor="h200",
-    secrets={"HF_TOKEN": token},
-    timeout="2h",
-)
-print(f"Job URL: {job.url}")
-print(f"Job ID:  {job.id}")

pyproject.toml

@@ -155,7 +155,7 @@ accelerate-dep = ["accelerate>=1.14.0,<2.0.0"]
 can-dep = ["python-can>=4.2.0,<5.0.0"]
 peft-dep = ["peft>=0.18.0,<1.0.0"]
 scipy-dep = ["scipy>=1.14.0,<2.0.0"]
-diffusers-dep = ["diffusers>=0.27.2,<0.36.0"]
+diffusers-dep = ["diffusers>=0.38.0,<0.40.0"]
 qwen-vl-utils-dep = ["qwen-vl-utils>=0.0.11,<0.1.0"]
 matplotlib-dep = ["matplotlib>=3.10.3,<4.0.0", "contourpy>=1.3.0,<2.0.0"] # NOTE: Explicitly listing contourpy helps the resolver converge faster.
 pyserial-dep = ["pyserial>=3.5,<4.0"]
@@ -494,6 +494,19 @@ ignore_errors = true
 module = "lerobot.envs.*"
 ignore_errors = false
 
+[[tool.mypy.overrides]]
+module = "lerobot.annotations.*"
+ignore_errors = false
+disallow_untyped_defs = true
+disallow_incomplete_defs = true
+check_untyped_defs = true
+
+[[tool.mypy.overrides]]
+module = "lerobot.transforms.*"
+ignore_errors = false
+disallow_untyped_defs = true
+disallow_incomplete_defs = true
+check_untyped_defs = true
 
 # [[tool.mypy.overrides]]
 # module = "lerobot.utils.*"

src/lerobot/annotations/steerable_pipeline/config.py

@@ -20,6 +20,29 @@ from dataclasses import dataclass, field
 from pathlib import Path
 from typing import Any
 
+from lerobot.configs.default import JobConfig
+
+# The annotation pipeline boots its own vLLM server, so the pod starts from the
+# official vLLM runtime rather than the prebuilt `lerobot-gpu` training image;
+# `lerobot` is pip-installed on top (see `lerobot.jobs.annotate`).
+DEFAULT_ANNOTATE_JOB_IMAGE = "vllm/vllm-openai:latest"
+
+
+@dataclass
+class AnnotationJobConfig(JobConfig):
+    """`JobConfig` with the annotation runtime's defaults.
+
+    Adds `lerobot_ref` because the vLLM image ships no lerobot: the pod installs
+    it from git, and the ref decides which code actually annotates. Point it at a
+    branch/tag/SHA to try unmerged changes remotely.
+    """
+
+    image: str = DEFAULT_ANNOTATE_JOB_IMAGE
+    # Annotation is a bounded pass over a dataset; a tighter cap than training's
+    # "2d" keeps a wedged vLLM server from burning a day of GPU time.
+    timeout: str | None = "2h"
+    lerobot_ref: str = "main"
+
 
 @dataclass
 class PlanConfig:
@@ -207,6 +230,11 @@ class AnnotationPipelineConfig:
     vlm: VlmConfig = field(default_factory=VlmConfig)
     executor: ExecutorConfig = field(default_factory=ExecutorConfig)
 
+    # Where the annotation runs: omitted / "local" annotates on this machine, any
+    # other value is an HF Jobs flavor (e.g. "h200") and submits the run there.
+    # List flavors + pricing with `hf jobs hardware`.
+    job: AnnotationJobConfig = field(default_factory=AnnotationJobConfig)
+
     skip_validation: bool = False
     only_episodes: tuple[int, ...] | None = None
 

src/lerobot/annotations/steerable_pipeline/executor.py

@@ -30,8 +30,8 @@ Phase 3 is why the ``plan`` module must be re-entered after the
 timestamps.
 
 Distributed execution is provided by Hugging Face Jobs (see
-``examples/annotations/run_hf_job.py``); the runner inside the job
-invokes ``lerobot-annotate`` which uses this in-process executor.
+``lerobot.jobs.annotate``, reached via ``--job.target=<flavor>``); the pod
+inside the job invokes ``lerobot-annotate`` which uses this in-process executor.
 Episode-level concurrency is controlled by
 ``ExecutorConfig.episode_parallelism``.
 """

src/lerobot/annotations/steerable_pipeline/vlm_client.py

@@ -194,12 +194,13 @@ def make_vlm_client(config: VlmConfig) -> VlmClient:
     """Build the shared VLM client.
 
     Only the ``openai`` backend is supported for now. The shipped workflow
-    is Hugging Face Jobs (``examples/annotations/run_hf_job.py``): it boots
-    a vLLM server inside the ``vllm/vllm-openai`` image and the pipeline
-    talks to it over the OpenAI-compatible API (``--vlm.backend=openai``,
-    optionally auto-spawning the server via ``auto_serve`` /
-    ``serve_command``). The former in-process ``vllm`` / ``transformers``
-    backends were removed to keep the support surface to the HF Jobs path.
+    is Hugging Face Jobs (``lerobot-annotate --job.target=<flavor>``): it
+    boots a vLLM server inside the ``vllm/vllm-openai`` image and the
+    pipeline talks to it over the OpenAI-compatible API
+    (``--vlm.backend=openai``, optionally auto-spawning the server via
+    ``auto_serve`` / ``serve_command``). The former in-process ``vllm`` /
+    ``transformers`` backends were removed to keep the support surface to
+    the HF Jobs path.
 
     For ``stub``, construct :class:`StubVlmClient` directly with a responder
     callable; it is rejected here to make accidental misuse obvious.
@@ -213,8 +214,8 @@ def make_vlm_client(config: VlmConfig) -> VlmClient:
     if config.backend in {"vllm", "transformers"}:
         raise ValueError(
             f"backend={config.backend!r} (in-process local model) is not supported for now — "
-            "only backend='openai' (the Hugging Face Jobs flow) is. Run the pipeline via "
-            "examples/annotations/run_hf_job.py, which serves the model with vLLM in the "
+            "only backend='openai' (the Hugging Face Jobs flow) is. Run the pipeline with "
+            "`lerobot-annotate --job.target=<flavor>`, which serves the model with vLLM in the "
             "vllm/vllm-openai image and talks to it over the OpenAI-compatible API."
         )
     raise ValueError(f"Unknown VLM backend: {config.backend!r}")

src/lerobot/cameras/opencv/camera_opencv.py

@@ -120,14 +120,22 @@ class OpenCVCamera(Camera):
         self.rotation: int | None = get_cv2_rotation(config.rotation)
         self.backend: int = config.backend
 
-        if self.height and self.width:
-            self.capture_width, self.capture_height = self.width, self.height
-            if self.rotation in [cv2.ROTATE_90_CLOCKWISE, cv2.ROTATE_90_COUNTERCLOCKWISE]:
-                self.capture_width, self.capture_height = self.height, self.width
+        self.capture_width: int | None = None
+        self.capture_height: int | None = None
+        self._reset_connection_settings()
 
     def __str__(self) -> str:
         return f"{self.__class__.__name__}({self.index_or_path})"
 
+    def _reset_connection_settings(self) -> None:
+        """Restore settings that may have been auto-detected during a failed connection."""
+        self.fps = self.config.fps
+        self.width = self.config.width
+        self.height = self.config.height
+        self.capture_width, self.capture_height = self.width, self.height
+        if self.rotation in [cv2.ROTATE_90_CLOCKWISE, cv2.ROTATE_90_COUNTERCLOCKWISE]:
+            self.capture_width, self.capture_height = self.height, self.width
+
     @property
     def is_connected(self) -> bool:
         """Checks if the camera is currently connected and opened."""
@@ -164,6 +172,7 @@ class OpenCVCamera(Camera):
                 f"Failed to open {self}.Run `lerobot-find-cameras opencv` to find available cameras."
             )
 
+        try:
             self._configure_capture_settings()
             self._start_read_thread()
 
@@ -175,6 +184,13 @@ class OpenCVCamera(Camera):
                 with self.frame_lock:
                     if self.latest_frame is None:
                         raise ConnectionError(f"{self} failed to capture frames during warmup.")
+        except BaseException:
+            try:
+                self._cleanup_resources()
+            except Exception:
+                logger.exception(f"Failed to fully clean up {self} after connect() failed.")
+            self._reset_connection_settings()
+            raise
 
         logger.info(f"{self} connected.")
 
@@ -312,6 +328,7 @@ class OpenCVCamera(Camera):
 
         for target in targets_to_scan:
             camera = cv2.VideoCapture(target)
+            try:
                 if camera.isOpened():
                     default_width = int(camera.get(cv2.CAP_PROP_FRAME_WIDTH))
                     default_height = int(camera.get(cv2.CAP_PROP_FRAME_HEIGHT))
@@ -321,7 +338,9 @@ class OpenCVCamera(Camera):
                     # Get FOURCC code and convert to string
                     default_fourcc_code = camera.get(cv2.CAP_PROP_FOURCC)
                     default_fourcc_code_int = int(default_fourcc_code)
-                default_fourcc = "".join([chr((default_fourcc_code_int >> 8 * i) & 0xFF) for i in range(4)])
+                    default_fourcc = "".join(
+                        [chr((default_fourcc_code_int >> 8 * i) & 0xFF) for i in range(4)]
+                    )
 
                     camera_info = {
                         "name": f"OpenCV Camera @ {target}",
@@ -338,6 +357,7 @@ class OpenCVCamera(Camera):
                     }
 
                     found_cameras_info.append(camera_info)
+            finally:
                 camera.release()
 
         return found_cameras_info
@@ -496,6 +516,26 @@ class OpenCVCamera(Camera):
             self.latest_timestamp = None
             self.new_frame_event.clear()
 
+    def _cleanup_resources(self) -> None:
+        """Stop background reads and release the capture, including after partial setup."""
+        read_thread = self.thread
+        videocapture = self.videocapture
+
+        try:
+            self._stop_read_thread()
+        finally:
+            self.videocapture = None
+            try:
+                if videocapture is not None:
+                    videocapture.release()
+            finally:
+                # Releasing the device may unblock a hardware read that outlived
+                # the first bounded join in _stop_read_thread().
+                if read_thread is not None and read_thread.is_alive():
+                    read_thread.join(timeout=2.0)
+                    if read_thread.is_alive():  # pragma: no cover
+                        logger.warning(f"{self} read thread remained alive after releasing the capture.")
+
     @check_if_not_connected
     def async_read(self, timeout_ms: float = 200) -> NDArray[Any]:
         """
@@ -586,16 +626,6 @@ class OpenCVCamera(Camera):
         if not self.is_connected and self.thread is None:
             raise DeviceNotConnectedError(f"{self} not connected.")
 
-        if self.thread is not None:
-            self._stop_read_thread()
-
-        if self.videocapture is not None:
-            self.videocapture.release()
-            self.videocapture = None
-
-        with self.frame_lock:
-            self.latest_frame = None
-            self.latest_timestamp = None
-            self.new_frame_event.clear()
+        self._cleanup_resources()
 
         logger.info(f"{self} disconnected.")

src/lerobot/cameras/reachy2_camera/reachy2_camera.py

@@ -173,7 +173,8 @@ class Reachy2Camera(Camera):
             raise ValueError(
                 f"Invalid color mode '{self.color_mode}'. Expected {ColorMode.RGB} or {ColorMode.BGR}."
             )
-        if self.color_mode == ColorMode.RGB:
+        is_depth_frame = self.config.name == "depth" and self.config.image_type == "depth"
+        if not is_depth_frame and self.color_mode == ColorMode.RGB:
             frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)
 
         self.latest_frame = frame

src/lerobot/cameras/realsense/camera_realsense.py

@@ -121,6 +121,9 @@ class RealSenseCamera(Camera):
 
         self.config = config
 
+        self.width: int | None = config.width
+        self.height: int | None = config.height
+
         if config.serial_number_or_name.isdigit():
             self.serial_number = config.serial_number_or_name
         else:
@@ -131,6 +134,9 @@ class RealSenseCamera(Camera):
         self.use_rgb = config.use_rgb
         self.use_depth = config.use_depth
         self.warmup_s = config.warmup_s
+        self.exposure: int | None = config.exposure
+        self.gain: int | None = config.gain
+        self.white_balance: int | None = config.white_balance
 
         self.rs_pipeline: rs.pipeline | None = None
         self.rs_profile: rs.pipeline_profile | None = None
@@ -145,14 +151,23 @@ class RealSenseCamera(Camera):
 
         self.rotation: int | None = get_cv2_rotation(config.rotation)
 
-        if self.height and self.width:
-            self.capture_width, self.capture_height = self.width, self.height
-            if self.rotation in [cv2.ROTATE_90_CLOCKWISE, cv2.ROTATE_90_COUNTERCLOCKWISE]:
-                self.capture_width, self.capture_height = self.height, self.width
+        self.capture_width: int | None = None
+        self.capture_height: int | None = None
+        self._reset_connection_settings()
 
     def __str__(self) -> str:
         return f"{self.__class__.__name__}({self.serial_number})"
 
+    def _reset_connection_settings(self) -> None:
+        """Restore settings that may have been auto-detected during a failed connection."""
+        self.fps = self.config.fps
+        self.width = self.config.width
+        self.height = self.config.height
+        self.warmup_s = self.config.warmup_s
+        self.capture_width, self.capture_height = self.width, self.height
+        if self.rotation in [cv2.ROTATE_90_CLOCKWISE, cv2.ROTATE_90_COUNTERCLOCKWISE]:
+            self.capture_width, self.capture_height = self.height, self.width
+
     @property
     def is_connected(self) -> bool:
         """Checks if the camera pipeline is started and streams are active."""
@@ -172,7 +187,8 @@ class RealSenseCamera(Camera):
 
         Raises:
             DeviceAlreadyConnectedError: If the camera is already connected.
-            ValueError: If the configuration is invalid (e.g., missing serial/name, name not unique).
+            ValueError: If the configuration is invalid, a requested sensor option is unsupported,
+                or a requested sensor value is invalid.
             ConnectionError: If the camera is found but fails to start the pipeline or no RealSense devices are detected at all.
             RuntimeError: If the pipeline starts but fails to apply requested settings.
         """
@@ -190,7 +206,9 @@ class RealSenseCamera(Camera):
                 f"Failed to open {self}.Run `lerobot-find-cameras realsense` to find available cameras."
             ) from e
 
+        try:
             self._configure_capture_settings()
+            self._configure_sensor_options()
             self._start_read_thread()
 
             # NOTE(Steven/Caroline): Enforcing at least one second of warmup as RS cameras need a bit of time before the first read. If we don't wait, the first read from the warmup will raise.
@@ -206,6 +224,13 @@ class RealSenseCamera(Camera):
                     self.use_depth and self.latest_depth_frame is None
                 ):
                     raise ConnectionError(f"{self} failed to capture frames during warmup.")
+        except BaseException:
+            try:
+                self._cleanup_resources()
+            except Exception:
+                logger.exception(f"Failed to fully clean up {self} after connect() failed.")
+            self._reset_connection_settings()
+            raise
 
         logger.info(f"{self} connected.")
 
@@ -339,6 +364,111 @@ class RealSenseCamera(Camera):
         self.new_frame_event.clear()
         return self._async_read(timeout_ms=10000, read_depth=read_depth)
 
+    def _get_color_sensor(self) -> "rs.sensor":
+        """Returns the sensor that controls the color stream.
+
+        Most RealSense cameras expose "RGB Camera" for color. The D405 has no
+        separate RGB module — its color stream comes from "Stereo Module".
+        We try RGB Camera first, then fall back to Stereo Module.
+        """
+        if self.rs_profile is None:
+            raise RuntimeError(f"{self}: rs_profile must be initialized before use.")
+
+        device = self.rs_profile.get_device()
+        sensors = {s.get_info(rs.camera_info.name): s for s in device.query_sensors()}
+
+        for name in ("RGB Camera", "Stereo Module"):
+            if name in sensors:
+                return sensors[name]
+
+        available = list(sensors.keys())
+        raise RuntimeError(f"{self}: no color sensor found. Available sensors: {available}")
+
+    def _set_sensor_option(self, sensor: "rs.sensor", option: "rs.option", value: float, label: str) -> None:
+        """Sets a sensor option, re-raising range errors with actionable diagnostics."""
+        try:
+            sensor.set_option(option, value)
+        except Exception as e:
+            range_info = ""
+            try:
+                option_range = sensor.get_option_range(option)
+                range_info = (
+                    f" (supported range: min={option_range.min}, max={option_range.max}, "
+                    f"step={option_range.step}, default={option_range.default})"
+                )
+            except Exception:
+                range_info = " (option range unavailable)"
+            raise ValueError(
+                f"{self}: failed to set {label} to {value}{range_info}. Original error: {e}"
+            ) from e
+
+    def _configure_sensor_options(self) -> None:
+        """Applies manual sensor options (exposure, gain, white balance) to the color sensor.
+
+        When exposure or gain is set, auto-exposure is disabled first. When white_balance
+        is set, auto white balance is disabled first. An omitted option is left unchanged,
+        and configuration is skipped entirely if all options are omitted.
+
+        Raises:
+            ValueError: If the sensor does not support a requested option or a requested
+                value is invalid. Invalid-value errors include the option name, requested
+                value, and supported range when available.
+        """
+        if self.exposure is None and self.gain is None and self.white_balance is None:
+            return
+
+        color_sensor = self._get_color_sensor()
+
+        requested_options = (
+            (rs.option.exposure, self.exposure, "exposure"),
+            (rs.option.gain, self.gain, "gain"),
+            (rs.option.white_balance, self.white_balance, "white balance"),
+        )
+        unsupported_options = [
+            label
+            for option, value, label in requested_options
+            if value is not None and not color_sensor.supports(option)
+        ]
+        if unsupported_options:
+            raise ValueError(
+                f"{self}: color sensor does not support requested manual options: {unsupported_options}."
+            )
+
+        manual_exposure_requested = self.exposure is not None or self.gain is not None
+        if manual_exposure_requested:
+            if color_sensor.supports(rs.option.enable_auto_exposure):
+                self._set_sensor_option(color_sensor, rs.option.enable_auto_exposure, 0, "auto-exposure")
+                logger.info(f"{self} auto-exposure disabled.")
+            else:
+                logger.warning(
+                    f"{self} sensor does not support disabling auto-exposure; "
+                    "applying manual exposure/gain directly."
+                )
+
+        if self.exposure is not None:
+            self._set_sensor_option(color_sensor, rs.option.exposure, self.exposure, "exposure")
+            logger.info(f"{self} exposure set to {self.exposure}.")
+
+        if self.gain is not None:
+            self._set_sensor_option(color_sensor, rs.option.gain, self.gain, "gain")
+            logger.info(f"{self} gain set to {self.gain}.")
+
+        if self.white_balance is not None:
+            if color_sensor.supports(rs.option.enable_auto_white_balance):
+                self._set_sensor_option(
+                    color_sensor, rs.option.enable_auto_white_balance, 0, "auto white balance"
+                )
+                logger.info(f"{self} auto white balance disabled.")
+            else:
+                logger.warning(
+                    f"{self} sensor does not support disabling auto white balance; "
+                    "applying manual white balance directly."
+                )
+            self._set_sensor_option(
+                color_sensor, rs.option.white_balance, self.white_balance, "white balance"
+            )
+            logger.info(f"{self} white balance set to {self.white_balance}.")
+
     @check_if_not_connected
     def read_depth(self, timeout_ms: int = 200) -> NDArray[Any]:
         """
@@ -453,7 +583,7 @@ class RealSenseCamera(Camera):
             )
 
         processed_image = image
-        if self.color_mode == ColorMode.BGR:
+        if not depth_frame and self.color_mode == ColorMode.BGR:
             processed_image = cv2.cvtColor(image, cv2.COLOR_RGB2BGR)
 
         if self.rotation in [cv2.ROTATE_90_CLOCKWISE, cv2.ROTATE_90_COUNTERCLOCKWISE, cv2.ROTATE_180]:
@@ -541,6 +671,27 @@ class RealSenseCamera(Camera):
             self.latest_timestamp = None
             self.new_frame_event.clear()
 
+    def _cleanup_resources(self) -> None:
+        """Stop background reads and stop the pipeline, including after partial setup."""
+        read_thread = self.thread
+        rs_pipeline = self.rs_pipeline
+
+        try:
+            self._stop_read_thread()
+        finally:
+            self.rs_pipeline = None
+            self.rs_profile = None
+            try:
+                if rs_pipeline is not None:
+                    rs_pipeline.stop()
+            finally:
+                # Stopping the pipeline may unblock a hardware read that outlived
+                # the first bounded join in _stop_read_thread().
+                if read_thread is not None and read_thread.is_alive():
+                    read_thread.join(timeout=2.0)
+                    if read_thread.is_alive():  # pragma: no cover
+                        logger.warning(f"{self} read thread remained alive after stopping the pipeline.")
+
     def _async_read(self, timeout_ms: float, read_depth: bool = False) -> NDArray[Any]:
         """Shared helper for :meth:`async_read`/:meth:`async_read_depth`: return the latest buffered frame."""
         if self.thread is None or not self.thread.is_alive():
@@ -684,18 +835,5 @@ class RealSenseCamera(Camera):
                 f"Attempted to disconnect {self}, but it appears already disconnected."
             )
 
-        if self.thread is not None:
-            self._stop_read_thread()
-
-        if self.rs_pipeline is not None:
-            self.rs_pipeline.stop()
-            self.rs_pipeline = None
-            self.rs_profile = None
-
-        with self.frame_lock:
-            self.latest_color_frame = None
-            self.latest_depth_frame = None
-            self.latest_timestamp = None
-            self.new_frame_event.clear()
-
+        self._cleanup_resources()
         logger.info(f"{self} disconnected.")

src/lerobot/cameras/realsense/configuration_realsense.py

@@ -46,6 +46,17 @@ class RealSenseCameraConfig(CameraConfig):
         use_depth: Whether to enable depth stream. Defaults to False.
         rotation: Image rotation setting (0°, 90°, 180°, or 270°). Defaults to no rotation.
         warmup_s: Time reading frames before returning from connect (in seconds)
+        exposure: Manual exposure value for the color sensor. When set, auto-exposure is
+            disabled and this fixed value is used. Valid ranges are camera-model specific
+            and reported if the value is rejected. Defaults to None (leave unchanged).
+        gain: Manual gain value for the color sensor. When set, auto-exposure is disabled
+            and this fixed gain is used, which also freezes exposure at its current value
+            when no exposure is configured. Valid ranges are camera-model specific and
+            reported if the value is rejected. Defaults to None (leave unchanged).
+        white_balance: Manual white balance value for the color sensor. When set, auto
+            white balance is disabled and this fixed value is used. Valid ranges are
+            camera-model specific and reported if the value is rejected. Defaults to None
+            (leave unchanged).
 
     Note:
         - Either name or serial_number must be specified.
@@ -61,6 +72,9 @@ class RealSenseCameraConfig(CameraConfig):
     use_depth: bool = False
     rotation: Cv2Rotation = Cv2Rotation.NO_ROTATION
     warmup_s: int = 1
+    exposure: int | None = None
+    gain: int | None = None
+    white_balance: int | None = None
 
     def __post_init__(self) -> None:
         self.color_mode = ColorMode(self.color_mode)
@@ -69,6 +83,18 @@ class RealSenseCameraConfig(CameraConfig):
         if not self.use_rgb and not self.use_depth:
             raise ValueError("At least one of `use_rgb` or `use_depth` must be enabled.")
 
+        manual_color_options = {
+            "exposure": self.exposure,
+            "gain": self.gain,
+            "white_balance": self.white_balance,
+        }
+        configured_color_options = [name for name, value in manual_color_options.items() if value is not None]
+        if configured_color_options and not self.use_rgb:
+            raise ValueError(
+                "Manual color sensor options require `use_rgb=True`. "
+                f"Configured options: {configured_color_options}."
+            )
+
         values = (self.fps, self.width, self.height)
         if any(v is not None for v in values) and any(v is None for v in values):
             raise ValueError(

src/lerobot/configs/dataset.py

@@ -71,13 +71,19 @@ class DatasetRecordConfig:
     # Number of threads per encoder instance. None = auto (codec default).
     # Lower values reduce CPU usage, maps to 'lp' (via svtav1-params) for libsvtav1 and 'threads' for h264/hevc..
     encoder_threads: int | None = None
+    # Skip appending the date-time tag to repo_id, keeping the user-provided name as-is
+    # (e.g. self-managed versioned names intended for a later `lerobot-edit-dataset merge`).
+    no_stamp: bool = False
 
     def stamp_repo_id(self) -> None:
         """Append a date-time tag to ``repo_id`` so each recording session gets a unique name.
 
         Must be called explicitly at dataset *creation* time — not on resume,
         where the existing ``repo_id`` (already stamped) must be preserved.
+        No-op when ``no_stamp`` is set, preserving a user-managed ``repo_id``.
         """
+        if self.no_stamp:
+            return
         if self.repo_id:
             timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
             self.repo_id = f"{self.repo_id}_{timestamp}"

src/lerobot/configs/train.py

@@ -14,6 +14,7 @@
 import builtins
 import datetime as dt
 import json
+import multiprocessing
 import os
 import tempfile
 from dataclasses import dataclass, field
@@ -101,6 +102,12 @@ class TrainPipelineConfig(HubMixin):
     batch_size: int = 8
     prefetch_factor: int = 4
     persistent_workers: bool = True
+    # DataLoader worker start method. "spawn" is safer than "fork" with
+    # non-fork-safe libs (PyAV / torchcodec / ffmpeg), but adds some
+    # worker-startup time per run since workers re-import modules instead
+    # of inheriting parent state. Override with `--dataloader_multiprocessing_context=fork`
+    # when appropriate, or set it to `null` to use Python's platform default.
+    dataloader_multiprocessing_context: str | None = "spawn"
     steps: int = 100_000
     # Run policy in the simulation environment every N steps to measure reward/success (0 = disabled).
     env_eval_freq: int = 20_000
@@ -212,6 +219,17 @@ class TrainPipelineConfig(HubMixin):
             self.reward_model.pretrained_path = str(policy_dir)
 
     def validate(self) -> None:
+        available_contexts = multiprocessing.get_all_start_methods()
+        if (
+            self.dataloader_multiprocessing_context is not None
+            and self.dataloader_multiprocessing_context not in available_contexts
+        ):
+            raise ValueError(
+                "`dataloader_multiprocessing_context` must be None or one of "
+                f"{available_contexts} on this platform, got "
+                f"{self.dataloader_multiprocessing_context!r}."
+            )
+
         self._resolve_pretrained_from_cli()
 
         if self.policy is None and self.reward_model is None:

src/lerobot/datasets/aggregate.py

@@ -19,6 +19,7 @@ import copy
 import logging
 import shutil
 from pathlib import Path
+from typing import Any, NotRequired, TypedDict
 
 import datasets
 import pandas as pd
@@ -49,8 +50,32 @@ from .utils import (
 )
 from .video_utils import concatenate_video_files, get_video_duration_in_s
 
+logger = logging.getLogger(__name__)
 
-def merge_video_feature_info_for_aggregate(all_metadata: list[LeRobotDatasetMetadata]) -> dict[str, dict]:
+type FeatureDict = dict[str, dict[str, Any]]
+type ChunkFile = tuple[int, int]
+
+
+class IndexState(TypedDict):
+    chunk: int
+    file: int
+    src_to_dst: NotRequired[dict[ChunkFile, ChunkFile]]
+
+
+class VideoIndex(TypedDict):
+    chunk: int
+    file: int
+    latest_duration: float
+    episode_duration: float
+    src_to_offset: NotRequired[dict[ChunkFile, float]]
+    src_to_dst: NotRequired[dict[ChunkFile, ChunkFile]]
+    dst_file_durations: NotRequired[dict[ChunkFile, float]]
+
+
+type VideoIndexState = dict[str, VideoIndex]
+
+
+def merge_video_feature_info_for_aggregate(all_metadata: list[LeRobotDatasetMetadata]) -> FeatureDict:
     """Create a merged video feature info dictionary for aggregation. The video encoder info is merged field-by-field: each key is kept only when every source agrees; otherwise that key is set to ``null`` (or ``{}`` for ``video.extra_options``) and a warning is logged.
 
     Args:
@@ -59,14 +84,14 @@ def merge_video_feature_info_for_aggregate(all_metadata: list[LeRobotDatasetMeta
     Returns:
         dict: A dictionary of merged video feature info.
     """
-    merged_info = copy.deepcopy(all_metadata[0].features)
+    merged_info: FeatureDict = copy.deepcopy(all_metadata[0].features)
     video_keys = [k for k in merged_info if merged_info[k].get("dtype") == "video"]
 
     for vk in video_keys:
         video_infos = [m.features.get(vk, {}).get("info") or {} for m in all_metadata]
         base_video_info = video_infos[0]
 
-        merged_encoder_info: dict = {}
+        merged_encoder_info: dict[str, Any] = {}
         fallback_keys: list[str] = []
         for info_key in VIDEO_ENCODER_INFO_KEYS:
             values = [info.get(info_key, None) for info in video_infos]
@@ -80,7 +105,7 @@ def merge_video_feature_info_for_aggregate(all_metadata: list[LeRobotDatasetMeta
                 merged_encoder_info[info_key] = {} if info_key == "video.extra_options" else None
 
         if fallback_keys:
-            logging.warning(
+            logger.warning(
                 f"Merging heterogeneous or incomplete video encoder metadata for feature {vk}. "
                 f"Setting these keys to null: {fallback_keys}.",
             )
@@ -92,7 +117,7 @@ def merge_video_feature_info_for_aggregate(all_metadata: list[LeRobotDatasetMeta
     return merged_info
 
 
-def validate_all_metadata(all_metadata: list[LeRobotDatasetMetadata]):
+def validate_all_metadata(all_metadata: list[LeRobotDatasetMetadata]) -> tuple[int, str | None, FeatureDict]:
     """Validates that all dataset metadata have consistent properties.
 
     Ensures all datasets have the same fps, robot_type, and features to guarantee
@@ -129,7 +154,9 @@ def validate_all_metadata(all_metadata: list[LeRobotDatasetMetadata]):
     return fps, robot_type, features
 
 
-def update_data_df(df, src_meta, dst_meta):
+def update_data_df(
+    df: pd.DataFrame, src_meta: LeRobotDatasetMetadata, dst_meta: LeRobotDatasetMetadata
+) -> pd.DataFrame:
     """Updates a data DataFrame with new indices and task mappings for aggregation.
 
     Adjusts episode indices, frame indices, and task indices to account for
@@ -154,12 +181,12 @@ def update_data_df(df, src_meta, dst_meta):
 
 
 def update_meta_data(
-    df,
-    dst_meta,
-    meta_idx,
-    data_idx,
-    videos_idx,
-):
+    df: pd.DataFrame,
+    dst_meta: LeRobotDatasetMetadata,
+    meta_idx: IndexState,
+    data_idx: IndexState,
+    videos_idx: VideoIndexState,
+) -> pd.DataFrame:
     """Updates metadata DataFrame with new chunk, file, and timestamp indices.
 
     Adjusts all indices and timestamps to account for previously aggregated
@@ -289,7 +316,7 @@ def aggregate_datasets(
     chunk_size: int | None = None,
     concatenate_videos: bool = True,
     concatenate_data: bool = True,
-):
+) -> None:
     """Aggregates multiple LeRobot datasets into a single unified dataset.
 
     This is the main function that orchestrates the aggregation process by:
@@ -309,7 +336,7 @@ def aggregate_datasets(
         concatenate_videos: When False, keep one mp4 per source file instead of packing into shards.
         concatenate_data: When False, keep one parquet per source file instead of packing into shards.
     """
-    logging.info("Start aggregate_datasets")
+    logger.info("Start aggregate_datasets")
 
     if data_files_size_in_mb is None:
         data_files_size_in_mb = DEFAULT_DATA_FILE_SIZE_IN_MB
@@ -341,15 +368,15 @@ def aggregate_datasets(
         video_files_size_in_mb=video_files_size_in_mb,
     )
 
-    logging.info("Find all tasks")
+    logger.info("Find all tasks")
     unique_tasks = pd.concat([m.tasks for m in all_metadata]).index.unique()
     dst_meta.tasks = pd.DataFrame(
         {"task_index": range(len(unique_tasks))}, index=pd.Index(unique_tasks, name="task")
     )
 
-    meta_idx = {"chunk": 0, "file": 0}
-    data_idx = {"chunk": 0, "file": 0}
-    videos_idx = {
+    meta_idx: IndexState = {"chunk": 0, "file": 0}
+    data_idx: IndexState = {"chunk": 0, "file": 0}
+    videos_idx: VideoIndexState = {
         key: {"chunk": 0, "file": 0, "latest_duration": 0, "episode_duration": 0} for key in video_keys
     }
 
@@ -373,12 +400,17 @@ def aggregate_datasets(
         dst_meta.info.total_frames += src_meta.total_frames
 
     finalize_aggregation(dst_meta, all_metadata)
-    logging.info("Aggregation complete.")
+    logger.info("Aggregation complete.")
 
 
 def aggregate_videos(
-    src_meta, dst_meta, videos_idx, video_files_size_in_mb, chunk_size, concatenate_videos=True
-):
+    src_meta: LeRobotDatasetMetadata,
+    dst_meta: LeRobotDatasetMetadata,
+    videos_idx: VideoIndexState,
+    video_files_size_in_mb: float,
+    chunk_size: int,
+    concatenate_videos: bool = True,
+) -> VideoIndexState:
     """Aggregates video chunks from a source dataset into the destination dataset.
 
     Handles video file concatenation and rotation based on file size limits.
@@ -406,7 +438,8 @@ def aggregate_videos(
             videos_idx[key]["dst_file_durations"] = {}
 
     for key, video_idx in videos_idx.items():
-        unique_chunk_file_pairs = {
+        unique_chunk_file_pairs: list[ChunkFile] = sorted(
+            {
                 (chunk, file)
                 for chunk, file in zip(
                     src_meta.episodes[f"videos/{key}/chunk_index"],
@@ -414,7 +447,7 @@ def aggregate_videos(
                     strict=False,
                 )
             }
-        unique_chunk_file_pairs = sorted(unique_chunk_file_pairs)
+        )
 
         chunk_idx = video_idx["chunk"]
         file_idx = video_idx["file"]
@@ -489,7 +522,14 @@ def aggregate_videos(
     return videos_idx
 
 
-def aggregate_data(src_meta, dst_meta, data_idx, data_files_size_in_mb, chunk_size, concatenate_data=True):
+def aggregate_data(
+    src_meta: LeRobotDatasetMetadata,
+    dst_meta: LeRobotDatasetMetadata,
+    data_idx: IndexState,
+    data_files_size_in_mb: float,
+    chunk_size: int,
+    concatenate_data: bool = True,
+) -> IndexState:
     """Aggregates data chunks from a source dataset into the destination dataset.
 
     Reads source data files, updates indices to match the aggregated dataset,
@@ -510,14 +550,16 @@ def aggregate_data(src_meta, dst_meta, data_idx, data_files_size_in_mb, chunk_si
     Returns:
         dict: Updated data_idx with current chunk and file indices.
     """
-    unique_chunk_file_ids = {
+    unique_chunk_file_ids: list[ChunkFile] = sorted(
+        {
             (c, f)
             for c, f in zip(
-            src_meta.episodes["data/chunk_index"], src_meta.episodes["data/file_index"], strict=False
+                src_meta.episodes["data/chunk_index"],
+                src_meta.episodes["data/file_index"],
+                strict=False,
             )
         }
-
-    unique_chunk_file_ids = sorted(unique_chunk_file_ids)
+    )
     contains_images = len(dst_meta.image_keys) > 0
 
     # retrieve features schema for proper image typing in parquet
@@ -525,7 +567,7 @@ def aggregate_data(src_meta, dst_meta, data_idx, data_files_size_in_mb, chunk_si
 
     # Track source to destination file mapping for metadata update
     # This is critical for handling datasets that are already results of a merge
-    src_to_dst: dict[tuple[int, int], tuple[int, int]] = {}
+    src_to_dst: dict[ChunkFile, ChunkFile] = {}
 
     for src_chunk_idx, src_file_idx in unique_chunk_file_ids:
         src_path = src_meta.root / DEFAULT_DATA_PATH.format(
@@ -564,7 +606,13 @@ def aggregate_data(src_meta, dst_meta, data_idx, data_files_size_in_mb, chunk_si
     return data_idx
 
 
-def aggregate_metadata(src_meta, dst_meta, meta_idx, data_idx, videos_idx):
+def aggregate_metadata(
+    src_meta: LeRobotDatasetMetadata,
+    dst_meta: LeRobotDatasetMetadata,
+    meta_idx: IndexState,
+    data_idx: IndexState,
+    videos_idx: VideoIndexState,
+) -> IndexState:
     """Aggregates metadata from a source dataset into the destination dataset.
 
     Reads source metadata files, updates all indices and timestamps,
@@ -580,7 +628,8 @@ def aggregate_metadata(src_meta, dst_meta, meta_idx, data_idx, videos_idx):
     Returns:
         dict: Updated meta_idx with current chunk and file indices.
     """
-    chunk_file_ids = {
+    chunk_file_ids: list[ChunkFile] = sorted(
+        {
             (c, f)
             for c, f in zip(
                 src_meta.episodes["meta/episodes/chunk_index"],
@@ -588,8 +637,7 @@ def aggregate_metadata(src_meta, dst_meta, meta_idx, data_idx, videos_idx):
                 strict=False,
             )
         }
-
-    chunk_file_ids = sorted(chunk_file_ids)
+    )
     for chunk_idx, file_idx in chunk_file_ids:
         src_path = src_meta.root / DEFAULT_EPISODES_PATH.format(chunk_index=chunk_idx, file_index=file_idx)
         df = pd.read_parquet(src_path)
@@ -622,16 +670,16 @@ def aggregate_metadata(src_meta, dst_meta, meta_idx, data_idx, videos_idx):
 def append_or_create_parquet_file(
     df: pd.DataFrame,
     src_path: Path,
-    idx: dict[str, int],
+    idx: IndexState,
     max_mb: float,
     chunk_size: int,
     default_path: str,
     contains_images: bool = False,
-    aggr_root: Path = None,
+    aggr_root: Path | None = None,
     hf_features: datasets.Features | None = None,
     concatenate: bool = True,
     one_row_group_per_episode: bool = False,
-) -> tuple[dict[str, int], tuple[int, int]]:
+) -> tuple[IndexState, ChunkFile]:
     """Appends data to an existing parquet file or creates a new one based on size constraints.
 
     Manages file rotation when size limits are exceeded to prevent individual files
@@ -654,7 +702,13 @@ def append_or_create_parquet_file(
     Returns:
         tuple: (updated_idx, (dst_chunk, dst_file)) where updated_idx is the index dict
                and (dst_chunk, dst_file) is the actual destination file the data was written to.
+
+    Raises:
+        ValueError: If aggr_root is not provided.
     """
+    if aggr_root is None:
+        raise ValueError("aggr_root must be provided.")
+
     dst_chunk, dst_file = idx["chunk"], idx["file"]
     dst_path = aggr_root / default_path.format(chunk_index=dst_chunk, file_index=dst_file)
 
@@ -698,7 +752,9 @@ def append_or_create_parquet_file(
     return idx, (dst_chunk, dst_file)
 
 
-def finalize_aggregation(aggr_meta, all_metadata):
+def finalize_aggregation(
+    aggr_meta: LeRobotDatasetMetadata, all_metadata: list[LeRobotDatasetMetadata]
+) -> None:
     """Finalizes the dataset aggregation by writing summary files and statistics.
 
     Writes the tasks file, info file with total counts and splits, and
@@ -708,16 +764,16 @@ def finalize_aggregation(aggr_meta, all_metadata):
         aggr_meta: Aggregated dataset metadata.
         all_metadata: List of all source dataset metadata objects.
     """
-    logging.info("write tasks")
+    logger.info("write tasks")
     write_tasks(aggr_meta.tasks, aggr_meta.root)
 
-    logging.info("write info")
+    logger.info("write info")
     aggr_meta.info.total_tasks = len(aggr_meta.tasks)
     aggr_meta.info.total_episodes = sum(m.total_episodes for m in all_metadata)
     aggr_meta.info.total_frames = sum(m.total_frames for m in all_metadata)
     aggr_meta.info.splits = {"train": f"0:{sum(m.total_episodes for m in all_metadata)}"}
     write_info(aggr_meta.info, aggr_meta.root)
 
-    logging.info("write stats")
+    logger.info("write stats")
     aggr_meta.stats = aggregate_stats([m.stats for m in all_metadata])
     write_stats(aggr_meta.stats, aggr_meta.root)

src/lerobot/datasets/dataset_metadata.py

@@ -73,6 +73,8 @@ class LeRobotDatasetMetadata:
         revision: str | None = None,
         force_cache_sync: bool = False,
         metadata_buffer_size: int = 10,
+        *,
+        token: str | bool | None = None,
     ):
         """Load or download metadata for an existing LeRobot dataset.
 
@@ -94,6 +96,10 @@ class LeRobotDatasetMetadata:
                 even when local files exist.
             metadata_buffer_size: Number of episode metadata records to buffer
                 in memory before flushing to parquet.
+            token: Authentication token used for Hub requests. Pass a string
+                token, ``True`` to require the locally stored token, ``False``
+                to disable authentication, or ``None`` to use the Hugging Face
+                Hub default.
         """
         self.repo_id = repo_id
         self.revision = revision if revision else CODEBASE_VERSION
@@ -113,9 +119,12 @@ class LeRobotDatasetMetadata:
             self._load_metadata()
         except (FileNotFoundError, NotADirectoryError):
             if is_valid_version(self.revision):
+                if token is None:
                     self.revision = get_safe_version(self.repo_id, self.revision)
+                else:
+                    self.revision = get_safe_version(self.repo_id, self.revision, token=token)
 
-            self._pull_from_repo(allow_patterns="meta/")
+            self._pull_from_repo(allow_patterns="meta/", token=token)
             self._load_metadata()
 
     def _flush_metadata_buffer(self) -> None:
@@ -179,8 +188,8 @@ class LeRobotDatasetMetadata:
     def _load_metadata(self):
         self.info = load_info(self.root)
         check_version_compatibility(self.repo_id, self._version, CODEBASE_VERSION)
-        self.tasks = load_tasks(self.root)
-        self.episodes = load_episodes(self.root)
+        self.tasks = load_tasks(self.root) if self.total_tasks > 0 else None
+        self.episodes = load_episodes(self.root) if self.total_episodes > 0 else None
         self.stats = load_stats(self.root)
 
     def ensure_readable(self) -> None:
@@ -220,7 +229,10 @@ class LeRobotDatasetMetadata:
         self,
         allow_patterns: list[str] | str | None = None,
         ignore_patterns: list[str] | str | None = None,
+        *,
+        token: str | bool | None = None,
     ) -> None:
+        token_kwargs = {} if token is None else {"token": token}
         if self._requested_root is None:
             self.root = Path(
                 snapshot_download(
@@ -230,6 +242,7 @@ class LeRobotDatasetMetadata:
                     cache_dir=HF_LEROBOT_HUB_CACHE,
                     allow_patterns=allow_patterns,
                     ignore_patterns=ignore_patterns,
+                    **token_kwargs,
                 )
             )
             return
@@ -242,6 +255,7 @@ class LeRobotDatasetMetadata:
             local_dir=self._requested_root,
             allow_patterns=allow_patterns,
             ignore_patterns=ignore_patterns,
+            **token_kwargs,
         )
         self.root = self._requested_root
 

src/lerobot/datasets/dataset_writer.py

@@ -172,6 +172,23 @@ class DatasetWriter:
     def _get_image_file_dir(self, episode_index: int, image_key: str) -> Path:
         return self._get_image_file_path(episode_index, image_key, frame_index=0).parent
 
+    def _get_episode_buffer_index(self) -> int:
+        episode_index = self.episode_buffer["episode_index"]
+        # episode_index is `int` when freshly created, but becomes `np.ndarray` after
+        # save_episode() mutates the buffer. Handle both types here.
+        if isinstance(episode_index, np.ndarray):
+            episode_index = episode_index.item() if episode_index.size == 1 else episode_index[0]
+        return int(episode_index)
+
+    def _delete_camera_frame_dirs(self, camera_keys: list[str]) -> None:
+        if self.image_writer is not None:
+            self._wait_image_writer()
+        episode_index = self._get_episode_buffer_index()
+        for camera_key in camera_keys:
+            img_dir = self._get_image_file_dir(episode_index, camera_key)
+            if img_dir.is_dir():
+                shutil.rmtree(img_dir)
+
     def _save_image(
         self, image: torch.Tensor | np.ndarray | PIL.Image.Image, fpath: Path, compress_level: int = 1
     ) -> None:
@@ -369,7 +386,9 @@ class DatasetWriter:
                 self._episodes_since_last_encoding = 0
 
         if episode_data is None:
-            self.clear_episode_buffer(delete_images=len(self._meta.image_keys) > 0)
+            if len(self._meta.image_keys) > 0:
+                self._delete_camera_frame_dirs(self._meta.image_keys)
+            self.episode_buffer = self._create_episode_buffer()
 
     def _batch_save_episode_video(self, start_episode: int, end_episode: int | None = None) -> None:
         """Batch save videos for multiple episodes."""
@@ -561,10 +580,10 @@ class DatasetWriter:
         return metadata
 
     def clear_episode_buffer(self, delete_images: bool = True) -> None:
-        """Discard the current episode buffer and optionally delete temp images.
+        """Discard the current episode buffer and optionally delete temp camera frames.
 
         Args:
-            delete_images: If ``True``, remove temporary image directories
+            delete_images: If ``True``, remove temporary camera frame directories
                 written for the current episode.
         """
         # Cancel streaming encoder if active
@@ -572,17 +591,7 @@ class DatasetWriter:
             self._streaming_encoder.cancel_episode()
 
         if delete_images:
-            if self.image_writer is not None:
-                self._wait_image_writer()
-            episode_index = self.episode_buffer["episode_index"]
-            # episode_index is `int` when freshly created, but becomes `np.ndarray` after
-            # save_episode() mutates the buffer. Handle both types here.
-            if isinstance(episode_index, np.ndarray):
-                episode_index = episode_index.item() if episode_index.size == 1 else episode_index[0]
-            for cam_key in self._meta.image_keys:
-                img_dir = self._get_image_file_dir(episode_index, cam_key)
-                if img_dir.is_dir():
-                    shutil.rmtree(img_dir)
+            self._delete_camera_frame_dirs(self._meta.camera_keys)
 
         self.episode_buffer = self._create_episode_buffer()
 

src/lerobot/datasets/lerobot_dataset.py

@@ -65,6 +65,8 @@ class LeRobotDataset(torch.utils.data.Dataset):
         encoder_threads: int | None = None,
         streaming_encoding: bool = False,
         encoder_queue_maxsize: int = 30,
+        *,
+        token: str | bool | None = None,
     ):
         """
         2 modes are available for instantiating this class, depending on 2 different use cases:
@@ -197,6 +199,11 @@ class LeRobotDataset(torch.utils.data.Dataset):
                 instead of writing PNG images first. This makes save_episode() near-instant. Defaults to False.
             encoder_queue_maxsize (int, optional): Maximum number of frames to buffer per camera when using
                 streaming encoding. Defaults to 30 (~1s at 30fps).
+            token: Authentication token used while downloading this dataset
+                from the Hub. Pass a string token, ``True`` to require the
+                locally stored token, ``False`` to disable authentication, or
+                ``None`` to use the Hugging Face Hub default. The token is not
+                retained on the dataset instance after initialization.
 
         Note:
             Write-mode parameters (``streaming_encoding``, ``batch_encoding_size``) passed to
@@ -220,7 +227,11 @@ class LeRobotDataset(torch.utils.data.Dataset):
 
         # Load metadata (sets self.root once from the resolved metadata root)
         self.meta = LeRobotDatasetMetadata(
-            self.repo_id, self._requested_root, self.revision, force_cache_sync=force_cache_sync
+            self.repo_id,
+            self._requested_root,
+            self.revision,
+            force_cache_sync=force_cache_sync,
+            token=token,
         )
         self.root = self.meta.root
         self.revision = self.meta.revision
@@ -260,8 +271,11 @@ class LeRobotDataset(torch.utils.data.Dataset):
         # Load actual data
         if force_cache_sync or not self.reader.try_load():
             if is_valid_version(self.revision):
+                if token is None:
                     self.revision = get_safe_version(self.repo_id, self.revision)
-            self._download(download_videos)
+                else:
+                    self.revision = get_safe_version(self.repo_id, self.revision, token=token)
+            self._download(download_videos, token=token)
             self.reader.load_and_activate()
 
         # Detect write-mode params for backward compatibility
@@ -478,18 +492,19 @@ class LeRobotDataset(torch.utils.data.Dataset):
         """Return the number of frames in the selected episodes."""
         return self.num_frames
 
-    def __getitem__(self, idx) -> dict:
-        """Return a single frame by index, with all transforms applied.
+    def __getitem__(self, idx: int | slice) -> dict | list[dict]:
+        """Return one frame or a slice of frames, with all transforms applied.
 
         Loads the frame from the underlying HF dataset, expands delta-timestamp
         windows, decodes video frames, and applies image transforms. Delegates
-        the core logic to :meth:`DatasetReader.get_item`.
+        the core logic to :class:`DatasetReader`.
 
         Args:
-            idx: Index into the (possibly episode-filtered) dataset.
+            idx: Integer index or slice into the possibly episode-filtered dataset.
 
         Returns:
-            Dict mapping feature names to their tensor values for this frame.
+            A frame dictionary for an integer index, or a list of frame
+            dictionaries for a slice.
 
         Raises:
             RuntimeError: If the dataset is currently being recorded and
@@ -499,6 +514,9 @@ class LeRobotDataset(torch.utils.data.Dataset):
             raise RuntimeError(
                 "Cannot read from a dataset that is being recorded. Call finalize() first, then access items."
             )
+        if isinstance(idx, slice):
+            return [self[item_idx] for item_idx in range(*idx.indices(len(self)))]
+
         reader = self._ensure_reader()
         if reader.hf_dataset is None:
             # One-shot load after finalize()
@@ -622,10 +640,11 @@ class LeRobotDataset(torch.utils.data.Dataset):
                 hub_api.delete_tag(self.repo_id, tag=CODEBASE_VERSION, repo_type="dataset")
             hub_api.create_tag(self.repo_id, tag=CODEBASE_VERSION, revision=branch, repo_type="dataset")
 
-    def _download(self, download_videos: bool = True) -> None:
+    def _download(self, download_videos: bool = True, *, token: str | bool | None = None) -> None:
         """Downloads the dataset from the given 'repo_id' at the provided version."""
         ignore_patterns = None if download_videos else "videos/"
         files = None
+        token_kwargs = {} if token is None else {"token": token}
         if self.episodes is not None:
             # Reader is guaranteed to exist here (created in __init__ before _download)
             files = self.reader.get_episodes_file_paths()
@@ -639,6 +658,7 @@ class LeRobotDataset(torch.utils.data.Dataset):
                     cache_dir=HF_LEROBOT_HUB_CACHE,
                     allow_patterns=files,
                     ignore_patterns=ignore_patterns,
+                    **token_kwargs,
                 )
             )
         else:
@@ -650,6 +670,7 @@ class LeRobotDataset(torch.utils.data.Dataset):
                 local_dir=self._requested_root,
                 allow_patterns=files,
                 ignore_patterns=ignore_patterns,
+                **token_kwargs,
             )
             self.meta.root = self._requested_root
 
@@ -789,6 +810,8 @@ class LeRobotDataset(torch.utils.data.Dataset):
         image_writer_threads: int = 0,
         streaming_encoding: bool = False,
         encoder_queue_maxsize: int = 30,
+        *,
+        token: str | bool | None = None,
     ) -> "LeRobotDataset":
         """Resume recording on an existing dataset.
 
@@ -822,6 +845,8 @@ class LeRobotDataset(torch.utils.data.Dataset):
             streaming_encoding: If ``True``, encode video in real-time during
                 capture.
             encoder_queue_maxsize: Max buffered frames per camera for streaming.
+            token: Authentication token used if metadata must be downloaded
+                from the Hub. The token is not retained on the dataset instance.
 
         Returns:
             A :class:`LeRobotDataset` in write mode, ready to append episodes.
@@ -850,7 +875,11 @@ class LeRobotDataset(torch.utils.data.Dataset):
 
         # Load metadata (revision-safe when root is not provided)
         obj.meta = LeRobotDatasetMetadata(
-            obj.repo_id, obj._requested_root, obj.revision, force_cache_sync=force_cache_sync
+            obj.repo_id,
+            obj._requested_root,
+            obj.revision,
+            force_cache_sync=force_cache_sync,
+            token=token,
         )
 
         obj._encoder_threads = encoder_threads

src/lerobot/datasets/multi_dataset.py

@@ -48,6 +48,8 @@ class MultiLeRobotDataset(torch.utils.data.Dataset):
         tolerances_s: dict | None = None,
         download_videos: bool = True,
         video_backend: str | None = None,
+        *,
+        token: str | bool | None = None,
     ):
         super().__init__()
         self.repo_ids = repo_ids
@@ -65,6 +67,7 @@ class MultiLeRobotDataset(torch.utils.data.Dataset):
                 tolerance_s=self.tolerances_s[repo_id],
                 download_videos=download_videos,
                 video_backend=video_backend,
+                token=token,
             )
             for repo_id in repo_ids
         ]

src/lerobot/datasets/streaming_dataset.py

@@ -256,6 +256,8 @@ class StreamingLeRobotDataset(torch.utils.data.IterableDataset):
         shuffle: bool = True,
         return_uint8: bool = False,
         depth_output_unit: str = DEFAULT_DEPTH_UNIT,
+        *,
+        token: str | bool | None = None,
     ):
         """Initialize a StreamingLeRobotDataset.
 
@@ -278,6 +280,11 @@ class StreamingLeRobotDataset(torch.utils.data.IterableDataset):
             shuffle (bool, optional): Whether to shuffle the dataset across exhaustions. Defaults to True.
             depth_output_unit (str, optional): Physical unit depth maps are dequantized to ("m" or "mm").
                 Defaults to "mm".
+            token: Authentication token used while streaming this dataset from
+                the Hub. Pass a string token, ``True`` to require the locally
+                stored token, ``False`` to disable authentication, or ``None``
+                to use the Hugging Face Hub default. The token is not retained
+                on the dataset instance after initialization.
         """
         super().__init__()
         self.repo_id = repo_id
@@ -306,7 +313,11 @@ class StreamingLeRobotDataset(torch.utils.data.IterableDataset):
 
         # Load metadata
         self.meta = LeRobotDatasetMetadata(
-            self.repo_id, self._requested_root, self.revision, force_cache_sync=force_cache_sync
+            self.repo_id,
+            self._requested_root,
+            self.revision,
+            force_cache_sync=force_cache_sync,
+            token=token,
         )
         self.root = self.meta.root
         self.revision = self.meta.revision
@@ -334,12 +345,14 @@ class StreamingLeRobotDataset(torch.utils.data.IterableDataset):
             self.delta_timestamps = delta_timestamps
             self.delta_indices = get_delta_indices(self.delta_timestamps, self.fps)
 
+        token_kwargs = {} if token is None or self.streaming_from_local else {"token": token}
         self.hf_dataset: datasets.IterableDataset = load_dataset(
             self.repo_id if not self.streaming_from_local else str(self.root),
             split="train",
             streaming=self.streaming,
             data_files="data/*/*.parquet",
             revision=self.revision,
+            **token_kwargs,
         )
 
         self.num_shards = min(self.hf_dataset.num_shards, max_num_shards)

src/lerobot/datasets/utils.py

@@ -325,16 +325,19 @@ def check_version_compatibility(
         logging.warning(FUTURE_MESSAGE.format(repo_id=repo_id, version=v_check))
 
 
-def get_repo_versions(repo_id: str) -> list[packaging.version.Version]:
+def get_repo_versions(repo_id: str, *, token: str | bool | None = None) -> list[packaging.version.Version]:
     """Return available valid versions (branches and tags) on a given Hub repo.
 
     Args:
         repo_id (str): The repository ID on the Hugging Face Hub.
+        token: Authentication token used for Hub requests. Pass a string token,
+            ``True`` to require the locally stored token, ``False`` to disable
+            authentication, or ``None`` to use the Hugging Face Hub default.
 
     Returns:
         list[packaging.version.Version]: A list of valid versions found.
     """
-    api = HfApi()
+    api = HfApi() if token is None else HfApi(token=token)
     repo_refs = api.list_repo_refs(repo_id, repo_type="dataset")
     repo_refs = [b.name for b in repo_refs.branches + repo_refs.tags]
     repo_versions = []
@@ -345,7 +348,12 @@ def get_repo_versions(repo_id: str) -> list[packaging.version.Version]:
     return repo_versions
 
 
-def get_safe_version(repo_id: str, version: str | packaging.version.Version) -> str:
+def get_safe_version(
+    repo_id: str,
+    version: str | packaging.version.Version,
+    *,
+    token: str | bool | None = None,
+) -> str:
     """Return the specified version if available on repo, or the latest compatible one.
 
     If the exact version is not found, it looks for the latest version with the
@@ -354,6 +362,7 @@ def get_safe_version(repo_id: str, version: str | packaging.version.Version) ->
     Args:
         repo_id (str): The repository ID on the Hugging Face Hub.
         version (str | packaging.version.Version): The target version.
+        token: Authentication token forwarded to the Hub version lookup.
 
     Returns:
         str: The safe version string (e.g., "v1.2.3") to use as a revision.
@@ -366,7 +375,7 @@ def get_safe_version(repo_id: str, version: str | packaging.version.Version) ->
     target_version = (
         packaging.version.parse(version) if not isinstance(version, packaging.version.Version) else version
     )
-    hub_versions = get_repo_versions(repo_id)
+    hub_versions = get_repo_versions(repo_id) if token is None else get_repo_versions(repo_id, token=token)
 
     if not hub_versions:
         raise RevisionNotFoundError(

src/lerobot/envs/configs.py

@@ -322,7 +322,7 @@ class HILSerlRobotEnvConfig(EnvConfig):
 class LiberoEnv(EnvConfig):
     task: str = "libero_10"  # can also choose libero_spatial, libero_object, etc.
     task_ids: list[int] | None = None
-    fps: int = 30
+    fps: int = 20  # Must match robosuite's default control_freq (20 Hz)
     episode_length: int | None = None
     obs_type: str = "pixels_agent_pos"
     render_mode: str = "rgb_array"
@@ -354,6 +354,9 @@ class LiberoEnv(EnvConfig):
     control_mode: str = "relative"  # or "absolute"
 
     def __post_init__(self):
+        if self.fps <= 0:
+            raise ValueError(f"fps must be positive, got {self.fps}")
+
         if self.obs_type == "pixels":
             self.features[LIBERO_KEY_PIXELS_AGENTVIEW] = PolicyFeature(
                 type=FeatureType.VISUAL, shape=(self.observation_height, self.observation_width, 3)
@@ -412,6 +415,7 @@ class LiberoEnv(EnvConfig):
             "render_mode": self.render_mode,
             "observation_height": self.observation_height,
             "observation_width": self.observation_width,
+            "control_freq": self.fps,
         }
         if self.task_ids is not None:
             kwargs["task_ids"] = self.task_ids

src/lerobot/envs/libero.py

@@ -125,10 +125,13 @@ class LiberoEnv(gym.Env):
         n_envs: int = 1,
         camera_name_mapping: dict[str, str] | None = None,
         num_steps_wait: int = 10,
+        control_freq: int = 20,
         control_mode: str = "relative",
         is_libero_plus: bool = False,
     ):
         super().__init__()
+        if control_freq <= 0:
+            raise ValueError(f"control_freq must be positive, got {control_freq}")
         self.task_id = task_id
         self.is_libero_plus = is_libero_plus
         self.obs_type = obs_type
@@ -154,6 +157,7 @@ class LiberoEnv(gym.Env):
             }
         self.camera_name_mapping = camera_name_mapping
         self.num_steps_wait = num_steps_wait
+        self.control_freq = control_freq
         self.episode_index = episode_index
         self.episode_length = episode_length
         # Load once and keep
@@ -260,6 +264,7 @@ class LiberoEnv(gym.Env):
             bddl_file_name=self._task_bddl_file,
             camera_heights=self.observation_height,
             camera_widths=self.observation_width,
+            control_freq=self.control_freq,
         )
         env.reset()
         self._env = env
@@ -379,7 +384,12 @@ class LiberoEnv(gym.Env):
 
     def close(self):
         if self._env is not None:
+            try:
                 self._env.close()
+            finally:
+                # LIBERO deletes its inner env on close, so this wrapper must
+                # be recreated before the next reset.
+                self._env = None
 
 
 def _make_env_fns(

src/lerobot/envs/metaworld.py

@@ -155,6 +155,7 @@ class MetaworldEnv(gym.Env):
             env.model.cam_pos[2] = [0.75, 0.075, 0.7]
         env.reset()
         env._freeze_rand_vec = False  # otherwise no randomization
+        env.seeded_rand_vec = True  # use seeded RNG so reset(seed=X) controls object positions
         self._env = env
 
     def render(self) -> np.ndarray:
@@ -220,6 +221,8 @@ class MetaworldEnv(gym.Env):
         self._ensure_env()
         super().reset(seed=seed)
 
+        if seed is not None:
+            self._env.seed(seed)
         raw_obs, info = self._env.reset(seed=seed)
 
         observation = self._format_raw_obs(raw_obs)

src/lerobot/envs/robotwin.py

@@ -384,7 +384,9 @@ class RoboTwinEnv(gym.Env):
 
         self._env: Any | None = None  # deferred — created on first reset() inside worker
         self._step_count: int = 0
-        self._black_frame = np.zeros((self.observation_height, self.observation_width, 3), dtype=np.uint8)
+        self._black_frame: np.ndarray = np.zeros(
+            (self.observation_height, self.observation_width, 3), dtype=np.uint8
+        )
 
         image_spaces = {
             cam: spaces.Box(

src/lerobot/envs/vlabench.py

@@ -373,7 +373,7 @@ class VLABenchEnv(gym.Env):
 
         if action.shape[0] != 7:
             # Unknown layout — fall back to zero-pad so the sim doesn't crash.
-            padded = np.zeros(ctrl_dim, dtype=np.float64)
+            padded: np.ndarray = np.zeros(ctrl_dim, dtype=np.float64)
             padded[: min(action.shape[0], ctrl_dim)] = action[:ctrl_dim]
             return padded
 

src/lerobot/jobs/__init__.py

@@ -18,6 +18,7 @@ from lerobot.utils.import_utils import require_package
 # guard the optional dependency here so importing this package fails loudly if it's missing.
 require_package("datasets", extra="dataset")
 
+from .annotate import submit_annotate_to_hf
 from .hf import submit_to_hf
 
-__all__ = ["submit_to_hf"]
+__all__ = ["submit_annotate_to_hf", "submit_to_hf"]

src/lerobot/jobs/annotate.py

@@ -0,0 +1,176 @@
+# Copyright 2026 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+"""Run ``lerobot-annotate`` on HF Jobs (HuggingFace GPUs).
+
+Same shape as the training submitter in ``hf.py``, with one difference: the
+annotation pipeline serves its own VLM, so the pod starts from the official
+``vllm/vllm-openai`` image (which has no lerobot) instead of the prebuilt
+``lerobot-gpu`` image, and installs lerobot on top before running.
+
+Because there is no config repo to stage, the pod replays the user's own CLI
+flags — everything except the client-only ``--job.*`` and the host-local
+``--root``, which is replaced by ``--repo_id`` so the pod pulls the dataset
+from the Hub.
+"""
+
+from __future__ import annotations
+
+import shlex
+import sys
+from dataclasses import is_dataclass
+from typing import TYPE_CHECKING
+
+from huggingface_hub import HfApi, get_token, run_job
+
+from .dataset import ensure_dataset_available
+
+# Package-internal reuse of the training submitter's job plumbing: following a
+# submitted job and forwarding argv are identical for annotation runs.
+from .hf import _pod_forwarded_args, follow_job, resolve_job_tags
+
+if TYPE_CHECKING:
+    from lerobot.annotations.steerable_pipeline.config import AnnotationPipelineConfig
+
+LEROBOT_GIT_URL = "https://github.com/huggingface/lerobot.git"
+
+# Mirrors the pins in pyproject.toml. The vLLM image resolves dependencies on its
+# own otherwise, and pulls av 18 / datasets 5 / draccus 0.11 — each of which breaks
+# lerobot at import time. `--upgrade-strategy only-if-needed` keeps vLLM's own
+# (torch, transformers, ...) pins intact.
+_RUNTIME_REQUIREMENTS = (
+    "'datasets>=4.7.0,<5.0.0' 'pyarrow>=21.0.0,<30.0.0' 'av>=15.0.0,<16.0.0' 'draccus==0.10.0' "
+    "'pandas>=2.0.0,<3.0.0' jsonlines gymnasium torchcodec mergedeep pyyaml-include toml typing-inspect "
+    "openai"
+)
+
+# Flags the submitter resolves itself instead of forwarding verbatim: `--root`
+# names a directory only this machine has, `--repo_id` is re-emitted from the
+# config, and the config-file args name local files (rejected up front by
+# `submit_annotate_to_hf`). `--job.*` is dropped separately, by prefix; bare
+# `--job` is not, hence its entry here — it is the one arg that could smuggle a
+# remote `target` onto the pod and have the job recursively submit itself.
+_SUBMITTER_OWNED_ARGS = ("--root", "--repo_id", "--config_path", "--job")
+
+
+def _local_config_file_args(cfg: AnnotationPipelineConfig) -> list[str]:
+    """The CLI args that name a config file on the client's disk.
+
+    draccus exposes ``--config_path`` for the whole config plus a ``--<field>``
+    for every nested dataclass (``--vlm``, ``--plan``, ``--job``, ...). The pod has
+    none of those files, so a remote run has to reject them rather than silently
+    drop the settings they carry.
+    """
+    return ["--config_path", *(f"--{name}" for name in vars(cfg) if is_dataclass(getattr(cfg, name)))]
+
+
+def build_pod_setup(lerobot_ref: str) -> str:
+    """Shell prelude that turns the vLLM image into a ``lerobot-annotate`` runtime."""
+    spec = f"lerobot @ git+{LEROBOT_GIT_URL}@{lerobot_ref}"
+    return (
+        # git to install from the repo, ffmpeg to decode the dataset's videos.
+        "apt-get update -qq && apt-get install -y -qq git ffmpeg && "
+        f"pip install --no-deps {shlex.quote(spec)} && "
+        f"pip install --upgrade-strategy only-if-needed {_RUNTIME_REQUIREMENTS} && "
+        # vLLM's cudagraph memory estimate over-reserves and starves the KV cache;
+        # PyAV is the video backend the server can decode our frames with.
+        "export VLLM_MEMORY_PROFILER_ESTIMATE_CUDAGRAPHS=0 && "
+        "export VLLM_VIDEO_BACKEND=pyav"
+    )
+
+
+def build_pod_command(repo_id: str, lerobot_ref: str, argv: list[str]) -> list[str]:
+    """Build the ``bash -c`` command the pod runs: setup prelude, then annotation.
+
+    ``argv`` is the user's CLI (``sys.argv[1:]``) minus the flags in
+    ``_SUBMITTER_OWNED_ARGS``; ``--repo_id`` is re-added from the config so the pod
+    always annotates the dataset we just made sure is reachable on the Hub.
+    ``--job.target=local`` stops the pod from re-dispatching to itself.
+    """
+    forwarded = _pod_forwarded_args(argv, drop_names=_SUBMITTER_OWNED_ARGS, drop_prefixes=("--job.",))
+    annotate = shlex.join(["lerobot-annotate", f"--repo_id={repo_id}", *forwarded, "--job.target=local"])
+    return ["bash", "-c", f"{build_pod_setup(lerobot_ref)} && {annotate}"]
+
+
+def submit_annotate_to_hf(cfg: AnnotationPipelineConfig) -> None:
+    """Submit an annotation run to HF Jobs infrastructure.
+
+    Resolves credentials, makes sure the source dataset is reachable from the pod,
+    submits the job, then tails its logs until the job reaches a terminal stage —
+    or returns immediately with ``--job.detach``. Ctrl-C detaches without
+    cancelling the remote job.
+    """
+    token = get_token()
+    if not token:
+        raise RuntimeError("Not logged in to Hugging Face. Run `hf auth login` first.")
+
+    if cfg.repo_id is None:
+        raise ValueError(
+            "Remote annotation requires --repo_id: the pod downloads the dataset from the Hub, "
+            "and --root only names a directory on this machine."
+        )
+
+    argv = sys.argv[1:]
+    passed = {tok.split("=", 1)[0] for tok in argv}
+    used_config_files = sorted(passed.intersection(_local_config_file_args(cfg)))
+    if used_config_files:
+        raise ValueError(
+            f"{', '.join(used_config_files)} cannot be used with a remote --job.target: the pod "
+            "cannot read config files from this machine. Pass the settings as CLI flags instead."
+        )
+
+    if not cfg.push_to_hub:
+        # The pod's filesystem is discarded when the job ends, so without a push the
+        # run produces nothing. Warn rather than fail: a smoke test over
+        # --only_episodes that only inspects the logs is a legitimate use.
+        print(
+            "WARNING: --push_to_hub is off. The annotated dataset lives only on the pod and is "
+            "discarded when the job ends. Pass --push_to_hub=true to keep the result."
+        )
+
+    api = HfApi(token=token)
+    tags = resolve_job_tags(cfg.job.tags)
+    ensure_dataset_available(cfg.repo_id, api=api, tags=tags)
+
+    command = build_pod_command(cfg.repo_id, cfg.job.lerobot_ref, argv)
+
+    print(f"Submitting job to HF Jobs (flavor={cfg.job.target}, image={cfg.job.image}) ...")
+    job_info = run_job(
+        image=cfg.job.image,
+        command=command,
+        flavor=cfg.job.target,
+        secrets={"HF_TOKEN": token},
+        timeout=cfg.job.timeout,
+        # HF Jobs labels are key/value; expose each tag as a queryable label.
+        labels=dict.fromkeys(tags, "true"),
+    )
+    job_id = job_info.id
+    job_url = getattr(job_info, "url", None)
+    print(f"Job submitted: {job_id}")
+    if job_url:
+        print(f"  Job page:     {job_url}")
+    target_repo_id = cfg.new_repo_id or cfg.repo_id
+    if cfg.push_to_hub:
+        print(f"  Dataset repo: https://huggingface.co/datasets/{target_repo_id}")
+    print(f"  Monitor:      hf jobs logs {job_id}")
+    print(f"  Cancel:       hf jobs cancel {job_id}")
+
+    # No success marker: `lerobot-annotate` keeps working after the upload log line
+    # (dataset card, version tag), so completion has to be stage-based.
+    if not follow_job(job_id, detach=cfg.job.detach):
+        return
+
+    if cfg.push_to_hub:
+        print(f"\nAnnotation complete — dataset pushed to https://huggingface.co/datasets/{target_repo_id}")
+    else:
+        print("\nAnnotation complete. Note: --push_to_hub was off, so the result stayed on the pod.")

src/lerobot/jobs/hf.py

@@ -223,6 +223,74 @@ def _poll_until_done(
     return None
 
 
+def follow_job(job_id: str, *, detach: bool = False, success_marker: str | None = None) -> bool:
+    """Watch a submitted job to the end, streaming its logs to stdout.
+
+    Returns True when the job finished successfully and False when we stopped watching
+    without a verdict — `detach`, or the user pressing Ctrl-C, which detaches rather than
+    cancelling the remote job. Raises RuntimeError when the job reaches a terminal stage
+    other than COMPLETED.
+
+    `success_marker` finishes as soon as that string appears in the logs instead of waiting
+    out the platform's post-run finalization (~30s). Callers that have a log line meaning
+    "the artifact is on the Hub" should pass it; without one, completion is stage-based.
+    """
+    if detach:
+        return False
+
+    done = threading.Event()
+    detached = threading.Event()
+    marker_seen = threading.Event()
+    stage_holder: dict[str, str | None] = {}
+
+    def _poll() -> None:
+        stage_holder["stage"] = _poll_until_done(job_id, done, status_holder=stage_holder)
+
+    poll_thread = threading.Thread(target=_poll, daemon=True)
+    poll_thread.start()
+    log_thread = threading.Thread(
+        target=_tail_logs, args=(job_id, done, success_marker, marker_seen), daemon=True
+    )
+    log_thread.start()
+
+    def _detach(sig, frame):
+        detached.set()
+        done.set()
+        print("\nDetached. Job is still running.")
+        print(f"  Monitor: hf jobs logs {job_id}")
+        print(f"  Cancel:  hf jobs cancel {job_id}")
+
+    # signal.signal only works on the main thread; when called from a worker thread
+    # (e.g. an orchestration framework) skip the Ctrl-C-detaches-instead-of-cancels
+    # handler rather than crashing with ValueError.
+    install_sigint = threading.current_thread() is threading.main_thread()
+    original_sigint = signal.getsignal(signal.SIGINT) if install_sigint else None
+    if install_sigint:
+        signal.signal(signal.SIGINT, _detach)
+    try:
+        # Timeout-based join so SIGINT is delivered to the main thread promptly.
+        while poll_thread.is_alive():
+            poll_thread.join(timeout=0.5)
+        log_thread.join(timeout=5)
+    finally:
+        if install_sigint:
+            signal.signal(signal.SIGINT, original_sigint)
+
+    if detached.is_set():
+        return False
+    if marker_seen.is_set():
+        return True
+
+    stage = stage_holder.get("stage")
+    if stage != "COMPLETED":
+        message = stage_holder.get("message")
+        detail = f" ({message})" if message else ""
+        raise RuntimeError(
+            f"Job {job_id} ended with stage={stage}{detail}. Check logs: hf jobs logs {job_id}"
+        )
+    return True
+
+
 def _pod_forwarded_args(
     argv: list[str], drop_names: tuple[str, ...] = (), drop_prefixes: tuple[str, ...] = ()
 ) -> list[str]:
@@ -362,64 +430,11 @@ def submit_to_hf(cfg: TrainPipelineConfig) -> None:
     print(f"  Monitor:    hf jobs logs {job_id}")
     print(f"  Cancel:     hf jobs cancel {job_id}")
 
-    if cfg.job.detach:
-        return
-
-    done = threading.Event()
-    detached = threading.Event()
-    pushed_ok = threading.Event()
-    stage_holder: dict[str, str | None] = {}
-
-    def _poll() -> None:
-        stage_holder["stage"] = _poll_until_done(job_id, done, status_holder=stage_holder)
-
-    poll_thread = threading.Thread(target=_poll, daemon=True)
-    poll_thread.start()
     # Finish as soon as the model is pushed, rather than waiting out the platform's
     # post-run finalization before the job stage flips to COMPLETED. This matches the
     # exact log line emitted by PreTrainedPolicy.push_model_to_hub — the two must stay
     # in sync. If it ever stops matching we just fall back to stage-based completion
     # (~30s slower), so the contract is an optimization, not a correctness requirement.
     success_marker = f"Model pushed to https://huggingface.co/{repo_id}"
-    log_thread = threading.Thread(
-        target=_tail_logs, args=(job_id, done, success_marker, pushed_ok), daemon=True
-    )
-    log_thread.start()
-
-    def _detach(sig, frame):
-        detached.set()
-        done.set()
-        print("\nDetached. Job is still running.")
-        print(f"  Monitor: hf jobs logs {job_id}")
-        print(f"  Cancel:  hf jobs cancel {job_id}")
-
-    # signal.signal only works on the main thread; when called from a worker thread
-    # (e.g. an orchestration framework) skip the Ctrl-C-detaches-instead-of-cancels
-    # handler rather than crashing with ValueError.
-    install_sigint = threading.current_thread() is threading.main_thread()
-    original_sigint = signal.getsignal(signal.SIGINT) if install_sigint else None
-    if install_sigint:
-        signal.signal(signal.SIGINT, _detach)
-    try:
-        # Timeout-based join so SIGINT is delivered to the main thread promptly.
-        while poll_thread.is_alive():
-            poll_thread.join(timeout=0.5)
-        log_thread.join(timeout=5)
-    finally:
-        if install_sigint:
-            signal.signal(signal.SIGINT, original_sigint)
-
-    if detached.is_set():
-        return
-
-    if pushed_ok.is_set():
+    if follow_job(job_id, detach=cfg.job.detach, success_marker=success_marker):
         print(f"\nTraining complete — model pushed to https://huggingface.co/{repo_id}")
-        return
-
-    stage = stage_holder.get("stage")
-    if stage != "COMPLETED":
-        message = stage_holder.get("message")
-        detail = f" ({message})" if message else ""
-        raise RuntimeError(
-            f"Job {job_id} ended with stage={stage}{detail}. Check logs: hf jobs logs {job_id}"
-        )

src/lerobot/motors/damiao/damiao.py

@@ -20,7 +20,6 @@ import logging
 import time
 from contextlib import contextmanager
 from copy import deepcopy
-from functools import cached_property
 from typing import TYPE_CHECKING, Any, TypedDict
 
 from lerobot.utils.decorators import check_if_already_connected, check_if_not_connected
@@ -854,7 +853,7 @@ class DamiaoMotorsBus(MotorsBusBase):
         else:
             raise ValueError(f"Motor {motor_obj} doesn't have a valid recv_id (None).")
 
-    @cached_property
+    @property
     def is_calibrated(self) -> bool:
         """Check if motors are calibrated."""
         return bool(self.calibration)

src/lerobot/motors/dynamixel/tables.py

@@ -122,6 +122,9 @@ MODEL_ENCODING_TABLE = {
     "xm430-w350": X_SERIES_ENCODINGS_TABLE,
     "xm540-w270": X_SERIES_ENCODINGS_TABLE,
     "xc430-w150": X_SERIES_ENCODINGS_TABLE,
+    "xh540-w150": X_SERIES_ENCODINGS_TABLE,
+    "xc330-t288": X_SERIES_ENCODINGS_TABLE,
+    "xc330-t181": X_SERIES_ENCODINGS_TABLE,
 }
 
 # {model: model_resolution}
@@ -134,6 +137,9 @@ MODEL_RESOLUTION = {
     "xm430-w350": 4096,
     "xm540-w270": 4096,
     "xc430-w150": 4096,
+    "xh540-w150": 4096,
+    "xc330-t288": 4096,
+    "xc330-t181": 4096,
 }
 
 # {model: model_number}
@@ -145,6 +151,9 @@ MODEL_NUMBER_TABLE = {
     "xm430-w350": 1020,
     "xm540-w270": 1120,
     "xc430-w150": 1070,
+    "xh540-w150": 1110,
+    "xc330-t288": 1220,
+    "xc330-t181": 1210,
 }
 
 # {model: available_operating_modes}
@@ -156,6 +165,9 @@ MODEL_OPERATING_MODES = {
     "xm430-w350": [0, 1, 3, 4, 5, 16],
     "xm540-w270": [0, 1, 3, 4, 5, 16],
     "xc430-w150": [1, 3, 4, 16],
+    "xh540-w150": [0, 1, 3, 4, 5, 16],
+    "xc330-t288": [0, 1, 3, 4, 5, 16],
+    "xc330-t181": [0, 1, 3, 4, 5, 16],
 }
 
 MODEL_CONTROL_TABLE = {
@@ -166,6 +178,9 @@ MODEL_CONTROL_TABLE = {
     "xm430-w350": X_SERIES_CONTROL_TABLE,
     "xm540-w270": X_SERIES_CONTROL_TABLE,
     "xc430-w150": X_SERIES_CONTROL_TABLE,
+    "xh540-w150": X_SERIES_CONTROL_TABLE,
+    "xc330-t288": X_SERIES_CONTROL_TABLE,
+    "xc330-t181": X_SERIES_CONTROL_TABLE,
 }
 
 MODEL_BAUDRATE_TABLE = {
@@ -176,6 +191,9 @@ MODEL_BAUDRATE_TABLE = {
     "xm430-w350": X_SERIES_BAUDRATE_TABLE,
     "xm540-w270": X_SERIES_BAUDRATE_TABLE,
     "xc430-w150": X_SERIES_BAUDRATE_TABLE,
+    "xh540-w150": X_SERIES_BAUDRATE_TABLE,
+    "xc330-t288": X_SERIES_BAUDRATE_TABLE,
+    "xc330-t181": X_SERIES_BAUDRATE_TABLE,
 }
 
 AVAILABLE_BAUDRATES = [

src/lerobot/motors/motors_bus.py

@@ -23,6 +23,7 @@ from __future__ import annotations
 
 import abc
 import logging
+import time
 from collections.abc import Sequence
 from contextlib import contextmanager
 from dataclasses import dataclass
@@ -818,13 +819,13 @@ class SerialMotorsBus(MotorsBusBase):
         """
         motor_names = self._get_motors_list(motors)
 
-        start_positions = self.sync_read("Present_Position", motor_names, normalize=False)
+        start_positions = self.sync_read("Present_Position", motor_names, normalize=False, num_retry=5)
         mins = start_positions.copy()
         maxes = start_positions.copy()
 
         user_pressed_enter = False
         while not user_pressed_enter:
-            positions = self.sync_read("Present_Position", motor_names, normalize=False)
+            positions = self.sync_read("Present_Position", motor_names, normalize=False, num_retry=5)
             mins = {motor: min(positions[motor], min_) for motor, min_ in mins.items()}
             maxes = {motor: max(positions[motor], max_) for motor, max_ in maxes.items()}
 
@@ -837,9 +838,12 @@ class SerialMotorsBus(MotorsBusBase):
             if enter_pressed():
                 user_pressed_enter = True
 
-            if display_values and not user_pressed_enter:
+            if not user_pressed_enter:
+                if display_values:
                     # Move cursor up to overwrite the previous output
                     move_cursor_up(len(motor_names) + 3)
+                # Throttle reads even when the live table is disabled.
+                time.sleep(0.02)
 
         same_min_max = [motor for motor in motor_names if mins[motor] == maxes[motor]]
         if same_min_max:

src/lerobot/policies/diffusion/configuration_diffusion.py

@@ -79,6 +79,8 @@ class DiffusionConfig(PreTrainedConfig):
         use_film_scale_modulation: FiLM (https://huggingface.co/papers/1709.07871) is used for the Unet conditioning.
             Bias modulation is used be default, while this parameter indicates whether to also use scale
             modulation.
+        gradient_checkpointing: Whether to checkpoint the Unet residual blocks during training. This reduces
+            activation memory at the cost of recomputing those blocks during the backward pass.
         noise_scheduler_type: Name of the noise scheduler to use. Supported options: ["DDPM", "DDIM"].
         num_train_timesteps: Number of diffusion steps for the forward diffusion schedule.
         beta_schedule: Name of the diffusion beta schedule as per DDPMScheduler from Hugging Face diffusers.
@@ -132,6 +134,7 @@ class DiffusionConfig(PreTrainedConfig):
     n_groups: int = 8
     diffusion_step_embed_dim: int = 128
     use_film_scale_modulation: bool = True
+    gradient_checkpointing: bool = False
     # Noise scheduler.
     noise_scheduler_type: str = "DDPM"
     num_train_timesteps: int = 100

src/lerobot/policies/diffusion/modeling_diffusion.py

@@ -31,6 +31,7 @@ import torch
 import torch.nn.functional as F  # noqa: N812
 import torchvision
 from torch import Tensor, nn
+from torch.utils.checkpoint import checkpoint
 
 from lerobot.utils.constants import ACTION, OBS_ENV_STATE, OBS_IMAGES, OBS_STATE
 from lerobot.utils.import_utils import _diffusers_available, require_package
@@ -727,20 +728,33 @@ class DiffusionConditionalUnet1d(nn.Module):
         else:
             global_feature = timesteps_embed
 
+        use_gc = self.config.gradient_checkpointing and self.training
+
         # Run encoder, keeping track of skip features to pass to the decoder.
         encoder_skip_features: list[Tensor] = []
         for resnet, resnet2, downsample in self.down_modules:
+            if use_gc:
+                x = checkpoint(resnet, x, global_feature, use_reentrant=False)
+                x = checkpoint(resnet2, x, global_feature, use_reentrant=False)
+            else:
                 x = resnet(x, global_feature)
                 x = resnet2(x, global_feature)
             encoder_skip_features.append(x)
             x = downsample(x)
 
         for mid_module in self.mid_modules:
+            if use_gc:
+                x = checkpoint(mid_module, x, global_feature, use_reentrant=False)
+            else:
                 x = mid_module(x, global_feature)
 
         # Run decoder, using the skip features from the encoder.
         for resnet, resnet2, upsample in self.up_modules:
             x = torch.cat((x, encoder_skip_features.pop()), dim=1)
+            if use_gc:
+                x = checkpoint(resnet, x, global_feature, use_reentrant=False)
+                x = checkpoint(resnet2, x, global_feature, use_reentrant=False)
+            else:
                 x = resnet(x, global_feature)
                 x = resnet2(x, global_feature)
             x = upsample(x)

src/lerobot/policies/evo1/processor_evo1.py

@@ -302,6 +302,33 @@ def _pad_evo1_stats(
     return padded_stats
 
 
+def _refresh_evo1_normalization_steps(
+    config: Evo1Config,
+    preprocessor: PolicyProcessorPipeline,
+    postprocessor: PolicyProcessorPipeline,
+) -> None:
+    """Re-pad checkpoint-loaded (un)normalizer stats/features to EVO1's fixed widths.
+
+    Loading a checkpoint injects the raw dataset stats (unpadded to max_state_dim/max_action_dim)
+    into the (un)normalizer via the generic override path in make_pre_post_processors. Those stats
+    and their declared features must be re-padded/reshaped to EVO1's fixed widths, otherwise
+    normalization fails against the padded state/action tensors (e.g. state padded to 24 vs. 8-dim
+    LIBERO stats). Padding is a no-op when stats are already at the target width.
+    """
+    normalization_features = _evo1_normalization_features(config)
+    action_features = _evo1_action_features(config)
+    for step in preprocessor.steps:
+        if isinstance(step, NormalizerProcessorStep):
+            step.features = normalization_features
+            step.stats = _pad_evo1_stats(config, step.stats)
+            step.to(device=step.device, dtype=step.dtype)
+    for step in postprocessor.steps:
+        if isinstance(step, UnnormalizerProcessorStep):
+            step.features = action_features
+            step.stats = _pad_evo1_stats(config, step.stats)
+            step.to(device=step.device, dtype=step.dtype)
+
+
 def reconcile_evo1_processors(
     config: Evo1Config,
     preprocessor: PolicyProcessorPipeline,
@@ -309,16 +336,19 @@ def reconcile_evo1_processors(
 ) -> tuple[PolicyProcessorPipeline, PolicyProcessorPipeline]:
     """Reconcile checkpoint-loaded pipelines with the current EVO1 config.
 
-    Two things cannot be restored from a serialized pipeline alone: the EVO1 batch converter
-    (converters are plain functions and are never serialized), and eval-time CLI overrides of the
-    action postprocessing flags (`postprocess_action_dim`, `binarize_gripper`, `gripper_*`). This
-    restores the converter and rebuilds the action step from the current config so those overrides
-    take effect.
+    Three things cannot be restored from a serialized pipeline alone: the EVO1 batch converter
+    (converters are plain functions and are never serialized), eval-time CLI overrides of the
+    action postprocessing flags (`postprocess_action_dim`, `binarize_gripper`, `gripper_*`), and the
+    (un)normalizer stats/features when the generic override path injects raw, unpadded dataset
+    stats. This restores the converter, re-pads the normalization stats to EVO1's fixed widths, and
+    rebuilds the action step from the current config so those overrides take effect.
     """
     # Pipelines reloaded from a checkpoint come back with the default batch converter, which drops
     # non-observation extras (embodiment_id, state_mask, custom task fields) needed by EVO1.
     preprocessor.to_transition = evo1_batch_to_transition
 
+    _refresh_evo1_normalization_steps(config, preprocessor, postprocessor)
+
     action_step = Evo1ActionProcessorStep(
         action_dim=_evo1_action_dim(config),
         binarize_gripper=config.binarize_gripper,

src/lerobot/policies/factory.py

@@ -44,12 +44,19 @@ from lerobot.utils.constants import (
     POLICY_PREPROCESSOR_DEFAULT_NAME,
 )
 from lerobot.utils.feature_utils import dataset_to_policy_features
+from lerobot.utils.import_utils import _peft_available, require_package
 
 from .evo1.configuration_evo1 import Evo1Config
 from .groot.configuration_groot import GrootConfig
 from .pretrained import PreTrainedPolicy
 from .utils import validate_visual_features_consistency
 
+if TYPE_CHECKING or _peft_available:
+    from peft import PeftConfig, PeftModel
+else:
+    PeftConfig = None
+    PeftModel = None
+
 
 def _reconnect_relative_absolute_steps(
     preprocessor: PolicyProcessorPipeline, postprocessor: PolicyProcessorPipeline
@@ -177,6 +184,7 @@ def make_pre_post_processors(
             return make_groot_pre_post_processors_from_pretrained(
                 config=policy_cfg,
                 pretrained_path=pretrained_path,
+                revision=pretrained_revision,
                 dataset_stats=kwargs.get("dataset_stats"),
                 dataset_meta=kwargs.get("dataset_meta"),
                 preprocessor_overrides=kwargs.get("preprocessor_overrides"),
@@ -333,12 +341,15 @@ def make_policy(
         # Load a pretrained PEFT model on top of the policy. The pretrained path points to the folder/repo
         # of the adapter and the adapter's config contains the path to the base policy. So we need the
         # adapter config first, then load the correct policy and then apply PEFT.
-        from peft import PeftConfig, PeftModel
+        require_package("peft", extra="peft")
 
         logging.info("Loading policy's PEFT adapter.")
 
         peft_pretrained_path = str(cfg.pretrained_path)
-        peft_config = PeftConfig.from_pretrained(peft_pretrained_path)
+        peft_config = PeftConfig.from_pretrained(
+            peft_pretrained_path,
+            revision=cfg.pretrained_revision,
+        )
 
         kwargs["pretrained_name_or_path"] = peft_config.base_model_name_or_path
         if not kwargs["pretrained_name_or_path"]:
@@ -349,9 +360,14 @@ def make_policy(
                 "the adapter was trained."
             )
 
+        kwargs["revision"] = peft_config.revision
         policy = policy_cls.from_pretrained(**kwargs)
         policy = PeftModel.from_pretrained(
-            policy, peft_pretrained_path, config=peft_config, is_trainable=True
+            policy,
+            peft_pretrained_path,
+            config=peft_config,
+            revision=cfg.pretrained_revision,
+            is_trainable=True,
         )
 
     else:

src/lerobot/policies/gaussian_actor/configuration_gaussian_actor.py

@@ -37,13 +37,19 @@ def is_image_feature(key: str) -> bool:
 @dataclass
 class ConcurrencyConfig:
     """Configuration for the concurrency of the actor and learner.
+
     Possible values are:
     - "threads": Use threads for the actor and learner.
     - "processes": Use processes for the actor and learner.
+
+    ``multiprocessing_context`` selects the process-wide start method when
+    processes are used. Set it to ``None`` to preserve Python's default or a
+    method already selected by the embedding application.
     """
 
     actor: str = "threads"
     learner: str = "threads"
+    multiprocessing_context: str | None = "spawn"
 
 
 @dataclass

src/lerobot/policies/groot/processor_groot.py

@@ -475,6 +475,7 @@ def make_groot_pre_post_processors_from_pretrained(
     config: GrootConfig,
     pretrained_path: str,
     *,
+    revision: str | None = None,
     dataset_stats: dict[str, dict[str, torch.Tensor]] | None = None,
     dataset_meta: Any | None = None,
     preprocessor_overrides: dict[str, Any] | None = None,
@@ -511,6 +512,7 @@ def make_groot_pre_post_processors_from_pretrained(
 
     preprocessor, postprocessor = _load_groot_processor_pipelines(
         pretrained_path,
+        revision=revision,
         preprocessor_overrides=preprocessor_overrides,
         postprocessor_overrides=postprocessor_overrides,
         preprocessor_config_filename=preprocessor_config_filename,
@@ -526,6 +528,7 @@ def make_groot_pre_post_processors_from_pretrained(
 def _load_groot_processor_pipelines(
     pretrained_path: str,
     *,
+    revision: str | None,
     preprocessor_overrides: dict[str, Any],
     postprocessor_overrides: dict[str, Any],
     preprocessor_config_filename: str,
@@ -540,6 +543,7 @@ def _load_groot_processor_pipelines(
     preprocessor = PolicyProcessorPipeline.from_pretrained(
         pretrained_model_name_or_path=pretrained_path,
         config_filename=preprocessor_config_filename,
+        revision=revision,
         overrides=preprocessor_overrides,
         to_transition=batch_to_transition,
         to_output=transition_to_batch,
@@ -547,6 +551,7 @@ def _load_groot_processor_pipelines(
     postprocessor = PolicyProcessorPipeline.from_pretrained(
         pretrained_model_name_or_path=pretrained_path,
         config_filename=postprocessor_config_filename,
+        revision=revision,
         overrides=postprocessor_overrides,
         to_transition=policy_action_to_transition,
         to_output=transition_to_policy_action,

src/lerobot/policies/molmoact2/modeling_molmoact2.py

@@ -43,11 +43,22 @@ from torch.distributions import Beta
 
 from lerobot.policies.pretrained import PreTrainedPolicy
 from lerobot.utils.constants import ACTION
-from lerobot.utils.import_utils import _scipy_available, _transformers_available, require_package
+from lerobot.utils.import_utils import (
+    _peft_available,
+    _scipy_available,
+    _transformers_available,
+    require_package,
+)
 
 from ..rtc.modeling_rtc import RTCProcessor
 from .configuration_molmoact2 import MolmoAct2Config
 
+if TYPE_CHECKING or _peft_available:
+    from peft import LoraConfig, get_peft_model
+else:
+    LoraConfig = None
+    get_peft_model = None
+
 logger = logging.getLogger(__name__)
 
 
@@ -1731,13 +1742,11 @@ class MolmoAct2Policy(PreTrainedPolicy):
 
     def _build_inner_lora_config(self):
         require_package("peft", extra="molmoact2")
-        from peft import LoraConfig
 
         return LoraConfig(**self._get_inner_peft_targets())
 
     def _apply_lora_adapters(self) -> None:
         require_package("peft", extra="molmoact2")
-        from peft import get_peft_model
 
         peft_config = self._build_inner_lora_config()
         self._validate_peft_config(peft_config)

src/lerobot/policies/pi05/modeling_pi05.py

@@ -524,8 +524,6 @@ class PI05Pytorch(nn.Module):  # see openpi `PI0Pytorch`
 
     def embed_suffix(self, noisy_actions, timestep):
         """Embed noisy_actions, timestep to prepare for Expert Gemma processing."""
-        embs = []
-        pad_masks = []
         att_masks = []
 
         # Embed timestep using sine-cosine positional encoding
@@ -551,23 +549,17 @@ class PI05Pytorch(nn.Module):  # see openpi `PI0Pytorch`
             return F.silu(x)
 
         time_emb = self._apply_checkpoint(time_mlp_func, time_emb)
-        action_time_emb = action_emb
         adarms_cond = time_emb
 
-        embs.append(action_time_emb)
-        bsize, action_time_dim = action_time_emb.shape[:2]
-        action_time_mask = torch.ones(bsize, action_time_dim, dtype=torch.bool, device=timestep.device)
-        pad_masks.append(action_time_mask)
+        bsize, action_time_dim = action_emb.shape[:2]
+        pad_masks = torch.ones(bsize, action_time_dim, dtype=torch.bool, device=timestep.device)
 
         # Set attention masks so that image, language and state inputs do not attend to action tokens
         att_masks += [1] + ([0] * (self.config.chunk_size - 1))
-
-        embs = torch.cat(embs, dim=1)
-        pad_masks = torch.cat(pad_masks, dim=1)
-        att_masks = torch.tensor(att_masks, dtype=embs.dtype, device=embs.device)
+        att_masks = torch.tensor(att_masks, dtype=action_emb.dtype, device=action_emb.device)
         att_masks = att_masks[None, :].expand(bsize, len(att_masks))
 
-        return embs, pad_masks, att_masks, adarms_cond
+        return action_emb, pad_masks, att_masks, adarms_cond
 
     def forward(self, images, img_masks, tokens, masks, actions, noise, time) -> Tensor:
         """Do a full training forward pass and compute the loss."""

src/lerobot/policies/pretrained.py

@@ -34,14 +34,22 @@ from lerobot.configs import PreTrainedConfig
 from lerobot.configs.train import TrainPipelineConfig
 from lerobot.utils.device_utils import resolve_safetensors_device
 from lerobot.utils.hub import HubMixin
+from lerobot.utils.import_utils import _peft_available, require_package
 
 from .utils import log_model_loading_keys
 
-T = TypeVar("T", bound="PreTrainedPolicy")
+if TYPE_CHECKING or _peft_available:
+    from peft import PEFT_TYPE_TO_CONFIG_MAPPING, PeftType, get_peft_model
+else:
+    PEFT_TYPE_TO_CONFIG_MAPPING = None
+    PeftType = None
+    get_peft_model = None
 
 if TYPE_CHECKING:
     from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
 
+T = TypeVar("T", bound="PreTrainedPolicy")
+
 
 def _build_card_context(
     cfg: TrainPipelineConfig | None,
@@ -384,7 +392,7 @@ class PreTrainedPolicy(nn.Module, HubMixin, abc.ABC):
             peft_cli_overrides: Optional dict of CLI overrides (method_type, target_modules, r, etc.)
                 These are merged with policy defaults to build the final config.
         """
-        from peft import get_peft_model
+        require_package("peft", extra="peft")
 
         # If user provided a complete config, use it directly (with overrides)
         if peft_config is not None:
@@ -455,7 +463,7 @@ class PreTrainedPolicy(nn.Module, HubMixin, abc.ABC):
         Returns:
             Preprocessed dict with renamed keys and init_type mapped to method-specific key.
         """
-        from peft import PeftType
+        require_package("peft", extra="peft")
 
         cli_overrides = cli_overrides.copy()
 
@@ -480,7 +488,7 @@ class PreTrainedPolicy(nn.Module, HubMixin, abc.ABC):
 
     def _build_peft_config(self, cli_overrides: dict):
         """Build a PEFT config from policy defaults and CLI overrides."""
-        from peft import PEFT_TYPE_TO_CONFIG_MAPPING, PeftType
+        require_package("peft", extra="peft")
 
         # Determine PEFT method type (default to LORA)
         method_type_str = cli_overrides.get("method_type") or "lora"
@@ -507,7 +515,7 @@ class PreTrainedPolicy(nn.Module, HubMixin, abc.ABC):
 
     def _apply_peft_cli_overrides(self, peft_config, cli_overrides: dict):
         """Apply CLI overrides to an existing PEFT config."""
-        from peft import PEFT_TYPE_TO_CONFIG_MAPPING, PeftType
+        require_package("peft", extra="peft")
 
         # Get method type from existing config or CLI override
         method_type_str = cli_overrides.get("method_type")

src/lerobot/policies/smolvla/modeling_smolvla.py

@@ -61,9 +61,15 @@ import torch.nn.functional as F  # noqa: N812
 from torch import Tensor, nn
 
 from lerobot.utils.constants import ACTION, OBS_LANGUAGE_ATTENTION_MASK, OBS_LANGUAGE_TOKENS, OBS_STATE
-from lerobot.utils.device_utils import get_safe_dtype
 from lerobot.utils.import_utils import require_package
 
+from ..common.flow_matching import euler_integrate, sample_noise, sample_time_beta
+from ..common.vla_utils import (
+    create_sinusoidal_pos_embedding,
+    make_att_2d_masks,
+    pad_vector,
+    resize_with_pad,
+)
 from ..pretrained import PreTrainedPolicy
 from ..rtc.modeling_rtc import RTCProcessor
 from ..utils import (
@@ -79,96 +85,6 @@ class ActionSelectKwargs(TypedDict, total=False):
     execution_horizon: int | None
 
 
-def create_sinusoidal_pos_embedding(
-    time: torch.tensor, dimension: int, min_period: float, max_period: float, device="cpu"
-) -> Tensor:
-    """Computes sine-cosine positional embedding vectors for scalar positions."""
-    if dimension % 2 != 0:
-        raise ValueError(f"dimension ({dimension}) must be divisible by 2")
-
-    if time.ndim != 1:
-        raise ValueError("The time tensor is expected to be of shape `(batch_size, )`.")
-
-    dtype = get_safe_dtype(torch.float64, device.type)
-    fraction = torch.linspace(0.0, 1.0, dimension // 2, dtype=dtype, device=device)
-    period = min_period * (max_period / min_period) ** fraction
-
-    # Compute the outer product
-    scaling_factor = 1.0 / period * 2 * math.pi
-    sin_input = scaling_factor[None, :] * time[:, None]
-    pos_emb = torch.cat([torch.sin(sin_input), torch.cos(sin_input)], dim=1)
-    return pos_emb
-
-
-def make_att_2d_masks(pad_masks, att_masks):
-    """Copied from big_vision.
-
-    Tokens can attend to valid inputs tokens which have a cumulative mask_ar
-    smaller or equal to theirs. This way `mask_ar` int[B, N] can be used to
-    setup several types of attention, for example:
-
-      [[1 1 1 1 1 1]]: pure causal attention.
-
-      [[0 0 0 1 1 1]]: prefix-lm attention. The first 3 tokens can attend between
-          themselves and the last 3 tokens have a causal attention. The first
-          entry could also be a 1 without changing behaviour.
-
-      [[1 0 1 0 1 0 0 1 0 0]]: causal attention between 4 blocks. Tokens of a
-          block can attend all previous blocks and all tokens on the same block.
-
-    Args:
-      input_mask: bool[B, N] true if its part of the input, false if padding.
-      mask_ar: int32[B, N] mask that's 1 where previous tokens cannot depend on
-        it and 0 where it shares the same attention mask as the previous token.
-    """
-    if att_masks.ndim != 2:
-        raise ValueError(att_masks.ndim)
-    if pad_masks.ndim != 2:
-        raise ValueError(pad_masks.ndim)
-
-    cumsum = torch.cumsum(att_masks, dim=1)
-    att_2d_masks = cumsum[:, None, :] <= cumsum[:, :, None]
-    pad_2d_masks = pad_masks[:, None, :] * pad_masks[:, :, None]
-    att_2d_masks = att_2d_masks & pad_2d_masks
-    return att_2d_masks
-
-
-def resize_with_pad(img, width, height, pad_value=-1):
-    # assume no-op when width height fits already
-    if img.ndim != 4:
-        raise ValueError(f"(b,c,h,w) expected, but {img.shape}")
-
-    cur_height, cur_width = img.shape[2:]
-
-    ratio = max(cur_width / width, cur_height / height)
-    resized_height = int(cur_height / ratio)
-    resized_width = int(cur_width / ratio)
-    resized_img = F.interpolate(
-        img, size=(resized_height, resized_width), mode="bilinear", align_corners=False
-    )
-
-    pad_height = max(0, int(height - resized_height))
-    pad_width = max(0, int(width - resized_width))
-
-    # pad on left and top of image
-    padded_img = F.pad(resized_img, (pad_width, 0, pad_height, 0), value=pad_value)
-    return padded_img
-
-
-def pad_vector(vector, new_dim):
-    """Can be (batch_size x sequence_length x features_dimension)
-    or (batch_size x features_dimension)
-    """
-    if vector.shape[-1] == new_dim:
-        return vector
-    shape = list(vector.shape)
-    current_dim = shape[-1]
-    shape[-1] = new_dim
-    new_vector = torch.zeros(*shape, dtype=vector.dtype, device=vector.device)
-    new_vector[..., :current_dim] = vector
-    return new_vector
-
-
 def normalize(x, min_val, max_val):
     return (x - min_val) / (max_val - min_val)
 
@@ -429,7 +345,13 @@ class SmolVLAPolicy(PreTrainedPolicy):
         for key in present_img_keys:
             img = batch[key][:, -1, :, :, :] if batch[key].ndim == 5 else batch[key]
             if self.config.resize_imgs_with_padding is not None:
-                img = resize_with_pad(img, *self.config.resize_imgs_with_padding, pad_value=0)
+                # SmolVLA stores the target as (width, height); the shared helper expects (height, width).
+                img = resize_with_pad(
+                    img,
+                    self.config.resize_imgs_with_padding[1],
+                    self.config.resize_imgs_with_padding[0],
+                    pad_value=0,
+                )
 
             # Normalize from range [0,1] to [-1,1] as expacted by siglip
             img = img * 2.0 - 1.0
@@ -619,20 +541,10 @@ class VLAFlowMatching(nn.Module):
             params.requires_grad = self.config.train_state_proj
 
     def sample_noise(self, shape, device):
-        noise = torch.normal(
-            mean=0.0,
-            std=1.0,
-            size=shape,
-            dtype=torch.float32,
-            device=device,
-        )
-        return noise
+        return sample_noise(shape, device)
 
     def sample_time(self, bsize, device):
-        beta_dist = torch.distributions.Beta(concentration1=1.5, concentration0=1.0)
-        time_beta = beta_dist.sample((bsize,)).to(device=device, dtype=torch.float32)
-        time = time_beta * 0.999 + 0.001
-        return time
+        return sample_time_beta(bsize, device, alpha=1.5, beta=1.0, scale=0.999, offset=0.001)
 
     def embed_prefix(
         self, images, img_masks, lang_tokens, lang_masks, state: torch.Tensor = None
@@ -800,7 +712,6 @@ class VLAFlowMatching(nn.Module):
             past_key_values=None,
             inputs_embeds=[prefix_embs, suffix_embs],
             use_cache=False,
-            fill_kv_cache=False,
         )
         suffix_out = suffix_out[:, -self.config.chunk_size :]
         # Original openpi code, upcast attention output
@@ -839,47 +750,25 @@ class VLAFlowMatching(nn.Module):
             past_key_values=None,
             inputs_embeds=[prefix_embs, None],
             use_cache=self.config.use_cache,
-            fill_kv_cache=True,
         )
         num_steps = self.config.num_steps
-        dt = -1.0 / num_steps
 
-        x_t = noise
-        for step in range(num_steps):
-            time = 1.0 + step * dt
-            time_tensor = torch.tensor(time, dtype=torch.float32, device=device).expand(bsize)
-
-            def denoise_step_partial_call(input_x_t, current_timestep=time_tensor):
-                return self.denoise_step(
+        return euler_integrate(
+            lambda input_x_t, current_timestep: self.denoise_step(
                 x_t=input_x_t,
                 prefix_pad_masks=prefix_pad_masks,
                 past_key_values=past_key_values,
                 timestep=current_timestep,
+            ),
+            noise,
+            num_steps,
+            rtc_processor=self.rtc_processor,
+            rtc_enabled=self._rtc_enabled(),
+            inference_delay=kwargs.get("inference_delay"),
+            prev_chunk_left_over=kwargs.get("prev_chunk_left_over"),
+            execution_horizon=kwargs.get("execution_horizon"),
         )
 
-            if self._rtc_enabled():
-                inference_delay = kwargs.get("inference_delay")
-                prev_chunk_left_over = kwargs.get("prev_chunk_left_over")
-                execution_horizon = kwargs.get("execution_horizon")
-
-                v_t = self.rtc_processor.denoise_step(
-                    x_t=x_t,
-                    prev_chunk_left_over=prev_chunk_left_over,
-                    inference_delay=inference_delay,
-                    time=time,
-                    original_denoise_step_partial=denoise_step_partial_call,
-                    execution_horizon=execution_horizon,
-                )
-            else:
-                v_t = denoise_step_partial_call(x_t)
-
-            x_t = x_t + dt * v_t
-
-            if self.rtc_processor is not None and self.rtc_processor.is_debug_enabled():
-                self.rtc_processor.track(time=time, x_t=x_t, v_t=v_t)
-
-        return x_t
-
     def denoise_step(
         self,
         prefix_pad_masks,
@@ -907,8 +796,10 @@ class VLAFlowMatching(nn.Module):
             past_key_values=past_key_values,
             inputs_embeds=[None, suffix_embs],
             use_cache=self.config.use_cache,
-            fill_kv_cache=False,
         )
+        if past_key_values is not None:
+            # Self-attention layers append suffix K/V in place; restore the prefix for the next step.
+            past_key_values.crop(prefix_len)
         suffix_out = outputs_embeds[1]
         suffix_out = suffix_out[:, -self.config.chunk_size :]
         suffix_out = suffix_out.to(dtype=torch.float32)

src/lerobot/policies/smolvla/smolvlm_with_expert.py

@@ -26,6 +26,7 @@ if TYPE_CHECKING or _transformers_available:
         AutoModel,
         AutoModelForImageTextToText,
         AutoProcessor,
+        DynamicCache,
         SmolVLMForConditionalGeneration,
     )
 else:
@@ -33,6 +34,7 @@ else:
     AutoModel = None
     AutoModelForImageTextToText = None
     AutoProcessor = None
+    DynamicCache = None
     SmolVLMForConditionalGeneration = None
 
 
@@ -216,9 +218,8 @@ class SmolVLMWithExpertModel(nn.Module):
         batch_size,
         head_dim,
         use_cache: bool = True,
-        fill_kv_cache: bool = True,
-        past_key_values=None,
-    ) -> list[torch.Tensor]:
+        past_key_values: "DynamicCache | None" = None,
+    ) -> "tuple[list[torch.Tensor], DynamicCache | None]":
         query_states = []
         key_states = []
         value_states = []
@@ -259,22 +260,16 @@ class SmolVLMWithExpertModel(nn.Module):
         query_states = apply_rope(query_states, position_ids_)
         key_states = apply_rope(key_states, position_ids_)
 
-        if use_cache and past_key_values is None:
-            past_key_values = {}
-
         if use_cache:
-            if fill_kv_cache:
-                past_key_values[layer_idx] = {
-                    "key_states": key_states,
-                    "value_states": value_states,
-                }
-            else:
-                # TODO here, some optimization can be done - similar to a `StaticCache` we can declare the `max_len` before.
-                # so we create an empty cache, with just one cuda malloc, and if (in autoregressive case) we reach
-                # the max len, then we (for instance) double the cache size. This implementation already exists
-                # in `transformers`. (molbap)
-                key_states = torch.cat([past_key_values[layer_idx]["key_states"], key_states], dim=1)
-                value_states = torch.cat([past_key_values[layer_idx]["value_states"], value_states], dim=1)
+            # `DynamicCache` stores tensors as [batch, heads, seq, head_dim]; this module works with
+            # [batch, seq, heads, head_dim]. During prefix prefill this stores the (post-RoPE) K/V and
+            # returns them unchanged; during denoising it appends the suffix K/V and returns
+            # [prefix; suffix], exactly like the previous hand-rolled dict cache.
+            key_states, value_states = past_key_values.update(
+                key_states.transpose(1, 2), value_states.transpose(1, 2), layer_idx
+            )
+            key_states = key_states.transpose(1, 2)
+            value_states = value_states.transpose(1, 2)
 
         attention_interface = self.get_attention_interface()
 
@@ -293,13 +288,12 @@ class SmolVLMWithExpertModel(nn.Module):
         batch_size,
         head_dim,
         use_cache: bool = True,
-        fill_kv_cache: bool = True,
-        past_key_values=None,
-    ) -> list[torch.Tensor]:
+        past_key_values: "DynamicCache | None" = None,
+    ) -> "tuple[list[torch.Tensor], DynamicCache | None]":
         attention_interface = self.get_attention_interface()
 
         att_outputs = []
-        assert len(inputs_embeds) == 2 or (use_cache and past_key_values is not None and not fill_kv_cache), (
+        assert len(inputs_embeds) == 2 or (use_cache and past_key_values is not None), (
             f"Both len(inputs_embeds) == {len(inputs_embeds)} and past_key_values is {past_key_values}"
         )
 
@@ -332,22 +326,13 @@ class SmolVLMWithExpertModel(nn.Module):
         else:
             expert_position_id = position_ids
 
-        if use_cache and past_key_values is None:
-            past_key_values = {}
-
-        if use_cache:
-            if fill_kv_cache:
-                past_key_values[layer_idx] = {
-                    "key_states": key_states,
-                    "value_states": value_states,
-                }
-            else:
-                # TODO here, some optimization can be done - similar to a `StaticCache` we can declare the `max_len` before.
-                # so we create an empty cache, with just one cuda malloc, and if (in autoregressive case) we reach
-                # the max len, then we (for instance) double the cache size. This implementation already exists
-                # in `transformers`. (molbap)
-                key_states = past_key_values[layer_idx]["key_states"]
-                value_states = past_key_values[layer_idx]["value_states"]
+        if use_cache and past_key_values is not None:
+            # Cross-attention layers never fill the cache themselves: during the prefix prefill every
+            # layer goes through `forward_attn_layer`, which stores the (post-RoPE) VLM K/V for this
+            # layer index. Here we only read them back (no concatenation: the expert cross-attends to
+            # the fixed prefix). `DynamicCache` stores [batch, heads, seq, head_dim]; transpose back.
+            key_states = past_key_values.layers[layer_idx].keys.transpose(1, 2)
+            value_states = past_key_values.layers[layer_idx].values.transpose(1, 2)
 
         # Expert
         expert_layer = model_layers[1][layer_idx]
@@ -360,14 +345,15 @@ class SmolVLMWithExpertModel(nn.Module):
             expert_hidden_states = expert_hidden_states.to(dtype=expert_layer.self_attn.q_proj.weight.dtype)
             expert_query_state = expert_layer.self_attn.q_proj(expert_hidden_states).view(expert_hidden_shape)
 
-            _key_states = key_states.to(dtype=expert_layer.self_attn.k_proj.weight.dtype).view(
+            # reshape (not view): K/V read back from the cache are transposed, hence non-contiguous
+            _key_states = key_states.to(dtype=expert_layer.self_attn.k_proj.weight.dtype).reshape(
                 *key_states.shape[:2], -1
             )
             expert_key_states = expert_layer.self_attn.k_proj(_key_states).view(
                 *_key_states.shape[:-1], -1, expert_layer.self_attn.head_dim
             )  # k_proj should have same dim as kv
 
-            _value_states = value_states.to(dtype=expert_layer.self_attn.v_proj.weight.dtype).view(
+            _value_states = value_states.to(dtype=expert_layer.self_attn.v_proj.weight.dtype).reshape(
                 *value_states.shape[:2], -1
             )
             expert_value_states = expert_layer.self_attn.v_proj(_value_states).view(
@@ -416,10 +402,9 @@ class SmolVLMWithExpertModel(nn.Module):
         self,
         attention_mask: torch.Tensor | None = None,
         position_ids: torch.LongTensor | None = None,
-        past_key_values: list[torch.FloatTensor] | None = None,
+        past_key_values: "DynamicCache | None" = None,
         inputs_embeds: list[torch.FloatTensor] = None,
         use_cache: bool | None = None,
-        fill_kv_cache: bool | None = None,
     ):
         models = [self.get_vlm_model().text_model, self.lm_expert]
         model_layers = self.get_model_layers(models)
@@ -431,6 +416,13 @@ class SmolVLMWithExpertModel(nn.Module):
                 continue
             batch_size = hidden_states.shape[0]
 
+        # Prefix prefill: no cache was passed, so create one and fill it (every layer runs
+        # self-attention over the prefix). When a filled cache is passed (denoising), layers
+        # read from it instead.
+        fill_kv_cache = use_cache and past_key_values is None
+        if fill_kv_cache:
+            past_key_values = DynamicCache()
+
         # RMSNorm
         num_layers = self.num_vlm_layers
         head_dim = self.vlm.config.text_config.head_dim
@@ -449,7 +441,6 @@ class SmolVLMWithExpertModel(nn.Module):
                     batch_size,
                     head_dim,
                     use_cache=use_cache,
-                    fill_kv_cache=fill_kv_cache,
                     past_key_values=past_key_values,
                 )
             else:
@@ -462,7 +453,6 @@ class SmolVLMWithExpertModel(nn.Module):
                     batch_size,
                     head_dim,
                     use_cache=use_cache,
-                    fill_kv_cache=fill_kv_cache,
                     past_key_values=past_key_values,
                 )
             outputs_embeds = []

src/lerobot/policies/vla_jepa/action_head.py

@@ -282,7 +282,6 @@ class VLAJEPAActionHead(nn.Module):
         actions: torch.Tensor,
         state: torch.Tensor | None = None,
         action_is_pad: torch.Tensor | None = None,
-        reduction: str = "mean",
     ) -> torch.Tensor:
         noise = torch.randn_like(actions)
         t = self.sample_time(actions.shape[0], actions.device, actions.dtype)
@@ -303,10 +302,6 @@ class VLAJEPAActionHead(nn.Module):
 
         loss = F.mse_loss(pred_actions, velocity, reduction="none")  # [B, T, action_dim]
         valid_mask = ~action_is_pad.unsqueeze(-1)  # [B, T, 1]
-        if reduction == "none":
-            # Per-sample loss (B,) for sample weighting (RA-BC): mask-average over T and action_dim.
-            per_sample_valid = valid_mask.sum(dim=(1, 2)) * loss.shape[-1]  # [B]
-            return (loss * valid_mask).sum(dim=(1, 2)) / per_sample_valid.clamp_min(1)
         num_valid = valid_mask.sum() * loss.shape[-1]
         return (loss * valid_mask).sum() / num_valid.clamp_min(1)
 

src/lerobot/policies/vla_jepa/configuration_vla_jepa.py

@@ -56,14 +56,6 @@ class VLAJEPAConfig(PreTrainedConfig):
     action_dim: int = 7
     state_dim: int = 8
 
-    # Relative actions: converts absolute actions to relative (action -= state) during
-    # preprocessing, and reverses it at postprocessing. Requires `state_dim` (OBS_STATE).
-    use_relative_actions: bool = False
-    # Joint names to keep absolute (not converted to relative). Empty list = all dims relative.
-    relative_exclude_joints: list[str] = field(default_factory=lambda: ["gripper"])
-    # Populated at runtime from dataset metadata by make_policy (used to build the exclude mask).
-    action_feature_names: list[str] | None = None
-
     num_action_tokens_per_timestep: int = 8
     num_embodied_action_tokens_per_instruction: int = 32
     num_inference_timesteps: int = 4

src/lerobot/policies/vla_jepa/modeling_vla_jepa.py

@@ -26,7 +26,6 @@ from torch import Tensor, nn
 from lerobot.policies.pretrained import PreTrainedPolicy, T
 from lerobot.policies.utils import populate_queues
 from lerobot.utils.constants import ACTION, OBS_STATE
-from lerobot.utils.device_utils import get_autocast_context
 from lerobot.utils.import_utils import _transformers_available, require_package
 
 if TYPE_CHECKING or _transformers_available:
@@ -184,7 +183,7 @@ class VLAJEPAModel(nn.Module):
             action_idx = action_mask.nonzero(as_tuple=True)
 
         device_type = next(self.parameters()).device.type
-        with get_autocast_context(device_type, torch.bfloat16):
+        with torch.autocast(device_type=device_type, dtype=torch.bfloat16):
             last_hidden = self._qwen_last_decoder_hidden(qwen_inputs)  # [B, seq_len, H]
             b, _, h = last_hidden.shape
             embodied_action_tokens = last_hidden[embodied_idx[0], embodied_idx[1], :].view(b, -1, h)
@@ -195,12 +194,8 @@ class VLAJEPAModel(nn.Module):
             )
         return embodied_action_tokens, action_tokens
 
-    def _world_model_loss(self, videos: Tensor, action_tokens: Tensor, reduction: str = "mean") -> Tensor:
-        """JEPA encode + predictor L1 loss. `videos` is [B, V, T, C, H, W] float in [0, 1].
-
-        `reduction="none"` returns a per-sample loss (B,) for sample weighting (RA-BC);
-        "mean" returns the scalar loss.
-        """
+    def _world_model_loss(self, videos: Tensor, action_tokens: Tensor) -> Tensor:
+        """JEPA encode + predictor L1 loss. `videos` is [B, V, T, C, H, W] float in [0, 1]."""
         # Match the world model's expected view count: pad with the first view, or trim extras.
         num_views = self.config.jepa_tubelet_size
         if videos.shape[1] < num_views:
@@ -228,8 +223,7 @@ class VLAJEPAModel(nn.Module):
         # num_video_frames raw frames → t_enc_total temporal positions after tubelet compression
         t_enc_total = self.config.num_video_frames // tubelet_size
         if t_enc_total < 2:
-            zero_shape = (video_embeddings.shape[0],) if reduction == "none" else ()
-            return torch.zeros(zero_shape, device=video_embeddings.device)
+            return torch.zeros((), device=video_embeddings.device)
 
         # Shift-by-one JEPA split: input_states = positions 0..T-2, gt_states = positions 1..T-1
         t_enc_ctx = t_enc_total - 1
@@ -245,10 +239,6 @@ class VLAJEPAModel(nn.Module):
         predicted_states = self.video_predictor(
             input_states.float(), action_tokens[:, :expected_actions].float()
         )
-        if reduction == "none":
-            # Per-sample loss (B,): mean over all non-batch dims (tokens, feature).
-            elementwise = F.l1_loss(predicted_states, gt_states.float(), reduction="none")
-            return elementwise.mean(dim=tuple(range(1, elementwise.ndim)))
         return F.l1_loss(predicted_states, gt_states.float(), reduction="mean")
 
     def _action_loss(
@@ -257,27 +247,17 @@ class VLAJEPAModel(nn.Module):
         actions: Tensor,
         state: Tensor | None,
         action_is_pad: Tensor | None,
-        reduction: str = "mean",
     ) -> Tensor:
-        """Flow-matching action-head loss, repeated over `repeated_diffusion_steps`.
-
-        `reduction="none"` returns a per-sample loss (B,) — the `repeated_diffusion_steps`
-        independent noise draws are averaged back per original sample — for RA-BC weighting.
-        """
+        """Flow-matching action-head loss, repeated over `repeated_diffusion_steps`."""
         device_type = next(self.parameters()).device.type
-        with get_autocast_context(device_type, torch.float32):
+        with torch.autocast(device_type=device_type, dtype=torch.float32):
             r = self.config.repeated_diffusion_steps
             horizon = self.config.chunk_size
-            b = embodied_action_tokens.shape[0]
             actions_target = actions[:, -horizon:, :].to(torch.float32).repeat(r, 1, 1)
             embodied = embodied_action_tokens.repeat(r, 1, 1)
             state_rep = state.to(embodied_action_tokens.dtype).repeat(r, 1, 1) if state is not None else None
             pad_rep = action_is_pad[:, -horizon:].repeat(r, 1) if action_is_pad is not None else None
-            loss = self.action_model(embodied, actions_target, state_rep, pad_rep, reduction=reduction)
-            if reduction == "none":
-                # `.repeat(r, 1, 1)` tiles as [rep0(b0..b_{B-1}), rep1(...), ...] → (r, B); mean over reps.
-                return loss.view(r, b).mean(dim=0)
-            return loss
+            return self.action_model(embodied, actions_target, state_rep, pad_rep)
 
     def forward(
         self,
@@ -287,29 +267,21 @@ class VLAJEPAModel(nn.Module):
         actions: Tensor | None = None,
         state: Tensor | None = None,
         action_is_pad: Tensor | None = None,
-        reduction: str = "mean",
     ) -> dict[str, Tensor]:
-        """Native forward: Qwen encode → optional world-model loss → optional action-head loss.
-
-        `reduction="none"` makes both loss terms per-sample (B,) for RA-BC weighting; "mean"
-        returns scalar losses.
-        """
+        """Native forward: Qwen encode → optional world-model loss → optional action-head loss."""
         embodied_action_tokens, action_tokens = self._encode_qwen(
             images, instructions, need_action_tokens=self.config.enable_world_model
         )
 
         if self.config.enable_world_model and videos is not None:
-            wm_loss = self._world_model_loss(videos, action_tokens, reduction=reduction)
+            wm_loss = self._world_model_loss(videos, action_tokens)
         else:
-            zero_shape = (embodied_action_tokens.shape[0],) if reduction == "none" else ()
-            wm_loss = torch.zeros(zero_shape, device=embodied_action_tokens.device)
+            wm_loss = torch.zeros((), device=embodied_action_tokens.device)
 
         if actions is None:
             return {"wm_loss": wm_loss}
 
-        action_loss = self._action_loss(
-            embodied_action_tokens, actions, state, action_is_pad, reduction=reduction
-        )
+        action_loss = self._action_loss(embodied_action_tokens, actions, state, action_is_pad)
         return {"action_loss": action_loss, "wm_loss": wm_loss * self.config.world_model_loss_weight}
 
     # ---- Native predict_action (follows original VLA_JEPA.predict_action) ----
@@ -395,19 +367,12 @@ class VLAJEPAPolicy(PreTrainedPolicy):
         batch_size = batch[image_keys[0]].shape[0]
 
         # Current-frame image per view ([B, C, H, W]); regroup per sample for Qwen messages.
-        # Resize to config.resize_images_to (as predict_action does) so training and inference feed
-        # Qwen the same resolution. Critical for memory: native camera frames (e.g. 720x1280) would
-        # otherwise blow up the Qwen3-VL vision-tower attention (patch count grows with resolution).
-        resize_hw = tuple(self.config.resize_images_to) if self.config.resize_images_to else None
         frames = []
         for key in image_keys:
             t = batch[key]
             if t.ndim == 5:  # [B, T, C, H, W] -> current observation (delta=0)
                 t = t[:, 0]
-            px = self.model.qwen.to_pixel_values(t)  # [B, C, H, W]
-            if resize_hw is not None and tuple(px.shape[-2:]) != resize_hw:
-                px = F.interpolate(px.float(), size=resize_hw, mode="area")
-            frames.append(px)
+            frames.append(self.model.qwen.to_pixel_values(t))
         images = [[frame[b] for frame in frames] for b in range(batch_size)]
 
         tasks = batch.get("task")
@@ -423,26 +388,7 @@ class VLAJEPAPolicy(PreTrainedPolicy):
         # Videos [B, V, T, C, H, W] - only assembled during training when the world model consumes them.
         if self.model.config.enable_world_model and training:
             views = [batch[k].unsqueeze(1) if batch[k].ndim == 4 else batch[k] for k in image_keys]
-            # The world model consumes a SINGLE stacked [B, V, T, C, H, W] tensor, so all camera
-            # views must share a spatial size. Cameras can differ (e.g. base 480x640 vs wrist
-            # 720x1280), so resize each view to a common size before stacking — config.resize_images_to
-            # if set (same target predict_action uses), else the first view's size (a no-op when all
-            # views already match, preserving behavior for single-resolution datasets). The vjepa video
-            # processor does the final resize to the encoder resolution downstream.
-            cfg = self.model.config
-            target_hw = tuple(cfg.resize_images_to) if cfg.resize_images_to else tuple(views[0].shape[-2:])
-            resized = []
-            for v in views:
-                if tuple(v.shape[-2:]) != target_hw:
-                    b, t, c = v.shape[0], v.shape[1], v.shape[2]
-                    v = F.interpolate(
-                        v.reshape(b * t, c, v.shape[3], v.shape[4]).float(),
-                        size=target_hw,
-                        mode="bilinear",
-                        align_corners=False,
-                    ).reshape(b, t, c, target_hw[0], target_hw[1])
-                resized.append(v)
-            inputs["videos"] = self.model.qwen.to_pixel_values(torch.stack(resized, dim=1))
+            inputs["videos"] = self.model.qwen.to_pixel_values(torch.stack(views, dim=1))
 
         actions = batch.get(ACTION)
         if actions is not None:
@@ -460,17 +406,15 @@ class VLAJEPAPolicy(PreTrainedPolicy):
 
     # ---- LeRobot Policy Interface ----
 
-    def forward(self, batch: dict[str, Tensor], reduction: str = "mean") -> tuple[Tensor, dict]:
+    def forward(self, batch: dict[str, Tensor]) -> tuple[Tensor, dict]:
         """LeRobot train forward: convert → native forward → aggregate losses."""
-        native_output = self.model.forward(
-            **self._prepare_model_inputs(batch, training=True), reduction=reduction
-        )
+        native_output = self.model.forward(**self._prepare_model_inputs(batch, training=True))
 
         ref = next(iter(native_output.values()))
-        zero = torch.zeros_like(ref)
+        zero = torch.zeros((), device=ref.device, dtype=ref.dtype)
         total_loss = native_output.get("action_loss", zero) + native_output.get("wm_loss", zero)
-        logs = {k: v.detach().mean().item() for k, v in native_output.items()}
-        logs["loss"] = total_loss.detach().mean().item()
+        logs = {k: v.detach().item() for k, v in native_output.items()}
+        logs["loss"] = total_loss.detach().item()
         return total_loss, logs
 
     def get_optim_params(self) -> dict:

src/lerobot/policies/vla_jepa/processor_vla_jepa.py

@@ -17,19 +17,14 @@ from __future__ import annotations
 from typing import Any
 
 import torch
-import torch.nn.functional as F  # noqa: N812
 
-from lerobot.configs import PipelineFeatureType, PolicyFeature
 from lerobot.policies.vla_jepa.configuration_vla_jepa import VLAJEPAConfig
 from lerobot.processor import (
-    AbsoluteActionsProcessorStep,
     EnvTransition,
-    ObservationProcessorStep,
     PolicyAction,
     PolicyProcessorPipeline,
     ProcessorStep,
     ProcessorStepRegistry,
-    RelativeActionsProcessorStep,
     TransitionKey,
     UnnormalizerProcessorStep,
     make_default_policy_processor_steps,
@@ -37,69 +32,6 @@ from lerobot.processor import (
 )
 
 
-@ProcessorStepRegistry.register(name="vla_jepa_image_prep")
-class ImagePrepProcessorStep(ObservationProcessorStep):
-    """Prepares image observations for the VLA-JEPA model: float cast, 1->3 channel expand, resize.
-
-    This makes explicit (in the serialized pipeline) the image prep the model used to do
-    internally. The model keeps the same operations as idempotent guards, so:
-      - checkpoints saved WITHOUT this step (older uploads) are unaffected — the model still
-        does the prep;
-      - checkpoints saved WITH this step get it done here, and the model-side guards no-op.
-
-    Mirrors `Qwen3VLInterface.to_pixel_values` + the `F.interpolate(mode="area")` resize in
-    `VLAJEPAPolicy._prepare_model_inputs`/`predict_action`. Deliberately does NOT clamp (the
-    model path doesn't), so values stay bit-identical. Handles [C,H,W], [B,C,H,W]/[T,C,H,W]
-    and [B,T,C,H,W] image tensors.
-    """
-
-    def __init__(self, resize_to: tuple[int, int] | None = None, expand_channels: bool = True):
-        self.resize_to = tuple(resize_to) if resize_to is not None else None
-        self.expand_channels = expand_channels
-
-    def observation(self, observation: dict) -> dict:
-        new_observation = dict(observation)
-        for key in observation:
-            if "image" not in key:
-                continue
-            image = observation[key].float()
-            if self.expand_channels and image.shape[-3] == 1:
-                repeats = [1] * image.ndim
-                repeats[-3] = 3
-                image = image.repeat(*repeats)
-            if self.resize_to is not None and tuple(image.shape[-2:]) != self.resize_to:
-                device = image.device
-                # NOTE: no "area" kernel on mps; resize on cpu then move back.
-                if device.type == "mps":
-                    image = image.cpu()
-                lead = image.shape[:-3]
-                c, h, w = image.shape[-3:]
-                flat = image.reshape(-1, c, h, w)
-                flat = F.interpolate(flat, size=self.resize_to, mode="area")
-                image = flat.reshape(*lead, c, *self.resize_to).to(device)
-            new_observation[key] = image
-        return new_observation
-
-    def get_config(self) -> dict[str, Any]:
-        return {
-            "resize_to": list(self.resize_to) if self.resize_to is not None else None,
-            "expand_channels": self.expand_channels,
-        }
-
-    def transform_features(self, features):
-        for key in features[PipelineFeatureType.OBSERVATION]:
-            if "image" not in key:
-                continue
-            feat = features[PipelineFeatureType.OBSERVATION][key]
-            # Match `to_pixel_values`: only a single channel is expanded to 3.
-            nb_channel = 3 if (self.expand_channels and feat.shape[0] == 1) else feat.shape[0]
-            spatial = self.resize_to if self.resize_to is not None else tuple(feat.shape[1:])
-            features[PipelineFeatureType.OBSERVATION][key] = PolicyFeature(
-                type=feat.type, shape=(nb_channel, *spatial)
-            )
-        return features
-
-
 @ProcessorStepRegistry.register(name="vla_jepa_clip_actions")
 class ClipActionsProcessorStep(ProcessorStep):
     """Clips action tensor to [-1, 1] before unnormalization."""
@@ -177,24 +109,10 @@ def make_vla_jepa_pre_post_processors(
 ]:
     features = {**config.input_features, **config.output_features}
     steps = make_default_policy_processor_steps(config, dataset_stats)
-
-    # Shared relative-action step (OpenPI order: raw -> relative -> normalize -> model ->
-    # unnormalize -> absolute). The SAME instance is passed to AbsoluteActionsProcessorStep
-    # below so its cached raw state (set during preprocessing) flows to postprocessing.
-    relative_step = RelativeActionsProcessorStep(
-        enabled=config.use_relative_actions,
-        exclude_joints=getattr(config, "relative_exclude_joints", []),
-        action_names=getattr(config, "action_feature_names", None),
-    )
-
     input_steps = [
         steps.rename_observations,
         steps.add_batch_dim,
         steps.to_device,
-        ImagePrepProcessorStep(
-            resize_to=tuple(config.resize_images_to) if config.resize_images_to else None,
-        ),
-        relative_step,
         steps.normalize,
     ]
     output_steps: list[ProcessorStep] = []
@@ -213,11 +131,6 @@ def make_vla_jepa_pre_post_processors(
             stats=dataset_stats,
         )
     )
-    # Reverse the relative conversion on the unnormalized action, before gripper binarization.
-    # gripper is kept absolute by relative_exclude_joints, so the two steps touch disjoint dims.
-    output_steps.append(
-        AbsoluteActionsProcessorStep(enabled=config.use_relative_actions, relative_step=relative_step)
-    )
     if config.binarize_gripper_action:
         output_steps.append(
             BinarizeGripperProcessorStep(gripper_dim=config.gripper_dim, threshold=config.gripper_threshold)

src/lerobot/processor/delta_action_processor.py

@@ -132,10 +132,20 @@ class MapDeltaActionToRobotActionStep(RobotActionProcessorStep):
     def transform_features(
         self, features: dict[PipelineFeatureType, dict[str, PolicyFeature]]
     ) -> dict[PipelineFeatureType, dict[str, PolicyFeature]]:
-        for axis in ["x", "y", "z", "gripper"]:
+        for axis in ["x", "y", "z"]:
             features[PipelineFeatureType.ACTION].pop(f"delta_{axis}", None)
+        features[PipelineFeatureType.ACTION].pop("gripper", None)
 
-        for feat in ["enabled", "target_x", "target_y", "target_z", "target_wx", "target_wy", "target_wz"]:
+        for feat in [
+            "enabled",
+            "target_x",
+            "target_y",
+            "target_z",
+            "target_wx",
+            "target_wy",
+            "target_wz",
+            "gripper_vel",
+        ]:
             features[PipelineFeatureType.ACTION][f"{feat}"] = PolicyFeature(
                 type=FeatureType.ACTION, shape=(1,)
             )

src/lerobot/processor/pipeline.py

@@ -713,6 +713,8 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
             ProcessorMigrationError: If the model requires migration to processor format.
         """
         model_id = str(pretrained_model_name_or_path)
+        model_path = Path(model_id)
+        is_local_source = model_path.is_dir() or model_path.is_file()
         hub_download_kwargs = {
             "force_download": force_download,
             "resume_download": resume_download,
@@ -731,7 +733,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
 
         # 3. Build steps with overrides
         steps, validated_overrides = cls._build_steps_with_overrides(
-            loaded_config, overrides or {}, model_id, base_path, hub_download_kwargs
+            loaded_config, overrides or {}, model_id, base_path, hub_download_kwargs, is_local_source
         )
 
         # 4. Validate that all overrides were used
@@ -921,6 +923,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
         model_id: str,
         base_path: Path | None,
         hub_download_kwargs: dict[str, Any],
+        is_local_source: bool = False,
     ) -> tuple[list[ProcessorStep], set[str]]:
         """Build all processor steps with overrides and state loading.
 
@@ -944,7 +947,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
         3. **State Loading** (via _load_step_state):
            - **If step has "state_file"**: Load tensor state from .safetensors
            - **Local first**: Check base_path/state_file.safetensors
-           - **Hub fallback**: Download state file if not found locally
+           - **Hub fallback**: Download state file if the pipeline was loaded from the Hub
            - **Optional**: Only load if step has load_state_dict method
 
         4. **Override Tracking**:
@@ -962,6 +965,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
             model_id: The model identifier (needed for Hub state file downloads)
             base_path: Local directory path for finding state files
             hub_download_kwargs: Parameters for hf_hub_download (tokens, cache, etc.)
+            is_local_source: Whether model_id resolved to a local directory or config file.
 
         Returns:
             Tuple of (instantiated_steps_list, unused_override_keys)
@@ -975,7 +979,9 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
         steps, remaining_override_keys = cls._build_steps_from_config(loaded_config, overrides)
 
         for step_instance, step_entry in zip(steps, loaded_config["steps"], strict=True):
-            cls._load_step_state(step_instance, step_entry, model_id, base_path, hub_download_kwargs)
+            cls._load_step_state(
+                step_instance, step_entry, model_id, base_path, hub_download_kwargs, is_local_source
+            )
 
         return steps, remaining_override_keys
 
@@ -1139,6 +1145,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
         model_id: str,
         base_path: Path | None,
         hub_download_kwargs: dict[str, Any],
+        is_local_source: bool = False,
     ) -> None:
         """Load state dictionary for a processor step if available.
 
@@ -1157,7 +1164,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
            - **Use case**: Loading from local saved model directory
 
         2. **Hub download fallback**: Download state file from repository
-           - **When triggered**: Local file not found or base_path is None
+           - **When triggered**: Local file not found and the pipeline source is a Hub repo
            - **Process**: Use hf_hub_download with same parameters as config
            - **Example**: Download "normalize_step_0.safetensors" from "user/repo"
            - **Result**: Downloaded to local cache, path returned
@@ -1178,6 +1185,7 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
             model_id: The model identifier (used for Hub downloads if needed)
             base_path: Local directory path for finding state files (None for Hub-only)
             hub_download_kwargs: Parameters for hf_hub_download (tokens, cache, etc.)
+            is_local_source: Whether model_id resolved to a local directory or config file.
 
         Note:
             This method modifies step_instance in-place and returns None.
@@ -1191,6 +1199,12 @@ class DataProcessorPipeline[TInput, TOutput](HubMixin):
         # Try local file first
         if base_path and (base_path / state_filename).exists():
             state_path = str(base_path / state_filename)
+        elif is_local_source:
+            state_path = base_path / state_filename if base_path else Path(state_filename)
+            raise FileNotFoundError(
+                f"State file '{state_filename}' was not found for local processor pipeline "
+                f"'{model_id}' at '{state_path}'."
+            )
         else:
             # Download from Hub
             state_path = hf_hub_download(

src/lerobot/rl/actor.py

@@ -91,7 +91,7 @@ from lerobot.robots import so_follower  # noqa: F401
 from lerobot.teleoperators import gamepad, so_leader  # noqa: F401
 from lerobot.teleoperators.utils import TeleopEvents
 from lerobot.utils.device_utils import get_safe_torch_device
-from lerobot.utils.process import ProcessSignalHandler
+from lerobot.utils.process import ProcessSignalHandler, ensure_multiprocessing_start_method
 from lerobot.utils.random_utils import set_seed
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.transition import (
@@ -124,9 +124,7 @@ def actor_cli(cfg: TrainRLServerPipelineConfig):
     cfg.validate()
     display_pid = False
     if not use_threads(cfg):
-        import torch.multiprocessing as mp
-
-        mp.set_start_method("spawn")
+        ensure_multiprocessing_start_method(cfg.policy.concurrency.multiprocessing_context)
         display_pid = True
 
     # Create logs directory to ensure it exists

src/lerobot/rl/buffer.py

@@ -18,7 +18,7 @@ import functools
 import threading
 from collections.abc import Callable, Sequence
 from contextlib import suppress
-from typing import TypedDict
+from typing import NotRequired, TypedDict
 
 import torch
 import torch.nn.functional as F  # noqa: N812
@@ -36,7 +36,7 @@ class BatchTransition(TypedDict):
     next_state: dict[str, torch.Tensor]
     done: torch.Tensor
     truncated: torch.Tensor
-    complementary_info: dict[str, torch.Tensor | float | int] | None = None
+    complementary_info: NotRequired[dict[str, torch.Tensor | float | int] | None]
 
 
 def random_crop_vectorized(images: torch.Tensor, output_size: tuple) -> torch.Tensor:

src/lerobot/rl/learner.py

@@ -102,7 +102,7 @@ from lerobot.utils.constants import (
 )
 from lerobot.utils.device_utils import get_safe_torch_device
 from lerobot.utils.io_utils import load_json, write_json
-from lerobot.utils.process import ProcessSignalHandler
+from lerobot.utils.process import ProcessSignalHandler, ensure_multiprocessing_start_method
 from lerobot.utils.random_utils import set_seed
 from lerobot.utils.utils import (
     format_big_number,
@@ -123,9 +123,7 @@ def train_cli(cfg: TrainRLServerPipelineConfig):
     # Fail fast with a friendly error if the optional ``hilserl`` extra is missing.
     require_package("grpcio", extra="hilserl", import_name="grpc")
     if not use_threads(cfg):
-        import torch.multiprocessing as mp
-
-        mp.set_start_method("spawn")
+        ensure_multiprocessing_start_method(cfg.policy.concurrency.multiprocessing_context)
 
     # Use the job_name from the config
     train(

src/lerobot/robots/bi_so_follower/bi_so_follower.py

@@ -58,6 +58,9 @@ class BiSOFollower(BimanualMixin, Robot):
             port=config.left_arm_config.port,
             disable_torque_on_disconnect=config.left_arm_config.disable_torque_on_disconnect,
             max_relative_target=config.left_arm_config.max_relative_target,
+            position_p_coefficient=config.left_arm_config.position_p_coefficient,
+            position_i_coefficient=config.left_arm_config.position_i_coefficient,
+            position_d_coefficient=config.left_arm_config.position_d_coefficient,
             use_degrees=config.left_arm_config.use_degrees,
             cameras=left_arm_cameras,
         )
@@ -68,6 +71,9 @@ class BiSOFollower(BimanualMixin, Robot):
             port=config.right_arm_config.port,
             disable_torque_on_disconnect=config.right_arm_config.disable_torque_on_disconnect,
             max_relative_target=config.right_arm_config.max_relative_target,
+            position_p_coefficient=config.right_arm_config.position_p_coefficient,
+            position_i_coefficient=config.right_arm_config.position_i_coefficient,
+            position_d_coefficient=config.right_arm_config.position_d_coefficient,
             use_degrees=config.right_arm_config.use_degrees,
             cameras=config.right_arm_config.cameras,
         )

src/lerobot/robots/lekiwi/lekiwi_client.py

@@ -323,6 +323,10 @@ class LeKiwiClient(Robot):
             np.ndarray: the action sent to the motors, potentially clipped.
         """
 
+        # Action values may be torch tensors (e.g. replayed from a dataset) or numpy
+        # scalars; json.dumps only serializes Python primitives, so coerce each value to a
+        # plain float before sending.
+        action = {key: float(value) for key, value in action.items()}
         self.zmq_cmd_socket.send_string(json.dumps(action))  # action is in motor space
 
         # TODO(Steven): Remove the np conversion when it is possible to record a non-numpy array value

src/lerobot/robots/openarm_follower/openarm_follower.py

@@ -150,9 +150,6 @@ class OpenArmFollower(Robot):
 
         self.configure()
 
-        if self.is_calibrated:
-            self.bus.set_zero_position()
-
         self.bus.enable_torque()
 
         logger.info(f"{self} connected.")

src/lerobot/robots/so_follower/config_so_follower.py

@@ -41,6 +41,17 @@ class SOFollowerConfig:
     # Set to `True` for backward compatibility with previous policies/dataset
     use_degrees: bool = True
 
+    # Position-mode PID gains written to Feetech STS3215 motors at connect time.
+    position_p_coefficient: int = 16
+    position_i_coefficient: int = 0
+    position_d_coefficient: int = 32
+
+    # Number of extra attempts when a `sync_read` of the motors fails. Feetech buses can occasionally
+    # return a corrupted status packet ("Incorrect status packet!"), especially when several joints move
+    # at once, which otherwise aborts the control loop. Retries are immediate (no sleep) and only happen on
+    # failure, so the steady-state read cost is unchanged.
+    num_read_retries: int = 2
+
 
 @RobotConfig.register_subclass("so101_follower")
 @RobotConfig.register_subclass("so100_follower")

src/lerobot/robots/so_follower/robot_kinematic_processor.py

@@ -510,10 +510,10 @@ class ForwardKinematicsJointsToEEAction(RobotActionProcessorStep):
         # We only use the ee pose in the dataset, so we don't need the joint positions
         for n in self.motor_names:
             features[PipelineFeatureType.ACTION].pop(f"{n}.pos", None)
-        # We specify the dataset features of this step that we want to be stored in the dataset
+        # Store end-effector features as actions in the dataset schema
         for k in ["x", "y", "z", "wx", "wy", "wz", "gripper_pos"]:
             features[PipelineFeatureType.ACTION][f"ee.{k}"] = PolicyFeature(
-                type=FeatureType.STATE, shape=(1,)
+                type=FeatureType.ACTION, shape=(1,)
             )
         return features
 

src/lerobot/robots/so_follower/so_follower.py

@@ -161,11 +161,9 @@ class SOFollower(Robot):
             self.bus.configure_motors()
             for motor in self.bus.motors:
                 self.bus.write("Operating_Mode", motor, OperatingMode.POSITION.value)
-                # Set P_Coefficient to lower value to avoid shakiness (Default is 32)
-                self.bus.write("P_Coefficient", motor, 16)
-                # Set I_Coefficient and D_Coefficient to default value 0 and 32
-                self.bus.write("I_Coefficient", motor, 0)
-                self.bus.write("D_Coefficient", motor, 32)
+                self.bus.write("P_Coefficient", motor, self.config.position_p_coefficient)
+                self.bus.write("I_Coefficient", motor, self.config.position_i_coefficient)
+                self.bus.write("D_Coefficient", motor, self.config.position_d_coefficient)
 
                 if motor == "gripper":
                     self.bus.write("Max_Torque_Limit", motor, 500)  # 50% of max torque to avoid burnout
@@ -182,7 +180,7 @@ class SOFollower(Robot):
     def get_observation(self) -> RobotObservation:
         # Read arm position
         start = time.perf_counter()
-        obs_dict = self.bus.sync_read("Present_Position")
+        obs_dict = self.bus.sync_read("Present_Position", num_retry=self.config.num_read_retries)
         obs_dict = {f"{motor}.pos": val for motor, val in obs_dict.items()}
         dt_ms = (time.perf_counter() - start) * 1e3
         logger.debug(f"{self} read state: {dt_ms:.1f}ms")
@@ -223,7 +221,7 @@ class SOFollower(Robot):
         # Cap goal position when too far away from present position.
         # /!\ Slower fps expected due to reading from the follower.
         if self.config.max_relative_target is not None:
-            present_pos = self.bus.sync_read("Present_Position")
+            present_pos = self.bus.sync_read("Present_Position", num_retry=self.config.num_read_retries)
             goal_present_pos = {key: (g_pos, present_pos[key]) for key, g_pos in goal_pos.items()}
             goal_pos = ensure_safe_goal_position(goal_present_pos, self.config.max_relative_target)
 

src/lerobot/rollout/configs.py

@@ -326,8 +326,17 @@ class RolloutConfig:
 
         policy_path = parser.get_path_arg("policy")
         if policy_path:
-            cli_overrides = parser.get_cli_overrides("policy")
-            self.policy = PreTrainedConfig.from_pretrained(policy_path, cli_overrides=cli_overrides)
+            yaml_overrides = parser.get_yaml_overrides("policy")
+            cli_overrides = parser.get_cli_overrides("policy") or []
+            policy_overrides = yaml_overrides + cli_overrides
+            pretrained_revision = parser.parse_arg("pretrained_revision", cli_overrides)
+            if pretrained_revision is None:
+                pretrained_revision = parser.parse_arg("pretrained_revision", yaml_overrides)
+            self.policy = PreTrainedConfig.from_pretrained(
+                policy_path,
+                revision=pretrained_revision,
+                cli_overrides=policy_overrides,
+            )
             self.policy.pretrained_path = policy_path
         if self.policy is None:
             raise ValueError("--policy.path is required for rollout")

src/lerobot/rollout/context.py

@@ -24,10 +24,11 @@ from __future__ import annotations
 import logging
 from dataclasses import dataclass, field
 from threading import Event
+from typing import TYPE_CHECKING
 
 import torch
 
-from lerobot.configs import FeatureType
+from lerobot.configs import FeatureType, PreTrainedConfig
 from lerobot.datasets import (
     LeRobotDataset,
     aggregate_pipeline_dataset_features,
@@ -47,6 +48,7 @@ from lerobot.processor.relative_action_processor import RelativeActionsProcessor
 from lerobot.robots import make_robot_from_config
 from lerobot.teleoperators import Teleoperator, make_teleoperator_from_config
 from lerobot.utils.feature_utils import combine_feature_dicts, hw_to_dataset_features
+from lerobot.utils.import_utils import _peft_available, require_package
 
 from .configs import BaseStrategyConfig, DAggerStrategyConfig, RolloutConfig
 from .inference import (
@@ -57,6 +59,12 @@ from .inference import (
 )
 from .robot_wrapper import ThreadSafeRobot
 
+if TYPE_CHECKING or _peft_available:
+    from peft import PeftConfig, PeftModel
+else:
+    PeftConfig = None
+    PeftModel = None
+
 logger = logging.getLogger(__name__)
 
 
@@ -159,6 +167,35 @@ class RolloutContext:
 # ---------------------------------------------------------------------------
 
 
+def _load_pretrained_policy(policy_config: PreTrainedConfig) -> PreTrainedPolicy:
+    """Load policy weights, keeping adapter and base-model revisions independent."""
+    pretrained_revision = policy_config.pretrained_revision
+    policy_class = get_policy_class(policy_config.type)
+
+    if not policy_config.use_peft:
+        return policy_class.from_pretrained(
+            policy_config.pretrained_path,
+            config=policy_config,
+            revision=pretrained_revision,
+        )
+
+    require_package("peft", extra="peft")
+
+    peft_path = policy_config.pretrained_path
+    peft_config = PeftConfig.from_pretrained(peft_path, revision=pretrained_revision)
+    policy = policy_class.from_pretrained(
+        pretrained_name_or_path=peft_config.base_model_name_or_path,
+        config=policy_config,
+        revision=peft_config.revision,
+    )
+    return PeftModel.from_pretrained(
+        policy,
+        peft_path,
+        config=peft_config,
+        revision=pretrained_revision,
+    )
+
+
 def build_rollout_context(
     cfg: RolloutConfig,
     shutdown_event: Event,
@@ -176,7 +213,6 @@ def build_rollout_context(
     # --- 1. Policy (heavy I/O, but no hardware yet) -------------------
     logger.info("Loading policy from '%s'...", cfg.policy.pretrained_path)
     policy_config = cfg.policy
-    policy_class = get_policy_class(policy_config.type)
 
     if hasattr(policy_config, "compile_model"):
         policy_config.compile_model = cfg.use_torch_compile
@@ -187,17 +223,7 @@ def build_rollout_context(
             "Please use `cpu` or `cuda` backend."
         )
 
-    if policy_config.use_peft:
-        from peft import PeftConfig, PeftModel
-
-        peft_path = policy_config.pretrained_path
-        peft_config = PeftConfig.from_pretrained(peft_path)
-        policy = policy_class.from_pretrained(
-            pretrained_name_or_path=peft_config.base_model_name_or_path, config=policy_config
-        )
-        policy = PeftModel.from_pretrained(policy, peft_path, config=peft_config)
-    else:
-        policy = policy_class.from_pretrained(policy_config.pretrained_path, config=policy_config)
+    policy = _load_pretrained_policy(policy_config)
 
     if is_rtc:
         policy.config.rtc_config = cfg.inference.rtc
@@ -276,12 +302,22 @@ def build_rollout_context(
     # ``observation_features`` values are either a tuple (camera shape) or the
     # ``float`` type itself used as a sentinel for scalar motor features —
     # see ``dict[str, type | tuple]`` annotation on ``Robot.observation_features``.
+    # Keep cameras (tuple) plus both joint-position (.pos) and base-velocity (.vel)
+    # scalar state features. LeKiwi's observation.state is 9-dim (6 arm .pos +
+    # x/y/theta.vel) and the policy was trained/normalized on all 9; the old .pos-only
+    # filter fed a 6-dim state into a 9-dim normalizer → RuntimeError (size 6 vs 9).
+    # Pure-arm robots have no .vel state keys, so this is a no-op for them.
     observation_features_hw = {
         k: v
         for k, v in all_obs_features.items()
-        if isinstance(v, tuple) or (v is float and k.endswith(".pos"))
+        if isinstance(v, tuple) or (v is float and k.endswith((".pos", ".vel")))
     }
-    action_features_hw = {k: v for k, v in robot.action_features.items() if k.endswith(".pos")}
+    # Keep both joint-position (.pos) and base-velocity (.vel) action features so
+    # mobile manipulators command the base too (e.g. LeKiwi: 6 arm .pos +
+    # x/y/theta.vel = 9-dim action). Pure-arm robots have no .vel keys, so this is
+    # a no-op for them. Without the .vel keys the base velocities are silently
+    # dropped from dataset_features[ACTION]/ordered_action_keys and the base never moves.
+    action_features_hw = {k: v for k, v in robot.action_features.items() if k.endswith((".pos", ".vel"))}
 
     # The action side is always needed: sync inference reads action names from
     # ``dataset_features[ACTION]`` to map policy tensors back to robot actions.
@@ -392,6 +428,7 @@ def build_rollout_context(
     preprocessor, postprocessor = make_pre_post_processors(
         policy_cfg=policy_config,
         pretrained_path=cfg.policy.pretrained_path,
+        pretrained_revision=policy_config.pretrained_revision,
         dataset_stats=dataset_stats,
         preprocessor_overrides={
             "device_processor": {"device": cfg.device},

src/lerobot/scripts/augment_dataset_quantile_stats.py

@@ -36,6 +36,7 @@ python src/lerobot/scripts/augment_dataset_quantile_stats.py \
 import argparse
 import concurrent.futures
 import logging
+import os
 from pathlib import Path
 
 import numpy as np
@@ -52,6 +53,7 @@ from lerobot.datasets import (
     get_feature_stats,
     write_stats,
 )
+from lerobot.datasets.compute_stats import sample_indices
 from lerobot.utils.utils import init_logging
 
 
@@ -77,12 +79,14 @@ def has_quantile_stats(stats: dict[str, dict] | None, quantile_list_keys: list[s
     return False
 
 
-def process_single_episode(dataset: LeRobotDataset, episode_idx: int) -> dict:
+def process_single_episode(dataset: LeRobotDataset, episode_idx: int, use_sampling: bool = True) -> dict:
     """Process a single episode and return its statistics.
 
     Args:
         dataset: The LeRobot dataset
         episode_idx: Index of the episode to process
+        use_sampling: If True, sub-sample image/video frames per episode to bound
+            memory. If False, use every frame (exact, higher memory).
 
     Returns:
         Dictionary containing episode statistics
@@ -92,16 +96,31 @@ def process_single_episode(dataset: LeRobotDataset, episode_idx: int) -> dict:
     start_idx = dataset.meta.episodes[episode_idx]["dataset_from_index"]
     end_idx = dataset.meta.episodes[episode_idx]["dataset_to_index"]
 
-    collected_data: dict[str, list] = {}
-    for idx in range(start_idx, end_idx):
-        item = dataset[idx]
-        for key, value in item.items():
-            if key not in dataset.features:
-                continue
+    episode_len = end_idx - start_idx
 
-            if key not in collected_data:
-                collected_data[key] = []
-            collected_data[key].append(value)
+    # Images/video are the memory hog, so sub-sample those frames per episode;
+    # numeric columns are cheap, so read them in full (exact).
+    image_keys = [k for k in dataset.features if dataset.features[k]["dtype"] in ("image", "video")]
+    numeric_keys = [
+        k for k in dataset.features if dataset.features[k]["dtype"] not in ("image", "video", "string")
+    ]
+
+    collected_data: dict[str, list] = {}
+
+    # Numeric features: every frame, read directly from the underlying table.
+    if numeric_keys:
+        numeric_cols = dataset.hf_dataset.select_columns(numeric_keys)[start_idx:end_idx]
+        for key in numeric_keys:
+            collected_data[key] = [torch.as_tensor(v) for v in numeric_cols[key]]
+
+    # Image/video features: decode only a sampled subset of frames.
+    if image_keys:
+        sampled_offsets = sample_indices(episode_len) if use_sampling else list(range(episode_len))
+        for offset in sampled_offsets:
+            item = dataset[start_idx + offset]
+            for key in image_keys:
+                if key in item:
+                    collected_data.setdefault(key, []).append(item[key])
 
     ep_stats = {}
     for key, data_list in collected_data.items():
@@ -131,11 +150,13 @@ def process_single_episode(dataset: LeRobotDataset, episode_idx: int) -> dict:
     return ep_stats
 
 
-def compute_quantile_stats_for_dataset(dataset: LeRobotDataset) -> dict[str, dict]:
+def compute_quantile_stats_for_dataset(dataset: LeRobotDataset, use_sampling: bool = True) -> dict[str, dict]:
     """Compute quantile statistics for all episodes in the dataset.
 
     Args:
         dataset: The LeRobot dataset to compute statistics for
+        use_sampling: If True, sub-sample image/video frames per episode to bound
+            memory. If False, use every frame (exact, higher memory).
 
     Returns:
         Dictionary containing aggregated statistics with quantiles
@@ -153,15 +174,15 @@ def compute_quantile_stats_for_dataset(dataset: LeRobotDataset) -> dict[str, dic
     if has_videos:
         logging.info("Dataset contains video keys - using sequential processing for thread safety")
         for episode_idx in tqdm(range(dataset.num_episodes), desc="Processing episodes"):
-            ep_stats = process_single_episode(dataset, episode_idx)
+            ep_stats = process_single_episode(dataset, episode_idx, use_sampling)
             episode_stats_list.append(ep_stats)
     else:
         logging.info("Dataset has no video keys - using parallel processing for better performance")
-        max_workers = min(dataset.num_episodes, 16)
+        max_workers = min(dataset.num_episodes, int(os.environ.get("LEROBOT_STATS_MAX_WORKERS", 16)))
 
         with concurrent.futures.ThreadPoolExecutor(max_workers=max_workers) as executor:
             future_to_episode = {
-                executor.submit(process_single_episode, dataset, episode_idx): episode_idx
+                executor.submit(process_single_episode, dataset, episode_idx, use_sampling): episode_idx
                 for episode_idx in range(dataset.num_episodes)
             }
 
@@ -188,6 +209,7 @@ def augment_dataset_with_quantile_stats(
     repo_id: str,
     root: str | Path | None = None,
     overwrite: bool = False,
+    use_sampling: bool = True,
 ) -> None:
     """Augment a dataset with quantile statistics if they are missing.
 
@@ -195,6 +217,8 @@ def augment_dataset_with_quantile_stats(
         repo_id: Repository ID of the dataset
         root: Local root directory for the dataset
         overwrite: Overwrite existing quantile statistics if they already exist
+        use_sampling: If True, sub-sample image/video frames per episode to bound
+            memory. If False, use every frame (exact, higher memory).
     """
     logging.info(f"Loading dataset: {repo_id}")
     dataset = LeRobotDataset(
@@ -208,7 +232,7 @@ def augment_dataset_with_quantile_stats(
 
     logging.info("Dataset does not contain quantile statistics. Computing them now...")
 
-    new_stats = compute_quantile_stats_for_dataset(dataset)
+    new_stats = compute_quantile_stats_for_dataset(dataset, use_sampling=use_sampling)
 
     logging.info("Updating dataset metadata with new quantile statistics")
     dataset.meta.stats = new_stats
@@ -248,6 +272,14 @@ def main():
         action="store_true",
         help="Overwrite existing quantile statistics if they already exist",
     )
+    parser.add_argument(
+        "--no-sampling",
+        action="store_true",
+        help=(
+            "Compute stats over every frame (exact, higher memory). By default, "
+            "image/video frames are sub-sampled per episode to bound memory."
+        ),
+    )
 
     args = parser.parse_args()
     root = Path(args.root) if args.root else None
@@ -258,6 +290,7 @@ def main():
         repo_id=args.repo_id,
         root=root,
         overwrite=args.overwrite,
+        use_sampling=not args.no_sampling,
     )
 
 

src/lerobot/scripts/convert_dataset_v21_to_v30.py

@@ -61,6 +61,7 @@ import pyarrow as pa
 import tqdm
 from datasets import Dataset, Features, Image
 from huggingface_hub import HfApi, snapshot_download
+from huggingface_hub.errors import RevisionNotFoundError
 from requests import HTTPError
 
 from lerobot.datasets import CODEBASE_VERSION, LeRobotDataset, aggregate_stats
@@ -93,6 +94,8 @@ from lerobot.datasets.video_utils import concatenate_video_files, get_video_dura
 from lerobot.utils.constants import HF_LEROBOT_HOME
 from lerobot.utils.utils import flatten_dict, init_logging
 
+logger = logging.getLogger(__name__)
+
 V21 = "v2.1"
 V30 = "v3.0"
 
@@ -475,11 +478,11 @@ def convert_dataset(
     # First check if the dataset already has a v3.0 version
     if root is None and not force_conversion:
         try:
-            print("Trying to download v3.0 version of the dataset from the hub...")
+            logger.info("Trying to download v3.0 version of the dataset from the hub...")
             snapshot_download(repo_id, repo_type="dataset", revision=V30, local_dir=HF_LEROBOT_HOME / repo_id)
             return
         except Exception:
-            print("Dataset does not have an uploaded v3.0 version. Continuing with conversion.")
+            logger.info("Dataset does not have an uploaded v3.0 version. Continuing with conversion.")
 
     # Set root based on whether local dataset path is provided
     use_local_dataset = False
@@ -487,7 +490,7 @@ def convert_dataset(
     if root.exists():
         validate_local_dataset_version(root)
         use_local_dataset = True
-        print(f"Using local dataset at {root}")
+        logger.info(f"Using local dataset at {root}")
 
     old_root = root.parent / f"{root.name}_old"
     new_root = root.parent / f"{root.name}_v30"
@@ -521,8 +524,8 @@ def convert_dataset(
         hub_api = HfApi()
         try:
             hub_api.delete_tag(repo_id, tag=CODEBASE_VERSION, repo_type="dataset")
-        except HTTPError as e:
-            print(f"tag={CODEBASE_VERSION} probably doesn't exist. Skipping exception ({e})")
+        except (HTTPError, RevisionNotFoundError) as e:
+            logger.warning(f"tag={CODEBASE_VERSION} probably doesn't exist. Skipping exception ({e})")
             pass
         hub_api.delete_files(
             delete_patterns=["data/chunk*/episode_*", "meta/*.jsonl", "videos/chunk*"],

src/lerobot/scripts/lerobot_annotate.py

@@ -24,7 +24,14 @@ Example:
       --root=/path/to/dataset \\
       --vlm.model_id=Qwen/Qwen2.5-VL-7B-Instruct
 
-For distributed runs, see ``examples/annotations/run_hf_job.py``.
+Pass ``--job.target=<flavor>`` to run the same command on a Hugging Face
+Jobs GPU instead of this machine (see ``lerobot.jobs.annotate``):
+
+  uv run lerobot-annotate \\
+      --repo_id=user/dataset \\
+      --new_repo_id=user/dataset_annotated \\
+      --push_to_hub=true \\
+      --job.target=h200
 """
 
 import logging
@@ -69,6 +76,14 @@ def _resolve_root(cfg: AnnotationPipelineConfig) -> Path:
 def annotate(cfg: AnnotationPipelineConfig) -> None:
     """Run the steerable annotation pipeline against a dataset."""
     logging.basicConfig(level=logging.INFO, format="%(asctime)s %(levelname)s %(message)s")
+
+    if cfg.job.is_remote:
+        # Imported lazily: the submitter pulls in LeRobotDataset (the `dataset`
+        # extra), which a local annotation run over --root doesn't need.
+        from lerobot.jobs.annotate import submit_annotate_to_hf
+
+        return submit_annotate_to_hf(cfg)
+
     root = _resolve_root(cfg)
     logger.info("annotate: root=%s", root)
 
@@ -139,14 +154,14 @@ def _push_to_hub(root: Path, cfg: AnnotationPipelineConfig) -> None:
     repo_id = cfg.new_repo_id or cfg.repo_id
     commit_message = cfg.push_commit_message or "Add steerable annotations (lerobot-annotate)"
     api = HfApi()
-    print(f"[lerobot-annotate] creating/locating dataset repo {repo_id}...", flush=True)
+    logger.info(f"[lerobot-annotate] creating/locating dataset repo {repo_id}...")
     api.create_repo(
         repo_id=repo_id,
         repo_type="dataset",
         private=cfg.push_private,
         exist_ok=True,
     )
-    print(f"[lerobot-annotate] uploading {root} -> {repo_id}...", flush=True)
+    logger.info(f"[lerobot-annotate] uploading {root} -> {repo_id}...")
     commit_info = api.upload_folder(
         folder_path=str(root),
         repo_id=repo_id,
@@ -157,7 +172,7 @@ def _push_to_hub(root: Path, cfg: AnnotationPipelineConfig) -> None:
         # at the source dataset; a fresh card is generated below instead.
         ignore_patterns=[".annotate_staging/**", "**/.DS_Store", "README.md"],
     )
-    print(f"[lerobot-annotate] uploaded to https://huggingface.co/datasets/{repo_id}", flush=True)
+    logger.info(f"[lerobot-annotate] uploaded to https://huggingface.co/datasets/{repo_id}")
 
     dataset_info = load_info(root)
     card = create_lerobot_dataset_card(dataset_info=dataset_info, license="apache-2.0", repo_id=repo_id)
@@ -185,14 +200,13 @@ def _push_to_hub(root: Path, cfg: AnnotationPipelineConfig) -> None:
         with suppress(RevisionNotFoundError):
             api.delete_tag(repo_id, tag=version_tag, repo_type="dataset")
         api.create_tag(**tag_kwargs)
-        print(f"[lerobot-annotate] tagged {repo_id} as {version_tag}", flush=True)
+        logger.info(f"[lerobot-annotate] tagged {repo_id} as {version_tag}")
     except Exception as exc:  # noqa: BLE001
-        print(
+        logger.warning(
             f"[lerobot-annotate] WARNING: could not create tag {version_tag!r} on {repo_id}: {exc}. "
             "Dataset is uploaded but ``LeRobotDataset`` won't be able to load it until it's tagged. "
             "Run: from huggingface_hub import HfApi; "
-            f"HfApi().create_tag({repo_id!r}, tag={version_tag!r}, repo_type='dataset', exist_ok=True)",
-            flush=True,
+            f"HfApi().create_tag({repo_id!r}, tag={version_tag!r}, repo_type='dataset', exist_ok=True)"
         )
 
 

src/lerobot/scripts/lerobot_dataset_viz.py

@@ -87,8 +87,15 @@ import tqdm
 from lerobot.configs import DEPTH_MILLIMETER_UNIT
 from lerobot.datasets import LeRobotDataset
 from lerobot.utils.constants import ACTION, DONE, OBS_STATE, REWARD, SUCCESS
+from lerobot.utils.dataset_visualization_utils import (
+    get_extra_scalar_keys,
+    is_scalar_like,
+    scalar_to_float,
+)
 from lerobot.utils.utils import init_logging
 
+logger = logging.getLogger(__name__)
+
 DEFAULT_FOXGLOVE_PORT = 8765
 DEFAULT_RERUN_PORT = 9090
 
@@ -151,6 +158,8 @@ def build_blueprint_from_dataset(dataset: LeRobotDataset):
     for key in (DONE, REWARD, SUCCESS):
         if key in dataset.features:
             views.append(rrb.TimeSeriesView(origin=key, name=key))
+    for key in get_extra_scalar_keys(dataset):
+        views.append(rrb.TimeSeriesView(origin=key, name=key))
 
     return rrb.Blueprint(rrb.Grid(*views))
 
@@ -242,6 +251,8 @@ def visualize_dataset(
         hi = stats["q99"] if "q99" in stats else stats["max"]
         depth_ranges[key] = (float(np.asarray(lo).item()), float(np.asarray(hi).item()))
 
+    extra_scalar_keys = get_extra_scalar_keys(dataset)
+
     first_index = None
     for batch in tqdm.tqdm(dataloader, total=len(dataloader)):
         if first_index is None:
@@ -285,6 +296,10 @@ def visualize_dataset(
             if SUCCESS in batch:
                 rr.log(SUCCESS, rr.Scalars(batch[SUCCESS][i].item()))
 
+            for key in extra_scalar_keys:
+                if key in batch and is_scalar_like(batch[key][i]):
+                    rr.log(key, rr.Scalars(scalar_to_float(batch[key][i])))
+
     # save .rrd locally
     if mode == "local" and save:
         output_dir.mkdir(parents=True, exist_ok=True)
@@ -299,7 +314,7 @@ def visualize_dataset(
             while True:
                 time.sleep(1)
         except KeyboardInterrupt:
-            print("Ctrl-C received. Exiting.")
+            logger.info("Ctrl-C received. Exiting.")
 
 
 def main():

src/lerobot/scripts/lerobot_eval.py

@@ -62,7 +62,7 @@ from dataclasses import asdict
 from functools import partial
 from pathlib import Path
 from pprint import pformat
-from typing import Any, TypedDict
+from typing import TYPE_CHECKING, Any, TypedDict
 
 import einops
 import gymnasium as gym
@@ -87,7 +87,7 @@ from lerobot.processor import PolicyProcessorPipeline
 from lerobot.types import PolicyAction
 from lerobot.utils.constants import ACTION, DONE, OBS_IMAGE, OBS_IMAGES, OBS_STR, REWARD
 from lerobot.utils.device_utils import get_safe_torch_device
-from lerobot.utils.import_utils import register_third_party_plugins
+from lerobot.utils.import_utils import _peft_available, register_third_party_plugins, require_package
 from lerobot.utils.io_utils import write_video
 from lerobot.utils.random_utils import set_seed
 from lerobot.utils.utils import (
@@ -95,6 +95,14 @@ from lerobot.utils.utils import (
     inside_slurm,
 )
 
+if TYPE_CHECKING or _peft_available:
+    from peft import PeftModel
+else:
+    PeftModel = None
+
+
+logger = logging.getLogger(__name__)
+
 
 def _env_features_to_dataset_features(env_features: dict) -> dict:
     """Convert EnvConfig.features to the dict format expected by LeRobotDataset.create()."""
@@ -444,15 +452,16 @@ def eval_policy(
         exc = ValueError(
             f"Policy of type 'PreTrainedPolicy' is expected, but type '{type(policy)}' was provided."
         )
-        try:
-            from peft import PeftModel
-
+        if not _peft_available:
+            raise exc
+        require_package("peft", extra="peft")
         if not isinstance(policy, PeftModel):
             raise exc
-        except ImportError:
-            raise exc from None
 
     start = time.time()
+    # Preserve the mode for direct callers. eval_policy_all scopes the mode
+    # around all tasks so parallel evaluations cannot race with each other.
+    was_training = policy.training
     policy.eval()
 
     # Determine how many batched rollouts we need to get n_episodes. Note that if n_episodes is not evenly
@@ -555,7 +564,7 @@ def eval_policy(
         if seeds:
             all_seeds.extend(seeds)
         else:
-            all_seeds.append(None)
+            all_seeds.extend([None] * env.num_envs)
 
         # FIXME: episode_data is either None or it doesn't exist
         if return_episode_data:
@@ -674,6 +683,8 @@ def eval_policy(
     if save_predicted_video:
         info["predicted_video_paths"] = predicted_video_paths
 
+    policy.train(was_training)
+
     return info
 
 
@@ -791,13 +802,13 @@ def eval_main(cfg: EvalPipelineConfig):
             recording_repo_id=cfg.eval.recording_repo_id,
             recording_private=cfg.eval.recording_private,
         )
-        print("Overall Aggregated Metrics:")
-        print(info["overall"])
+        logger.info("Overall Aggregated Metrics:")
+        logger.info(info["overall"])
 
         # Print per-suite stats
         for task_group, task_group_info in info.items():
-            print(f"\nAggregated Metrics for {task_group}:")
-            print(task_group_info)
+            logger.info(f"\nAggregated Metrics for {task_group}:")
+            logger.info(task_group_info)
     # Close all vec envs
     close_envs(envs)
 
@@ -1010,6 +1021,12 @@ def eval_policy_all(
         recording_private=recording_private,
     )
 
+    # Set the shared policy's mode before launching any workers. Restoring it
+    # inside individual tasks would let one task enable training mode while
+    # another task is still evaluating.
+    was_training = policy.training
+    policy.eval()
+    try:
         if max_parallel_tasks <= 1:
             prefetch_thread: threading.Thread | None = None
             for i, (task_group, task_id, env) in enumerate(tasks):
@@ -1044,6 +1061,8 @@ def eval_policy_all(
                         per_task_infos.append({"task_group": tg, "task_id": tid, "metrics": metrics})
                     finally:
                         env.close()
+    finally:
+        policy.train(was_training)
 
     # compute aggregated metrics helper (robust to lists/scalars)
     def _agg_from_list(xs):

src/lerobot/scripts/lerobot_find_cameras.py

@@ -28,7 +28,6 @@ lerobot-find-cameras
 # NOTE(Steven): macOS cameras sometimes report different FPS at init time, not an issue here as we don't specify FPS when opening the cameras, but the information displayed might not be truthful.
 
 import argparse
-import concurrent.futures
 import logging
 import time
 from pathlib import Path
@@ -40,6 +39,7 @@ from PIL import Image
 from lerobot.cameras import ColorMode
 from lerobot.cameras.opencv import OpenCVCamera, OpenCVCameraConfig
 from lerobot.cameras.realsense import RealSenseCamera, RealSenseCameraConfig
+from lerobot.utils.utils import init_logging
 
 logger = logging.getLogger(__name__)
 
@@ -132,7 +132,7 @@ def save_image(
     camera_identifier: str | int,
     images_dir: Path,
     camera_type: str,
-):
+) -> None:
     """
     Saves a single image to disk using Pillow. Handles color conversion if necessary.
     """
@@ -151,7 +151,7 @@ def save_image(
         logger.error(f"Failed to save image for camera {camera_identifier} (type {camera_type}): {e}")
 
 
-def create_camera_instance(cam_meta: dict[str, Any]) -> dict[str, Any] | None:
+def create_camera_instance(cam_meta: dict[str, Any], *, warmup_s: int = 1) -> dict[str, Any] | None:
     """Create and connect to a camera instance based on metadata."""
     cam_type = cam_meta.get("type")
     cam_id = cam_meta.get("id")
@@ -164,12 +164,14 @@ def create_camera_instance(cam_meta: dict[str, Any]) -> dict[str, Any] | None:
             cv_config = OpenCVCameraConfig(
                 index_or_path=cam_id,
                 color_mode=ColorMode.RGB,
+                warmup_s=warmup_s,
             )
             instance = OpenCVCamera(cv_config)
         elif cam_type == "RealSense":
             rs_config = RealSenseCameraConfig(
                 serial_number_or_name=cam_id,
                 color_mode=ColorMode.RGB,
+                warmup_s=warmup_s,
             )
             instance = RealSenseCamera(rs_config)
         else:
@@ -187,9 +189,7 @@ def create_camera_instance(cam_meta: dict[str, Any]) -> dict[str, Any] | None:
         return None
 
 
-def process_camera_image(
-    cam_dict: dict[str, Any], output_dir: Path, current_time: float
-) -> concurrent.futures.Future | None:
+def process_camera_image(cam_dict: dict[str, Any], output_dir: Path, current_time: float) -> None:
     """Capture and process an image from a single camera."""
     cam = cam_dict["instance"]
     meta = cam_dict["meta"]
@@ -199,7 +199,7 @@ def process_camera_image(
     try:
         image_data = cam.read()
 
-        return save_image(
+        save_image(
             image_data,
             cam_id_str,
             output_dir,
@@ -214,10 +214,9 @@ def process_camera_image(
     return None
 
 
-def cleanup_cameras(cameras_to_use: list[dict[str, Any]]):
+def cleanup_camera(cam_dict: dict[str, Any]) -> None:
     """Disconnect all cameras."""
-    logger.info(f"Disconnecting {len(cameras_to_use)} cameras...")
-    for cam_dict in cameras_to_use:
+    logger.info(f"Disconnecting camera with ID {cam_dict['meta'].get('id')}...")
     try:
         if cam_dict["instance"] and cam_dict["instance"].is_connected:
             cam_dict["instance"].disconnect()
@@ -229,6 +228,7 @@ def save_images_from_all_cameras(
     output_dir: Path,
     record_time_s: float = 2.0,
     camera_type: str | None = None,
+    warmup_s: int = 1,
 ):
     """
     Connects to detected cameras (optionally filtered by type) and saves images from each.
@@ -239,6 +239,7 @@ def save_images_from_all_cameras(
         record_time_s: Duration in seconds to record images.
         camera_type: Optional string to filter cameras ("realsense" or "opencv").
                             If None, uses all detected cameras.
+        warmup_s: Duration in seconds to warmup camera before recording images.
     """
     output_dir.mkdir(parents=True, exist_ok=True)
     logger.info(f"Saving images to {output_dir}")
@@ -248,47 +249,32 @@ def save_images_from_all_cameras(
         logger.warning("No cameras detected matching the criteria. Cannot save images.")
         return
 
-    cameras_to_use = []
-    for cam_meta in all_camera_metadata:
-        camera_instance = create_camera_instance(cam_meta)
-        if camera_instance:
-            cameras_to_use.append(camera_instance)
+    logger.info(
+        f"Starting image capture for {record_time_s} seconds from {len(all_camera_metadata)} cameras."
+    )
 
-    if not cameras_to_use:
-        logger.warning("No cameras could be connected. Aborting image save.")
-        return
-
-    logger.info(f"Starting image capture for {record_time_s} seconds from {len(cameras_to_use)} cameras.")
-    start_time = time.perf_counter()
-
-    with concurrent.futures.ThreadPoolExecutor(max_workers=len(cameras_to_use) * 2) as executor:
     try:
+        for cam_meta in all_camera_metadata:
+            cam_dict = create_camera_instance(cam_meta, warmup_s=warmup_s)
+            if cam_dict is None:
+                continue
+            start_time = time.perf_counter()
             while time.perf_counter() - start_time < record_time_s:
-                futures = []
                 current_capture_time = time.perf_counter()
-
-                for cam_dict in cameras_to_use:
-                    future = process_camera_image(cam_dict, output_dir, current_capture_time)
-                    if future:
-                        futures.append(future)
-
-                if futures:
-                    concurrent.futures.wait(futures)
-
+                process_camera_image(cam_dict, output_dir, current_capture_time)
+            cleanup_camera(cam_dict)
     except KeyboardInterrupt:
         logger.info("Capture interrupted by user.")
     finally:
-            print("\nFinalizing image saving...")
-            executor.shutdown(wait=True)
-            cleanup_cameras(cameras_to_use)
         print(f"Image capture finished. Images saved to {output_dir}")
 
 
 def main():
+    init_logging()
+
     parser = argparse.ArgumentParser(
         description="Unified camera utility script for listing cameras and capturing images."
     )
-
     parser.add_argument(
         "camera_type",
         type=str,
@@ -306,8 +292,14 @@ def main():
     parser.add_argument(
         "--record-time-s",
         type=float,
-        default=6.0,
-        help="Time duration to attempt capturing frames. Default: 6 seconds.",
+        default=2.0,
+        help="Time duration to attempt capturing frames. Default: 2 seconds.",
+    )
+    parser.add_argument(
+        "--warmup-s",
+        type=int,
+        default=1,
+        help="Time duration to warmup camera before attempting to capture frames. Default: 1 second.",
     )
     args = parser.parse_args()
     save_images_from_all_cameras(**vars(args))

src/lerobot/scripts/lerobot_record.py

@@ -453,9 +453,11 @@ def record(
                 encoder_queue_maxsize=cfg.dataset.encoder_queue_maxsize,
             )
 
-        robot.connect()
+        # Connect the teleoperator before the robot so the robot isn't left idle (and possibly
+        # tripping a firmware watchdog) during teleop init. Matches lerobot_teleoperate.py.
         if teleop is not None:
             teleop.connect()
+        robot.connect()
 
         listener, events = init_keyboard_listener()
 

src/lerobot/scripts/lerobot_replay.py

@@ -61,6 +61,7 @@ from lerobot.robots import (  # noqa: F401
     earthrover_mini_plus,
     hope_jr,
     koch_follower,
+    lekiwi,
     make_robot_from_config,
     omx_follower,
     openarm_follower,

src/lerobot/scripts/lerobot_rollout.py

@@ -165,6 +165,7 @@ from lerobot.robots import (  # noqa: F401
     earthrover_mini_plus,
     hope_jr,
     koch_follower,
+    lekiwi,
     omx_follower,
     openarm_follower,
     reachy2,

src/lerobot/scripts/lerobot_setup_motors.py

@@ -51,19 +51,7 @@ from lerobot.teleoperators import (  # noqa: F401
     rebot_102_leader,
     so_leader,
 )
-
-COMPATIBLE_DEVICES = [
-    "koch_follower",
-    "koch_leader",
-    "omx_follower",
-    "omx_leader",
-    "openarm_mini",
-    "so100_follower",
-    "so100_leader",
-    "so101_follower",
-    "so101_leader",
-    "lekiwi",
-]
+from lerobot.utils.import_utils import register_third_party_plugins
 
 
 @dataclass
@@ -80,18 +68,19 @@ class SetupConfig:
 
 @draccus.wrap()
 def setup_motors(cfg: SetupConfig):
-    if cfg.device.type not in COMPATIBLE_DEVICES:
-        raise NotImplementedError
-
     if isinstance(cfg.device, RobotConfig):
         device = make_robot_from_config(cfg.device)
     else:
         device = make_teleoperator_from_config(cfg.device)
 
-    device.setup_motors()
+    setup = getattr(device, "setup_motors", None)
+    if not callable(setup):
+        raise NotImplementedError(f"Device type '{cfg.device.type}' does not support motor setup.")
+    setup()
 
 
 def main():
+    register_third_party_plugins()
     setup_motors()
 
 

src/lerobot/scripts/lerobot_train.py

@@ -22,7 +22,8 @@ import dataclasses
 import logging
 import sys
 import time
-from contextlib import nullcontext
+from collections.abc import Iterator
+from contextlib import contextmanager, nullcontext
 from pprint import pformat
 from typing import TYPE_CHECKING, Any
 
@@ -57,7 +58,7 @@ from lerobot.optim.factory import make_optimizer_and_scheduler
 from lerobot.policies import PreTrainedPolicy, make_policy, make_pre_post_processors
 from lerobot.rewards import make_reward_pre_post_processors
 from lerobot.utils.collate import lerobot_collate_fn
-from lerobot.utils.import_utils import register_third_party_plugins
+from lerobot.utils.import_utils import _peft_available, register_third_party_plugins, require_package
 from lerobot.utils.logging_utils import AverageMeter, MetricsTracker
 from lerobot.utils.random_utils import set_seed
 from lerobot.utils.utils import (
@@ -68,9 +69,38 @@ from lerobot.utils.utils import (
     inside_slurm,
 )
 
+if TYPE_CHECKING or _peft_available:
+    from peft import PeftModel
+else:
+    PeftModel = None
+
 from .lerobot_eval import eval_policy_all
 
 
+@contextmanager
+def _make_eval_envs(cfg: TrainPipelineConfig) -> Iterator[dict[str, dict[int, Any]]]:
+    """Create evaluation environments for one run and always dispose of them."""
+    envs = make_env(
+        cfg.env,
+        n_envs=cfg.eval.batch_size,
+        use_async_envs=cfg.eval.use_async_envs,
+    )
+    try:
+        yield envs
+    finally:
+        close_envs(envs)
+
+
+def _dataloader_worker_kwargs(cfg: TrainPipelineConfig) -> dict[str, Any]:
+    """Return worker-only DataLoader options, disabling them for single-process loading."""
+    workers_enabled = cfg.num_workers > 0
+    return {
+        "prefetch_factor": cfg.prefetch_factor if workers_enabled else None,
+        "persistent_workers": cfg.persistent_workers and workers_enabled,
+        "multiprocessing_context": cfg.dataloader_multiprocessing_context if workers_enabled else None,
+    }
+
+
 def update_policy(
     train_metrics: MetricsTracker,
     policy: PreTrainedPolicy,
@@ -197,8 +227,6 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
     if cfg.job.is_remote:
         return submit_to_hf(cfg)
 
-    from lerobot.utils.import_utils import require_package
-
     require_package("accelerate", extra="training")
     from accelerate import Accelerator
     from accelerate.utils import DistributedDataParallelKwargs, DistributedType
@@ -267,14 +295,6 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
     if not is_main_process:
         dataset, eval_dataset = make_train_eval_datasets(cfg)
 
-    # Create environment used for evaluating checkpoints during training on simulation data.
-    # On real-world data, no need to create an environment as evaluations are done outside train.py,
-    # using the eval.py instead, with gym_dora environment and dora-rs.
-    eval_env = None
-    if cfg.env_eval_freq > 0 and cfg.env is not None and is_main_process:
-        logging.info("Creating env")
-        eval_env = make_env(cfg.env, n_envs=cfg.eval.batch_size, use_async_envs=cfg.eval.use_async_envs)
-
     if cfg.is_reward_model_training:
         if is_main_process:
             logging.info("Creating reward model")
@@ -302,7 +322,7 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
     if cfg.peft is not None:
         if cfg.is_reward_model_training:
             raise ValueError("PEFT is only supported for policy training. ")
-        from peft import PeftModel
+        require_package("peft", extra="peft")
 
         if isinstance(policy, PeftModel):
             logging.info("PEFT adapter already loaded from checkpoint, skipping wrap_with_peft.")
@@ -473,8 +493,7 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
         pin_memory=device.type == "cuda",
         drop_last=False,
         collate_fn=collate_fn,
-        prefetch_factor=cfg.prefetch_factor if cfg.num_workers > 0 else None,
-        persistent_workers=cfg.persistent_workers and cfg.num_workers > 0,
+        **_dataloader_worker_kwargs(cfg),
     )
 
     # Build eval dataloader if a held-out split exists
@@ -500,8 +519,7 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
             pin_memory=device.type == "cuda",
             drop_last=False,
             collate_fn=eval_collate_fn,
-            prefetch_factor=cfg.prefetch_factor if cfg.num_workers > 0 else None,
-            persistent_workers=cfg.persistent_workers and cfg.num_workers > 0,
+            **_dataloader_worker_kwargs(cfg),
         )
 
     # Prepare everything with accelerator
@@ -684,7 +702,7 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
             if is_main_process:
                 step_id = get_step_identifier(step, cfg.steps)
                 logging.info(f"Eval policy at step {step}")
-                with torch.no_grad(), accelerator.autocast():
+                with _make_eval_envs(cfg) as eval_env, torch.no_grad(), accelerator.autocast():
                     eval_info = eval_policy_all(
                         envs=eval_env,  # dict[suite][task_id] -> vec_env
                         policy=accelerator.unwrap_model(policy),
@@ -732,9 +750,6 @@ def train(cfg: TrainPipelineConfig, accelerator: "Accelerator | None" = None):
     if is_main_process:
         progbar.close()
 
-    if eval_env:
-        close_envs(eval_env)
-
     is_fsdp = accelerator.distributed_type == DistributedType.FSDP
     model_state_dict = accelerator.get_state_dict(policy) if is_fsdp else None
     if is_main_process:

src/lerobot/scripts/lerobot_train_tokenizer.py

@@ -45,6 +45,7 @@ lerobot-train-tokenizer \
 """
 
 import json
+import logging
 from dataclasses import dataclass
 from pathlib import Path
 from typing import TYPE_CHECKING
@@ -63,6 +64,9 @@ else:
 from lerobot.configs import NormalizationMode, parser
 from lerobot.datasets import LeRobotDataset
 from lerobot.utils.constants import ACTION, OBS_STATE
+from lerobot.utils.utils import init_logging
+
+logger = logging.getLogger(__name__)
 
 
 @dataclass
@@ -274,11 +278,8 @@ def process_episode(args):
 
         return action_chunks
 
-    except Exception as e:
-        print(f"Error processing episode {ep_idx}: {e}")
-        import traceback
-
-        traceback.print_exc()
+    except Exception:
+        logger.exception("Error processing episode %s", ep_idx)
         return None
 
 
@@ -300,10 +301,10 @@ def train_fast_tokenizer(
     Returns:
         Trained FAST tokenizer
     """
-    print(f"Training FAST tokenizer on {len(action_chunks)} action chunks...")
-    print(f"Action chunk shape: {action_chunks.shape}")
-    print(f"Vocab size: {vocab_size}")
-    print(f"DCT scale: {scale}")
+    logger.info(f"Training FAST tokenizer on {len(action_chunks)} action chunks...")
+    logger.info(f"Action chunk shape: {action_chunks.shape}")
+    logger.info(f"Vocab size: {vocab_size}")
+    logger.info(f"DCT scale: {scale}")
 
     # download the tokenizer source code (not pretrained weights)
     # we'll train a new tokenizer on our own data
@@ -314,7 +315,7 @@ def train_fast_tokenizer(
 
     # train the new tokenizer on our action data using .fit()
     # this trains the BPE tokenizer on DCT coefficients
-    print("Training new tokenizer (this may take a few minutes)...")
+    logger.info("Training new tokenizer (this may take a few minutes)...")
     tokenizer = base_tokenizer.fit(
         action_data_list,
         scale=scale,
@@ -322,21 +323,21 @@ def train_fast_tokenizer(
         time_horizon=action_chunks.shape[1],  # action_horizon
         action_dim=action_chunks.shape[2],  # encoded dimensions
     )
-    print("✓ Tokenizer training complete!")
+    logger.info("✓ Tokenizer training complete!")
 
     # validate it works
     sample_chunk = action_chunks[0]
     encoded = tokenizer(sample_chunk[None])[0]
     if isinstance(encoded, list):
         encoded = np.array(encoded)
-    print(f"Sample encoding: {len(encoded)} tokens for chunk shape {sample_chunk.shape}")
+    logger.info(f"Sample encoding: {len(encoded)} tokens for chunk shape {sample_chunk.shape}")
 
     return tokenizer
 
 
 def compute_compression_stats(tokenizer, action_chunks: np.ndarray):
     """Compute compression statistics."""
-    print("\nComputing compression statistics...")
+    logger.info("\nComputing compression statistics...")
 
     # sample for stats (use max 1000 chunks for speed)
     sample_size = min(1000, len(action_chunks))
@@ -366,12 +367,12 @@ def compute_compression_stats(tokenizer, action_chunks: np.ndarray):
         "max_token_length": float(np.max(token_lengths)),
     }
 
-    print("Compression Statistics:")
-    print(f"  Average compression ratio: {stats['compression_ratio']:.2f}x")
-    print(f"  Mean token length: {stats['mean_token_length']:.1f}")
-    print(f"  P99 token length: {stats['p99_token_length']:.0f}")
-    print(f"  Min token length: {stats['min_token_length']:.0f}")
-    print(f"  Max token length: {stats['max_token_length']:.0f}")
+    logger.info("Compression Statistics:")
+    logger.info(f"  Average compression ratio: {stats['compression_ratio']:.2f}x")
+    logger.info(f"  Mean token length: {stats['mean_token_length']:.1f}")
+    logger.info(f"  P99 token length: {stats['p99_token_length']:.0f}")
+    logger.info(f"  Min token length: {stats['min_token_length']:.0f}")
+    logger.info(f"  Max token length: {stats['max_token_length']:.0f}")
 
     return stats
 
@@ -385,9 +386,9 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
         cfg: TokenizerTrainingConfig dataclass with all configuration parameters
     """
     # load dataset
-    print(f"Loading dataset: {cfg.repo_id}")
+    logger.info(f"Loading dataset: {cfg.repo_id}")
     dataset = LeRobotDataset(repo_id=cfg.repo_id, root=cfg.root)
-    print(f"Dataset loaded: {dataset.num_episodes} episodes, {dataset.num_frames} frames")
+    logger.info(f"Dataset loaded: {dataset.num_episodes} episodes, {dataset.num_frames} frames")
 
     # parse normalization mode
     try:
@@ -397,7 +398,7 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
             f"Invalid normalization_mode: {cfg.normalization_mode}. "
             f"Must be one of: {', '.join([m.value for m in NormalizationMode])}"
         ) from err
-    print(f"Normalization mode: {norm_mode.value}")
+    logger.info(f"Normalization mode: {norm_mode.value}")
 
     # parse encoded dimensions
     encoded_dim_ranges = []
@@ -406,38 +407,38 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
         encoded_dim_ranges.append((start, end))
 
     total_encoded_dims = sum(end - start for start, end in encoded_dim_ranges)
-    print(f"Encoding {total_encoded_dims} dimensions: {cfg.encoded_dims}")
+    logger.info(f"Encoding {total_encoded_dims} dimensions: {cfg.encoded_dims}")
 
     # parse relative dimensions
     relative_dim_list = None
     if cfg.relative_dims is not None and cfg.relative_dims.strip():
         relative_dim_list = [int(d.strip()) for d in cfg.relative_dims.split(",")]
-        print(f"Relative dimensions: {relative_dim_list}")
+        logger.info(f"Relative dimensions: {relative_dim_list}")
     else:
-        print("No relative dimensions specified")
+        logger.info("No relative dimensions specified")
 
-    print(f"Use relative transform: {cfg.use_relative_transform}")
+    logger.info(f"Use relative transform: {cfg.use_relative_transform}")
     if cfg.use_relative_transform and (relative_dim_list is None or len(relative_dim_list) == 0):
-        print(
+        logger.warning(
             "Warning: use_relative_transform=True but no relative_dims specified. "
             "No relative transform will be applied."
         )
 
-    print(f"Action horizon: {cfg.action_horizon}")
-    print(f"State key: {cfg.state_key}")
+    logger.info(f"Action horizon: {cfg.action_horizon}")
+    logger.info(f"State key: {cfg.state_key}")
 
     # determine episodes to process
     num_episodes = dataset.num_episodes
     if cfg.max_episodes is not None:
         num_episodes = min(cfg.max_episodes, num_episodes)
 
-    print(f"Processing {num_episodes} episodes...")
+    logger.info(f"Processing {num_episodes} episodes...")
 
     # process episodes sequentially (to avoid pickling issues with dataset)
     all_chunks = []
     for ep_idx in range(num_episodes):
         if ep_idx % 10 == 0:
-            print(f"  Processing episode {ep_idx}/{num_episodes}...")
+            logger.info(f"  Processing episode {ep_idx}/{num_episodes}...")
 
         chunks = process_episode(
             (
@@ -455,19 +456,19 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
 
     # concatenate all chunks
     all_chunks = np.concatenate(all_chunks, axis=0)
-    print(f"Collected {len(all_chunks)} action chunks")
+    logger.info(f"Collected {len(all_chunks)} action chunks")
 
     # extract only encoded dimensions FIRST (before normalization)
     encoded_chunks = []
     for start, end in encoded_dim_ranges:
         encoded_chunks.append(all_chunks[:, :, start:end])
     encoded_chunks = np.concatenate(encoded_chunks, axis=-1)  # [N, H, D_encoded]
-    print(f"Extracted {encoded_chunks.shape[-1]} encoded dimensions")
+    logger.info(f"Extracted {encoded_chunks.shape[-1]} encoded dimensions")
 
     # apply normalization to encoded dimensions
-    print("\nBefore normalization - overall stats:")
-    print(f"  Min: {np.min(encoded_chunks):.4f}, Max: {np.max(encoded_chunks):.4f}")
-    print(f"  Mean: {np.mean(encoded_chunks):.4f}, Std: {np.std(encoded_chunks):.4f}")
+    logger.info("\nBefore normalization - overall stats:")
+    logger.info(f"  Min: {np.min(encoded_chunks):.4f}, Max: {np.max(encoded_chunks):.4f}")
+    logger.info(f"  Mean: {np.mean(encoded_chunks):.4f}, Std: {np.std(encoded_chunks):.4f}")
 
     # get normalization stats from dataset
     norm_stats = dataset.meta.stats
@@ -489,9 +490,9 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
                     encoded_stats[stat_name] = stat_array[encoded_dim_indices]
 
         if encoded_stats:
-            print(f"\nNormalization stats for encoded dimensions (mode: {norm_mode.value}):")
+            logger.info(f"\nNormalization stats for encoded dimensions (mode: {norm_mode.value}):")
             for stat_name, stat_values in encoded_stats.items():
-                print(
+                logger.info(
                     f"  {stat_name}: shape={stat_values.shape}, "
                     f"range=[{np.min(stat_values):.4f}, {np.max(stat_values):.4f}]"
                 )
@@ -499,27 +500,27 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
             # apply normalization based on mode
             try:
                 encoded_chunks = apply_normalization(encoded_chunks, encoded_stats, norm_mode, eps=1e-8)
-                print(f"\nApplied {norm_mode.value} normalization")
+                logger.info(f"\nApplied {norm_mode.value} normalization")
             except ValueError as e:
-                print(f"Warning: {e}. Using raw actions without normalization.")
+                logger.warning(f"Warning: {e}. Using raw actions without normalization.")
 
-            print("\nAfter normalization - overall stats:")
-            print(f"  Min: {np.min(encoded_chunks):.4f}, Max: {np.max(encoded_chunks):.4f}")
-            print(f"  Mean: {np.mean(encoded_chunks):.4f}, Std: {np.std(encoded_chunks):.4f}")
+            logger.info("\nAfter normalization - overall stats:")
+            logger.info(f"  Min: {np.min(encoded_chunks):.4f}, Max: {np.max(encoded_chunks):.4f}")
+            logger.info(f"  Mean: {np.mean(encoded_chunks):.4f}, Std: {np.std(encoded_chunks):.4f}")
 
-            print("\nPer-dimension stats (after normalization):")
+            logger.info("\nPer-dimension stats (after normalization):")
             for d in range(encoded_chunks.shape[-1]):
                 dim_data = encoded_chunks[:, :, d]
-                print(
+                logger.info(
                     f"  Dim {d}: min={np.min(dim_data):7.4f}, max={np.max(dim_data):7.4f}, "
                     f"mean={np.mean(dim_data):7.4f}, std={np.std(dim_data):7.4f}"
                 )
         else:
-            print("Warning: Could not extract stats for encoded dimensions, using raw actions")
+            logger.warning("Warning: Could not extract stats for encoded dimensions, using raw actions")
     else:
-        print("Warning: No normalization stats found in dataset, using raw actions")
+        logger.warning("Warning: No normalization stats found in dataset, using raw actions")
 
-    print(f"Encoded chunks shape: {encoded_chunks.shape}")
+    logger.info(f"Encoded chunks shape: {encoded_chunks.shape}")
 
     # train FAST tokenizer
     tokenizer = train_fast_tokenizer(
@@ -561,8 +562,8 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
     with open(output_path / "metadata.json", "w") as f:
         json.dump(metadata, f, indent=2)
 
-    print(f"\nSaved FAST tokenizer to {output_path}")
-    print(f"Metadata: {json.dumps(metadata, indent=2)}")
+    logger.info(f"\nSaved FAST tokenizer to {output_path}")
+    logger.info(f"Metadata: {json.dumps(metadata, indent=2)}")
 
     # push to Hugging Face Hub if requested
     if cfg.push_to_hub:
@@ -570,10 +571,10 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
         hub_repo_id = cfg.hub_repo_id
         if hub_repo_id is None:
             hub_repo_id = output_path.name
-            print(f"\nNo hub_repo_id provided, using: {hub_repo_id}")
+            logger.info(f"\nNo hub_repo_id provided, using: {hub_repo_id}")
 
-        print(f"\nPushing tokenizer to Hugging Face Hub: {hub_repo_id}")
-        print(f"   Private: {cfg.hub_private}")
+        logger.info(f"\nPushing tokenizer to Hugging Face Hub: {hub_repo_id}")
+        logger.info(f"   Private: {cfg.hub_private}")
 
         try:
             # use the tokenizer's push_to_hub method
@@ -593,14 +594,15 @@ def train_tokenizer(cfg: TokenizerTrainingConfig):
                 commit_message="Upload tokenizer metadata",
             )
 
-            print(f"Successfully pushed tokenizer to: https://huggingface.co/{hub_repo_id}")
+            logger.info(f"Successfully pushed tokenizer to: https://huggingface.co/{hub_repo_id}")
         except Exception as e:
-            print(f"Error pushing to hub: {e}")
-            print("   Make sure you're logged in with `huggingface-cli login`")
+            logger.error(f"Error pushing to hub: {e}")
+            logger.error("   Make sure you're logged in with `huggingface-cli login`")
 
 
 def main():
     """CLI entry point that parses arguments and runs the tokenizer training."""
+    init_logging()
     train_tokenizer()
 
 

src/lerobot/teleoperators/gamepad/configuration_gamepad.py

@@ -23,3 +23,5 @@ from ..config import TeleoperatorConfig
 @dataclass
 class GamepadTeleopConfig(TeleoperatorConfig):
     use_gripper: bool = True
+    # Use hidapi instead of pygame for controllers that pygame cannot detect reliably.
+    hidapi_fallback: bool = False

src/lerobot/teleoperators/gamepad/teleop_gamepad.py

@@ -14,6 +14,7 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
+import logging
 import sys
 from enum import IntEnum
 from typing import Any
@@ -27,6 +28,8 @@ from ..teleoperator import Teleoperator
 from ..utils import TeleopEvents
 from .configuration_gamepad import GamepadTeleopConfig
 
+logger = logging.getLogger(__name__)
+
 
 class GripperAction(IntEnum):
     CLOSE = 0
@@ -56,6 +59,13 @@ class GamepadTeleop(Teleoperator):
 
         self.gamepad = None
 
+        self.hidapi_fallback = config.hidapi_fallback
+        if sys.platform == "darwin" and not self.hidapi_fallback:
+            logger.warning(
+                "On macOS, pygame may not reliably detect input from some controllers. "
+                "If you experience issues, set `hidapi_fallback=true`."
+            )
+
     @property
     def action_features(self) -> dict:
         if self.config.use_gripper:
@@ -76,9 +86,7 @@ class GamepadTeleop(Teleoperator):
         return {}
 
     def connect(self) -> None:
-        # use HidApi for macos
-        if sys.platform == "darwin":
-            # NOTE: On macOS, pygame doesn’t reliably detect input from some controllers so we fall back to hidapi
+        if self.hidapi_fallback:
             from .gamepad_utils import GamepadControllerHID as Gamepad
         else:
             from .gamepad_utils import GamepadController as Gamepad

src/lerobot/teleoperators/phone/teleop_phone.py

@@ -171,7 +171,13 @@ class IOSPhone(BasePhone, Teleoperator):
         # HEBI provides orientation in w, x, y, z format.
         # Scipy's Rotation expects x, y, z, w.
         quat_xyzw = np.concatenate((ar_quat[1:], [ar_quat[0]]))  # wxyz to xyzw
+        # ARKit can emit zero/NaN quaternions before tracking is ready or on a
+        # dropped packet. Rotation.from_quat now rejects those; degrade the same
+        # way as a missing pose so teleop stays alive mid-session.
+        try:
             rot = Rotation.from_quat(quat_xyzw)
+        except ValueError:
+            return False, None, None, None
         pos = ar_pos - rot.apply(self.config.camera_offset)
         return True, pos, rot, pose
 

src/lerobot/teleoperators/so_leader/config_so_leader.py

@@ -29,6 +29,12 @@ class SOLeaderConfig:
     # Whether to use degrees for angles
     use_degrees: bool = True
 
+    # Number of extra attempts when a `sync_read` of the motors fails. Feetech buses can occasionally
+    # return a corrupted status packet ("Incorrect status packet!"), especially when several joints move
+    # at once, which otherwise aborts the teleoperation loop. Retries are immediate (no sleep) and only
+    # happen on failure, so the steady-state read cost is unchanged.
+    num_read_retries: int = 2
+
 
 @TeleoperatorConfig.register_subclass("so101_leader")
 @TeleoperatorConfig.register_subclass("so100_leader")

src/lerobot/teleoperators/so_leader/so_leader.py

@@ -145,7 +145,7 @@ class SOLeader(Teleoperator):
     @check_if_not_connected
     def get_action(self) -> dict[str, float]:
         start = time.perf_counter()
-        action = self.bus.sync_read("Present_Position")
+        action = self.bus.sync_read("Present_Position", num_retry=self.config.num_read_retries)
         action = {f"{motor}.pos": val for motor, val in action.items()}
         dt_ms = (time.perf_counter() - start) * 1e3
         logger.debug(f"{self} read action: {dt_ms:.1f}ms")

src/lerobot/transforms/transforms.py

@@ -41,7 +41,7 @@ class RandomSubsetApply(Transform):
 
     def __init__(
         self,
-        transforms: Sequence[Callable],
+        transforms: Sequence[Callable[..., Any]],
         p: list[float] | None = None,
         n_subset: int | None = None,
         random_order: bool = False,
@@ -50,7 +50,7 @@ class RandomSubsetApply(Transform):
         if not isinstance(transforms, Sequence):
             raise TypeError("Argument transforms should be a sequence of callables")
         if p is None:
-            p = [1] * len(transforms)
+            p = [1.0] * len(transforms)
         elif len(p) != len(transforms):
             raise ValueError(
                 f"Length of p doesn't match the number of transforms: {len(p)} != {len(transforms)}"
@@ -69,7 +69,7 @@ class RandomSubsetApply(Transform):
         self.n_subset = n_subset
         self.random_order = random_order
 
-        self.selected_transforms = None
+        self.selected_transforms: list[Callable[..., Any]] = []
 
     def forward(self, *inputs: Any) -> Any:
         needs_unpacking = len(inputs) > 1
@@ -119,7 +119,7 @@ class SharpnessJitter(Transform):
         super().__init__()
         self.sharpness = self._check_input(sharpness)
 
-    def _check_input(self, sharpness):
+    def _check_input(self, sharpness: float | Sequence[float]) -> tuple[float, float]:
         if isinstance(sharpness, (int | float)):
             if sharpness < 0:
                 raise ValueError("If sharpness is a single number, it must be non negative.")
@@ -215,7 +215,7 @@ class ImageTransformsConfig:
     )
 
 
-def make_transform_from_config(cfg: ImageTransformConfig):
+def make_transform_from_config(cfg: ImageTransformConfig) -> Transform:
     if cfg.type == "SharpnessJitter":
         return SharpnessJitter(**cfg.kwargs)
 
@@ -236,8 +236,8 @@ class ImageTransforms(Transform):
         super().__init__()
         self._cfg = cfg
 
-        self.weights = []
-        self.transforms = {}
+        self.weights: list[float] = []
+        self.transforms: dict[str, Transform] = {}
         for tf_name, tf_cfg in cfg.tfs.items():
             if tf_cfg.weight <= 0.0:
                 continue

src/lerobot/utils/dataset_visualization_utils.py

@@ -0,0 +1,99 @@
+# Copyright 2024 The HuggingFace Inc. team. All rights reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""Shared helpers for visualizing scalar features from a LeRobot dataset."""
+
+from collections.abc import Iterable, Mapping
+from typing import TYPE_CHECKING
+
+import numpy as np
+import torch
+
+from .constants import ACTION, DEFAULT_FEATURES, DONE, OBS_STATE, REWARD, SUCCESS
+
+if TYPE_CHECKING:
+    from lerobot.datasets import LeRobotDataset
+
+
+METADATA_KEYS = {*DEFAULT_FEATURES, "task"}
+KNOWN_SCALAR_KEYS = {DONE, REWARD, SUCCESS}
+SCALAR_DTYPE_KINDS = {"b", "i", "u", "f"}
+
+
+def is_scalar_feature(feature: Mapping) -> bool:
+    """Return whether a feature schema describes a numeric or boolean scalar."""
+
+    dtype = feature.get("dtype")
+    if not isinstance(dtype, str):
+        return False
+    try:
+        dtype_kind = np.dtype(dtype).kind
+    except (TypeError, ValueError):
+        return False
+    if dtype_kind not in SCALAR_DTYPE_KINDS:
+        return False
+
+    shape = feature.get("shape")
+    if shape is None:
+        return True
+    if isinstance(shape, int):
+        return shape == 1
+    if not isinstance(shape, (list, tuple)):
+        return False
+    return len(shape) == 0 or (len(shape) == 1 and shape[0] == 1)
+
+
+def get_extra_scalar_keys(dataset: "LeRobotDataset", additional_known_keys: Iterable[str] = ()) -> list[str]:
+    """Return scalar feature keys not handled by the visualizer's standard paths."""
+
+    known_keys = {
+        ACTION,
+        OBS_STATE,
+        *KNOWN_SCALAR_KEYS,
+        *METADATA_KEYS,
+        *additional_known_keys,
+        *dataset.meta.camera_keys,
+    }
+    return [
+        key
+        for key, feature in dataset.features.items()
+        if key not in known_keys and is_scalar_feature(feature)
+    ]
+
+
+def is_scalar_like(value: object) -> bool:
+    """Return whether a runtime value contains exactly one numeric or boolean scalar."""
+
+    if isinstance(value, torch.Tensor):
+        return value.numel() == 1 and not value.is_complex()
+    if isinstance(value, np.ndarray):
+        return value.size == 1 and value.dtype.kind in SCALAR_DTYPE_KINDS
+    return np.isscalar(value) and np.asarray(value).dtype.kind in SCALAR_DTYPE_KINDS
+
+
+def scalar_to_float(value: object) -> float:
+    """Convert a scalar-like tensor, array, or Python value to ``float``."""
+
+    return float(value.item() if hasattr(value, "item") else value)
+
+
+def get_scalar_values(sample: Mapping, keys: Iterable[str]) -> dict[str, float]:
+    """Select and convert scalar-like values from ``sample`` for the requested keys."""
+
+    values = {}
+    for key in keys:
+        value = sample.get(key)
+        if value is not None and is_scalar_like(value):
+            values[key] = scalar_to_float(value)
+    return values

src/lerobot/utils/device_utils.py

@@ -15,7 +15,6 @@
 # limitations under the License.
 
 import logging
-from contextlib import nullcontext
 
 import torch
 
@@ -38,16 +37,25 @@ def auto_select_torch_device() -> torch.device:
 
 # TODO(Steven): Remove log. log shouldn't be an argument, this should be handled by the logger level
 def get_safe_torch_device(try_device: str, log: bool = False) -> torch.device:
-    """Given a string, return a torch.device with checks on whether the device is available."""
+    """Given a string, return a torch.device with checks on whether the device is available.
+
+    Raises:
+        ValueError: If the requested device family is known but not available on
+            this machine (``AssertionError`` was previously used and is easy to
+            mistake for a programmer bug under ``python -O`` where asserts vanish).
+    """
     try_device = str(try_device)
     if try_device.startswith("cuda"):
-        assert torch.cuda.is_available()
+        if not torch.cuda.is_available():
+            raise ValueError(f"Requested device {try_device!r} but CUDA is not available.")
         device = torch.device(try_device)
     elif try_device == "mps":
-        assert torch.backends.mps.is_available()
+        if not torch.backends.mps.is_available():
+            raise ValueError("Requested device 'mps' but MPS is not available.")
         device = torch.device("mps")
     elif try_device == "xpu":
-        assert torch.xpu.is_available()
+        if not torch.xpu.is_available():
+            raise ValueError("Requested device 'xpu' but XPU is not available.")
         device = torch.device("xpu")
     elif try_device == "cpu":
         device = torch.device("cpu")
@@ -122,25 +130,3 @@ def is_amp_available(device: str):
         return False
     else:
         raise ValueError(f"Unknown device '{device}.")
-
-
-def get_autocast_context(device_type: str, dtype: torch.dtype = torch.bfloat16):
-    """Return a device-safe autocast context manager.
-
-    Hardcoding `torch.autocast(dtype=torch.bfloat16)` breaks on backends without AMP
-    (MPS) and silently misbehaves on pre-Ampere CUDA GPUs that lack bf16 support. This
-    picks a safe context per device:
-      - no AMP support (e.g. mps): `nullcontext()` (run in the tensors' native dtype)
-      - CUDA requesting bf16 on compute capability < 8.0 (pre-Ampere): fall back to fp16
-      - otherwise: `torch.autocast(device_type, dtype)`
-    """
-    if not is_amp_available(device_type):
-        return nullcontext()
-    if (
-        device_type == "cuda"
-        and dtype == torch.bfloat16
-        and torch.cuda.is_available()
-        and torch.cuda.get_device_capability()[0] < 8
-    ):
-        dtype = torch.float16
-    return torch.autocast(device_type=device_type, dtype=dtype)

src/lerobot/utils/foxglove_visualization.py

@@ -23,6 +23,7 @@ importing from here directly. Requires the ``viz`` extra (``pip install 'lerobot
 import logging
 import numbers
 import time
+from collections.abc import Iterable, Mapping
 
 import cv2
 import numpy as np
@@ -41,6 +42,7 @@ from .constants import (
     SUCCESS,
     TRUNCATED,
 )
+from .dataset_visualization_utils import get_extra_scalar_keys, get_scalar_values
 from .import_utils import require_package
 
 # Static schema shared by all scalar topics. Each message carries a flat list of ``{label, value}``
@@ -405,6 +407,24 @@ def _frame_to_scalars(sample: dict, key: str, labels: list[str] | None = None) -
     return _labeled_scalars(name, arr.flatten(), labels)
 
 
+def _dataset_frame_scalar_groups(
+    sample: Mapping, extra_scalar_keys: Iterable[str]
+) -> tuple[dict[str, float], dict[str, float]]:
+    """Return standard episode scalars and custom scalar features for one dataset frame."""
+
+    episode_scalars = {}
+    for feature, label in (
+        (DONE, "done"),
+        (TRUNCATED, "truncated"),
+        (REWARD, "reward"),
+        (SUCCESS, "success"),
+    ):
+        value = sample.get(feature)
+        if value is not None:
+            episode_scalars[label] = float(value)
+    return episode_scalars, get_scalar_values(sample, extra_scalar_keys)
+
+
 def serve_foxglove_dataset_playback(
     dataset,
     episode_index: int,
@@ -452,6 +472,7 @@ def serve_foxglove_dataset_playback(
         raise ValueError("Cannot visualize an empty episode.")
     first_ns, last_ns = times_ns[0], times_ns[-1]
     camera_keys = list(dataset.meta.camera_keys)
+    extra_scalar_keys = get_extra_scalar_keys(dataset, additional_known_keys=(TRUNCATED,))
     # Dataset-wide q01/q99 depth bounds (fallback min/max) used to normalize depth to [0, 1].
     depth_ranges: dict[str, tuple[float, float]] = {}
     for key in dataset.meta.depth_keys:
@@ -500,17 +521,14 @@ def serve_foxglove_dataset_playback(
             channels=channels,
             log_time=log_time,
         )
-        episode_scalars = {}
-        for feat, label in (
-            (DONE, "done"),
-            (TRUNCATED, "truncated"),
-            (REWARD, "reward"),
-            (SUCCESS, "success"),
-        ):
-            v = sample.get(feat)
-            if v is not None:
-                episode_scalars[label] = float(v)
+        episode_scalars, extra_scalars = _dataset_frame_scalar_groups(sample, extra_scalar_keys)
         _log_foxglove_scalars("/episode/state", episode_scalars, channels=channels, log_time=log_time)
+        _log_foxglove_scalars(
+            "/episode/extras",
+            extra_scalars,
+            channels=channels,
+            log_time=log_time,
+        )
 
     lock = threading.Lock()
     stop_event = threading.Event()

src/lerobot/utils/io_utils.py

@@ -32,21 +32,21 @@ def load_json(fpath: Path) -> Any:
     Returns:
         Any: The data loaded from the JSON file.
     """
-    with open(fpath) as f:
+    with open(fpath, encoding="utf-8") as f:
         return json.load(f)
 
 
-def write_json(data: dict, fpath: Path) -> None:
-    """Write data to a JSON file.
+def write_json(data: JsonLike, fpath: Path) -> None:
+    """Write JSON-serializable data to a file.
 
     Creates parent directories if they don't exist.
 
     Args:
-        data (dict): The dictionary to write.
+        data: JSON-serializable data to write.
         fpath (Path): The path to the output JSON file.
     """
     fpath.parent.mkdir(exist_ok=True, parents=True)
-    with open(fpath, "w") as f:
+    with open(fpath, "w", encoding="utf-8") as f:
         json.dump(data, f, indent=4, ensure_ascii=False)