run precommit and fix issues

Fast tests currently fail because of the missing dependency.

.github/PULL_REQUEST_TEMPLATE.md

@@ -22,21 +22,20 @@ Short, imperative summary (e.g., "fix(robots): handle None in sensor parser"). S
 - Short, concrete bullets of the modifications (files/behaviour).
 - Short note if this introduces breaking changes and migration steps.
 
-## How was this tested (or how to run locally)
+## How was this tested
 
 - Tests added: list new tests or test files.
 - Manual checks / dataset runs performed.
-- Instructions for the reviewer
 
-Example:
+## How to run locally (reviewer)
 
-- Ran the relevant tests:
+- Run the relevant tests:
 
   ```bash
   pytest -q tests/ -k <keyword>
   ```
 
-- Reproduce with a quick example or CLI (if applicable):
+- Run a quick example or CLI (if applicable):
 
   ```bash
   lerobot-train --some.option=true

.github/workflows/documentation.yml

@@ -18,11 +18,6 @@ name: Documentation
 on:
   # Allows running this workflow manually from the Actions tab
   workflow_dispatch:
-    inputs:
-      version:
-        description: 'Version tag (e.g. v0.1.2) - Leave empty for standard main build'
-        required: false
-        type: string
 
   # Triggers the workflow on push events to main for the docs folder
   push:
@@ -38,9 +33,6 @@ on:
     paths:
       - "docs/**"
 
-  release:
-    types: [published]
-
 # Ensures that only the latest commit for a PR or branch is built, canceling older runs.
 concurrency:
   group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
@@ -51,7 +43,7 @@ jobs:
   build_main_docs:
     name: Build Main Docs
     if: >
-      (github.event_name == 'push' || github.event_name == 'workflow_dispatch' || github.event_name == 'release') &&
+      (github.event_name == 'push' || github.event_name == 'workflow_dispatch') &&
       github.repository == 'huggingface/lerobot'
     permissions:
       contents: read
@@ -59,13 +51,7 @@ jobs:
     with:
       commit_sha: ${{ github.sha }}
       package: lerobot
-      additional_args: >-
-        --not_python_module
-        ${{
-          (github.event_name == 'release' && format('--version {0}', github.event.release.tag_name)) ||
-          (inputs.version != '' && format('--version {0}', inputs.version)) ||
-          ''
-        }}
+      additional_args: --not_python_module
     secrets:
       token: ${{ secrets.HUGGINGFACE_PUSH }}
       hf_token: ${{ secrets.HF_DOC_BUILD_PUSH }}

.github/workflows/fast_tests.yml

@@ -27,7 +27,6 @@ on:
       - "tests/**"
       - ".github/workflows/**"
       - "pyproject.toml"
-      - "uv.lock"
       - "Makefile"
   push:
     branches:
@@ -37,7 +36,6 @@ on:
       - "tests/**"
       - ".github/workflows/**"
       - "pyproject.toml"
-      - "uv.lock"
       - "Makefile"
 
 permissions:
@@ -46,7 +44,7 @@ permissions:
 # Sets up the environment variables
 env:
   UV_VERSION: "0.8.0"
-  PYTHON_VERSION: "3.12"
+  PYTHON_VERSION: "3.10"
 
 # Ensures that only the latest commit for a PR or branch is built, canceling older runs.
 concurrency:
@@ -63,9 +61,8 @@ jobs:
       MUJOCO_GL: egl
       HF_HOME: /mnt/cache/.cache/huggingface
       HF_LEROBOT_HOME: /mnt/cache/.cache/huggingface/lerobot
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
         with:
           persist-credentials: false
           lfs: true
@@ -90,13 +87,7 @@ jobs:
           python-version: ${{ env.PYTHON_VERSION }}
 
       - name: Install lerobot with test extras
-        run: uv sync --locked --extra "test"
-
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          uv run hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          uv run hf auth whoami
+        run: uv sync --extra "test"
 
       - name: Run pytest
         run: uv run pytest tests -vv --maxfail=10

.github/workflows/full_tests.yml

@@ -29,7 +29,6 @@ on:
       - "tests/**"
       - ".github/workflows/**"
       - "pyproject.toml"
-      - "uv.lock"
       - "Makefile"
 
 permissions:
@@ -38,7 +37,7 @@ permissions:
 # Sets up the environment variables
 env:
   UV_VERSION: "0.8.0"
-  PYTHON_VERSION: "3.12"
+  PYTHON_VERSION: "3.10"
   DOCKER_IMAGE_NAME: huggingface/lerobot-gpu
 
 # Ensures that only the latest action is built, canceling older runs.
@@ -61,9 +60,8 @@ jobs:
       MUJOCO_GL: egl
       HF_HOME: /mnt/cache/.cache/huggingface
       HF_LEROBOT_HOME: /mnt/cache/.cache/huggingface/lerobot
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
         with:
           lfs: true
           persist-credentials: false
@@ -87,13 +85,7 @@ jobs:
           python-version: ${{ env.PYTHON_VERSION }}
 
       - name: Install lerobot with all extras
-        run: uv sync --locked --extra all # TODO(Steven): Make flash-attn optional
-
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          uv run hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          uv run hf auth whoami
+        run: uv sync --all-extras --no-extra groot --no-extra wallx # TODO(Steven): Make flash-attn optional
 
       - name: Run pytest (all extras)
         run: uv run pytest tests -vv --maxfail=10
@@ -109,11 +101,9 @@ jobs:
     runs-on:
       group: aws-general-8-plus
     if: |
-      github.repository == 'huggingface/lerobot' && (
-        (github.event_name == 'pull_request_review' && github.event.review.state == 'approved' && github.event.pull_request.head.repo.fork == false) ||
-        github.event_name == 'push' ||
-        github.event_name == 'workflow_dispatch'
-      )
+      (github.event_name == 'pull_request_review' && github.event.review.state == 'approved' && github.event.pull_request.head.repo.fork == false) ||
+      github.event_name == 'push' ||
+      github.event_name == 'workflow_dispatch'
     outputs:
       image_tag: ${{ steps.set_tag.outputs.image_tag }}
     env:
@@ -137,7 +127,7 @@ jobs:
           sudo apt-get update
           sudo apt-get install git-lfs
           git lfs install
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
         with:
           lfs: true
           persist-credentials: false
@@ -170,7 +160,6 @@ jobs:
       HF_LEROBOT_HOME: /home/user_lerobot/.cache/huggingface/lerobot
       TORCH_HOME: /home/user_lerobot/.cache/torch
       TRITON_CACHE_DIR: /home/user_lerobot/.cache/triton
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
     container:
       image: ${{ needs.build-and-push-docker.outputs.image_tag }} # zizmor: ignore[unpinned-images]
       options: --gpus all --shm-size "16gb"
@@ -182,13 +171,6 @@ jobs:
         shell: bash
         working-directory: /lerobot
     steps:
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          hf auth whoami
-      - name: Fix ptxas permissions
-        run: chmod +x /lerobot/.venv/lib/python3.12/site-packages/triton/backends/nvidia/bin/ptxas
       - name: Run pytest on GPU
         run: pytest tests -vv --maxfail=10
       - name: Run end-to-end tests
@@ -204,18 +186,15 @@ jobs:
     steps:
       - name: Get Docker Hub Token and Delete Image
         # zizmor: ignore[template-injection]
-        env:
-          DOCKERHUB_LEROBOT_USERNAME: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
-          DOCKERHUB_LEROBOT_PASSWORD: ${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}
-          IMAGE_FULL: ${{ needs.build-and-push-docker.outputs.image_tag }}
         run: |
-          IMAGE_NAME=$(echo "$IMAGE_FULL" | cut -d':' -f1)
-          IMAGE_TAG=$(echo "$IMAGE_FULL" | cut -d':' -f2-)
+          IMAGE_NAME=$(echo "${{ needs.build-and-push-docker.outputs.image_tag }}" | cut -d':' -f1)
+          IMAGE_TAG=$(echo "${{ needs.build-and-push-docker.outputs.image_tag }}" | cut -d':' -f2)
+
           echo "Attempting to delete image: $IMAGE_NAME:$IMAGE_TAG"
 
           TOKEN=$(curl -s -H "Content-Type: application/json" \
                        -X POST \
-                       -d "{\"username\": \"$DOCKERHUB_LEROBOT_USERNAME\", \"password\": \"$DOCKERHUB_LEROBOT_PASSWORD\"}" \
+                       -d '{"username": "${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}", "password": "${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}"}' \
                        https://hub.docker.com/v2/users/login/ | jq -r .token)
 
           if [ "$TOKEN" == "null" ] || [ -z "$TOKEN" ]; then
@@ -226,7 +205,7 @@ jobs:
           HTTP_RESPONSE=$(curl -s -o /dev/null -w "%{http_code}" \
                                -H "Authorization: JWT ${TOKEN}" \
                                -X DELETE \
-                               https://hub.docker.com/v2/repositories/${IMAGE_NAME}/tags/$IMAGE_TAG)
+                               https://hub.docker.com/v2/repositories/${IMAGE_NAME}/tags/${IMAGE_TAG}/)
 
           if [ "$HTTP_RESPONSE" -eq 204 ]; then
             echo "Successfully deleted Docker image tag: $IMAGE_NAME:$IMAGE_TAG"

.github/workflows/issue_labeler.yml

@@ -42,26 +42,38 @@ jobs:
 
             // Keyword Heuristics
 
-            if (matches(/\b(bug|error|crash|exception)\b/i)) labelsToAdd.add('bug');
-            if (matches(/\b(new feature|enhancement|improvement|proposal|feature request)\b/i)) labelsToAdd.add('enhancement');
-            if (matches(/\b(question|how to|clarify|explain|how do i|help me|question about)\b/i)) labelsToAdd.add('question');
-            if (matches(/\b(documentation|docs?|readme|tutorial|wiki|typo|docstring)\b/i)) labelsToAdd.add('documentation');
-            if (matches(/\b(example|sample|demo|notebook)s?\b/i)) labelsToAdd.add('examples');
+            // Domain Specific
+            if (matches(/\b(bug|error|issue|fault|crash|exception)\b/i)) labelsToAdd.add('bug');
+            if (matches(/\b(feature|enhancement|improvement|support|implement|proposal)\b/i)) labelsToAdd.add('enhancement');
+            if (matches(/\b(question|help|how to||clarify|explain|unclear)\b/i)) labelsToAdd.add('question');
+            if (matches(/\b(maintenance|documentation|docs|readme|tutorial|guide|wiki)\b/i)) labelsToAdd.add('documentation');
+            if (matches(/\b(example|script|sample|demo|notebook)s?\b/i)) labelsToAdd.add('examples');
             if (matches(/\b(datasets?|data loader|data augmentation|data preprocessing)\b/i)) labelsToAdd.add('dataset');
             if (matches(/\b(mujoco|isaac|simulation|sim)\b/i)) labelsToAdd.add('simulation');
-            if (matches(/\b(train|training|optimizer|gradient|wandb|sac)\b/i)) labelsToAdd.add('training');
-            if (matches(/\b(rerun|plot|render|rendering|visualizer)/i)) labelsToAdd.add('visualization');
-            if (matches(/\b(cameras?|opencv|realsense|lidars?|sensors?|imus?|microphones?|rgbd|encoders?)\b/i)) labelsToAdd.add('sensors');
-            if (matches(/\b(urdf|actuators?|calibration|end-effector|kinematics)\b/i)) labelsToAdd.add('robots');
+            if (matches(/\b(train|training|loss|optimizer|backward|gradient|wandb|sac)\b/i)) labelsToAdd.add('training');
+            if (matches(/\b(rerun|plot|video|render|visualiz|gif)/i)) labelsToAdd.add('visualization');
+            if (matches(/\b(camera|realsense|lidar|depth|sensor|imu|microphone|rgbd)\b/i)) labelsToAdd.add('sensors');
+            if (matches(/\b(aloha|koch|so-100|so100|mobile|teleop|manipulator|robots?)\b/i)) labelsToAdd.add('robots');
             if (matches(/\b(teleop|teleoperator|controller|leader|follower|joystick|gamepad)\b/i)) labelsToAdd.add('teleoperators');
-            if (matches(/\b(policy|policies|model?)\b/i)) labelsToAdd.add('policies');
-            if (matches(/\b(processor|pipeline|preprocessor|postprocessor)s?\b/i)) labelsToAdd.add('processor');
-            if (matches(/\b(eval|evaluate|evaluation|metrics?|score|benchmarks?)\b/i)) labelsToAdd.add('evaluation');
+            if (matches(/\b(policy|policies|p0licy)\b/i)) labelsToAdd.add('policies');
+            if (matches(/\b(processors?|pipeline)\b/i)) labelsToAdd.add('processor');
+            if (matches(/\b(eval|evaluate|evaluation|metrics?|score|benchmark)\b/i)) labelsToAdd.add('evaluation');
+
+            // Infrastructure & Code Quality
             if (matches(/\b(tests?|pytest|unittest|failing test)\b/i)) labelsToAdd.add('tests');
-            if (matches(/\b(ci|github actions?|github workflows?|gha|docker|pypi)\b/i)) labelsToAdd.add('CI');
-            if (matches(/\b(perf|latency|throughput|fps|speed|performance|slow|fast|slower|faster|memory usage)\b/i)) labelsToAdd.add('performance');
-            if (matches(/\b(dependency|dependencies|pip|install error|importerror|package not found|pyproject)\b/i)) labelsToAdd.add('dependencies');
-            if (matches(/\b(configuration|config|arguments?|input feature|dracuss)\b/i)) labelsToAdd.add('configuration');
+            if (matches(/\b(ci|github actions|workflow|gha|actions?|pipeline)\b/i)) {
+              labelsToAdd.add('CI');
+              labelsToAdd.add('github_actions');
+            }
+            if (matches(/\b(perf|latency|throughput|fps|speed|performance)\b/i)) labelsToAdd.add('performance');
+            if (matches(/\b(dependency|requirements|pip|conda|install error|importerror|package not found)\b/i)) labelsToAdd.add('dependencies');
+            if (matches(/\b(python|pyproject|requirements(\.txt)?|pip install|typing error)\b/i)) labelsToAdd.add('python');
+
+            // Documentation & Meta
+            if (matches(/\b(doc|documentation|docs|readme|typo|how to)\b/i)) labelsToAdd.add('documentation');
+            if (matches(/\b(refactor|cleanup|restructure|rename|modernize code)\b/i)) labelsToAdd.add('refactor');
+            if (matches(/\b(release|changelog|version bump|cut a release|tag v)\b/i)) labelsToAdd.add('release');
+            if (matches(/\b(breaking change|major change)\b/i)) labelsToAdd.add('breaking change');
 
             // Apply Labels
             const labels = Array.from(labelsToAdd).filter(Boolean);

.github/workflows/latest_deps_tests.yml

@@ -1,309 +0,0 @@
-# Copyright 2025 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.
-
-# This workflow tests the project against the latest upstream dependencies
-# (within pyproject.toml constraints) and opens a PR to update uv.lock
-# if the tests pass and the lockfile has changed.
-name: Latest Dependency Tests
-
-on:
-  # Allows running this workflow manually from the Actions tab
-  workflow_dispatch:
-
-  # Runs at 03:00 UTC
-  schedule:
-    - cron: "0 3 * * *"
-
-# Sets up the environment variables
-env:
-  UV_VERSION: "0.8.0"
-  PYTHON_VERSION: "3.12"
-  DOCKER_IMAGE_NAME: huggingface/lerobot-gpu:latest-deps
-
-# Ensures that only the latest run is active, canceling older runs.
-concurrency:
-  group: ${{ github.workflow }}
-  cancel-in-progress: true
-
-jobs:
-
-  # This job upgrades the lockfile and checks if dependencies have changed
-  upgrade-lock:
-    name: Upgrade Lockfile
-    runs-on: ubuntu-latest
-    if: github.repository == 'huggingface/lerobot'
-    permissions:
-      contents: read
-    outputs:
-      changed: ${{ steps.diff.outputs.changed }}
-    steps:
-      - uses: actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Setup uv and Python
-        uses: astral-sh/setup-uv@v6 # zizmor: ignore[unpinned-uses]
-        with:
-          version: ${{ env.UV_VERSION }}
-          python-version: ${{ env.PYTHON_VERSION }}
-
-      - name: Upgrade uv.lock
-        run: uv lock --upgrade
-
-      - name: Check for changes
-        id: diff
-        run: |
-          if git diff --quiet uv.lock; then
-            echo "changed=false" >> "$GITHUB_OUTPUT"
-            echo "uv.lock is up to date — no dependency changes."
-          else
-            echo "changed=true" >> "$GITHUB_OUTPUT"
-            echo "uv.lock has changed — running tests."
-          fi
-
-      - name: Upload updated lockfile
-        if: steps.diff.outputs.changed == 'true'
-        uses: actions/upload-artifact@v4 # zizmor: ignore[unpinned-uses]
-        with:
-          name: uv-lock
-          path: uv.lock
-
-  # This job runs the full test suite with the upgraded dependencies
-  cpu-tests:
-    name: CPU Tests (Latest Deps)
-    needs: [upgrade-lock]
-    if: needs.upgrade-lock.outputs.changed == 'true'
-    runs-on: ubuntu-latest
-    permissions:
-      contents: read
-    env:
-      MUJOCO_GL: egl
-      HF_HOME: /mnt/cache/.cache/huggingface
-      HF_LEROBOT_HOME: /mnt/cache/.cache/huggingface/lerobot
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
-    steps:
-      - uses: actions/checkout@v6
-        with:
-          lfs: true
-          persist-credentials: false
-
-      - name: Download updated lockfile
-        uses: actions/download-artifact@v4 # zizmor: ignore[unpinned-uses]
-        with:
-          name: uv-lock
-
-      # NOTE(Steven): Mount to `/mnt` to avoid the limited storage on `/home`. Consider cleaning default SDKs or using self-hosted runners for more space.
-      # (As of 2024-06-10, the runner's `/home` has only 6.2 GB free—8% of its 72 GB total.)
-      - name: Setup /mnt storage
-        run: sudo chown -R $USER:$USER /mnt
-
-      - name: Install apt dependencies
-        run: |
-          sudo apt-get update && sudo apt-get install -y build-essential \
-          git curl libglib2.0-0 libegl1-mesa-dev ffmpeg libusb-1.0-0-dev \
-          speech-dispatcher libgeos-dev portaudio19-dev
-
-      - name: Setup uv and Python
-        uses: astral-sh/setup-uv@v6 # zizmor: ignore[unpinned-uses]
-        with:
-          enable-cache: true
-          version: ${{ env.UV_VERSION }}
-          python-version: ${{ env.PYTHON_VERSION }}
-
-      - name: Install lerobot with all extras
-        run: uv sync --locked --extra all # TODO(Steven): Make flash-attn optional
-
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          uv run hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          uv run hf auth whoami
-
-      - name: Run pytest (all extras)
-        run: uv run pytest tests -vv --maxfail=10
-
-      - name: Run end-to-end tests
-        run: uv run make test-end-to-end
-
-  # This job builds a GPU-enabled Docker image with the upgraded dependencies
-  build-and-push-docker:
-    name: Build and Push Docker
-    needs: [upgrade-lock]
-    if: needs.upgrade-lock.outputs.changed == 'true'
-    permissions:
-      contents: read
-    runs-on:
-      group: aws-general-8-plus
-    outputs:
-      image_tag: ${{ env.DOCKER_IMAGE_NAME }}
-    steps:
-      - name: Install Git LFS
-        run: |
-          sudo apt-get update
-          sudo apt-get install git-lfs
-          git lfs install
-      - uses: actions/checkout@v6
-        with:
-          lfs: true
-          persist-credentials: false
-
-      - name: Download updated lockfile
-        uses: actions/download-artifact@v4 # zizmor: ignore[unpinned-uses]
-        with:
-          name: uv-lock
-
-      - name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v3 # zizmor: ignore[unpinned-uses]
-        with:
-          cache-binary: false
-      - name: Login to Docker Hub
-        uses: docker/login-action@v3 # zizmor: ignore[unpinned-uses]
-        with:
-          username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
-          password: ${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}
-      - name: Build and push Docker image
-        uses: docker/build-push-action@v6 # zizmor: ignore[unpinned-uses]
-        with:
-          context: .
-          file: ./docker/Dockerfile.internal
-          push: true
-          tags: ${{ env.DOCKER_IMAGE_NAME }}
-
-  # This job runs pytest with all extras on a GPU-enabled host
-  gpu-tests:
-    name: GPU Tests (Latest Deps)
-    needs: [build-and-push-docker]
-    permissions:
-      contents: read
-    runs-on:
-      group: aws-g6-4xlarge-plus
-    env:
-      HF_HOME: /home/user_lerobot/.cache/huggingface
-      HF_LEROBOT_HOME: /home/user_lerobot/.cache/huggingface/lerobot
-      TORCH_HOME: /home/user_lerobot/.cache/torch
-      TRITON_CACHE_DIR: /home/user_lerobot/.cache/triton
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
-    container:
-      image: ${{ needs.build-and-push-docker.outputs.image_tag }} # zizmor: ignore[unpinned-images]
-      options: --gpus all --shm-size "16gb"
-      credentials:
-        username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
-        password: ${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}
-    defaults:
-      run:
-        shell: bash
-        working-directory: /lerobot
-    steps:
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          hf auth whoami
-      - name: Fix ptxas permissions
-        run: chmod +x /lerobot/.venv/lib/python3.12/site-packages/triton/backends/nvidia/bin/ptxas
-      - name: Run pytest on GPU
-        run: pytest tests -vv --maxfail=10
-      - name: Run end-to-end tests
-        run: make test-end-to-end
-
-  # This job creates or updates a PR with the upgraded lockfile
-  open-pr:
-    name: Open PR
-    needs: [cpu-tests, gpu-tests, upgrade-lock]
-    if: success() && needs.upgrade-lock.outputs.changed == 'true'
-    runs-on: ubuntu-latest
-    permissions:
-      contents: write
-      pull-requests: write
-    env:
-      GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-    steps:
-      - uses: actions/checkout@v6
-        with:
-          persist-credentials: false
-
-      - name: Download updated lockfile
-        uses: actions/download-artifact@v4 # zizmor: ignore[unpinned-uses]
-        with:
-          name: uv-lock
-
-      - name: Create or update PR
-        run: |
-          set -euo pipefail
-          BRANCH="auto/update-uv-lock"
-
-          git config user.name "github-actions[bot]"
-          git config user.email "github-actions[bot]@users.noreply.github.com"
-          git remote set-url origin "https://x-access-token:${GH_TOKEN}@github.com/${{ github.repository }}.git"
-
-          git checkout -B "$BRANCH"
-          git add uv.lock
-          git commit -m "chore(dependencies): update uv.lock"
-          git push --force origin "$BRANCH"
-
-          # Create PR only if one doesn't already exist for this branch
-          EXISTING_PR=$(gh pr list --head "$BRANCH" --state open --json number --jq '.[0].number')
-          if [ -z "$EXISTING_PR" ]; then
-            gh pr create \
-              --title "chore(dependencies): update uv.lock" \
-              --body "Automated update of \`uv.lock\` after successful latest dependency tests (CPU + GPU).
-
-          This PR upgrades all dependencies to their latest versions within the ranges specified in \`pyproject.toml\`." \
-              --head "$BRANCH" \
-              --base main
-          else
-            echo "PR #$EXISTING_PR already exists, branch has been updated."
-          fi
-
-  # This job deletes the temporary Docker image after tests complete
-  cleanup-docker:
-    name: Cleanup Docker Image
-    needs: [gpu-tests, build-and-push-docker]
-    if: always() && needs.build-and-push-docker.result == 'success'
-    permissions:
-      contents: read
-    runs-on: ubuntu-latest
-    steps:
-      - name: Get Docker Hub Token and Delete Image
-        # zizmor: ignore[template-injection]
-        env:
-          DOCKERHUB_LEROBOT_USERNAME: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
-          DOCKERHUB_LEROBOT_PASSWORD: ${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}
-          IMAGE_FULL: ${{ needs.build-and-push-docker.outputs.image_tag }}
-        run: |
-          IMAGE_NAME=$(echo "$IMAGE_FULL" | cut -d':' -f1)
-          IMAGE_TAG=$(echo "$IMAGE_FULL" | cut -d':' -f2-)
-          echo "Attempting to delete image: $IMAGE_NAME:$IMAGE_TAG"
-
-          TOKEN=$(curl -s -H "Content-Type: application/json" \
-                       -X POST \
-                       -d "{\"username\": \"$DOCKERHUB_LEROBOT_USERNAME\", \"password\": \"$DOCKERHUB_LEROBOT_PASSWORD\"}" \
-                       https://hub.docker.com/v2/users/login/ | jq -r .token)
-
-          if [ "$TOKEN" == "null" ] || [ -z "$TOKEN" ]; then
-            echo "::error::Failed to get Docker Hub token."
-            exit 1
-          fi
-
-          HTTP_RESPONSE=$(curl -s -o /dev/null -w "%{http_code}" \
-                               -H "Authorization: JWT ${TOKEN}" \
-                               -X DELETE \
-                               https://hub.docker.com/v2/repositories/${IMAGE_NAME}/tags/$IMAGE_TAG)
-
-          if [ "$HTTP_RESPONSE" -eq 204 ]; then
-            echo "Successfully deleted Docker image tag: $IMAGE_NAME:$IMAGE_TAG"
-          else
-            echo "::error::Failed to delete Docker image. HTTP status: $HTTP_RESPONSE"
-            exit 1
-          fi

.github/workflows/nightly.yml

@@ -12,8 +12,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-# This workflow handles Docker image publishing & testing.
-name: Docker Publish & Test
+# This workflow handles nightly testing & docker images publishing.
+name: Nightly
 permissions:
   contents: read
 
@@ -28,7 +28,7 @@ on:
 # Sets up the environment variables
 env:
   UV_VERSION: "0.8.0"
-  PYTHON_VERSION: "3.12"
+  PYTHON_VERSION: "3.10"
   DOCKER_IMAGE_NAME_CPU: huggingface/lerobot-cpu:latest
   DOCKER_IMAGE_NAME_GPU: huggingface/lerobot-gpu:latest
 
@@ -39,8 +39,8 @@ concurrency:
 
 jobs:
   # This job builds a CPU image for testing & distribution
-  build-docker-cpu:
-    name: Build CPU Docker
+  build-docker-cpu-nightly:
+    name: Build CPU Docker for Nightly
     runs-on:
       group: aws-general-8-plus
     if: github.repository == 'huggingface/lerobot'
@@ -52,7 +52,7 @@ jobs:
           sudo apt-get update
           sudo apt-get install git-lfs
           git lfs install
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
         with:
           lfs: true
           persist-credentials: false
@@ -74,8 +74,8 @@ jobs:
           tags: ${{ env.DOCKER_IMAGE_NAME_CPU }}
 
   # This job builds a GPU image for testing & distribution
-  build-docker-gpu:
-    name: Build GPU Docker
+  build-docker-gpu-nightly:
+    name: Build GPU Docker for Nightly
     runs-on:
       group: aws-general-8-plus
     if: github.repository == 'huggingface/lerobot'
@@ -87,7 +87,7 @@ jobs:
           sudo apt-get update
           sudo apt-get install git-lfs
           git lfs install
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
         with:
           lfs: true
           persist-credentials: false
@@ -109,9 +109,9 @@ jobs:
           tags: ${{ env.DOCKER_IMAGE_NAME_GPU }}
 
   # This job runs the E2E tests + pytest with all extras in the CPU image
-  cpu-tests:
-    name: CPU Tests
-    needs: [build-docker-cpu]
+  nightly-cpu-tests:
+    name: Nightly CPU Tests
+    needs: [build-docker-cpu-nightly]
     runs-on:
       group: aws-g6-4xlarge-plus
     env:
@@ -119,9 +119,8 @@ jobs:
       HF_LEROBOT_HOME: /home/user_lerobot/.cache/huggingface/lerobot
       TORCH_HOME: /home/user_lerobot/.cache/torch
       TRITON_CACHE_DIR: /home/user_lerobot/.cache/triton
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
     container:
-      image: ${{ needs.build-docker-cpu.outputs.image_tag }} # zizmor: ignore[unpinned-images]
+      image: ${{ needs.build-docker-cpu-nightly.outputs.image_tag }} # zizmor: ignore[unpinned-images]
       options: --shm-size "16gb"
       credentials:
         username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
@@ -131,20 +130,15 @@ jobs:
         shell: bash
         working-directory: /lerobot
     steps:
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          hf auth whoami
       - name: Run pytest on CPU
         run: pytest tests -vv --maxfail=10
       - name: Run end-to-end tests
         run: make test-end-to-end
 
   # This job runs the E2E tests + pytest with all extras in the GPU image
-  gpu-tests:
-    name: GPU Tests
-    needs: [build-docker-gpu]
+  nightly-gpu-tests:
+    name: Nightly GPU Tests
+    needs: [build-docker-gpu-nightly]
     runs-on:
       group: aws-g6-4xlarge-plus
     env:
@@ -152,9 +146,8 @@ jobs:
       HF_LEROBOT_HOME: /home/user_lerobot/.cache/huggingface/lerobot
       TORCH_HOME: /home/user_lerobot/.cache/torch
       TRITON_CACHE_DIR: /home/user_lerobot/.cache/triton
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
     container:
-      image: ${{ needs.build-docker-gpu.outputs.image_tag }} # zizmor: ignore[unpinned-images]
+      image: ${{ needs.build-docker-gpu-nightly.outputs.image_tag }} # zizmor: ignore[unpinned-images]
       options: --gpus all --shm-size "16gb"
       credentials:
         username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
@@ -164,20 +157,15 @@ jobs:
         shell: bash
         working-directory: /lerobot
     steps:
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          hf auth whoami
       - name: Run pytest on GPU
         run: pytest tests -vv --maxfail=10
       - name: Run end-to-end tests
         run: make test-end-to-end
 
   # This job runs multi-GPU training tests with 4 GPUs
-  multi-gpu-tests:
-    name: Multi-GPU Tests
-    needs: [build-docker-gpu]
+  nightly-multi-gpu-tests:
+    name: Nightly Multi-GPU Tests
+    needs: [build-docker-gpu-nightly]
     runs-on:
       group: aws-g4dn-12xlarge  # Instance with 4 GPUs
     env:
@@ -186,9 +174,8 @@ jobs:
       TORCH_HOME: /home/user_lerobot/.cache/torch
       TRITON_CACHE_DIR: /home/user_lerobot/.cache/triton
       CUDA_VISIBLE_DEVICES: "0,1,2,3"
-      HF_USER_TOKEN: ${{ secrets.LEROBOT_HF_USER }}
     container:
-      image: ${{ needs.build-docker-gpu.outputs.image_tag }} # zizmor: ignore[unpinned-images]
+      image: ${{ needs.build-docker-gpu-nightly.outputs.image_tag }} # zizmor: ignore[unpinned-images]
       options: --gpus all --shm-size "16gb"
       credentials:
         username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
@@ -198,15 +185,12 @@ jobs:
         shell: bash
         working-directory: /lerobot
     steps:
-      - name: Login to Hugging Face
-        if: env.HF_USER_TOKEN != ''
-        run: |
-          hf auth login --token "$HF_USER_TOKEN" --add-to-git-credential
-          hf auth whoami
       - name: Verify GPU availability
         run: |
           nvidia-smi
           python -c "import torch; print(f'PyTorch CUDA available: {torch.cuda.is_available()}'); print(f'Number of GPUs: {torch.cuda.device_count()}')"
 
       - name: Run multi-GPU training tests
-        run: pytest -vv tests/training/
+      # TODO(Steven): Investigate why motors tests are failing in multi-GPU setup
+        run: pytest tests -vv --maxfail=10 --ignore=tests/motors/
+        timeout-minutes: 10

.github/workflows/quality.yml

@@ -43,14 +43,14 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
         with:
           persist-credentials: false
 
       - name: Set up Python
-        uses: actions/setup-python@v6
+        uses: actions/setup-python@v5
         with:
-          python-version: '3.12'
+          python-version: '3.10'
 
       - name: Run pre-commit hooks
         uses: pre-commit/action@v3.0.1 # zizmor: ignore[unpinned-uses]

.github/workflows/release.yml

@@ -22,7 +22,7 @@ on:
 # Sets up the environment variables
 env:
   UV_VERSION: "0.8.0"
-  PYTHON_VERSION: "3.12"
+  PYTHON_VERSION: "3.10"
 
 jobs:
   # This job builds the Python package and publishes it to PyPI
@@ -38,14 +38,14 @@ jobs:
 
     steps:
       - name: Checkout code
-        uses: actions/checkout@v6
+        uses: actions/checkout@v4
         with:
           persist-credentials: false
 
       - name: Set up Python
-        uses: actions/setup-python@v6
+        uses: actions/setup-python@v5
         with:
-          python-version: '3.12'
+          python-version: '3.10'
 
       - name: Extract Version
         id: extract_info
@@ -83,6 +83,14 @@ jobs:
             exit 1
           fi
 
+      - name: Remove Tags with Git dependencies
+        # TODO(Steven): Temporary patch to remove pi from PyPi 0.4.0 release due to its reliance on git dependencies.
+        run: |
+          echo "::info:: Checking for Git dependencies to remove from pyproject.toml..."
+          grep -E '@ git\+https|lerobot\[pi\]' pyproject.toml | sed 's/^/::warning:: Removing line: /' || true
+          sed -E -i '/@ git\+https|lerobot\[pi\]/d' pyproject.toml
+          echo "::info:: Git dependencies removed. Proceeding with build."
+
       - name: Install build dependencies
         run: python -m pip install build
 
@@ -127,7 +135,7 @@ jobs:
     env:
       MUJOCO_GL: egl
     steps:
-      - uses: actions/checkout@v6
+      - uses: actions/checkout@v4
         with:
           lfs: true
           persist-credentials: false
@@ -169,3 +177,4 @@ jobs:
 
 # TODO(Steven): Publish draft/pre-release and to test pypi weekly
 # TODO(Steven): Separate build and publish job
+# TODO(Steven): Tag documentation with the same version as the package

.github/workflows/security.yml

@@ -43,7 +43,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Checkout code
-        uses: actions/checkout@v6 # zizmor: ignore[unpinned-uses]
+        uses: actions/checkout@v4 # zizmor: ignore[unpinned-uses]
         with:
           fetch-depth: 0
           persist-credentials: false

.github/workflows/unbound_deps_tests.yml

@@ -0,0 +1,191 @@
+# Copyright 2025 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.
+
+# This workflow handles full testing with unboud dependencies versions.
+name: Unbound Dependency Tests
+
+on:
+  # Allows running this workflow manually from the Actions tab
+  workflow_dispatch:
+
+  # Run on the 1st and 15th of every month at 09:00 UTC
+  schedule:
+    - cron: '0 2 1,15 * *'
+
+permissions:
+  contents: read
+
+# Sets up the environment variables
+env:
+  UV_VERSION: "0.8.0"
+  PYTHON_VERSION: "3.10"
+  DOCKER_IMAGE_NAME: huggingface/lerobot-gpu:unbound
+
+# Ensures that only the latest action is built, canceling older runs.
+concurrency:
+  group: ${{ github.workflow }}-${{ github.head_ref || github.run_id }}
+  cancel-in-progress: true
+
+jobs:
+
+  # This job runs the E2E tests + pytest with all unbound extras
+  full-tests:
+    name: Full Unbound Tests
+    runs-on: ubuntu-latest
+    if: github.repository == 'huggingface/lerobot'
+    env:
+      MUJOCO_GL: egl
+      HF_HOME: /mnt/cache/.cache/huggingface
+      HF_LEROBOT_HOME: /mnt/cache/.cache/huggingface/lerobot
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          lfs: true
+          persist-credentials: false
+
+      # NOTE(Steven): Mount to `/mnt` to avoid the limited storage on `/home`. Consider cleaning default SDKs or using self-hosted runners for more space.
+      # (As of 2024-06-10, the runner's `/home` has only 6.2 GB free—8% of its 72 GB total.)
+      - name: Setup /mnt storage
+        run: sudo chown -R $USER:$USER /mnt
+
+      - name: Install apt dependencies
+        run: |
+          sudo apt-get update && sudo apt-get install -y build-essential \
+          git curl libglib2.0-0 libegl1-mesa-dev ffmpeg libusb-1.0-0-dev \
+          speech-dispatcher libgeos-dev portaudio19-dev
+
+      - name: Setup uv and Python
+        uses: astral-sh/setup-uv@v6 # zizmor: ignore[unpinned-uses]
+        with:
+          enable-cache: true
+          version: ${{ env.UV_VERSION }}
+          python-version: ${{ env.PYTHON_VERSION }}
+
+      - name: Unbound dependencies
+        run: |
+          sed -i 's/,[[:space:]]*<[0-9\.]*//g' pyproject.toml
+          echo "Dependencies unbound:" && cat pyproject.toml
+
+      - name: Install lerobot with all extras
+        run: uv sync --all-extras --no-extra groot --no-extra wallx # TODO(Steven): Make flash-attn optional
+
+      - name: Run pytest (all extras)
+        run: uv run pytest tests -vv
+
+      - name: Run end-to-end tests
+        run: uv run make test-end-to-end
+
+  # This job builds a GPU enabled image for testing
+  build-and-push-docker:
+    name: Build and Push Docker
+    runs-on:
+      group: aws-general-8-plus
+    outputs:
+      image_tag: ${{ env.DOCKER_IMAGE_NAME }}
+    env:
+      GITHUB_REF: ${{ github.ref }}
+    steps:
+      - name: Install Git LFS
+        run: |
+          sudo apt-get update
+          sudo apt-get install git-lfs
+          git lfs install
+      - uses: actions/checkout@v4
+        with:
+          lfs: true
+          persist-credentials: false
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3 # zizmor: ignore[unpinned-uses]
+        with:
+          cache-binary: false
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3 # zizmor: ignore[unpinned-uses]
+        with:
+          username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
+          password: ${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}
+      - name: Build and push Docker image
+        uses: docker/build-push-action@v6 # zizmor: ignore[unpinned-uses]
+        with:
+          context: .
+          file: ./docker/Dockerfile.internal
+          push: true
+          tags: ${{ env.DOCKER_IMAGE_NAME }}
+          build-args: |
+            UNBOUND_DEPS=true
+
+  # This job runs pytest with all unbound extras in a GPU enabled host
+  # It runs everytime a test image is created
+  gpu-tests:
+    name: GPU Unbound Tests
+    needs: [build-and-push-docker]
+    runs-on:
+      group: aws-g6-4xlarge-plus
+    env:
+      HF_HOME: /home/user_lerobot/.cache/huggingface
+      HF_LEROBOT_HOME: /home/user_lerobot/.cache/huggingface/lerobot
+      TORCH_HOME: /home/user_lerobot/.cache/torch
+      TRITON_CACHE_DIR: /home/user_lerobot/.cache/triton
+    container:
+      image: ${{ needs.build-and-push-docker.outputs.image_tag }} # zizmor: ignore[unpinned-images]
+      options: --gpus all --shm-size "16gb"
+      credentials:
+        username: ${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}
+        password: ${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}
+    defaults:
+      run:
+        shell: bash
+        working-directory: /lerobot
+    steps:
+      - name: Run pytest on GPU
+        run: pytest tests -vv
+      - name: Run end-to-end tests
+        run: make test-end-to-end
+
+  # This job deletes the test image recently created
+  # It runs everytime after the gpu-tests have finished
+  delete-unbound-image:
+    name: Delete Unbound Image
+    needs: [gpu-tests, build-and-push-docker]
+    if: always() && needs.build-and-push-docker.result == 'success'
+    runs-on: ubuntu-latest
+    steps:
+      - name: Get Docker Hub Token and Delete Image
+        # zizmor: ignore[template-injection]
+        run: |
+          IMAGE_NAME=$(echo "${{ needs.build-and-push-docker.outputs.image_tag }}" | cut -d':' -f1)
+          IMAGE_TAG=$(echo "${{ needs.build-and-push-docker.outputs.image_tag }}" | cut -d':' -f2)
+
+          echo "Attempting to delete image: $IMAGE_NAME:$IMAGE_TAG"
+
+          TOKEN=$(curl -s -H "Content-Type: application/json" \
+                       -X POST \
+                       -d '{"username": "${{ secrets.DOCKERHUB_LEROBOT_USERNAME }}", "password": "${{ secrets.DOCKERHUB_LEROBOT_PASSWORD }}"}' \
+                       https://hub.docker.com/v2/users/login/ | jq -r .token)
+
+          if [ "$TOKEN" == "null" ] || [ -z "$TOKEN" ]; then
+            echo "::error::Failed to get Docker Hub token."
+            exit 1
+          fi
+
+          HTTP_RESPONSE=$(curl -s -o /dev/null -w "%{http_code}" \
+                               -H "Authorization: JWT ${TOKEN}" \
+                               -X DELETE \
+                               https://hub.docker.com/v2/repositories/${IMAGE_NAME}/tags/${IMAGE_TAG}/)
+
+          if [ "$HTTP_RESPONSE" -eq 204 ]; then
+            echo "Successfully deleted Docker image tag: $IMAGE_NAME:$IMAGE_TAG"
+          else
+            echo "::error::Failed to delete Docker image. HTTP status: $HTTP_RESPONSE"
+            exit 1
+          fi

.gitignore

@@ -25,6 +25,7 @@ node_modules/
 
 # Lock files
 poetry.lock
+uv.lock
 Pipfile.lock
 
 ### Build & Distribution ###

.pre-commit-config.yaml

@@ -13,7 +13,7 @@
 # limitations under the License.
 
 default_language_version:
-    python: python3.12
+    python: python3.10
 
 exclude: "tests/artifacts/.*\\.safetensors$"
 
@@ -55,7 +55,7 @@ repos:
     rev: v3.21.0
     hooks:
     -   id: pyupgrade
-        args: [--py312-plus]
+        args: [--py310-plus]
 
   ##### Markdown Quality #####
   - repo: https://github.com/rbubley/mirrors-prettier

AI_POLICY.md

@@ -1,25 +0,0 @@
-# AI Usage Policy
-
-The LeRobot project welcomes contributions from everyone, and we have a few guidelines regarding AI usage to ensure high code quality, clear communication, and a healthy open-source ecosystem:
-
-- **Please disclose significant AI assistance.** If you used AI tools (e.g., Copilot, Claude, Cursor, ChatGPT) to generate a substantial portion of your code or text, let us know in your PR description. Transparency helps us review your changes more effectively.
-- **Own your code (The Human-in-the-Loop).** You must fully understand all the changes you are proposing. If you cannot explain what your AI-assisted code does or how it interacts with LeRobot's broader architecture, please take the time to learn and test it before submitting.
-- **Keep issues and discussions focused.** You are welcome to use AI to help draft issues or PR descriptions, but please review and edit them carefully before posting. AI can often be overly verbose; trimming the noise and getting straight to the point helps our maintainers address your needs faster.
-
-Our core maintainers also use AI tools to aid their workflows, but they do so while bringing deep contextual knowledge of the LeRobot codebase to validate the output. We ask all contributors to apply that same level of rigor.
-
-## Remember the Human Maintainers
-
-Please remember that LeRobot is maintained by a dedicated team of humans.
-
-Every discussion, issue, and pull request is read and reviewed by real people. While AI tools can generate thousands of lines of code in seconds, reviewing that code still takes human time and energy. Submitting unverified or low-effort AI output puts an unfair burden on our maintainers.
-
-Today, the quality of the AI output still heavily depends on the developer driving the tool. We ask that you respect our maintainers' time by thoroughly vetting, testing, and refining your submissions.
-
-## AI is Welcome Here
-
-LeRobot operates at the cutting edge of AI and robotics, and many of our maintainers actively embrace AI coding assistants as valuable productivity tools. We are a pro-AI project!
-
-Our reason for having an AI policy is not an anti-AI stance. Rather, it exists to ensure that AI is used to enhance human contributions, not replace them with unverified noise. It's about how the tools are used, not the tools themselves.
-
-We value the unique human insight you bring to the LeRobot community. Let AI empower your workflow, but always let your own judgment take the wheel.

CONTRIBUTING.md

@@ -1,83 +1,323 @@
-# How to contribute to 🤗 LeRobot
+# How to contribute to 🤗 LeRobot?
 
-Everyone is welcome to contribute, and we value everybody's contribution. Code is not the only way to help the community. Answering questions, helping others, reaching out, and improving the documentation are immensely valuable.
+Everyone is welcome to contribute, and we value everybody's contribution. Code
+is thus not the only way to help the community. Answering questions, helping
+others, reaching out and improving the documentations are immensely valuable to
+the community.
 
-Whichever way you choose to contribute, please be mindful to respect our [code of conduct](https://github.com/huggingface/lerobot/blob/main/CODE_OF_CONDUCT.md) and our [AI policy](https://github.com/huggingface/lerobot/blob/main/AI_POLICY.md).
+It also helps us if you spread the word: reference the library from blog posts
+on the awesome projects it made possible, shout out on Twitter when it has
+helped you, or simply ⭐️ the repo to say "thank you".
 
-## Ways to Contribute
+Whichever way you choose to contribute, please be mindful to respect our
+[code of conduct](https://github.com/huggingface/lerobot/blob/main/CODE_OF_CONDUCT.md).
 
-You can contribute in many ways:
+## You can contribute in so many ways!
 
-- **Fixing issues:** Resolve bugs or improve existing code.
-- **New features:** Develop new features.
-- **Extend:** Implement new models/policies, robots, or simulation environments and upload datasets to the Hugging Face Hub.
-- **Documentation:** Improve examples, guides, and docstrings.
-- **Feedback:** Submit tickets related to bugs or desired new features.
+Some of the ways you can contribute to 🤗 LeRobot:
 
-If you are unsure where to start, join our [Discord Channel](https://discord.gg/q8Dzzpym3f).
+- Fixing outstanding issues with the existing code.
+- Implementing new models, datasets or simulation environments.
+- Contributing to the examples or to the documentation.
+- Submitting issues related to bugs or desired new features.
 
-## Development Setup
+Following the guides below, feel free to open issues and PRs and to coordinate your efforts with the community on our [Discord Channel](https://discord.gg/VjFz58wn3R). For specific inquiries, reach out to [Remi Cadene](mailto:remi.cadene@huggingface.co).
 
-To contribute code, you need to set up a development environment.
+If you are not sure how to contribute or want to know the next features we working on, look on this project page: [LeRobot TODO](https://github.com/orgs/huggingface/projects/46)
 
-### 1. Fork and Clone
+## Submitting a new issue or feature request
 
-Fork the repository on GitHub, then clone your fork:
-
-```bash
-git clone https://github.com/<your-handle>/lerobot.git
-cd lerobot
-git remote add upstream https://github.com/huggingface/lerobot.git
-```
-
-### 2. Environment Installation
-
-Please follow our [Installation Guide](https://huggingface.co/docs/lerobot/installation) for the environment setup & installation from source.
-
-## Running Tests & Quality Checks
-
-### Code Style (Pre-commit)
-
-Install `pre-commit` hooks to run checks automatically before you commit:
-
-```bash
-pre-commit install
-```
-
-To run checks manually on all files:
-
-```bash
-pre-commit run --all-files
-```
-
-### Running Tests
-
-We use `pytest`. First, ensure you have test artifacts by installing **git-lfs**:
+Do your best to follow these guidelines when submitting an issue or a feature
+request. It will make it easier for us to come back to you quickly and with good
+feedback.
+
+### Did you find a bug?
+
+The 🤗 LeRobot library is robust and reliable thanks to the users who notify us of
+the problems they encounter. So thank you for reporting an issue.
+
+First, we would really appreciate it if you could **make sure the bug was not
+already reported** (use the search bar on Github under Issues).
+
+Did not find it? :( So we can act quickly on it, please follow these steps:
+
+- Include your **OS type and version**, the versions of **Python** and **PyTorch**.
+- A short, self-contained, code snippet that allows us to reproduce the bug in
+  less than 30s.
+- The full traceback if an exception is raised.
+- Attach any other additional information, like screenshots, you think may help.
+
+### Do you want a new feature?
+
+A good feature request addresses the following points:
+
+1. Motivation first:
+
+- Is it related to a problem/frustration with the library? If so, please explain
+  why. Providing a code snippet that demonstrates the problem is best.
+- Is it related to something you would need for a project? We'd love to hear
+  about it!
+- Is it something you worked on and think could benefit the community?
+  Awesome! Tell us what problem it solved for you.
+
+2. Write a _paragraph_ describing the feature.
+3. Provide a **code snippet** that demonstrates its future use.
+4. In case this is related to a paper, please attach a link.
+5. Attach any additional information (drawings, screenshots, etc.) you think may help.
+
+If your issue is well written we're already 80% of the way there by the time you
+post it.
+
+## Adding new policies, datasets or environments
+
+Look at our implementations for [datasets](./src/lerobot/datasets/), [policies](./src/lerobot/policies/),
+environments ([aloha](https://github.com/huggingface/gym-aloha),
+[pusht](https://github.com/huggingface/gym-pusht))
+and follow the same api design.
+
+When implementing a new dataset loadable with LeRobotDataset follow these steps:
+
+- Update `available_datasets_per_env` in `lerobot/__init__.py`
+
+When implementing a new environment (e.g. `gym_aloha`), follow these steps:
+
+- Update `available_tasks_per_env` and `available_datasets_per_env` in `lerobot/__init__.py`
+
+When implementing a new policy class (e.g. `DiffusionPolicy`) follow these steps:
+
+- Update `available_policies` and `available_policies_per_env`, in `lerobot/__init__.py`
+- Set the required `name` class attribute.
+- Update variables in `tests/test_available.py` by importing your new Policy class
+
+## Submitting a pull request (PR)
+
+Before writing code, we strongly advise you to search through the existing PRs or
+issues to make sure that nobody is already working on the same thing. If you are
+unsure, it is always a good idea to open an issue to get some feedback.
+
+You will need basic `git` proficiency to be able to contribute to
+🤗 LeRobot. `git` is not the easiest tool to use but it has the greatest
+manual. Type `git --help` in a shell and enjoy. If you prefer books, [Pro
+Git](https://git-scm.com/book/en/v2) is a very good reference.
+
+Follow these steps to start contributing:
+
+1. Fork the [repository](https://github.com/huggingface/lerobot) by
+   clicking on the 'Fork' button on the repository's page. This creates a copy of the code
+   under your GitHub user account.
+
+2. Clone your fork to your local disk, and add the base repository as a remote. The following command
+   assumes you have your public SSH key uploaded to GitHub. See the following guide for more
+   [information](https://docs.github.com/en/repositories/creating-and-managing-repositories/cloning-a-repository).
+
+   ```bash
+   git clone git@github.com:<your Github handle>/lerobot.git
+   cd lerobot
+   git remote add upstream https://github.com/huggingface/lerobot.git
+   ```
+
+3. Create a new branch to hold your development changes, and do this for every new PR you work on.
+
+   Start by synchronizing your `main` branch with the `upstream/main` branch (more details in the [GitHub Docs](https://docs.github.com/en/github/collaborating-with-issues-and-pull-requests/syncing-a-fork)):
+
+   ```bash
+   git checkout main
+   git fetch upstream
+   git rebase upstream/main
+   ```
+
+   Once your `main` branch is synchronized, create a new branch from it:
+
+   ```bash
+   git checkout -b a-descriptive-name-for-my-changes
+   ```
+
+   🚨 **Do not** work on the `main` branch.
+
+4. for development, we advise to use a tool like `poetry` or `uv` instead of just `pip` to easily track our dependencies.
+   Follow the instructions to [install poetry](https://python-poetry.org/docs/#installation) (use a version >=2.1.0) or to [install uv](https://docs.astral.sh/uv/getting-started/installation/#installation-methods) if you don't have one of them already.
+
+   Set up a development environment with conda:
+
+   ```bash
+   conda create -y -n lerobot-dev python=3.10 && conda activate lerobot-dev
+   ```
+
+   If you're using `uv`, it can manage python versions so you can instead do:
+
+   ```bash
+   uv venv --python 3.10 && source .venv/bin/activate
+   ```
+
+   To develop on 🤗 LeRobot, you will at least need to install the `dev` and `test` extras dependencies along with the core library:
+
+   using `poetry`
+
+   ```bash
+   poetry sync --extras "dev test"
+   ```
+
+   using `uv`
+
+   ```bash
+   uv sync --extra dev --extra test
+   ```
+
+   You can also install the project with all its dependencies (including environments):
+
+   using `poetry`
+
+   ```bash
+   poetry sync --all-extras
+   ```
+
+   using `uv`
+
+   ```bash
+   uv sync --all-extras
+   ```
+
+   > **Note:** If you don't install simulation environments with `--all-extras`, the tests that require them will be skipped when running the pytest suite locally. However, they _will_ be tested in the CI. In general, we advise you to install everything and test locally before pushing.
+
+   Whichever command you chose to install the project (e.g. `poetry sync --all-extras`), you should run it again when pulling code with an updated version of `pyproject.toml` and `poetry.lock` in order to synchronize your virtual environment with the new dependencies.
+
+   The equivalent of `pip install some-package`, would just be:
+
+   using `poetry`
+
+   ```bash
+   poetry add some-package
+   ```
+
+   using `uv`
+
+   ```bash
+   uv add some-package
+   ```
+
+   When making changes to the poetry sections of the `pyproject.toml`, you should run the following command to lock dependencies.
+   using `poetry`
+
+   ```bash
+   poetry lock
+   ```
+
+   using `uv`
+
+   ```bash
+   uv lock
+   ```
+
+5. Develop the features on your branch.
+
+   As you work on the features, you should make sure that the test suite
+   passes. You should run the tests impacted by your changes like this (see
+   below an explanation regarding the environment variable):
+
+   ```bash
+   pytest tests/<TEST_TO_RUN>.py
+   ```
+
+6. Follow our style.
+
+   `lerobot` relies on `ruff` to format its source code
+   consistently. Set up [`pre-commit`](https://pre-commit.com/) to run these checks
+   automatically as Git commit hooks.
+
+   Install `pre-commit` hooks:
+
+   ```bash
+   pre-commit install
+   ```
+
+   You can run these hooks whenever you need on staged files with:
+
+   ```bash
+   pre-commit
+   ```
+
+   Once you're happy with your changes, add changed files using `git add` and
+   make a commit with `git commit` to record your changes locally:
+
+   ```bash
+   git add modified_file.py
+   git commit
+   ```
+
+   Note, if you already committed some changes that have a wrong formatting, you can use:
+
+   ```bash
+   pre-commit run --all-files
+   ```
+
+   Please write [good commit messages](https://chris.beams.io/posts/git-commit/).
+
+   It is a good idea to sync your copy of the code with the original
+   repository regularly. This way you can quickly account for changes:
+
+   ```bash
+   git fetch upstream
+   git rebase upstream/main
+   ```
+
+   Push the changes to your account using:
+
+   ```bash
+   git push -u origin a-descriptive-name-for-my-changes
+   ```
+
+7. Once you are satisfied (**and the checklist below is happy too**), go to the
+   webpage of your fork on GitHub. Click on 'Pull request' to send your changes
+   to the project maintainers for review.
+
+8. It's ok if maintainers ask you for changes. It happens to core contributors
+   too! So everyone can see the changes in the Pull request, work in your local
+   branch and push the changes to your fork. They will automatically appear in
+   the pull request.
+
+### Checklist
+
+1. The title of your pull request should be a summary of its contribution;
+2. If your pull request addresses an issue, please mention the issue number in
+   the pull request description to make sure they are linked (and people
+   consulting the issue know you are working on it);
+3. To indicate a work in progress please prefix the title with `[WIP]`, or preferably mark
+   the PR as a draft PR. These are useful to avoid duplicated work, and to differentiate
+   it from PRs ready to be merged;
+4. Make sure existing tests pass;
+
+### Tests
+
+An extensive test suite is included to test the library behavior and several examples. Library tests can be found in the [tests folder](https://github.com/huggingface/lerobot/tree/main/tests).
+
+Install [git lfs](https://git-lfs.com/) to retrieve test artifacts (if you don't have it already).
+
+On Mac:
 
 ```bash
+brew install git-lfs
 git lfs install
+```
+
+On Ubuntu:
+
+```bash
+sudo apt-get install git-lfs
+git lfs install
+```
+
+Pull artifacts if they're not in [tests/artifacts](tests/artifacts)
+
+```bash
 git lfs pull
 ```
 
-Run the full suite (this may require extras installed):
+We use `pytest` in order to run the tests. From the root of the
+repository, here's how to run tests with `pytest` for the library:
 
 ```bash
-pytest -sv ./tests
+python -m pytest -sv ./tests
 ```
 
-Or run a specific test file during development:
-
-```bash
-pytest -sv tests/test_specific_feature.py
-```
-
-## Submitting Issues & Pull Requests
-
-Use the templates for required fields and examples.
-
-- **Issues:** Follow the [ticket template](https://github.com/huggingface/lerobot/blob/main/.github/ISSUE_TEMPLATE/bug-report.yml).
-- **Pull requests:** Rebase on `upstream/main`, use a descriptive branch (don't work on `main`), run `pre-commit` and tests locally, and follow the [PR template](https://github.com/huggingface/lerobot/blob/main/.github/PULL_REQUEST_TEMPLATE.md).
-
-One member of the LeRobot team will then review your contribution.
-
-Thank you for contributing to LeRobot!
+You can specify a smaller set of tests in order to test only the feature
+you're working on.

MANIFEST.in

@@ -1,3 +1,2 @@
 include src/lerobot/templates/lerobot_modelcard_template.md
 include src/lerobot/datasets/card_template.md
-include src/lerobot/envs/metaworld_config.json

README.md

@@ -4,13 +4,12 @@
 
 <div align="center">
 
-[![Tests](https://github.com/huggingface/lerobot/actions/workflows/docker_publish.yml/badge.svg?branch=main)](https://github.com/huggingface/lerobot/actions/workflows/docker_publish.yml?query=branch%3Amain)
+[![Tests](https://github.com/huggingface/lerobot/actions/workflows/nightly.yml/badge.svg?branch=main)](https://github.com/huggingface/lerobot/actions/workflows/nightly.yml?query=branch%3Amain)
 [![Python versions](https://img.shields.io/pypi/pyversions/lerobot)](https://www.python.org/downloads/)
 [![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://github.com/huggingface/lerobot/blob/main/LICENSE)
 [![Status](https://img.shields.io/pypi/status/lerobot)](https://pypi.org/project/lerobot/)
 [![Version](https://img.shields.io/pypi/v/lerobot)](https://pypi.org/project/lerobot/)
 [![Contributor Covenant](https://img.shields.io/badge/Contributor%20Covenant-v2.1-ff69b4.svg)](https://github.com/huggingface/lerobot/blob/main/CODE_OF_CONDUCT.md)
-[![Discord](https://img.shields.io/badge/Discord-Join_Us-5865F2?style=flat&logo=discord&logoColor=white)](https://discord.gg/q8Dzzpym3f)
 
 </div>
 
@@ -100,11 +99,11 @@ lerobot-train \
   --dataset.repo_id=lerobot/aloha_mobile_cabinet
 ```
 
-| Category                   | Models                                                                                                                                                                                                                  |
-| -------------------------- | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| **Imitation Learning**     | [ACT](./docs/source/policy_act_README.md), [Diffusion](./docs/source/policy_diffusion_README.md), [VQ-BeT](./docs/source/policy_vqbet_README.md), [Multitask DiT Policy](./docs/source/policy_multi_task_dit_README.md) |
-| **Reinforcement Learning** | [HIL-SERL](./docs/source/hilserl.mdx), [TDMPC](./docs/source/policy_tdmpc_README.md) & QC-FQL (coming soon)                                                                                                             |
-| **VLAs Models**            | [Pi0Fast](./docs/source/pi0fast.mdx), [Pi0.5](./docs/source/pi05.mdx), [GR00T N1.5](./docs/source/policy_groot_README.md), [SmolVLA](./docs/source/policy_smolvla_README.md), [XVLA](./docs/source/xvla.mdx)            |
+| Category                   | Models                                                                                                                                                                 |
+| -------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
+| **Imitation Learning**     | [ACT](./docs/source/policy_act_README.md), [Diffusion](./docs/source/policy_diffusion_README.md), [VQ-BeT](./docs/source/policy_vqbet_README.md)                       |
+| **Reinforcement Learning** | [HIL-SERL](./docs/source/hilserl.mdx), [TDMPC](./docs/source/policy_tdmpc_README.md) & QC-FQL (coming soon)                                                            |
+| **VLAs Models**            | [Pi0.5](./docs/source/pi05.mdx), [GR00T N1.5](./docs/source/policy_groot_README.md), [SmolVLA](./docs/source/policy_smolvla_README.md), [XVLA](./docs/source/xvla.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
 
@@ -128,14 +127,13 @@ Learn how to implement your own simulation environment or benchmark and distribu
 ## Resources
 
 - **[Documentation](https://huggingface.co/docs/lerobot/index):** The complete guide to tutorials & API.
-- **[Chinese Tutorials: LeRobot+SO-ARM101中文教程-同济子豪兄](https://zihao-ai.feishu.cn/wiki/space/7589642043471924447)** Detailed doc for assembling, teleoperate, dataset, train, deploy. Verified by Seed Studio and 5 global hackathon players.
-- **[Discord](https://discord.gg/q8Dzzpym3f):** Join the `LeRobot` server to discuss with the community.
+- **[Discord](https://discord.gg/3gxM6Avj):** Join the `LeRobot` server to discuss with the community.
 - **[X](https://x.com/LeRobotHF):** Follow us on X to stay up-to-date with the latest developments.
 - **[Robot Learning Tutorial](https://huggingface.co/spaces/lerobot/robot-learning-tutorial):** A free, hands-on course to learn robot learning using LeRobot.
 
 ## Citation
 
-If you use LeRobot in your project, please cite the GitHub repository to acknowledge the ongoing development and contributors:
+If you use LeRobot in your research, please cite:
 
 ```bibtex
 @misc{cadene2024lerobot,
@@ -146,26 +144,9 @@ If you use LeRobot in your project, please cite the GitHub repository to acknowl
 }
 ```
 
-If you are referencing our research or the academic paper, please also cite our ICLR publication:
-
-<details>
-<summary><b>ICLR 2026 Paper</b></summary>
-
-```bibtex
-@inproceedings{cadenelerobot,
-  title={LeRobot: An Open-Source Library for End-to-End Robot Learning},
-  author={Cadene, Remi and Alibert, Simon and Capuano, Francesco and Aractingi, Michel and Zouitine, Adil and Kooijmans, Pepijn and Choghari, Jade and Russi, Martino and Pascal, Caroline and Palma, Steven and Shukor, Mustafa and Moss, Jess and Soare, Alexander and Aubakirova, Dana and Lhoest, Quentin and Gallou\'edec, Quentin and Wolf, Thomas},
-  booktitle={The Fourteenth International Conference on Learning Representations},
-  year={2026},
-  url={https://arxiv.org/abs/2602.22818}
-}
-```
-
-</details>
-
 ## Contribute
 
-We welcome contributions from everyone in the community! To get started, please read our [CONTRIBUTING.md](https://github.com/huggingface/lerobot/blob/main/CONTRIBUTING.md) guide. Whether you're adding a new feature, improving documentation, or fixing a bug, your help and feedback are invaluable. We're incredibly excited about the future of open-source robotics and can't wait to work with you on what's next—thank you for your support!
+We welcome contributions from everyone in the community! To get started, please read our [CONTRIBUTING.md](./CONTRIBUTING.md) guide. Whether you're adding a new feature, improving documentation, or fixing a bug, your help and feedback are invaluable. We're incredibly excited about the future of open-source robotics and can't wait to work with you on what's next—thank you for your support!
 
 <p align="center">
   <img alt="SO101 Video" src="./media/readme/so100_video.webp" width="640px">

SECURITY.md

@@ -1,48 +0,0 @@
-# Security Policy
-
-## Project Status & Philosophy
-
-`lerobot` has so far been primarily a research and prototyping tool, which is why deployment security hasn’t been a strong focus until now. As `lerobot` continues to be adopted and deployed in production, we are paying much closer attention to these kinds of issues.
-
-Fortunately, being an open-source project, the community can also help by reporting and fixing vulnerabilities. We appreciate your efforts to responsibly disclose your findings and will make every effort to acknowledge your contributions.
-
-## Reporting a Vulnerability
-
-To report a security issue, please use the GitHub Security Advisory ["Report a Vulnerability"](https://github.com/huggingface/lerobot/security/advisories/new) tab.
-
-The `lerobot` team will send a response indicating the next steps in handling your report. After the initial reply to your report, the security team will keep you informed of the progress towards a fix and full announcement, and may ask for additional information or guidance.
-
-#### Hugging Face Security Team
-
-Since this project is part of the Hugging Face ecosystem, feel free to submit vulnerability reports directly to: **[security@huggingface.co](mailto:security@huggingface.co)**. Someone from the HF security team will review the report and recommend next steps.
-
-#### Open Source Disclosures
-
-If reporting a vulnerability specific to the open-source codebase (and not the underlying Hub infrastructure), you may also use [Huntr](https://huntr.com), a vulnerability disclosure program for open source software.
-
-## Supported Versions
-
-Currently, we treat `lerobot` as a rolling release. We prioritize security updates for the latest available version (`main` branch).
-
-| Version  | Supported |
-| -------- | --------- |
-| Latest   | ✅        |
-| < Latest | ❌        |
-
-## Secure Usage Guidelines
-
-`lerobot` is tightly coupled to the Hugging Face Hub for sharing data and pretrained policies. When downloading artifacts uploaded by others, you expose yourself to risks. Please read below for recommendations to keep your runtime and robot environment safe.
-
-### Remote Artefacts (Weights & Policies)
-
-Models and policies uploaded to the Hugging Face Hub come in different formats. We heavily recommend uploading and downloading models in the [`safetensors`](https://github.com/huggingface/safetensors) format.
-
-`safetensors` was developed specifically to prevent arbitrary code execution on your system, which is critical when running software on physical hardware/robots.
-
-To avoid loading models from unsafe formats (e.g., `pickle`), you should ensure you are prioritizing `safetensors` files.
-
-### Remote Code
-
-Some models or environments on the Hub may require `trust_remote_code=True` to run custom architecture code.
-
-Please **always** verify the content of the modeling files when using this argument. We recommend setting a specific `revision` (commit hash) when loading remote code to ensure you protect yourself from unverified updates to the repository.

benchmarks/video/README.md

@@ -28,9 +28,9 @@ We don't expect the same optimal settings for a dataset of images from a simulat
 For these reasons, we run this benchmark on four representative datasets:
 
 - `lerobot/pusht_image`: (96 x 96 pixels) simulation with simple geometric shapes, fixed camera.
-- `lerobot/aloha_mobile_shrimp_image`: (480 x 640 pixels) real-world indoor, moving camera.
-- `lerobot/paris_street`: (720 x 1280 pixels) real-world outdoor, moving camera.
-- `lerobot/kitchen`: (1080 x 1920 pixels) real-world indoor, fixed camera.
+- `aliberts/aloha_mobile_shrimp_image`: (480 x 640 pixels) real-world indoor, moving camera.
+- `aliberts/paris_street`: (720 x 1280 pixels) real-world outdoor, moving camera.
+- `aliberts/kitchen`: (1080 x 1920 pixels) real-world indoor, fixed camera.
 
 Note: The datasets used for this benchmark need to be image datasets, not video datasets.
 
@@ -179,7 +179,7 @@ python benchmark/video/run_video_benchmark.py \
     --output-dir outputs/video_benchmark \
     --repo-ids \
         lerobot/pusht_image \
-        lerobot/aloha_mobile_shrimp_image \
+        aliberts/aloha_mobile_shrimp_image \
     --vcodec libx264 libx265 \
     --pix-fmt yuv444p yuv420p \
     --g 2 20 None \
@@ -203,9 +203,9 @@ python benchmark/video/run_video_benchmark.py \
     --output-dir outputs/video_benchmark \
     --repo-ids \
         lerobot/pusht_image \
-        lerobot/aloha_mobile_shrimp_image \
-        lerobot/paris_street \
-        lerobot/kitchen \
+        aliberts/aloha_mobile_shrimp_image \
+        aliberts/paris_street \
+        aliberts/kitchen \
     --vcodec libx264 libx265 \
     --pix-fmt yuv444p yuv420p \
     --g 1 2 3 4 5 6 10 15 20 40 None \
@@ -221,9 +221,9 @@ python benchmark/video/run_video_benchmark.py \
     --output-dir outputs/video_benchmark \
     --repo-ids \
         lerobot/pusht_image \
-        lerobot/aloha_mobile_shrimp_image \
-        lerobot/paris_street \
-        lerobot/kitchen \
+        aliberts/aloha_mobile_shrimp_image \
+        aliberts/paris_street \
+        aliberts/kitchen \
     --vcodec libsvtav1 \
     --pix-fmt yuv420p \
     --g 1 2 3 4 5 6 10 15 20 40 None \
@@ -252,37 +252,37 @@ Since we're using av1 encoding, we're choosing the `pyav` decoder as `video_read
 
 These tables show the results for `g=2` and `crf=30`, using `timestamps-modes=6_frames` and `backend=pyav`
 
-| video_images_size_ratio           | vcodec     | pix_fmt |           |           |           |
-| --------------------------------- | ---------- | ------- | --------- | --------- | --------- |
-|                                   | libx264    |         | libx265   |           | libsvtav1 |
-| repo_id                           | yuv420p    | yuv444p | yuv420p   | yuv444p   | yuv420p   |
-| lerobot/pusht_image               | **16.97%** | 17.58%  | 18.57%    | 18.86%    | 22.06%    |
-| lerobot/aloha_mobile_shrimp_image | 2.14%      | 2.11%   | 1.38%     | **1.37%** | 5.59%     |
-| lerobot/paris_street              | 2.12%      | 2.13%   | **1.54%** | **1.54%** | 4.43%     |
-| lerobot/kitchen                   | 1.40%      | 1.39%   | **1.00%** | **1.00%** | 2.52%     |
+| video_images_size_ratio            | vcodec     | pix_fmt |           |           |           |
+| ---------------------------------- | ---------- | ------- | --------- | --------- | --------- |
+|                                    | libx264    |         | libx265   |           | libsvtav1 |
+| repo_id                            | yuv420p    | yuv444p | yuv420p   | yuv444p   | yuv420p   |
+| lerobot/pusht_image                | **16.97%** | 17.58%  | 18.57%    | 18.86%    | 22.06%    |
+| aliberts/aloha_mobile_shrimp_image | 2.14%      | 2.11%   | 1.38%     | **1.37%** | 5.59%     |
+| aliberts/paris_street              | 2.12%      | 2.13%   | **1.54%** | **1.54%** | 4.43%     |
+| aliberts/kitchen                   | 1.40%      | 1.39%   | **1.00%** | **1.00%** | 2.52%     |
 
-| video_images_load_time_ratio      | vcodec  | pix_fmt |          |         |           |
-| --------------------------------- | ------- | ------- | -------- | ------- | --------- |
-|                                   | libx264 |         | libx265  |         | libsvtav1 |
-| repo_id                           | yuv420p | yuv444p | yuv420p  | yuv444p | yuv420p   |
-| lerobot/pusht_image               | 6.45    | 5.19    | **1.90** | 2.12    | 2.47      |
-| lerobot/aloha_mobile_shrimp_image | 11.80   | 7.92    | 0.71     | 0.85    | **0.48**  |
-| lerobot/paris_street              | 2.21    | 2.05    | 0.36     | 0.49    | **0.30**  |
-| lerobot/kitchen                   | 1.46    | 1.46    | 0.28     | 0.51    | **0.26**  |
+| video_images_load_time_ratio       | vcodec  | pix_fmt |          |         |           |
+| ---------------------------------- | ------- | ------- | -------- | ------- | --------- |
+|                                    | libx264 |         | libx265  |         | libsvtav1 |
+| repo_id                            | yuv420p | yuv444p | yuv420p  | yuv444p | yuv420p   |
+| lerobot/pusht_image                | 6.45    | 5.19    | **1.90** | 2.12    | 2.47      |
+| aliberts/aloha_mobile_shrimp_image | 11.80   | 7.92    | 0.71     | 0.85    | **0.48**  |
+| aliberts/paris_street              | 2.21    | 2.05    | 0.36     | 0.49    | **0.30**  |
+| aliberts/kitchen                   | 1.46    | 1.46    | 0.28     | 0.51    | **0.26**  |
 
-|                                   |          | vcodec   | pix_fmt      |          |           |              |
-| --------------------------------- | -------- | -------- | ------------ | -------- | --------- | ------------ |
-|                                   |          | libx264  |              | libx265  |           | libsvtav1    |
-| repo_id                           | metric   | yuv420p  | yuv444p      | yuv420p  | yuv444p   | yuv420p      |
-| lerobot/pusht_image               | avg_mse  | 2.90E-04 | **2.03E-04** | 3.13E-04 | 2.29E-04  | 2.19E-04     |
-|                                   | avg_psnr | 35.44    | 37.07        | 35.49    | **37.30** | 37.20        |
-|                                   | avg_ssim | 98.28%   | **98.85%**   | 98.31%   | 98.84%    | 98.72%       |
-| lerobot/aloha_mobile_shrimp_image | avg_mse  | 2.76E-04 | 2.59E-04     | 3.17E-04 | 3.06E-04  | **1.30E-04** |
-|                                   | avg_psnr | 35.91    | 36.21        | 35.88    | 36.09     | **40.17**    |
-|                                   | avg_ssim | 95.19%   | 95.18%       | 95.00%   | 95.05%    | **97.73%**   |
-| lerobot/paris_street              | avg_mse  | 6.89E-04 | 6.70E-04     | 4.03E-03 | 4.02E-03  | **3.09E-04** |
-|                                   | avg_psnr | 33.48    | 33.68        | 32.05    | 32.15     | **35.40**    |
-|                                   | avg_ssim | 93.76%   | 93.75%       | 89.46%   | 89.46%    | **95.46%**   |
-| lerobot/kitchen                   | avg_mse  | 2.50E-04 | 2.24E-04     | 4.28E-04 | 4.18E-04  | **1.53E-04** |
-|                                   | avg_psnr | 36.73    | 37.33        | 36.56    | 36.75     | **39.12**    |
-|                                   | avg_ssim | 95.47%   | 95.58%       | 95.52%   | 95.53%    | **96.82%**   |
+|                                    |          | vcodec   | pix_fmt      |          |           |              |
+| ---------------------------------- | -------- | -------- | ------------ | -------- | --------- | ------------ |
+|                                    |          | libx264  |              | libx265  |           | libsvtav1    |
+| repo_id                            | metric   | yuv420p  | yuv444p      | yuv420p  | yuv444p   | yuv420p      |
+| lerobot/pusht_image                | avg_mse  | 2.90E-04 | **2.03E-04** | 3.13E-04 | 2.29E-04  | 2.19E-04     |
+|                                    | avg_psnr | 35.44    | 37.07        | 35.49    | **37.30** | 37.20        |
+|                                    | avg_ssim | 98.28%   | **98.85%**   | 98.31%   | 98.84%    | 98.72%       |
+| aliberts/aloha_mobile_shrimp_image | avg_mse  | 2.76E-04 | 2.59E-04     | 3.17E-04 | 3.06E-04  | **1.30E-04** |
+|                                    | avg_psnr | 35.91    | 36.21        | 35.88    | 36.09     | **40.17**    |
+|                                    | avg_ssim | 95.19%   | 95.18%       | 95.00%   | 95.05%    | **97.73%**   |
+| aliberts/paris_street              | avg_mse  | 6.89E-04 | 6.70E-04     | 4.03E-03 | 4.02E-03  | **3.09E-04** |
+|                                    | avg_psnr | 33.48    | 33.68        | 32.05    | 32.15     | **35.40**    |
+|                                    | avg_ssim | 93.76%   | 93.75%       | 89.46%   | 89.46%    | **95.46%**   |
+| aliberts/kitchen                   | avg_mse  | 2.50E-04 | 2.24E-04     | 4.28E-04 | 4.18E-04  | **1.53E-04** |
+|                                    | avg_psnr | 36.73    | 37.33        | 36.56    | 36.75     | **39.12**    |
+|                                    | avg_ssim | 95.47%   | 95.58%       | 95.52%   | 95.53%    | **96.82%**   |

docker/Dockerfile.internal

@@ -24,7 +24,7 @@ ARG OS_VERSION=22.04
 FROM nvidia/cuda:${CUDA_VERSION}-base-ubuntu${OS_VERSION}
 
 # Define Python version argument
-ARG PYTHON_VERSION=3.12
+ARG PYTHON_VERSION=3.10
 
 # Configure environment variables
 ENV DEBIAN_FRONTEND=noninteractive \
@@ -73,12 +73,17 @@ ENV HOME=/home/user_lerobot \
 RUN uv venv --python python${PYTHON_VERSION}
 
 # Install Python dependencies for caching
-COPY --chown=user_lerobot:user_lerobot setup.py pyproject.toml uv.lock README.md MANIFEST.in ./
+COPY --chown=user_lerobot:user_lerobot pyproject.toml README.md MANIFEST.in ./
 COPY --chown=user_lerobot:user_lerobot src/ src/
 
-RUN uv sync --locked --extra all --no-cache
+ARG UNBOUND_DEPS=false
 
-RUN chmod +x /lerobot/.venv/lib/python${PYTHON_VERSION}/site-packages/triton/backends/nvidia/bin/ptxas
+RUN if [ "$UNBOUND_DEPS" = "true" ]; then \
+    sed -i 's/,[[:space:]]*<[0-9\.]*//g' pyproject.toml; \
+    echo "Dependencies unbound:" && cat pyproject.toml; \
+    fi
+
+RUN uv pip install --no-cache ".[all]"
 
 # Copy the rest of the application source code
 # Make sure to have the git-LFS files for testing

docker/Dockerfile.user

@@ -18,10 +18,8 @@
 # docker build -f docker/Dockerfile.user -t lerobot-user .
 # docker run -it --rm lerobot-user
 
-# With USB physical access : docker run -it --device=/dev/ -v /dev/:/dev/ --rm lerobot-user
-
 # Configure the base image
-ARG PYTHON_VERSION=3.12
+ARG PYTHON_VERSION=3.10
 FROM python:${PYTHON_VERSION}-slim
 
 # Configure environment variables
@@ -61,10 +59,17 @@ ENV HOME=/home/user_lerobot \
 RUN uv venv
 
 # Install Python dependencies for caching
-COPY --chown=user_lerobot:user_lerobot setup.py pyproject.toml uv.lock README.md MANIFEST.in ./
+COPY --chown=user_lerobot:user_lerobot pyproject.toml README.md MANIFEST.in ./
 COPY --chown=user_lerobot:user_lerobot src/ src/
 
-RUN uv sync --locked --extra all --no-cache
+ARG UNBOUND_DEPS=false
+
+RUN if [ "$UNBOUND_DEPS" = "true" ]; then \
+    sed -i 's/,[[:space:]]*<[0-9\.]*//g' pyproject.toml; \
+    echo "Dependencies unbound:" && cat pyproject.toml; \
+    fi
+
+RUN uv pip install --no-cache ".[all]"
 
 # Copy the rest of the application code
 # Make sure to have the git-LFS files for testing

docker/README.md

@@ -1,77 +0,0 @@
-# Docker
-
-This directory contains Dockerfiles for running LeRobot in containerized environments. Both images are **built nightly from `main`** and published to Docker Hub with the full environment pre-baked — no dependency setup required.
-
-## Pre-built Images
-
-```bash
-# CPU-only image (based on Dockerfile.user)
-docker pull huggingface/lerobot-cpu:latest
-
-# GPU image with CUDA support (based on Dockerfile.internal)
-docker pull huggingface/lerobot-gpu:latest
-```
-
-## Quick Start
-
-The fastest way to start training is to pull the GPU image and run `lerobot-train` directly. This is the same environment used for all of our CI, so it is a well-tested, batteries-included setup.
-
-```bash
-docker run -it --rm --gpus all --shm-size 16gb huggingface/lerobot-gpu:latest
-
-# inside the container:
-lerobot-train --policy.type=act --dataset.repo_id=lerobot/aloha_sim_transfer_cube_human
-```
-
-## Dockerfiles
-
-### `Dockerfile.user` (CPU)
-
-A lightweight image based on `python:3.12-slim`. Includes all Python dependencies and system libraries but does not include CUDA — there is no GPU support. Useful for exploring the codebase, running scripts, or working with robots, but not practical for training.
-
-### `Dockerfile.internal` (GPU)
-
-A CUDA-enabled image based on `nvidia/cuda`. This is the image for training — mostly used for internal interactions with the GPU cluster.
-
-## Usage
-
-### Running a pre-built image
-
-```bash
-# CPU
-docker run -it --rm huggingface/lerobot-cpu:latest
-
-# GPU
-docker run -it --rm --gpus all --shm-size 16gb huggingface/lerobot-gpu:latest
-```
-
-### Building locally
-
-From the repo root:
-
-```bash
-# CPU
-docker build -f docker/Dockerfile.user -t lerobot-user .
-docker run -it --rm lerobot-user
-
-# GPU
-docker build -f docker/Dockerfile.internal -t lerobot-internal .
-docker run -it --rm --gpus all --shm-size 16gb lerobot-internal
-```
-
-### Multi-GPU training
-
-To select specific GPUs, set `CUDA_VISIBLE_DEVICES` when launching the container:
-
-```bash
-# Use 4 GPUs
-docker run -it --rm --gpus all --shm-size 16gb \
-  -e CUDA_VISIBLE_DEVICES=0,1,2,3 \
-  huggingface/lerobot-gpu:latest
-```
-
-### USB device access (e.g. robots, cameras)
-
-```bash
-docker run -it --device=/dev/ -v /dev/:/dev/ --rm huggingface/lerobot-cpu:latest
-```

docs/source/_toctree.yml

@@ -7,6 +7,8 @@
 - sections:
   - local: il_robots
     title: Imitation Learning for Robots
+  - local: cameras
+    title: Cameras
   - local: bring_your_own_policies
     title: Bring Your Own Policies
   - local: integrate_hardware
@@ -17,12 +19,8 @@
     title: Train RL in Simulation
   - local: multi_gpu_training
     title: Multi GPU training
-  - local: hil_data_collection
-    title: Human In the Loop Data Collection
   - local: peft_training
     title: Training with PEFT (e.g., LoRA)
-  - local: rename_map
-    title: Using Rename Map and Empty Cameras
   title: "Tutorials"
 - sections:
   - local: lerobot-dataset-v3
@@ -31,10 +29,6 @@
     title: Porting Large Datasets
   - local: using_dataset_tools
     title: Using the Dataset Tools
-  - local: dataset_subtask
-    title: Using Subtasks in the Dataset
-  - local: streaming_video_encoding
-    title: Streaming Video Encoding
   title: "Datasets"
 - sections:
   - local: act
@@ -43,18 +37,12 @@
     title: SmolVLA
   - local: pi0
     title: π₀ (Pi0)
-  - local: pi0fast
-    title: π₀-FAST (Pi0Fast)
   - local: pi05
     title: π₀.₅ (Pi05)
   - local: groot
     title: NVIDIA GR00T N1.5
   - local: xvla
     title: X-VLA
-  - local: multi_task_dit
-    title: Multitask DiT Policy
-  - local: walloss
-    title: WALL-OSS
   title: "Policies"
 - sections:
   - local: sarm
@@ -71,17 +59,11 @@
     title: Environments from the Hub
   - local: envhub_leisaac
     title: Control & Train Robots in Sim (LeIsaac)
-  title: "Simulation"
-- sections:
-  - local: adding_benchmarks
-    title: Adding a New Benchmark
   - local: libero
-    title: LIBERO
+    title: Using Libero
   - local: metaworld
-    title: Meta-World
-  - local: envhub_isaaclab_arena
-    title: NVIDIA IsaacLab Arena Environments
-  title: "Benchmarks"
+    title: Using MetaWorld
+  title: "Simulation"
 - sections:
   - local: introduction_processors
     title: Introduction to Robot Processors
@@ -93,8 +75,6 @@
     title: Processors for Robots and Teleoperators
   - local: env_processor
     title: Environment Processors
-  - local: action_representations
-    title: Action Representations
   title: "Robot Processors"
 - sections:
   - local: so101
@@ -113,19 +93,11 @@
     title: Unitree G1
   - local: earthrover_mini_plus
     title: Earth Rover Mini
-  - local: omx
-    title: OMX
-  - local: openarm
-    title: OpenArm
   title: "Robots"
 - sections:
   - local: phone_teleop
     title: Phone
   title: "Teleoperators"
-- sections:
-  - local: cameras
-    title: Cameras
-  title: "Sensors"
 - sections:
   - local: torch_accelerators
     title: PyTorch accelerators
@@ -135,8 +107,6 @@
     title: Notebooks
   - local: feetech
     title: Updating Feetech Firmware
-  - local: damiao
-    title: Damiao Motors and CAN Bus
   title: "Resources"
 - sections:
   - local: contributing

docs/source/act.mdx

@@ -88,8 +88,5 @@ lerobot-record \
   --dataset.repo_id=${HF_USER}/eval_act_your_dataset \
   --dataset.num_episodes=10 \
   --dataset.single_task="Your task description" \
-  --dataset.streaming_encoding=true \
-  --dataset.encoder_threads=2 \
-  # --dataset.vcodec=auto \
   --policy.path=${HF_USER}/act_policy
 ```

docs/source/action_representations.mdx

@@ -1,223 +0,0 @@
-# Action Representations
-
-This guide explains the different ways robot actions can be represented in LeRobot, how they relate to each other, and when to use each one.
-
-## Joint Space vs End-Effector Space
-
-Before discussing action representations, it helps to understand the two coordinate spaces actions can live in.
-
-### Joint Space
-
-Joint-space actions directly specify target positions for each motor. For a 6-DOF arm with a gripper, a joint-space action might look like:
-
-```
-action = [shoulder_pan: 45.0, shoulder_lift: -20.0, elbow: -30.0, wrist_pitch: 10.0, wrist_roll: 0.0, wrist_yaw: 5.0, gripper: 0.8]
-```
-
-Joint space is the default in LeRobot. It is simple, requires no kinematics model, and maps directly to motor commands. Most beginner setups (SO-100, Koch) use joint-space actions.
-
-### End-Effector (EE) Space
-
-End-effector-space actions specify the desired position and orientation of the robot's tool tip (gripper) in Cartesian coordinates:
-
-```
-action = [x: 0.25, y: -0.10, z: 0.15, wx: 0.0, wy: 0.0, wz: 0.1, gripper: 0.8]
-```
-
-EE space is more intuitive for tasks like pick-and-place because it directly describes where the gripper should go, but it requires a kinematics model (URDF) to convert between EE poses and joint angles.
-
-### Converting Between Spaces
-
-LeRobot provides processor steps for converting between joint and EE spaces using forward and inverse kinematics. These are built on top of `RobotKinematics`, which loads a URDF model of your robot.
-
-```python
-from lerobot.model.kinematics import RobotKinematics
-from lerobot.robots.so_follower.robot_kinematic_processor import (
-    ForwardKinematicsJointsToEE,
-    InverseKinematicsEEToJoints,
-)
-
-kinematics = RobotKinematics(
-    urdf_path="./SO101/so101_new_calib.urdf",
-    target_frame_name="gripper_frame_link",
-    joint_names=["shoulder", "elbow", "wrist_pitch", "wrist_roll", "wrist_yaw"],
-)
-
-# Joints → EE (for observations: "where is my gripper?")
-fk_step = ForwardKinematicsJointsToEE(kinematics=kinematics, motor_names=[...])
-
-# EE → Joints (for actions: "move my gripper here")
-ik_step = InverseKinematicsEEToJoints(kinematics=kinematics, motor_names=[...])
-```
-
-See [`examples/so100_to_so100_EE/`](https://github.com/huggingface/lerobot/tree/main/examples/so100_to_so100_EE) for a complete working example of recording, replaying, and evaluating with EE-space actions on an SO-100 arm.
-
-## Absolute, Relative, and Delta Actions
-
-Regardless of whether you work in joint space or EE space, the action values can be expressed in three different ways. The terminology follows [UMI (Chi et al., 2024)](https://arxiv.org/abs/2402.10329).
-
-### Absolute Actions (LeRobot default)
-
-Each action specifies the target position directly.
-
-**Example** (joint space, chunk of 4):
-
-```
-current_state = [45.0, -30.0, 10.0]
-
-action_chunk = [
-    [46.0, -29.0, 11.0],   # go to 46, -29, 11
-    [47.5, -27.0, 12.0],   # go to 47.5, -27, 12
-    [49.0, -25.0, 13.5],   # go to 49, -25, 13.5
-    [50.0, -24.0, 15.0],   # go to 50, -24, 15
-]
-```
-
-Each value is a target position in the robot's coordinate frame. Simple and direct, but requires a consistent global coordinate frame. This is the default in LeRobot.
-
-### Relative Actions (used by OpenPI / pi0)
-
-Each action in the chunk is an offset from the **current state at the moment of prediction**. All actions in the chunk share the same reference point:
-
-```
-current_state = [45.0, -30.0, 10.0]
-
-relative_chunk = [
-    [1.0,  1.0, 1.0],   # +1 from current → target 46, -29, 11
-    [2.5,  3.0, 2.0],   # +2.5 from current → target 47.5, -27, 12
-    [4.0,  5.0, 3.5],   # +4 from current → target 49, -25, 13.5
-    [5.0,  6.0, 5.0],   # +5 from current → target 50, -24, 15
-]
-```
-
-The conversion is straightforward: `relative = absolute - current_state`. To recover absolute: `absolute = relative + current_state`.
-
-**Why use relative actions?** The model learns to predict offsets centered around zero, which is easier to normalize and leads to more stable training. Because every chunk references the same current state, there is no error accumulation across chunks.
-
-### Delta Actions (sequential differences)
-
-Each action is an offset from the **previous action** (or from the current state for the first step):
-
-```
-current_state = [45.0, -30.0, 10.0]
-
-delta_chunk = [
-    [1.0,  1.0, 1.0],   # current → 46, -29, 11
-    [1.5,  2.0, 1.0],   # previous action → 47.5, -27, 12
-    [1.5,  2.0, 1.5],   # previous action → 49, -25, 13.5
-    [1.0,  1.0, 1.5],   # previous action → 50, -24, 15
-]
-```
-
-Here each step is relative to the one before it. To recover absolute positions you must sum all previous deltas, which means errors accumulate over time. UMI explicitly argues against this representation for this reason.
-
-### Visual Comparison
-
-The figure below (based on a figure from [UMI, Chi et al., 2024](https://arxiv.org/abs/2402.10329)) illustrates the key difference. With **relative trajectory**, every action in the chunk points back to the same origin (current state), so a new inference step cleanly resets the reference. With **delta**, each action depends on the previous one, so errors accumulate. **Absolute** actions require a consistent global coordinate frame.
-
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/action_representations_umi.png"
-  alt="Relative Trajectory as Action Representation (UMI, Chi et al., 2024)"
-  width="85%"
-/>
-
-## Using Relative Actions in LeRobot
-
-LeRobot provides `RelativeActionsProcessorStep` to convert between absolute and relative actions inside the processor pipeline. This is how pi0, pi0.5, and pi0_fast support relative actions.
-
-> **Note:** All pi models (pi0, pi0.5, pi0*fast) apply relative conversion \_before* normalization (`relative → normalize`), so the normalizer always sees delta (relative) values. This means **relative action stats are required** for all of them when training with `use_relative_actions=true`. In pi0_fast the `RelativeActionsProcessorStep` only modifies the action — the state observation is unchanged — so `NormalizerProcessorStep` still runs before the state tokenizer and the tokenizer continues to receive normalized state as expected.
-
-### How it works
-
-During **training** (preprocessing), actions are converted from absolute to relative before the model sees them:
-
-```
-raw absolute action → RelativeActionsProcessorStep → normalize → model
-```
-
-During **inference** (postprocessing), model predictions are converted back to absolute before being sent to the robot:
-
-```
-model output → unnormalize → AbsoluteActionsProcessorStep → robot
-```
-
-The `AbsoluteActionsProcessorStep` reads the cached current state from its paired `RelativeActionsProcessorStep`, so the two must be wired together (handled automatically by the policy factory).
-
-### Enabling relative actions for the pi family (pi0, pi0.5, pi0_fast)
-
-**Step 1**: Precompute relative action statistics for your dataset:
-
-```bash
-lerobot-edit-dataset \
-    --repo_id your_dataset \
-    --operation.type recompute_stats \
-    --operation.relative_action true \
-    --operation.chunk_size 50 \
-    --operation.relative_exclude_joints "['gripper']"
-```
-
-**Step 2**: Train with relative actions enabled:
-
-```bash
-lerobot-train \
-    --dataset.repo_id=your_dataset \
-    --policy.type=pi0 \
-    --policy.use_relative_actions=true \
-    --policy.relative_exclude_joints='["gripper"]'
-```
-
-The `relative_exclude_joints` parameter specifies joints that should remain in absolute space. For example, gripper commands are typically binary (open/close) and don't benefit from relative encoding.
-
-### Combining relative actions with RTC
-
-[RTC](https://arxiv.org/abs/2506.07339) runs policy inference at high frequency and sends actions to the robot as they are predicted rather than waiting for a full chunk. Relative actions and RTC are fully compatible: because every chunk in relative mode references the **same** current state (captured at the start of inference), each predicted action in the chunk remains a valid offset even if the robot has already moved. No special handling is needed — `RelativeActionsProcessorStep` caches the state once per inference call and `AbsoluteActionsProcessorStep` applies it to every action in the streamed output.
-
-### Combining relative actions with EE space
-
-Relative actions work in both joint space and EE space. For example, if your dataset stores EE actions, relative encoding converts them to offsets from the current EE pose:
-
-```
-current_ee_state = [x: 0.25, y: -0.10, z: 0.15, gripper: 0.8]
-
-absolute_ee_chunk = [
-    [0.26, -0.09, 0.16, 0.8],
-    [0.28, -0.07, 0.18, 0.8],
-]
-
-relative_ee_chunk = [
-    [0.01,  0.01, 0.01, 0.0],   # offset from current EE pose
-    [0.03,  0.03, 0.03, 0.0],   # offset from current EE pose
-]
-```
-
-## Processing Pipeline Summary
-
-Here is how the different processors compose. Each arrow is a processor step, and they can be chained in a `RobotProcessorPipeline` or `PolicyProcessorPipeline`:
-
-```
-                    ┌─────────────────────────────────────────┐
-   Action Space     │   Joint Space  ←──IK──→  EE Space      │
-                    │   ForwardKinematicsJointsToEE           │
-                    │   InverseKinematicsEEToJoints           │
-                    └─────────────────────────────────────────┘
-
-                    ┌─────────────────────────────────────────┐
-   Representation   │   Absolute  ←────→  Relative            │
-                    │   RelativeActionsProcessorStep (pre)    │
-                    │   AbsoluteActionsProcessorStep (post)   │
-                    └─────────────────────────────────────────┘
-
-                    ┌─────────────────────────────────────────┐
-   Normalization    │   Raw  ←────→  Normalized               │
-                    │   NormalizerProcessorStep (pre)         │
-                    │   UnnormalizerProcessorStep (post)      │
-                    └─────────────────────────────────────────┘
-```
-
-A typical training preprocessor might chain: `raw absolute joint actions → relative → normalize`. A typical inference postprocessor: `unnormalize → absolute → (optionally IK to joints)`.
-
-## References
-
-- [Universal Manipulation Interface (UMI)](https://arxiv.org/abs/2402.10329) - Chi et al., 2024. Defines the relative trajectory action representation and compares it with absolute and delta actions.
-- [Introduction to Processors](./introduction_processors) - How processor pipelines work in LeRobot.
-- [`examples/so100_to_so100_EE/`](https://github.com/huggingface/lerobot/tree/main/examples/so100_to_so100_EE) - Complete example of recording and evaluating with EE-space actions.

docs/source/adding_benchmarks.mdx

@@ -1,329 +0,0 @@
-# Adding a New Benchmark
-
-This guide walks you through adding a new simulation benchmark to LeRobot. Follow the steps in order and use the existing benchmarks as templates.
-
-A benchmark in LeRobot is a set of [Gymnasium](https://gymnasium.farama.org/) environments that wrap a third-party simulator (like LIBERO or Meta-World) behind a standard `gym.Env` interface. The `lerobot-eval` CLI then runs evaluation uniformly across all benchmarks.
-
-## Existing benchmarks at a glance
-
-Before diving in, here is what is already integrated:
-
-| Benchmark      | Env file            | Config class       | Tasks               | Action dim   | Processor                    |
-| -------------- | ------------------- | ------------------ | ------------------- | ------------ | ---------------------------- |
-| LIBERO         | `envs/libero.py`    | `LiberoEnv`        | 130 across 5 suites | 7            | `LiberoProcessorStep`        |
-| Meta-World     | `envs/metaworld.py` | `MetaworldEnv`     | 50 (MT50)           | 4            | None                         |
-| IsaacLab Arena | Hub-hosted          | `IsaaclabArenaEnv` | Configurable        | Configurable | `IsaaclabArenaProcessorStep` |
-
-Use `src/lerobot/envs/libero.py` and `src/lerobot/envs/metaworld.py` as reference implementations.
-
-## How it all fits together
-
-### Data flow
-
-During evaluation, data moves through four stages:
-
-```
-1. gym.Env  ──→  raw observations (numpy dicts)
-
-2. Preprocessing  ──→  standard LeRobot keys + task description
-   (preprocess_observation, add_envs_task in envs/utils.py)
-
-3. Processors  ──→  env-specific then policy-specific transforms
-   (env_preprocessor, policy_preprocessor)
-
-4. Policy  ──→  select_action()  ──→  action tensor
-   then reverse: policy_postprocessor → env_postprocessor → numpy action → env.step()
-```
-
-Most benchmarks only need to care about stage 1 (producing observations in the right format) and optionally stage 3 (if env-specific transforms are needed).
-
-### Environment structure
-
-`make_env()` returns a nested dict of vectorized environments:
-
-```python
-dict[str, dict[int, gym.vector.VectorEnv]]
-#    ^suite       ^task_id
-```
-
-A single-task env (e.g. PushT) looks like `{"pusht": {0: vec_env}}`.
-A multi-task benchmark (e.g. LIBERO) looks like `{"libero_spatial": {0: vec0, 1: vec1, ...}, ...}`.
-
-### How evaluation runs
-
-All benchmarks are evaluated the same way by `lerobot-eval`:
-
-1. `make_env()` builds the nested `{suite: {task_id: VectorEnv}}` dict.
-2. `eval_policy_all()` iterates over every suite and task.
-3. For each task, it runs `n_episodes` rollouts via `rollout()`.
-4. Results are aggregated hierarchically: episode, task, suite, overall.
-5. Metrics include `pc_success` (success rate), `avg_sum_reward`, and `avg_max_reward`.
-
-The critical piece: your env must return `info["is_success"]` on every `step()` call. This is how the eval loop knows whether a task was completed.
-
-## What your environment must provide
-
-LeRobot does not enforce a strict observation schema. Instead it relies on a set of conventions that all benchmarks follow.
-
-### Env attributes
-
-Your `gym.Env` must set these attributes:
-
-| Attribute            | Type  | Why                                                  |
-| -------------------- | ----- | ---------------------------------------------------- |
-| `_max_episode_steps` | `int` | `rollout()` uses this to cap episode length          |
-| `task_description`   | `str` | Passed to VLA policies as a language instruction     |
-| `task`               | `str` | Fallback identifier if `task_description` is not set |
-
-### Success reporting
-
-Your `step()` and `reset()` must include `"is_success"` in the `info` dict:
-
-```python
-info = {"is_success": True}   # or False
-return observation, reward, terminated, truncated, info
-```
-
-### Observations
-
-The simplest approach is to map your simulator's outputs to the standard keys that `preprocess_observation()` already understands. Do this inside your `gym.Env` (e.g. in a `_format_raw_obs()` helper):
-
-| Your env should output    | LeRobot maps it to         | What it is                            |
-| ------------------------- | -------------------------- | ------------------------------------- |
-| `"pixels"` (single array) | `observation.image`        | Single camera image, HWC uint8        |
-| `"pixels"` (dict)         | `observation.images.<cam>` | Multiple cameras, each HWC uint8      |
-| `"agent_pos"`             | `observation.state`        | Proprioceptive state vector           |
-| `"environment_state"`     | `observation.env_state`    | Full environment state (e.g. PushT)   |
-| `"robot_state"`           | `observation.robot_state`  | Nested robot state dict (e.g. LIBERO) |
-
-If your simulator uses different key names, you have two options:
-
-1. **Recommended:** Rename them to the standard keys inside your `gym.Env` wrapper.
-2. **Alternative:** Write an env processor to transform observations after `preprocess_observation()` runs (see step 4 below).
-
-### Actions
-
-Actions are continuous numpy arrays in a `gym.spaces.Box`. The dimensionality depends on your benchmark (7 for LIBERO, 4 for Meta-World, etc.). Policies adapt to different action dimensions through their `input_features` / `output_features` config.
-
-### Feature declaration
-
-Each `EnvConfig` subclass declares two dicts that tell the policy what to expect:
-
-- `features` — maps feature names to `PolicyFeature(type, shape)` (e.g. action dim, image shape).
-- `features_map` — maps raw observation keys to LeRobot convention keys (e.g. `"agent_pos"` to `"observation.state"`).
-
-## Step by step
-
-<Tip>
-  At minimum, you need three files: a **gym.Env wrapper**, an **EnvConfig
-  subclass**, and a **factory dispatch branch**. Everything else is optional or
-  documentation.
-</Tip>
-
-### Checklist
-
-| File                                     | Required | Why                                       |
-| ---------------------------------------- | -------- | ----------------------------------------- |
-| `src/lerobot/envs/<benchmark>.py`        | Yes      | Wraps the simulator as a standard gym.Env |
-| `src/lerobot/envs/configs.py`            | Yes      | Registers your benchmark for the CLI      |
-| `src/lerobot/envs/factory.py`            | Yes      | Tells `make_env()` how to build your envs |
-| `src/lerobot/processor/env_processor.py` | Optional | Custom observation/action transforms      |
-| `src/lerobot/envs/utils.py`              | Optional | Only if you need new raw observation keys |
-| `pyproject.toml`                         | Yes      | Declares benchmark-specific dependencies  |
-| `docs/source/<benchmark>.mdx`            | Yes      | User-facing documentation page            |
-| `docs/source/_toctree.yml`               | Yes      | Adds your page to the docs sidebar        |
-
-### 1. The gym.Env wrapper (`src/lerobot/envs/<benchmark>.py`)
-
-Create a `gym.Env` subclass that wraps the third-party simulator:
-
-```python
-class MyBenchmarkEnv(gym.Env):
-    metadata = {"render_modes": ["rgb_array"], "render_fps": <fps>}
-
-    def __init__(self, task_suite, task_id, ...):
-        super().__init__()
-        self.task = <task_name_string>
-        self.task_description = <natural_language_instruction>
-        self._max_episode_steps = <max_steps>
-        self.observation_space = spaces.Dict({...})
-        self.action_space = spaces.Box(low=..., high=..., shape=(...,), dtype=np.float32)
-
-    def reset(self, seed=None, **kwargs):
-        ...  # return (observation, info) — info must contain {"is_success": False}
-
-    def step(self, action: np.ndarray):
-        ...  # return (obs, reward, terminated, truncated, info) — info must contain {"is_success": <bool>}
-
-    def render(self):
-        ...  # return RGB image as numpy array
-
-    def close(self):
-        ...
-```
-
-Also provide a factory function that returns the nested dict structure:
-
-```python
-def create_mybenchmark_envs(
-    task: str,
-    n_envs: int,
-    gym_kwargs: dict | None = None,
-    env_cls: type | None = None,
-) -> dict[str, dict[int, Any]]:
-    """Create {suite_name: {task_id: VectorEnv}} for MyBenchmark."""
-    ...
-```
-
-See `create_libero_envs()` (multi-suite, multi-task) and `create_metaworld_envs()` (difficulty-grouped tasks) for reference.
-
-### 2. The config (`src/lerobot/envs/configs.py`)
-
-Register a config dataclass so users can select your benchmark with `--env.type=<name>`:
-
-```python
-@EnvConfig.register_subclass("<benchmark_name>")
-@dataclass
-class MyBenchmarkEnvConfig(EnvConfig):
-    task: str = "<default_task>"
-    fps: int = <fps>
-    obs_type: str = "pixels_agent_pos"
-
-    features: dict[str, PolicyFeature] = field(default_factory=lambda: {
-        ACTION: PolicyFeature(type=FeatureType.ACTION, shape=(<action_dim>,)),
-    })
-    features_map: dict[str, str] = field(default_factory=lambda: {
-        ACTION: ACTION,
-        "agent_pos": OBS_STATE,
-        "pixels": OBS_IMAGE,
-    })
-
-    def __post_init__(self):
-        ...  # populate features based on obs_type
-
-    @property
-    def gym_kwargs(self) -> dict:
-        return {"obs_type": self.obs_type, "render_mode": self.render_mode}
-```
-
-Key points:
-
-- The `register_subclass` name is what users pass on the CLI (`--env.type=<name>`).
-- `features` tells the policy what the environment produces.
-- `features_map` maps raw observation keys to LeRobot convention keys.
-
-### 3. The factory dispatch (`src/lerobot/envs/factory.py`)
-
-Add a branch in `make_env()` to call your factory function:
-
-```python
-elif "<benchmark_name>" in cfg.type:
-    from lerobot.envs.<benchmark> import create_<benchmark>_envs
-
-    if cfg.task is None:
-        raise ValueError("<BenchmarkName> requires a task to be specified")
-
-    return create_<benchmark>_envs(
-        task=cfg.task,
-        n_envs=n_envs,
-        gym_kwargs=cfg.gym_kwargs,
-        env_cls=env_cls,
-    )
-```
-
-If your benchmark needs an env processor, add it in `make_env_pre_post_processors()`:
-
-```python
-if isinstance(env_cfg, MyBenchmarkEnvConfig) or "<benchmark_name>" in env_cfg.type:
-    preprocessor_steps.append(MyBenchmarkProcessorStep())
-```
-
-### 4. Env processor (optional — `src/lerobot/processor/env_processor.py`)
-
-Only needed if your benchmark requires observation transforms beyond what `preprocess_observation()` handles (e.g. image flipping, coordinate conversion):
-
-```python
-@dataclass
-@ProcessorStepRegistry.register(name="<benchmark>_processor")
-class MyBenchmarkProcessorStep(ObservationProcessorStep):
-    def _process_observation(self, observation):
-        processed = observation.copy()
-        # your transforms here
-        return processed
-
-    def transform_features(self, features):
-        return features  # update if shapes change
-
-    def observation(self, observation):
-        return self._process_observation(observation)
-```
-
-See `LiberoProcessorStep` for a full example (image rotation, quaternion-to-axis-angle conversion).
-
-### 5. Dependencies (`pyproject.toml`)
-
-Add a new optional-dependency group:
-
-```toml
-mybenchmark = ["my-benchmark-pkg==1.2.3", "lerobot[scipy-dep]"]
-```
-
-Pinning rules:
-
-- **Always pin** benchmark packages to exact versions for reproducibility (e.g. `metaworld==3.0.0`).
-- **Add platform markers** when needed (e.g. `; sys_platform == 'linux'`).
-- **Pin fragile transitive deps** if known (e.g. `gymnasium==1.1.0` for Meta-World).
-- **Document constraints** in your benchmark doc page.
-
-Users install with:
-
-```bash
-pip install -e ".[mybenchmark]"
-```
-
-### 6. Documentation (`docs/source/<benchmark>.mdx`)
-
-Write a user-facing page following the template in the next section. See `docs/source/libero.mdx` and `docs/source/metaworld.mdx` for full examples.
-
-### 7. Table of contents (`docs/source/_toctree.yml`)
-
-Add your benchmark to the "Benchmarks" section:
-
-```yaml
-- sections:
-    - local: libero
-      title: LIBERO
-    - local: metaworld
-      title: Meta-World
-    - local: envhub_isaaclab_arena
-      title: NVIDIA IsaacLab Arena Environments
-    - local: <your_benchmark>
-      title: <Your Benchmark Name>
-  title: "Benchmarks"
-```
-
-## Verifying your integration
-
-After completing the steps above, confirm that everything works:
-
-1. **Install** — `pip install -e ".[mybenchmark]"` and verify the dependency group installs cleanly.
-2. **Smoke test env creation** — call `make_env()` with your config in Python, check that the returned dict has the expected `{suite: {task_id: VectorEnv}}` shape, and that `reset()` returns observations with the right keys.
-3. **Run a full eval** — `lerobot-eval --env.type=<name> --env.task=<task> --eval.n_episodes=1 --eval.batch_size=1 --policy.path=<any_compatible_policy>` to exercise the full pipeline end-to-end.
-4. **Check success detection** — verify that `info["is_success"]` flips to `True` when the task is actually completed. This is what the eval loop uses to compute success rates.
-
-## Writing a benchmark doc page
-
-Each benchmark `.mdx` page should include:
-
-- **Title and description** — 1-2 paragraphs on what the benchmark tests and why it matters.
-- **Links** — paper, GitHub repo, project website (if available).
-- **Overview image or GIF.**
-- **Available tasks** — table of task suites with counts and brief descriptions.
-- **Installation** — `pip install -e ".[<benchmark>]"` plus any extra steps (env vars, system packages).
-- **Evaluation** — recommended `lerobot-eval` command with `n_episodes` and `batch_size` for reproducible results. Include single-task and multi-task examples if applicable.
-- **Policy inputs and outputs** — observation keys with shapes, action space description.
-- **Recommended evaluation episodes** — how many episodes per task is standard.
-- **Training** — example `lerobot-train` command.
-- **Reproducing published results** — link to pretrained model, eval command, results table (if available).
-
-See `docs/source/libero.mdx` and `docs/source/metaworld.mdx` for complete examples.

docs/source/async.mdx

@@ -48,7 +48,7 @@ python -m lerobot.async_inference.robot_client \
     --task="dummy" \ # POLICY: The task to run the policy on (`Fold my t-shirt`). Not necessarily defined for all policies, such as `act`
     --policy_type=your_policy_type \ # POLICY: the type of policy to run (smolvla, act, etc)
     --pretrained_name_or_path=user/model \ # POLICY: the model name/path on server to the checkpoint to run (e.g., lerobot/smolvla_base)
-    --policy_device=mps \ # POLICY: the device to run the policy on, on the server (cuda, mps, xpu, cpu)
+    --policy_device=mps \ # POLICY: the device to run the policy on, on the server
     --actions_per_chunk=50 \ # POLICY: the number of actions to output at once
     --chunk_size_threshold=0.5 \ # CLIENT: the threshold for the chunk size before sending a new observation to the server
     --aggregate_fn_name=weighted_average \ # CLIENT: the function to aggregate actions on overlapping portions
@@ -169,7 +169,7 @@ python -m lerobot.async_inference.robot_client \
 <!-- prettier-ignore-start -->
 ```python
 import threading
-from lerobot.robots.so_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower import SO100FollowerConfig
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
 from lerobot.async_inference.configs import RobotClientConfig
 from lerobot.async_inference.robot_client import RobotClient
@@ -195,7 +195,6 @@ client_cfg = RobotClientConfig(
     robot=robot_cfg,
     server_address="localhost:8080",
     policy_device="mps",
-    client_device="cpu",
     policy_type="smolvla",
     pretrained_name_or_path="<user>/smolvla_async",
     chunk_size_threshold=0.5,
@@ -310,4 +309,4 @@ Asynchronous inference represents a significant advancement in real-time robotic
 - **Universal Compatibility**: Works with all LeRobot-supported policies, from lightweight ACT models to vision-language models like SmolVLA
 
 Start experimenting with the default parameters, monitor your action queue sizes, and iteratively refine your setup to achieve optimal performance for your specific use case.
-If you want to discuss this further, hop into our [Discord community](https://discord.gg/s3KuuzsPFb), or open an issue on our [GitHub repository](https://github.com/huggingface/lerobot/issues).
+If you want to discuss this further, hop into our [Discord community](https://discord.gg/s3KuuzsPFb), or open an issue on our [GitHub repository](https://github.com/lerobot/lerobot/issues).

docs/source/bring_your_own_policies.mdx

@@ -32,7 +32,7 @@ version = "0.1.0"
 dependencies = [
     # your policy-specific dependencies
 ]
-requires-python = ">= 3.12"
+requires-python = ">= 3.11"
 
 [build-system]
 build-backend = # your-build-backend
@@ -41,15 +41,13 @@ requires = # your-build-system
 
 ## Step 2: Define the Policy Configuration
 
-Create a configuration class that inherits from [`PreTrainedConfig`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/configs/policies.py) and registers your policy type:
-Here is a template to get you started, customize the parameters and methods as needed for your policy's architecture and training requirements.
+Create a configuration class that inherits from `PreTrainedConfig` and registers your policy type:
 
 ```python
 # configuration_my_custom_policy.py
 from dataclasses import dataclass, field
 from lerobot.configs.policies import PreTrainedConfig
-from lerobot.optim.optimizers import AdamWConfig
-from lerobot.optim.schedulers import CosineDecayWithWarmupSchedulerConfig
+from lerobot.configs.types import NormalizationMode
 
 @PreTrainedConfig.register_subclass("my_custom_policy")
 @dataclass
@@ -63,131 +61,61 @@ class MyCustomPolicyConfig(PreTrainedConfig):
         hidden_dim: Hidden dimension for the policy network
         # Add your policy-specific parameters here
     """
-
-    horizon: int = 50
-    n_action_steps: int = 50
-    hidden_dim: int = 256
-
-    optimizer_lr: float = 1e-4
-    optimizer_weight_decay: float = 1e-4
+    # ...PreTrainedConfig fields...
+    pass
 
     def __post_init__(self):
         super().__post_init__()
-        if self.n_action_steps > self.horizon:
-            raise ValueError("n_action_steps cannot exceed horizon")
+        # Add any validation logic here
 
     def validate_features(self) -> None:
         """Validate input/output feature compatibility."""
-        if not self.image_features:
-            raise ValueError("MyCustomPolicy requires at least one image feature.")
-        if self.action_feature is None:
-            raise ValueError("MyCustomPolicy requires 'action' in output_features.")
-
-    def get_optimizer_preset(self) -> AdamWConfig:
-        return AdamWConfig(lr=self.optimizer_lr, weight_decay=self.optimizer_weight_decay)
-
-    def get_scheduler_preset(self):
-        return None
-
-    @property
-    def observation_delta_indices(self) -> list[int] | None:
-        """Relative timestep offsets the dataset loader provides per observation.
-
-        Return `None` for single-frame policies. For temporal policies that consume
-        multiple past or future frames, return a list of offsets, e.g. `[-20, -10, 0, 10]` for
-        3 past frames at stride 10 and 1 future frame at stride 10.
-        """
-        return None
-
-    @property
-    def action_delta_indices(self) -> list[int]:
-        """Relative timestep offsets for the action chunk the dataset loader returns.
-        """
-        return list(range(self.horizon))
-
-    @property
-    def reward_delta_indices(self) -> None:
-        return None
+        # Implement validation logic for your policy's requirements
+        pass
 ```
 
 ## Step 3: Implement the Policy Class
 
-Create your policy implementation by inheriting from [`PreTrainedPolicy`](https://github.com/huggingface/lerobot/blob/main/src/lerobot/policies/pretrained.py):
+Create your policy implementation by inheriting from LeRobot's base `PreTrainedPolicy` class:
 
 ```python
 # modeling_my_custom_policy.py
 import torch
 import torch.nn as nn
-from typing import Any
+from typing import Dict, Any
 
 from lerobot.policies.pretrained import PreTrainedPolicy
-from lerobot.utils.constants import ACTION
 from .configuration_my_custom_policy import MyCustomPolicyConfig
 
 class MyCustomPolicy(PreTrainedPolicy):
-    config_class = MyCustomPolicyConfig  # must match the string in @register_subclass
+    config_class = MyCustomPolicyConfig
     name = "my_custom_policy"
 
-    def __init__(self, config: MyCustomPolicyConfig, dataset_stats: dict[str, Any] = None):
+    def __init__(self, config: MyCustomPolicyConfig, dataset_stats: Dict[str, Any] = None):
         super().__init__(config, dataset_stats)
-        config.validate_features()  # not called automatically by the base class
-        self.config = config
-        self.model = ...  # your nn.Module here
-
-    def reset(self):
-        """Reset episode state."""
         ...
-
-    def get_optim_params(self) -> dict:
-        """Return parameters to pass to the optimizer (e.g. with per-group lr/wd)."""
-        return {"params": self.parameters()}
-
-    def predict_action_chunk(self, batch: dict[str, torch.Tensor], **kwargs) -> torch.Tensor:
-        """Return the full action chunk (B, chunk_size, action_dim) for the current observation."""
-        ...
-
-    def select_action(self, batch: dict[str, torch.Tensor], **kwargs) -> torch.Tensor:
-        """Return a single action for the current timestep (called at inference)."""
-        ...
-
-    def forward(self, batch: dict[str, torch.Tensor]) -> dict[str, torch.Tensor]:
-        """Compute the training loss.
-
-        `batch["action_is_pad"]` is a bool mask of shape (B, horizon) that marks
-        timesteps padded because the episode ended before `horizon` steps, you
-        can exclude those from your loss.
-        """
-        actions = batch[ACTION]
-        action_is_pad = batch.get("action_is_pad")
-        ...
-        return {"loss": ...}
 ```
 
 ## Step 4: Add Data Processors
 
-Create processor functions. 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).
+Create processor functions:
 
 ```python
 # processor_my_custom_policy.py
-from typing import Any
+from typing import Dict, Any
 import torch
 
-from lerobot.processor import PolicyAction, PolicyProcessorPipeline
-
 
 def make_my_custom_policy_pre_post_processors(
     config,
-    dataset_stats: dict[str, dict[str, torch.Tensor]] | None = None,
 ) -> tuple[
     PolicyProcessorPipeline[dict[str, Any], dict[str, Any]],
     PolicyProcessorPipeline[PolicyAction, PolicyAction],
 ]:
-    preprocessor = ...   # build your PolicyProcessorPipeline for inputs
-    postprocessor = ...  # build your PolicyProcessorPipeline for outputs
-    return preprocessor, postprocessor
-```
+    """Create preprocessing and postprocessing functions for your policy."""
+    pass  # Define your preprocessing and postprocessing logic here
 
-**Important - function naming:** LeRobot discovers your processor by name. The function **must** be called `make_{policy_name}_pre_post_processors` (matching the string you passed to `@PreTrainedConfig.register_subclass`).
+```
 
 ## Step 5: Package Initialization
 

docs/source/cameras.mdx

@@ -1,22 +1,12 @@
 # Cameras
 
-LeRobot offers multiple options for video capture:
+LeRobot offers multiple options for video capture, including phone cameras, built-in laptop cameras, external webcams, and Intel RealSense cameras. To efficiently record frames from most cameras, you can use either the `OpenCVCamera` or `RealSenseCamera` class. For additional compatibility details on the `OpenCVCamera` class, refer to the [Video I/O with OpenCV Overview](https://docs.opencv.org/4.x/d0/da7/videoio_overview.html).
 
-| Class             | Supported Cameras                   |
-| ----------------- | ----------------------------------- |
-| `OpenCVCamera`    | Phone, built-in laptop, USB webcams |
-| `ZMQCamera`       | Network-connected cameras           |
-| `RealSenseCamera` | Intel RealSense (with depth)        |
-| `Reachy2Camera`   | Reachy 2 robot cameras              |
+### Finding your camera
 
-> [!TIP]
-> For `OpenCVCamera` compatibility details, see the [Video I/O with OpenCV Overview](https://docs.opencv.org/4.x/d0/da7/videoio_overview.html).
+To instantiate a camera, you need a camera identifier. This identifier might change if you reboot your computer or re-plug your camera, a behavior mostly dependant on your operating system.
 
-### Find your camera
-
-Every camera requires a unique identifier to be instantiated, allowing you to distinguish between multiple connected devices.
-
-`OpenCVCamera` and `RealSenseCamera` support auto-discovery. Run the command below to list available devices and their identifiers. Note that these identifiers may change after rebooting your computer or re-plugging the camera, depending on your operating system.
+To find the camera indices of the cameras plugged into your system, run the following script:
 
 ```bash
 lerobot-find-cameras opencv # or realsense for Intel Realsense cameras
@@ -24,7 +14,7 @@ lerobot-find-cameras opencv # or realsense for Intel Realsense cameras
 
 The output will look something like this if you have two cameras connected:
 
-```bash
+```
 --- Detected Cameras ---
 Camera #0:
   Name: OpenCV Camera @ 0
@@ -43,37 +33,13 @@ Camera #0:
 > [!WARNING]
 > When using Intel RealSense cameras in `macOS`, you could get this [error](https://github.com/IntelRealSense/librealsense/issues/12307): `Error finding RealSense cameras: failed to set power state`, this can be solved by running the same command with `sudo` permissions. Note that using RealSense cameras in `macOS` is unstable.
 
-`ZMQCamera` and `Reachy2Camera` do not support auto-discovery. They must be configured manually by providing their network address and port or robot SDK settings.
+## Use Cameras
 
-## Use cameras
+Below are two examples, demonstrating how to work with the API.
 
-### Frame access modes
-
-All camera classes implement three access modes for capturing frames:
-
-| Method                    | Behavior                                                                                                                                                   | Blocks?        | Best For                                 |
-| ------------------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------- | ---------------------------------------- |
-| `read()`                  | Waits for the camera hardware to return a frame. May block for a long time depending on the camera and SDK.                                                | Yes            | Simple scripts, sequential capture       |
-| `async_read(timeout_ms)`  | Returns the latest unconsumed frame from background thread. Blocks only if buffer is empty, up to `timeout_ms`. Raises `TimeoutError` if no frame arrives. | With a timeout | Control loops synchronized to camera FPS |
-| `read_latest(max_age_ms)` | Peeks at the most recent frame in buffer (may be stale). Raises `TimeoutError` if frame is older than `max_age_ms`.                                        | No             | UI visualization, logging, monitoring    |
-
-### Usage examples
-
-The following examples show how to use the camera API to configure and capture frames from different camera types.
-
-- **Blocking and non-blocking frame capture** using an OpenCV-based camera
+- **Asynchronous frame capture** using an OpenCV-based camera
 - **Color and depth capture** using an Intel RealSense camera
 
-> [!WARNING]
-> Failing to cleanly disconnect cameras can cause resource leaks. Use the context manager protocol to ensure automatic cleanup:
->
-> ```python
-> with OpenCVCamera(config) as camera:
->     ...
-> ```
->
-> You can also call `connect()` and `disconnect()` manually, but always use a `finally` block for the latter.
-
 <hfoptions id="shell_restart">
 <hfoption id="Open CV Camera">
 
@@ -94,30 +60,16 @@ config = OpenCVCameraConfig(
 )
 
 # Instantiate and connect an `OpenCVCamera`, performing a warm-up read (default).
-with OpenCVCamera(config) as camera:
-
-    # Read a frame synchronously — blocks until hardware delivers a new frame
-    frame = camera.read()
-    print(f"read() call returned frame with shape:", frame.shape)
-
-    # Read a frame asynchronously with a timeout — returns the latest unconsumed frame or waits up to timeout_ms for a new one
-    try:
-        for i in range(10):
-            frame = camera.async_read(timeout_ms=200)
-            print(f"async_read call returned frame {i} with shape:", frame.shape)
-    except TimeoutError as e:
-        print(f"No frame received within timeout: {e}")
-
-    # Instantly return a frame - returns the most recent frame captured by the camera
-    try:
-        initial_frame = camera.read_latest(max_age_ms=1000)
-        for i in range(10):
-            frame = camera.read_latest(max_age_ms=1000)
-            print(f"read_latest call returned frame {i} with shape:", frame.shape)
-            print(f"Was a new frame received by the camera? {not (initial_frame == frame).any()}")
-    except TimeoutError as e:
-        print(f"Frame too old: {e}")
+camera = OpenCVCamera(config)
+camera.connect()
 
+# Read frames asynchronously in a loop via `async_read(timeout_ms)`
+try:
+    for i in range(10):
+        frame = camera.async_read(timeout_ms=200)
+        print(f"Async frame {i} shape:", frame.shape)
+finally:
+    camera.disconnect()
 ```
 <!-- prettier-ignore-end -->
 
@@ -159,10 +111,10 @@ finally:
 </hfoption>
 </hfoptions>
 
-## Use your phone's camera
+## Use your phone
 
 <hfoptions id="use phone">
-<hfoption id="iPhone & macOS">
+<hfoption id="Mac">
 
 To use your iPhone as a camera on macOS, enable the Continuity Camera feature:
 
@@ -172,49 +124,83 @@ To use your iPhone as a camera on macOS, enable the Continuity Camera feature:
 
 For more details, visit [Apple support](https://support.apple.com/en-gb/guide/mac-help/mchl77879b8a/mac).
 
+Your iPhone should be detected automatically when running the camera setup script in the next section.
+
 </hfoption>
-<hfoption id="OBS virtual camera">
+<hfoption id="Linux">
 
-If you want to use your phone as a camera using OBS, follow these steps to set up a virtual camera.
+If you want to use your phone as a camera on Linux, follow these steps to set up a virtual camera
 
-1. _(Linux only) Install `v4l2loopback-dkms` and `v4l-utils`_. These packages create virtual camera devices and verify their settings. Install with:
+1. _Install `v4l2loopback-dkms` and `v4l-utils`_. Those packages are required to create virtual camera devices (`v4l2loopback`) and verify their settings with the `v4l2-ctl` utility from `v4l-utils`. Install them using:
 
-```bash
+<!-- prettier-ignore-start -->
+```python
 sudo apt install v4l2loopback-dkms v4l-utils
 ```
+<!-- prettier-ignore-end -->
 
-2. _Install the [DroidCam app](https://droidcam.app) on your phone_. This app is available for both iOS and Android.
-3. _Download and install [OBS Studio](https://obsproject.com)_.
-4. _Download and install the [DroidCam OBS plugin](https://droidcam.app/obs)_.
-5. _Start OBS Studio_.
+2. _Install [DroidCam](https://droidcam.app) on your phone_. This app is available for both iOS and Android.
+3. _Install [OBS Studio](https://obsproject.com)_. This software will help you manage the camera feed. Install it using [Flatpak](https://flatpak.org):
 
-6. _Add your phone as a source_. Follow the instructions [here](https://droidcam.app/obs/usage). Be sure to set the resolution to `640x480` to avoid the watermarks.
-7. _Adjust resolution settings_. In OBS Studio, go to `File > Settings > Video` or `OBS > Preferences... > Video`. Change the `Base(Canvas) Resolution` and the `Output(Scaled) Resolution` to `640x480` by manually typing it.
+<!-- prettier-ignore-start -->
+```python
+flatpak install flathub com.obsproject.Studio
+```
+<!-- prettier-ignore-end -->
+
+4. _Install the DroidCam OBS plugin_. This plugin integrates DroidCam with OBS Studio. Install it with:
+
+<!-- prettier-ignore-start -->
+```python
+flatpak install flathub com.obsproject.Studio.Plugin.DroidCam
+```
+<!-- prettier-ignore-end -->
+
+5. _Start OBS Studio_. Launch with:
+
+<!-- prettier-ignore-start -->
+```python
+flatpak run com.obsproject.Studio
+```
+<!-- prettier-ignore-end -->
+
+6. _Add your phone as a source_. Follow the instructions [here](https://droidcam.app/obs/usage). Be sure to set the resolution to `640x480`.
+7. _Adjust resolution settings_. In OBS Studio, go to `File > Settings > Video`. Change the `Base(Canvas) Resolution` and the `Output(Scaled) Resolution` to `640x480` by manually typing it in.
 8. _Start virtual camera_. In OBS Studio, follow the instructions [here](https://obsproject.com/kb/virtual-camera-guide).
-9. _Verify the virtual camera setup and resolution_.
-   - **Linux**: Use `v4l2-ctl` to list devices and check resolution:
-     ```bash
-     v4l2-ctl --list-devices  # find VirtualCam and note its /dev/videoX path
-     v4l2-ctl -d /dev/videoX --get-fmt-video  # replace with your VirtualCam path
-     ```
-     You should see `VirtualCam` listed and resolution `640x480`.
-   - **macOS**: Open Photo Booth or FaceTime and select "OBS Virtual Camera" as the input.
-   - **Windows**: The native Camera app doesn't support virtual cameras. Use a video conferencing app (Zoom, Teams) or run `lerobot-find-cameras opencv` directly to verify.
+9. _Verify the virtual camera setup_. Use `v4l2-ctl` to list the devices:
 
-<details>
-<summary><strong>Troubleshooting</strong></summary>
+<!-- prettier-ignore-start -->
+```python
+v4l2-ctl --list-devices
+```
+<!-- prettier-ignore-end -->
 
-> The virtual camera resolution is incorrect.
+You should see an entry like:
 
-Delete the virtual camera source and recreate it. The resolution cannot be changed after creation.
+```
+VirtualCam (platform:v4l2loopback-000):
+/dev/video1
+```
 
-> Error reading frame in background thread for OpenCVCamera(X): OpenCVCamera(X) frame width=640 or height=480 do not match configured width=1920 or height=1080.
+10. _Check the camera resolution_. Use `v4l2-ctl` to ensure that the virtual camera output resolution is `640x480`. Change `/dev/video1` to the port of your virtual camera from the output of `v4l2-ctl --list-devices`.
 
-This error is caused by OBS Virtual Camera advertising a `1920x1080` resolution despite rescaling. The only fix for now is to comment out the width and height check in `_postprocess_image()`.
+<!-- prettier-ignore-start -->
+```python
+v4l2-ctl -d /dev/video1 --get-fmt-video
+```
+<!-- prettier-ignore-end -->
 
-</details>
+You should see an entry like:
+
+```
+>>> Format Video Capture:
+>>>	Width/Height      : 640/480
+>>>	Pixel Format      : 'YUYV' (YUYV 4:2:2)
+```
+
+Troubleshooting: If the resolution is not correct you will have to delete the Virtual Camera port and try again as it cannot be changed.
+
+If everything is set up correctly, you can proceed with the rest of the tutorial.
 
 </hfoption>
 </hfoptions>
-
-If everything is set up correctly, your phone will appear as a standard OpenCV camera and can be used with `OpenCVCamera`.

docs/source/damiao.mdx

@@ -1,165 +0,0 @@
-# Damiao Motors and CAN Bus
-
-This guide covers setup and usage of Damiao motors with LeRobot via CAN bus communication.
-
-Currently, only Linux is supported, as the OpenArms CAN adapter only has drivers for Linux.
-
-## Linux CAN Setup
-
-Before using Damiao motors, you need to set up the CAN interface on your Linux system.
-
-### Install CAN Utilities
-
-```bash
-sudo apt-get install can-utils
-```
-
-### Configure CAN Interface (Manual)
-
-For standard CAN FD (recommended for OpenArms):
-
-```bash
-sudo ip link set can0 down
-sudo ip link set can0 type can bitrate 1000000 dbitrate 5000000 fd on
-sudo ip link set can0 up
-```
-
-For standard CAN (without FD):
-
-```bash
-sudo ip link set can0 down
-sudo ip link set can0 type can bitrate 1000000
-sudo ip link set can0 up
-```
-
-### Configure CAN Interface (Using LeRobot)
-
-LeRobot provides a utility script to setup and test CAN interfaces:
-
-```bash
-# Setup multiple interfaces (e.g., OpenArms Followers with 2 CAN buses)
-lerobot-setup-can --mode=setup --interfaces=can0,can1
-```
-
-## Debugging CAN Communication
-
-Use the built-in debug tools to test motor communication:
-
-```bash
-# Test motors on all interfaces
-lerobot-setup-can --mode=test --interfaces=can0,can1
-
-# Run speed/latency test
-lerobot-setup-can --mode=speed --interfaces=can0
-```
-
-The test mode will scan for motors (IDs 0x01-0x08) and report which ones respond. Example output:
-
-```
-can0: UP (CAN FD)
-  Motor 0x01 (joint_1): ✓ FOUND
-    → Response 0x11 [FD]: 00112233...
-  Motor 0x02 (joint_2): ✓ FOUND
-  Motor 0x03 (joint_3): ✗ No response
-  ...
-  Summary: 2/8 motors found
-```
-
-## Usage
-
-### Basic Setup
-
-```python
-from lerobot.motors import Motor
-from lerobot.motors.damiao import DamiaoMotorsBus
-
-# Define your motors with send/receive CAN IDs
-motors = {
-    "joint_1": Motor(id=0x01, motor_type_str="dm8009", recv_id=0x11),
-    "joint_2": Motor(id=0x02, motor_type_str="dm4340", recv_id=0x12),
-    "joint_3": Motor(id=0x03, motor_type_str="dm4310", recv_id=0x13),
-}
-
-# Create the bus
-bus = DamiaoMotorsBus(
-    port="can0",  # Linux socketcan interface
-    motors=motors,
-)
-
-# Connect
-bus.connect()
-```
-
-### Reading Motor States
-
-```python
-# Read single motor position (degrees)
-position = bus.read("Present_Position", "joint_1")
-
-# Read from multiple motors
-positions = bus.sync_read("Present_Position")  # All motors
-positions = bus.sync_read("Present_Position", ["joint_1", "joint_2"])
-
-# Read all states at once (position, velocity, torque)
-states = bus.sync_read_all_states()
-# Returns: {'joint_1': {'position': 45.2, 'velocity': 1.3, 'torque': 0.5}, ...}
-```
-
-### Writing Motor Commands
-
-```python
-# Enable torque
-bus.enable_torque()
-
-# Set goal position (degrees)
-bus.write("Goal_Position", "joint_1", 45.0)
-
-# Set positions for multiple motors
-bus.sync_write("Goal_Position", {
-    "joint_1": 45.0,
-    "joint_2": -30.0,
-    "joint_3": 90.0,
-})
-
-# Disable torque
-bus.disable_torque()
-```
-
-## Configuration Options
-
-| Parameter      | Default   | Description                                                 |
-| -------------- | --------- | ----------------------------------------------------------- |
-| `port`         | -         | CAN interface (`can0`) or serial port (`/dev/cu.usbmodem*`) |
-| `use_can_fd`   | `True`    | Enable CAN FD for higher data rates                         |
-| `bitrate`      | `1000000` | Nominal bitrate (1 Mbps)                                    |
-| `data_bitrate` | `5000000` | CAN FD data bitrate (5 Mbps)                                |
-
-## Motor Configuration
-
-Each motor requires:
-
-- `id`: CAN ID for sending commands
-- `motor_type`: One of the supported motor types (e.g., `"dm8009"`, `"dm4340"`)
-- `recv_id`: CAN ID for receiving responses
-
-OpenArms default IDs follow the pattern: send ID `0x0N`, receive ID `0x1N` where N is the joint number.
-
-## Troubleshooting
-
-### No Response from Motors
-
-1. **Check power**
-2. **Verify CAN wiring**: Check CAN-H, CAN-L, and GND connections
-3. **Check motor IDs**: Use Damiao Debugging Tools to verify/configure IDs
-4. **Test CAN interface**: Run `candump can0` to see if messages are being received
-5. **Run diagnostics**: `lerobot-setup-can --mode=test --interfaces=can0`
-
-### Motor Timeout Parameter
-
-If motors were configured with timeout=0, they won't respond to commands. Use Damiao Debugging Tools to set a non-zero timeout value.
-
-### Verify CAN FD Status
-
-```bash
-ip -d link show can0 | grep fd
-```

docs/source/dataset_subtask.mdx

@@ -1,278 +0,0 @@
-# Using Subtasks in LeRobot Datasets
-
-Subtask support in robotics datasets has proven effective in improving robot reasoning and understanding. Subtasks are particularly useful for:
-
-- **Hierarchical policies**: Building policies that include subtask predictions to visualize robot reasoning in real time
-- **Reward modeling**: Helping reward models understand task progression (e.g., SARM-style stage-aware reward models)
-- **Task decomposition**: Breaking down complex manipulation tasks into atomic, interpretable steps
-
-LeRobotDataset now supports subtasks as part of its dataset structure, alongside tasks.
-
-## What are Subtasks?
-
-While a **task** describes the overall goal (e.g., "Pick up the apple and place it in the basket"), **subtasks** break down the execution into finer-grained steps:
-
-1. "Approach the apple"
-2. "Grasp the apple"
-3. "Lift the apple"
-4. "Move to basket"
-5. "Release the apple"
-
-Each frame in the dataset can be annotated with its corresponding subtask, enabling models to learn and predict these intermediate stages.
-
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/subtask-asset.png"
-  alt="An overview of subtask annotation showing how frames are labeled with intermediate subtask stages"
-  width="80%"
-/>
-
-<p>
-  <em>Figure: Overview of subtask annotation.</em>
-</p>
-
-**Reference:** _Subtask-learning based for robot self-assembly in flexible collaborative assembly in manufacturing_, Original Article, Published: 19 April 2022.
-
-## Dataset Structure
-
-Subtask information is stored in the dataset metadata:
-
-```
-my-dataset/
-├── data/
-│   └── ...
-├── meta/
-│   ├── info.json
-│   ├── stats.json
-│   ├── tasks.parquet
-│   ├── subtasks.parquet      # Subtask index → subtask string mapping
-│   └── episodes/
-│       └── ...
-└── videos/
-    └── ...
-```
-
-### Subtasks Parquet File
-
-The `meta/subtasks.parquet` file maps subtask indices to their natural language descriptions:
-
-| subtask_index | subtask (index column) |
-| ------------- | ---------------------- |
-| 0             | "Approach the apple"   |
-| 1             | "Grasp the apple"      |
-| 2             | "Lift the apple"       |
-| ...           | ...                    |
-
-### Frame-Level Annotations
-
-Each frame in the dataset can include a `subtask_index` field that references the subtasks parquet file:
-
-```python
-# Example frame data in the parquet file
-{
-    "index": 42,
-    "timestamp": 1.4,
-    "episode_index": 0,
-    "task_index": 0,
-    "subtask_index": 2,  # References "Lift the apple"
-    "observation.state": [...],
-    "action": [...],
-}
-```
-
-## Annotating Datasets with Subtasks
-
-We provide a HuggingFace Space for easily annotating any LeRobotDataset with subtasks:
-
-**[https://huggingface.co/spaces/lerobot/annotate](https://huggingface.co/spaces/lerobot/annotate)**
-
-After completing your annotation:
-
-1. Click "Push to Hub" to upload your annotated dataset
-2. You can also run the annotation space locally by following the instructions at [github.com/huggingface/lerobot-annotate](https://github.com/huggingface/lerobot-annotate)
-
-## Loading Datasets with Subtasks
-
-When you load a dataset with subtask annotations, the subtask information is automatically available:
-
-```python
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
-
-# Load a dataset with subtask annotations
-dataset = LeRobotDataset("jadechoghari/collect-fruit-annotated")
-
-# Access a sample
-sample = dataset[100]
-
-# The sample includes both task and subtask information
-print(sample["task"])        # "Collect the fruit"
-print(sample["subtask"])     # "Grasp the apple"
-print(sample["task_index"])  # tensor(0)
-print(sample["subtask_index"])  # tensor(2)
-```
-
-### Checking for Subtask Support
-
-You can check if a dataset has subtask annotations:
-
-```python
-# Check if subtasks are available
-has_subtasks = (
-    "subtask_index" in dataset.features
-    and dataset.meta.subtasks is not None
-)
-
-if has_subtasks:
-    print(f"Dataset has {len(dataset.meta.subtasks)} unique subtasks")
-    print("Subtasks:", list(dataset.meta.subtasks.index))
-```
-
-## Using Subtasks for Training
-
-### With the Tokenizer Processor
-
-The `TokenizerProcessor` automatically handles subtask tokenization for Vision-Language Action (VLA) models:
-
-```python
-from lerobot.processor.tokenizer_processor import TokenizerProcessor
-from lerobot.processor.pipeline import ProcessorPipeline
-
-# Create a tokenizer processor
-tokenizer_processor = TokenizerProcessor(
-    tokenizer_name_or_path="google/paligemma-3b-pt-224",
-    padding="max_length",
-    max_length=64,
-)
-
-# The processor will automatically tokenize subtasks if present in the batch
-# and add them to the observation under:
-# - "observation.subtask.tokens"
-# - "observation.subtask.attention_mask"
-```
-
-When subtasks are available in the batch, the tokenizer processor adds:
-
-- `observation.subtask.tokens`: Tokenized subtask text
-- `observation.subtask.attention_mask`: Attention mask for the subtask tokens
-
-### DataLoader with Subtasks
-
-```python
-import torch
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
-
-dataset = LeRobotDataset("jadechoghari/collect-fruit-annotated")
-
-dataloader = torch.utils.data.DataLoader(
-    dataset,
-    batch_size=16,
-    shuffle=True,
-)
-
-for batch in dataloader:
-    # Access subtask information in the batch
-    subtasks = batch["subtask"]  # List of subtask strings
-    subtask_indices = batch["subtask_index"]  # Tensor of subtask indices
-
-    # Use for training hierarchical policies or reward models
-    print(f"Batch subtasks: {set(subtasks)}")
-```
-
-## Example Datasets with Subtask Annotations
-
-Try loading a dataset with subtask annotations:
-
-```python
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
-
-# Example dataset with subtask annotations
-dataset = LeRobotDataset("jadechoghari/collect-fruit-annotated")
-
-# Explore the subtasks
-print("Available subtasks:")
-for subtask_name in dataset.meta.subtasks.index:
-    print(f"  - {subtask_name}")
-
-# Get subtask distribution
-subtask_counts = {}
-for i in range(len(dataset)):
-    sample = dataset[i]
-    subtask = sample["subtask"]
-    subtask_counts[subtask] = subtask_counts.get(subtask, 0) + 1
-
-print("\nSubtask distribution:")
-for subtask, count in sorted(subtask_counts.items(), key=lambda x: -x[1]):
-    print(f"  {subtask}: {count} frames")
-```
-
-## Use Cases
-
-### 1. Hierarchical Policy Training
-
-Train policies that predict both actions and current subtask:
-
-```python
-class HierarchicalPolicy(nn.Module):
-    def __init__(self, num_subtasks):
-        super().__init__()
-        self.action_head = nn.Linear(hidden_dim, action_dim)
-        self.subtask_head = nn.Linear(hidden_dim, num_subtasks)
-
-    def forward(self, observations):
-        features = self.encoder(observations)
-        actions = self.action_head(features)
-        subtask_logits = self.subtask_head(features)
-        return actions, subtask_logits
-```
-
-### 2. Stage-Aware Reward Modeling (SARM)
-
-Build reward models that understand task progression:
-
-```python
-# SARM predicts:
-# - Stage: Which subtask is being executed (discrete)
-# - Progress: How far along the subtask (continuous 0-1)
-
-class SARMRewardModel(nn.Module):
-    def forward(self, observations):
-        features = self.encoder(observations)
-        stage_logits = self.stage_classifier(features)
-        progress = self.progress_regressor(features)
-        return stage_logits, progress
-```
-
-### 3. Progress Visualization
-
-Monitor robot execution by tracking subtask progression:
-
-```python
-def visualize_execution(model, observations):
-    for t, obs in enumerate(observations):
-        action, subtask_logits = model(obs)
-        predicted_subtask = subtask_names[subtask_logits.argmax()]
-        print(f"t={t}: Executing '{predicted_subtask}'")
-```
-
-## API Reference
-
-### LeRobotDataset Properties
-
-| Property                    | Type                   | Description                                |
-| --------------------------- | ---------------------- | ------------------------------------------ |
-| `meta.subtasks`             | `pd.DataFrame \| None` | DataFrame mapping subtask names to indices |
-| `features["subtask_index"]` | `dict`                 | Feature spec for subtask_index if present  |
-
-### Sample Keys
-
-When subtasks are available, each sample includes:
-
-| Key             | Type           | Description                          |
-| --------------- | -------------- | ------------------------------------ |
-| `subtask_index` | `torch.Tensor` | Integer index of the current subtask |
-| `subtask`       | `str`          | Natural language subtask description |
-
-## Related Resources
-
-- [SARM Paper](https://arxiv.org/pdf/2509.25358) - Stage-Aware Reward Modeling for Long Horizon Robot Manipulation
-- [LeRobot Annotate Space](https://huggingface.co/spaces/lerobot/annotate) - Interactive annotation tool
-- [LeRobotDataset v3.0](./lerobot-dataset-v3) - Dataset format documentation

docs/source/earthrover_mini_plus.mdx

@@ -1,11 +1,5 @@
 # EarthRover Mini Plus
 
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/Earth_Rover_Mini_5_240c9adc-4f9e-44b7-982f-5d1dc24af1d8.png.webp"
-  alt="EarthRover Mini Plus"
-  width="70%"
-/>
-
 The EarthRover Mini Plus is a fully open source mobile robot that connects through the cloud using the Frodobots SDK. This lets you control the robot and record datasets for training AI models.
 
 ## What You Need
@@ -13,47 +7,28 @@ The EarthRover Mini Plus is a fully open source mobile robot that connects throu
 ### Hardware
 
 - EarthRover Mini robot
-- Computer with Python 3.12 or newer
+- Computer with Python 3.10 or newer
 - Internet connection
 
 ### Setting Up the Frodobots SDK
 
-The robot needs the [Frodobots SDK](https://github.com/frodobots-org/earth-rovers-sdk) running on your computer. Here's how:
+The robot needs the [Frodobots SDK](https://github.com/Frodobots/earth-rovers-sdk) running on your computer. Here's how:
 
 1. Download and install the SDK:
 
 ```bash
-git clone https://github.com/frodobots-org/earth-rovers-sdk.git
+git clone https://github.com/Frodobots/earth-rovers-sdk.git
 cd earth-rovers-sdk
 pip install -r requirements.txt
 ```
 
-2. Save Credentials:
-
-Write your .env variables with the SDK API key and bot name provided by the Frodobots team.
-
-```bash
-SDK_API_TOKEN=your_sdk_api_token_here
-BOT_SLUG=your_bot_slug_here
-CHROME_EXECUTABLE_PATH=/path/to/chrome_or_chromium
-# Default value is MAP_ZOOM_LEVEL=18 https://wiki.openstreetmap.org/wiki/Zoom_levels
-MAP_ZOOM_LEVEL=18
-MISSION_SLUG=your_mission_slug_here
-# Image quality between 0.1 and 1.0 (default: 0.8)
-# Recommended: 0.8 for better performance
-IMAGE_QUALITY=0.8
-# Image format: jpeg, png or webp (default: png)
-# Recommended: jpeg for better performance and lower bandwidth usage
-IMAGE_FORMAT=jpeg
-```
-
-3. Start the SDK:
+2. Start the SDK:
 
 ```bash
 hypercorn main:app --reload
 ```
 
-4. Open your web browser and go to `http://localhost:8000`, then click "Join"
+3. Open your web browser and go to `http://localhost:8000`, then click "Join"
 
 The SDK gives you:
 
@@ -170,13 +145,13 @@ Once you can drive the robot well, you can start recording data to train AI mode
 We use Hugging Face to store your data online. First, log in with your token from [Hugging Face settings](https://huggingface.co/settings/tokens):
 
 ```bash
-hf auth login --token ${HUGGINGFACE_TOKEN} --add-to-git-credential
+huggingface-cli login --token ${HUGGINGFACE_TOKEN} --add-to-git-credential
 ```
 
 Store your Hugging Face username:
 
 ```bash
-HF_USER=$(hf auth whoami | awk -F': *' 'NR==1 {print $2}')
+HF_USER=$(huggingface-cli whoami | head -n 1)
 echo $HF_USER
 ```
 
@@ -185,16 +160,13 @@ echo $HF_USER
 Use the standard recording command:
 
 ```bash
-lerobot-record \
+python src/lerobot/scripts/lerobot_record.py \
     --robot.type=earthrover_mini_plus \
     --teleop.type=keyboard_rover \
     --dataset.repo_id=your_username/dataset_name \
     --dataset.num_episodes=2 \
     --dataset.fps=10 \
     --dataset.single_task="Navigate around obstacles" \
-    --dataset.streaming_encoding=true \
-    --dataset.encoder_threads=2 \
-    # --dataset.vcodec=auto \
     --display_data=true
 ```
 
@@ -204,26 +176,22 @@ Replace `your_username/dataset_name` with your Hugging Face username and a name
 
 Your dataset includes:
 
-**Your Actions (2 features)**:
+**Your Actions (2 things)**:
 
-- `linear_velocity`: How much you moved forward/backward
-- `angular_velocity`: How much you turned left/right
+- How much you moved forward/backward
+- How much you turned left/right
 
-**Robot Observations (24 features)**:
+**Robot Observations (12 things)**:
 
 - Front camera video
 - Rear camera video
 - Current speed
 - Battery level
-- Orientation
-- GPS (latitude, longitude, signal strength)
+- Which way the robot is facing
+- GPS location (latitude, longitude, signal strength)
 - Network signal strength
 - Vibration level
-- Lamp state (on/off)
-- Accelerometer (x, y, z)
-- Gyroscope (x, y, z)
-- Magnetometer (x, y, z)
-- Wheel RPMs (4 wheels)
+- Lamp status (on/off)
 
 ### Where Your Data Goes
 

docs/source/envhub.mdx

@@ -2,32 +2,14 @@
 
 The **EnvHub** feature allows you to load simulation environments directly from the Hugging Face Hub with a single line of code. This unlocks a powerful new model for collaboration: instead of environments being locked away inside monolithic libraries, anyone can publish custom environments and share them with the community.
 
-## What is EnvHub?
+## Overview
 
-EnvHub lets you create custom robotics simulation environments with your own robot models and scenarios, and make them easily usable by anyone through the LeRobot framework.
+With EnvHub, you can:
 
-EnvHub packages are stored on the Hugging Face Hub, and can be seamlessly pulled and used in your AI robotics projects through LeRobot with a single line of code.
-
-Thanks to EnvHub, you can:
-
-1. **Create and publish environments** to the Hugging Face Hub as Git repositories, and distribute complex physics simulations without packaging hassles
-2. **Load environments** dynamically, without installing them as packages
-3. **Version and track** environment changes using Git semantics
-4. **Discover** new simulation tasks shared by the community
-
-This design means you can go from discovering an interesting environment on the Hub to running experiments in seconds, or create your own custom robot and environment without worrying about dependency conflicts or complex installation procedures.
-
-When you create an EnvHub package, you can build anything you want inside it and use any simulation tool you like: this is your own space to play with. The only requirement is that the package contains an `env.py` file that defines the environment and allows LeRobot to load and use your EnvHub package.
-
-This `env.py` file needs to expose a small API so LeRobot can load and run it. In particular, you must provide a `make_env(n_envs: int = 1, use_async_envs: bool = False)` or `make_env(n_envs: int = 1, use_async_envs: bool = False, cfg: EnvConfig)` function, which is the main entry point for LeRobot. It should return one of:
-
-- A `gym.vector.VectorEnv` (most common)
-- A single `gym.Env` (will be automatically wrapped)
-- A dict mapping `{suite_name: {task_id: VectorEnv}}` (for multi-task benchmarks)
-
-You can also pass an `EnvConfig` object to `make_env` to configure the environment (e.g. the number of environments, task, camera name, initial states, control mode, episode length, etc.).
-
-Finally, your environment must implement the standard `gym.vector.VectorEnv` interface so it works with LeRobot, including methods like `reset` and `step`.
+- Load environments from the Hub instantly
+- Share your custom simulation tasks with the community
+- Version control your environments using Git
+- Distribute complex physics simulations without packaging hassles
 
 ## Quick Start
 
@@ -47,6 +29,17 @@ env = make_env("lerobot/cartpole-env", trust_remote_code=True)
   hash for reproducibility and security.
 </Tip>
 
+## What is EnvHub?
+
+EnvHub is a framework that allows researchers and developers to:
+
+1. **Publish environments** to the Hugging Face Hub as Git repositories
+2. **Load environments** dynamically without installing them as packages
+3. **Version and track** environment changes using Git semantics
+4. **Discover** new simulation tasks shared by the community
+
+This design means you can go from discovering an interesting environment on the Hub to running experiments in seconds, without worrying about dependency conflicts or complex installation procedures.
+
 ## Repository Structure
 
 To make your environment loadable from the Hub, your repository must contain at minimum:
@@ -155,10 +148,10 @@ Upload your repository to Hugging Face:
 pip install huggingface_hub
 
 # Login to Hugging Face
-hf auth login
+huggingface-cli login
 
 # Create a new repository
-hf repo create my-org/my-custom-env
+huggingface-cli repo create my-custom-env --type space --org my-org
 
 # Initialize git and push
 git init

docs/source/envhub_isaaclab_arena.mdx

@@ -1,510 +0,0 @@
-# NVIDIA IsaacLab Arena & LeRobot
-
-LeRobot EnvHub now supports **GPU-accelerated simulation** with IsaacLab Arena for policy evaluation at scale.
-Train and evaluate imitation learning policies with high-fidelity simulation — all integrated into the LeRobot ecosystem.
-
-<img
-  src="https://huggingface.co/nvidia/isaaclab-arena-envs/resolve/main/assets/Gr1OpenMicrowaveEnvironment.png"
-  alt="IsaacLab Arena - GR1 Microwave Environment"
-  style={{ maxWidth: "100%", borderRadius: "8px", marginBottom: "1rem" }}
-/>
-
-[IsaacLab Arena](https://github.com/isaac-sim/IsaacLab-Arena) integrates with NVIDIA IsaacLab to provide:
-
-- 🤖 **Humanoid embodiments**: GR1, G1, Galileo with various configurations
-- 🎯 **Manipulation & loco-manipulation tasks**: Door opening, pick-and-place, button pressing, and more
-- ⚡ **GPU-accelerated rollouts**: Parallel environment execution on NVIDIA GPUs
-- 🖼️ **RTX Rendering**: Evaluate vision-based policies with realistic rendering, reflections and refractions
-- 📦 **LeRobot-compatible datasets**: Ready for training with GR00T N1x, PI0, SmolVLA, ACT, and Diffusion policies
-- 🔄 **EnvHub integration**: Load environments from HuggingFace EnvHub with one line
-
-## Installation
-
-### Prerequisites
-
-Hardware requirements are shared with Isaac Sim, and are detailed in [Isaac Sim Requirements](https://docs.isaacsim.omniverse.nvidia.com/5.1.0/installation/requirements.html).
-
-- NVIDIA GPU with CUDA support
-- NVIDIA driver compatible with IsaacSim 5.1.0
-- Linux (Ubuntu 22.04 / 24.04)
-
-### Setup
-
-```bash
-# 1. Create conda environment
-conda create -y -n lerobot-arena python=3.11
-conda activate lerobot-arena
-conda install -y -c conda-forge ffmpeg=7.1.1
-
-# 2. Install Isaac Sim 5.1.0
-pip install "isaacsim[all,extscache]==5.1.0" --extra-index-url https://pypi.nvidia.com
-
-# Accept NVIDIA EULA (required)
-export ACCEPT_EULA=Y
-export PRIVACY_CONSENT=Y
-
-# 3. Install IsaacLab 2.3.0
-git clone https://github.com/isaac-sim/IsaacLab.git
-cd IsaacLab
-git checkout v2.3.0
-./isaaclab.sh -i
-cd ..
-
-# 4. Install IsaacLab Arena
-git clone https://github.com/isaac-sim/IsaacLab-Arena.git
-cd IsaacLab-Arena
-git checkout release/0.1.1
-pip install -e .
-cd ..
-
-
-# 5. Install LeRobot
-git clone https://github.com/huggingface/lerobot.git
-cd lerobot
-pip install -e .
-cd ..
-
-
-# 6. Install additional dependencies
-pip install onnxruntime==1.23.2 lightwheel-sdk==1.0.1 vuer[all]==0.0.70 qpsolvers==4.8.1
-pip install numpy==1.26.0 # Isaac Sim 5.1 depends on numpy==1.26.0, this will be fixed in next release
-```
-
-## Evaluating Policies
-
-### Pre-trained Policies
-
-The following trained policies are available:
-
-| Policy                      | Architecture | Task          | Link                                                                     |
-| :-------------------------- | :----------- | :------------ | :----------------------------------------------------------------------- |
-| pi05-arena-gr1-microwave    | PI0.5        | GR1 Microwave | [HuggingFace](https://huggingface.co/nvidia/pi05-arena-gr1-microwave)    |
-| smolvla-arena-gr1-microwave | SmolVLA      | GR1 Microwave | [HuggingFace](https://huggingface.co/nvidia/smolvla-arena-gr1-microwave) |
-
-### Evaluate SmolVLA
-
-```bash
-pip install -e ".[smolvla]"
-pip install numpy==1.26.0 # revert numpy to version 1.26
-```
-
-```bash
-lerobot-eval \
-    --policy.path=nvidia/smolvla-arena-gr1-microwave \
-    --env.type=isaaclab_arena \
-    --env.hub_path=nvidia/isaaclab-arena-envs \
-    --rename_map='{"observation.images.robot_pov_cam_rgb": "observation.images.robot_pov_cam"}' \
-    --policy.device=cuda \
-    --env.environment=gr1_microwave \
-    --env.embodiment=gr1_pink \
-    --env.object=mustard_bottle \
-    --env.headless=false \
-    --env.enable_cameras=true \
-    --env.video=true \
-    --env.video_length=10 \
-    --env.video_interval=15 \
-    --env.state_keys=robot_joint_pos \
-    --env.camera_keys=robot_pov_cam_rgb \
-    --trust_remote_code=True \
-    --eval.batch_size=1
-```
-
-### Evaluate PI0.5
-
-```bash
-pip install -e ".[pi]"
-pip install numpy==1.26.0 # revert numpy to version 1.26
-```
-
-<Tip>PI0.5 requires disabling torch compile for evaluation:</Tip>
-
-```bash
-TORCH_COMPILE_DISABLE=1 TORCHINDUCTOR_DISABLE=1 lerobot-eval \
-    --policy.path=nvidia/pi05-arena-gr1-microwave \
-    --env.type=isaaclab_arena \
-    --env.hub_path=nvidia/isaaclab-arena-envs \
-    --rename_map='{"observation.images.robot_pov_cam_rgb": "observation.images.robot_pov_cam"}' \
-    --policy.device=cuda \
-    --env.environment=gr1_microwave \
-    --env.embodiment=gr1_pink \
-    --env.object=mustard_bottle \
-    --env.headless=false \
-    --env.enable_cameras=true \
-    --env.video=true \
-    --env.video_length=15 \
-    --env.video_interval=15 \
-    --env.state_keys=robot_joint_pos \
-    --env.camera_keys=robot_pov_cam_rgb \
-    --trust_remote_code=True \
-    --eval.batch_size=1
-```
-
-<Tip>
-  To change the number of parallel environments, use the ```--eval.batch_size```
-  flag.
-</Tip>
-
-### What to Expect
-
-During evaluation, you will see a progress bar showing the running success rate:
-
-```
-Stepping through eval batches:   8%|██████▍    | 4/50 [00:45<08:06, 10.58s/it, running_success_rate=25.0%]
-```
-
-### Video Recording
-
-To enable video recording during evaluation, add the following flags to your command:
-
-```bash
---env.video=true \
---env.video_length=15 \
---env.video_interval=15
-```
-
-For more details on video recording, see the [IsaacLab Recording Documentation](https://isaac-sim.github.io/IsaacLab/main/source/how-to/record_video.html).
-
-<Tip>
-When running headless with `--env.headless=true`, you must also enable cameras explicitly for camera enabled environments:
-
-```bash
---env.headless=true --env.enable_cameras=true
-```
-
-</Tip>
-
-### Output Directory
-
-Evaluation videos are saved to the output directory with the following structure:
-
-```
-outputs/eval/<date>/<timestamp>_<env>_<policy>/videos/<task>_<env_id>/eval_episode_<n>.mp4
-```
-
-For example:
-
-```
-outputs/eval/2026-01-02/14-38-01_isaaclab_arena_smolvla/videos/gr1_microwave_0/eval_episode_0.mp4
-```
-
-## Training Policies
-
-To learn more about training policies with LeRobot, please refer to the training documentation:
-
-- [SmolVLA](./smolvla)
-- [Pi0.5](./pi05)
-- [GR00T N1.5](./groot)
-
-Sample IsaacLab Arena datasets are available on HuggingFace Hub for experimentation:
-
-| Dataset                                                                                                   | Description                | Frames |
-| :-------------------------------------------------------------------------------------------------------- | :------------------------- | :----- |
-| [Arena-GR1-Manipulation-Task](https://huggingface.co/datasets/nvidia/Arena-GR1-Manipulation-Task-v3)      | GR1 microwave manipulation | ~4K    |
-| [Arena-G1-Loco-Manipulation-Task](https://huggingface.co/datasets/nvidia/Arena-G1-Loco-Manipulation-Task) | G1 loco-manipulation       | ~4K    |
-
-## Environment Configuration
-
-### Full Configuration Options
-
-```python
-from lerobot.envs.configs import IsaaclabArenaEnv
-
-config = IsaaclabArenaEnv(
-    # Environment selection
-    environment="gr1_microwave",      # Task environment
-    embodiment="gr1_pink",            # Robot embodiment
-    object="power_drill",             # Object to manipulate
-
-    # Simulation settings
-    episode_length=300,               # Max steps per episode
-    headless=True,                    # Run without GUI
-    device="cuda:0",                  # GPU device
-    seed=42,                          # Random seed
-
-    # Observation configuration
-    state_keys="robot_joint_pos",     # State observation keys (comma-separated)
-    camera_keys="robot_pov_cam_rgb",  # Camera observation keys (comma-separated)
-    state_dim=54,                     # Expected state dimension
-    action_dim=36,                    # Expected action dimension
-    camera_height=512,                # Camera image height
-    camera_width=512,                 # Camera image width
-    enable_cameras=True,              # Enable camera observations
-
-    # Video recording
-    video=False,                      # Enable video recording
-    video_length=100,                 # Frames per video
-    video_interval=200,               # Steps between recordings
-
-    # Advanced
-    mimic=False,                      # Enable mimic mode
-    teleop_device=None,               # Teleoperation device
-    disable_fabric=False,             # Disable fabric optimization
-    enable_pinocchio=True,            # Enable Pinocchio for IK
-)
-```
-
-### Using Environment Hub directly for advanced usage
-
-Create a file called `test_env_load_arena.py` or [download from the EnvHub](https://huggingface.co/nvidia/isaaclab-arena-envs/blob/main/tests/test_env_load_arena.py):
-
-```python
-import logging
-from dataclasses import asdict
-from pprint import pformat
-import torch
-import tqdm
-from lerobot.configs import parser
-from lerobot.configs.eval import EvalPipelineConfig
-
-
-@parser.wrap()
-def main(cfg: EvalPipelineConfig):
-    """Run random action rollout for IsaacLab Arena environment."""
-    logging.info(pformat(asdict(cfg)))
-
-    from lerobot.envs.factory import make_env
-
-    env_dict = make_env(
-        cfg.env,
-        n_envs=cfg.env.num_envs,
-        trust_remote_code=True,
-    )
-    env = next(iter(env_dict.values()))[0]
-    env.reset()
-    for _ in tqdm.tqdm(range(cfg.env.episode_length)):
-        with torch.inference_mode():
-            actions = env.action_space.sample()
-            obs, rewards, terminated, truncated, info = env.step(actions)
-            if terminated.any() or truncated.any():
-                obs, info = env.reset()
-    env.close()
-
-
-if __name__ == "__main__":
-    main()
-```
-
-Run with:
-
-```bash
-python test_env_load_arena.py \
-    --env.environment=g1_locomanip_pnp \
-    --env.embodiment=gr1_pink \
-    --env.object=cracker_box \
-    --env.num_envs=4 \
-    --env.enable_cameras=true \
-    --env.seed=1000 \
-    --env.video=true \
-    --env.video_length=10 \
-    --env.video_interval=15 \
-    --env.headless=false \
-    --env.hub_path=nvidia/isaaclab-arena-envs \
-    --env.type=isaaclab_arena
-```
-
-## Creating New Environments
-
-First create a new IsaacLab Arena environment by following the [IsaacLab Arena Documentation](https://isaac-sim.github.io/IsaacLab-Arena/release/0.1.1/index.html).
-
-Clone our EnvHub repo:
-
-```bash
-git clone https://huggingface.co/nvidia/isaaclab-arena-envs
-```
-
-Modify the `example_envs.yaml` file based on your new environment.
-[Upload](./envhub#step-3-upload-to-the-hub) your modified repo to HuggingFace EnvHub.
-
-<Tip>
-  Your IsaacLab Arena environment code must be locally available during
-  evaluation. Users can clone your environment repository separately, or you can
-  bundle the environment code and assets directly in your EnvHub repo.
-</Tip>
-
-Then, when evaluating, use your new environment:
-
-```bash
-lerobot-eval \
-    --env.hub_path=<your-env-hub-path>/isaaclab-arena-envs \
-    --env.environment=<your new environment> \
-    ...other flags...
-```
-
-We look forward to your contributions!
-
-## Troubleshooting
-
-### CUDA out of memory
-
-Reduce `batch_size` or use a GPU with more VRAM:
-
-```bash
---eval.batch_size=1
-```
-
-### EULA not accepted
-
-Set environment variables before running:
-
-```bash
-export ACCEPT_EULA=Y
-export PRIVACY_CONSENT=Y
-```
-
-### Video recording not working
-
-Enable cameras when running headless:
-
-```bash
---env.video=true --env.enable_cameras=true --env.headless=true
-```
-
-### Policy output dimension mismatch
-
-Ensure `action_dim` matches your policy:
-
-```bash
---env.action_dim=36
-```
-
-### libGLU.so.1 Errors during Isaac Sim initialization
-
-Ensure you have the following dependencies installed, this is likely to happen on headless machines.
-
-```bash
-sudo apt update && sudo apt install -y libglu1-mesa libxt6
-```
-
-## See Also
-
-- [EnvHub Documentation](./envhub.mdx) - General EnvHub usage
-- [IsaacLab Arena GitHub](https://github.com/isaac-sim/IsaacLab-Arena)
-- [IsaacLab Documentation](https://isaac-sim.github.io/IsaacLab/)
-
-## Lightwheel LW-BenchHub
-
-[Lightwheel](https://www.lightwheel.ai) is bringing `Lightwheel-Libero-Tasks` and `Lightwheel-RoboCasa-Tasks` with 268 tasks to the LeRobot ecosystem.
-LW-BenchHub collects and generates large-scale datasets via teleoperation that comply with the LeRobot specification, enabling out-of-the-box training and evaluation workflows.
-With the unified interface provided by EnvHub, developers can quickly build end-to-end experimental pipelines.
-
-### Install
-
-Assuming you followed the [Installation](#installation) steps, you can install LW-BenchHub with:
-
-```bash
-conda install pinocchio -c conda-forge -y
-pip install numpy==1.26.0 # revert numpy to version 1.26
-
-sudo apt-get install git-lfs && git lfs install
-
-git clone https://github.com/LightwheelAI/lw_benchhub
-git lfs pull # Ensure LFS files (e.g., .usd assets) are downloaded
-
-cd lw_benchhub
-pip install -e .
-```
-
-For more detailed instructions, please refer to the [LW-BenchHub Documentation](https://docs.lightwheel.net/lw_benchhub/usage/Installation).
-
-### Lightwheel Tasks Dataset
-
-LW-BenchHub datasets are available on HuggingFace Hub:
-
-| Dataset                                                                                                       | Description             | Tasks | Frames |
-| :------------------------------------------------------------------------------------------------------------ | :---------------------- | :---- | :----- |
-| [Lightwheel-Tasks-X7S](https://huggingface.co/datasets/LightwheelAI/Lightwheel-Tasks-X7S)                     | X7S LIBERO and RoboCasa | 117   | ~10.3M |
-| [Lightwheel-Tasks-Double-Piper](https://huggingface.co/datasets/LightwheelAI/Lightwheel-Tasks-Double-Piper)   | Double-Piper LIBERO     | 130   | ~6.0M  |
-| [Lightwheel-Tasks-G1-Controller](https://huggingface.co/datasets/LightwheelAI/Lightwheel-Tasks-G1-Controller) | G1-Controller LIBERO    | 62    | ~2.7M  |
-| [Lightwheel-Tasks-G1-WBC](https://huggingface.co/datasets/LightwheelAI/Lightwheel-Tasks-G1-WBC)               | G1-WBC RoboCasa         | 32    | ~1.5M  |
-
-For training policies, refer to the [Training Policies](#training-policies) section.
-
-### Evaluating Policies
-
-#### Pre-trained Policies
-
-The following trained policies are available:
-
-| Policy                   | Architecture | Task                           | Layout     | Robot           | Link                                                                                  |
-| :----------------------- | :----------- | :----------------------------- | :--------- | :-------------- | :------------------------------------------------------------------------------------ |
-| smolvla-double-piper-pnp | SmolVLA      | L90K1PutTheBlackBowlOnThePlate | libero-1-1 | DoublePiper-Abs | [HuggingFace](https://huggingface.co/LightwheelAI/smolvla-double-piper-pnp/tree/main) |
-
-#### Evaluate SmolVLA
-
-```bash
-lerobot-eval \
-  --policy.path=LightwheelAI/smolvla-double-piper-pnp \
-  --env.type=isaaclab_arena \
-  --rename_map='{"observation.images.left_hand_camera_rgb": "observation.images.left_hand", "observation.images.right_hand_camera_rgb": "observation.images.right_hand", "observation.images.first_person_camera_rgb": "observation.images.first_person"}' \
-  --env.hub_path=LightwheelAI/lw_benchhub_env \
-  --env.kwargs='{"config_path": "configs/envhub/example.yml"}' \
-  --trust_remote_code=true \
-  --env.state_keys=joint_pos \
-  --env.action_dim=12 \
-  --env.camera_keys=left_hand_camera_rgb,right_hand_camera_rgb,first_person_camera_rgb \
-  --policy.device=cuda \
-  --eval.batch_size=10 \
-  --eval.n_episodes=100
-```
-
-### Environment Configuration
-
-Evaluation can be quickly launched by modifying the `robot`, `task`, and `layout` settings in the configuration file.
-
-#### Full Configuration Options
-
-```yml
-# =========================
-# Basic Settings
-# =========================
-disable_fabric: false
-device: cuda:0
-sensitivity: 1.0
-step_hz: 50
-enable_cameras: true
-execute_mode: eval
-episode_length_s: 20.0 # Episode length in seconds, increase if episodes timeout during eval
-
-# =========================
-# Robot Settings
-# =========================
-robot: DoublePiper-Abs # Robot type, DoublePiper-Abs, X7S-Abs, G1-Controller or G1-Controller-DecoupledWBC
-robot_scale: 1.0
-
-# =========================
-# Task & Scene Settings
-# =========================
-task: L90K1PutTheBlackBowlOnThePlate # Task name
-scene_backend: robocasa
-task_backend: robocasa
-debug_assets: null
-layout: libero-1-1 # Layout and style ID
-sources:
-  - objaverse
-  - lightwheel
-  - aigen_objs
-object_projects: []
-usd_simplify: false
-seed: 42
-
-# =========================
-# Object Placement Retry Settings
-# =========================
-max_scene_retry: 4
-max_object_placement_retry: 3
-
-resample_objects_placement_on_reset: true
-resample_robot_placement_on_reset: true
-
-# =========================
-# Replay Configuration Settings
-# =========================
-replay_cfgs:
-  add_camera_to_observation: true
-  render_resolution: [640, 480]
-```
-
-### See Also
-
-- [LW-BenchHub GitHub](https://github.com/LightwheelAI/LW-BenchHub)
-- [LW-BenchHub Documentation](https://docs.lightwheel.net/lw_benchhub/)

docs/source/envhub_leisaac.mdx

@@ -137,8 +137,7 @@ from lerobot.teleoperators import (  # noqa: F401
     Teleoperator,
     TeleoperatorConfig,
     make_teleoperator_from_config,
-    so_leader,
-    bi_so_leader,
+    so101_leader,
 )
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.utils import init_logging
@@ -197,7 +196,7 @@ def teleop_loop(teleop: Teleoperator, env: gym.Env, fps: int):
             obs, info = env.reset()
 
         dt_s = time.perf_counter() - loop_start
-        precise_sleep(max(1 / fps - dt_s, 0.0))
+        precise_sleep(1 / fps - dt_s)
         loop_s = time.perf_counter() - loop_start
         print(f"\ntime: {loop_s * 1e3:.2f}ms ({1 / loop_s:.0f} Hz)")
 
@@ -223,7 +222,7 @@ def teleoperate(cfg: TeleoperateConfig):
 
 def main():
     teleoperate(TeleoperateConfig(
-        teleop=so_leader.SO101LeaderConfig(
+        teleop=so101_leader.SO101LeaderConfig(
             port="/dev/ttyACM0",
             id='leader',
             use_degrees=False,

docs/source/groot.mdx

@@ -12,12 +12,6 @@ Developers and researchers can post-train GR00T N1.5 with their own real or synt
 
 GR00T N1.5 (specifically the GR00T-N1.5-3B model) is built using pre-trained vision and language encoders. It utilizes a flow matching action transformer to model a chunk of actions, conditioned on vision, language, and proprioception.
 
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/lerobot-groot-paper1%20(1).png"
-  alt="An overview of GR00T"
-  width="80%"
-/>
-
 Its strong performance comes from being trained on an expansive and diverse humanoid dataset, which includes:
 
 - Real captured data from robots.
@@ -109,7 +103,7 @@ Once you have trained your model using your parameters you can run inference in
 
 ```bash
 lerobot-record \
-  --robot.type=bi_so_follower \
+  --robot.type=bi_so100_follower \
   --robot.left_arm_port=/dev/ttyACM1 \
   --robot.right_arm_port=/dev/ttyACM0 \
   --robot.id=bimanual_follower \
@@ -120,15 +114,12 @@ lerobot-record \
   --display_data=true \
   --dataset.repo_id=<user>/eval_groot-bimanual  \
   --dataset.num_episodes=10 \
-  --dataset.single_task="Grab and handover the red cube to the other arm" \
-  --dataset.streaming_encoding=true \
-  --dataset.encoder_threads=2 \
-  # --dataset.vcodec=auto \
-  --policy.path=<user>/groot-bimanual \ # your trained model
-  --dataset.episode_time_s=30 \
+  --dataset.single_task="Grab and handover the red cube to the other arm"
+  --policy.path=<user>/groot-bimanual # your trained model
+  --dataset.episode_time_s=30
   --dataset.reset_time_s=10
 ```
 
 ## License
 
-This model follows NVIDIA's proprietary license, consistent with the original [GR00T repository](https://github.com/NVIDIA/Isaac-GR00T). Future versions (starting from N1.7) will follow **Apache 2.0 License**.
+This model follows the **Apache 2.0 License**, consistent with the original [GR00T repository](https://github.com/NVIDIA/Isaac-GR00T).

docs/source/hil_data_collection.mdx

@@ -1,269 +0,0 @@
-# Human-In-the-Loop Data Collection
-
-Human-In-the-Loop (HIL) data collection lets you improve a trained policy by deploying it on a real robot while a human operator monitors and intervenes when needed. The intervention data (recovery movements and corrections) is recorded alongside autonomous segments, producing a richer training dataset that teaches the policy how to handle failures.
-
----
-
-## Why Human-In-the-Loop?
-
-Standard behavioral cloning trains policies on successful demonstrations only. During deployment, small errors can compound and push the robot into states never seen during training (distribution shift). HIL data collection addresses this by:
-
-- Running the trained policy on the real robot
-- Having a human intervene when the robot is about to fail
-- Recording the human's recovery and correction as training data
-- Fine-tuning the policy on the combined dataset
-
-This produces a policy that not only knows how to perform the task, but also how to recover when things go wrong.
-
----
-
-## How It Works
-
-During a HIL session, the human operator follows this loop within each episode:
-
-1. **Watch** the policy run autonomously
-2. **Pause** when failure is imminent, the robot holds its position
-3. **Take control** and teleoperate the robot back to a good state (recovery), then correct the behavior
-4. **Return control to the policy**, the policy resumes autonomous execution
-5. Repeat steps 2–4 as many times as needed during the episode
-6. **End the episode** when the task is complete, save and move on to the next rollout
-
-Both autonomous and human-controlled segments are recorded. The policy and human can alternate control multiple times within a single episode, and the episode continues from the current state after each handoff (no reset required just because intervention happened). This captures autonomous execution, recovery, and correction in one continuous trajectory. After collection, the combined dataset (original demonstrations + HIL data) is used to fine-tune the policy.
-
-This process can be repeated iteratively: deploy, collect, fine-tune, repeat. Each round targets the current policy's failure modes.
-
-```
-┌─────────────────────────────────────────────────────────────────────────┐
-│  Policy v0 (trained on demos)                                           │
-│       ↓                                                                 │
-│  HIL Collection (target current failure modes) → Fine-tune → Policy v1  │
-│       ↓                                                                 │
-│  HIL Collection (target new failure modes) → Fine-tune → Policy v2      │
-│       ↓                                                                 │
-│  ... (repeat until satisfactory performance)                            │
-└─────────────────────────────────────────────────────────────────────────┘
-```
-
----
-
-## Hardware Requirements
-
-### Teleoperator Requirements
-
-The `examples/hil` HIL scripts require **teleoperators with active motors** that can:
-
-- Enable/disable torque programmatically
-- Move to target positions (to mirror the robot state when pausing)
-
-**Compatible teleoperators in the current `examples/hil` scripts:**
-
-- `openarm_mini` - OpenArm Mini
-- `so_leader` - SO100 / SO101 leader arm
-
-> [!IMPORTANT]
-> The provided `examples/hil` commands default to `bi_openarm_follower` + `openarm_mini`.
-> `so_follower` + `so_leader` configs are also registered and can be used via CLI flags.
-
----
-
-## Script
-
-A single script handles both synchronous and RTC-based inference. Toggle RTC with `--rtc.enabled=true`:
-
-| Mode                     | Flag                 | Models                |
-| ------------------------ | -------------------- | --------------------- |
-| Standard (default)       | _(no flag needed)_   | ACT, Diffusion Policy |
-| Real-Time Chunking (RTC) | `--rtc.enabled=true` | Pi0, Pi0.5, SmolVLA   |
-
----
-
-## Step-by-Step Guide
-
-### Step 1: Pre-train a Base Policy
-
-First, train a policy on your demonstration dataset:
-
-```bash
-python src/lerobot/scripts/lerobot_train.py \
-    --dataset.repo_id=your-username/demo-dataset \
-    --policy.type=pi0 \
-    --output_dir=outputs/pretrain \
-    --batch_size=32 \
-    --steps=50000
-```
-
-### Step 2: Collect HIL Data
-
-**Standard inference (ACT, Diffusion Policy):**
-
-```bash
-python examples/hil/hil_data_collection.py \
-    --robot.type=bi_openarm_follower \
-    --robot.left_arm_config.port=can1 \
-    --robot.left_arm_config.side=left \
-    --robot.right_arm_config.port=can0 \
-    --robot.right_arm_config.side=right \
-    --robot.cameras='{left_wrist: {type: opencv, index_or_path: "/dev/video0", width: 1280, height: 720, fps: 30}, right_wrist: {type: opencv, index_or_path: "/dev/video4", width: 1280, height: 720, fps: 30}, base: {type: opencv, index_or_path: "/dev/video2", width: 640, height: 480, fps: 30}}' \
-    --teleop.type=openarm_mini \
-    --teleop.port_left=/dev/ttyACM0 \
-    --teleop.port_right=/dev/ttyACM1 \
-    --policy.path=outputs/pretrain/checkpoints/last/pretrained_model \
-    --dataset.repo_id=your-username/hil-dataset \
-    --dataset.single_task="Fold the T-shirt properly" \
-    --dataset.fps=30 \
-    --dataset.episode_time_s=1000 \
-    --dataset.num_episodes=50 \
-    --interpolation_multiplier=2
-```
-
-**With RTC for large models (Pi0, Pi0.5, SmolVLA):**
-
-For models with high inference latency, enable RTC for smooth execution:
-
-```bash
-python examples/hil/hil_data_collection.py \
-    --rtc.enabled=true \
-    --rtc.execution_horizon=20 \
-    --rtc.max_guidance_weight=5.0 \
-    --rtc.prefix_attention_schedule=LINEAR \
-    --robot.type=bi_openarm_follower \
-    --robot.left_arm_config.port=can1 \
-    --robot.left_arm_config.side=left \
-    --robot.right_arm_config.port=can0 \
-    --robot.right_arm_config.side=right \
-    --robot.cameras='{left_wrist: {type: opencv, index_or_path: "/dev/video0", width: 1280, height: 720, fps: 30}, right_wrist: {type: opencv, index_or_path: "/dev/video4", width: 1280, height: 720, fps: 30}, base: {type: opencv, index_or_path: "/dev/video2", width: 640, height: 480, fps: 30}}' \
-    --teleop.type=openarm_mini \
-    --teleop.port_left=/dev/ttyACM0 \
-    --teleop.port_right=/dev/ttyACM1 \
-    --policy.path=outputs/pretrain/checkpoints/last/pretrained_model \
-    --dataset.repo_id=your-username/hil-rtc-dataset \
-    --dataset.single_task="Fold the T-shirt properly" \
-    --dataset.fps=30 \
-    --dataset.episode_time_s=1000 \
-    --dataset.num_episodes=50 \
-    --interpolation_multiplier=3
-```
-
-**Controls (Conceptual):**
-
-The interaction model is:
-
-- **Pause input**: pause autonomous policy execution
-- **Takeover input**: transfer control to the human operator and record intervention data
-- **Return-to-policy input**: hand control back to the policy and continue the same episode
-- **Episode control inputs**: save/re-record/stop/reset as needed
-
-Exact key/pedal bindings can differ across scripts and hardware integrations. Use each script's printed controls as the source of truth for the concrete mapping on your setup.
-
-**The HIL Protocol:**
-
-1. Watch the policy run autonomously (teleop is idle/free)
-2. When you see imminent failure, trigger the **pause input**
-   - Policy stops
-   - Teleoperator moves to match robot position (torque enabled)
-   - No frames recorded during pause
-3. Trigger the **takeover input** to take control
-   - Teleoperator torque disabled, free to move
-   - **Recovery**: Teleoperate the robot back to a good state
-   - **Correction**: Correct the behavior
-   - All movements are recorded
-4. Trigger the **return-to-policy input**
-   - Policy resumes autonomous execution from the current state
-   - You can intervene again at any time (repeat steps 2–4)
-5. End and save the episode when the task is complete (or episode time limit is reached)
-6. **Reset**: Teleop moves to robot position, you can move the robot to the starting position
-7. Start the next episode
-
-**Foot Pedal Setup (Linux):**
-
-If using a USB foot pedal (PCsensor FootSwitch), ensure access:
-
-```bash
-sudo setfacl -m u:$USER:rw /dev/input/by-id/usb-PCsensor_FootSwitch-event-kbd
-```
-
-### Step 3: Fine-tune the Policy
-
-Fine-tune on the **combined** dataset (`demo-dataset` + `hil-dataset` merged together):
-
-```bash
-python src/lerobot/scripts/lerobot_train.py \
-    --dataset.repo_id=your-username/hil-dataset \
-    --policy.type=pi0 \
-    --policy.pretrained_path=outputs/pretrain/checkpoints/last/pretrained_model \
-    --output_dir=outputs/hil_finetune \
-    --steps=20000
-```
-
-Then deploy the fine-tuned policy and repeat from Step 2 to target its remaining failure modes.
-
----
-
-## Tips for Effective HIL Collection
-
-### When to Intervene
-
-Intervene when you see:
-
-- Robot about to make an irreversible mistake
-- Robot hesitating or showing uncertain behavior
-- Robot deviating from the expected trajectory
-
-### Recovery: Teleoperating Back to a Good State
-
-During recovery, teleoperate the robot back to a state where:
-
-- The robot is in a familiar, in-distribution configuration
-- The current subtask can still be completed
-- The recovery trajectory itself is informative training data
-
-### Quality of Corrections
-
-During correction:
-
-- Provide **confident, clean** trajectories
-- Complete the current subtask fully
-- Don't overcorrect or add unnecessary movements
-
----
-
-## Related Work
-
-This HIL data collection approach builds on ideas from interactive imitation learning:
-
-- **DAgger** (Ross et al., 2011) introduced the core idea: instead of only training on expert demonstrations, query the expert for corrections on states the _learner_ visits. This breaks the compounding-error cycle of standard behavioral cloning by iteratively collecting on-policy data.
-
-- **HG-DAgger** (Kelly et al., 2019) made this practical for robotics: a human expert monitors the robot and only intervenes when needed, rather than labeling every state. The gating between autonomous and human control is exactly the pause → takeover → return-to-policy loop used in the scripts here.
-
-- **RaC** (Hu et al., 2025) scales this loop to long-horizon tasks by explicitly decomposing interventions into **recovery** (teleoperating back to a good state) and **correction** (demonstrating the right behavior from there). This decomposition is the protocol followed by the HIL scripts in `examples/hil`.
-
-- **π0.6/RECAP** (Physical Intelligence, 2025) applies the same iterative collect-and-finetune loop at scale with VLA models, showing that even large pretrained policies benefit substantially from targeted human corrections on their own failure modes. π0.6 is trained using RECAP.
-
-```bibtex
-@article{ross2011dagger,
-  title={A Reduction of Imitation Learning and Structured Prediction to No-Regret Online Learning},
-  author={Ross, Stéphane and Gordon, Geoffrey and Bagnell, Drew},
-  journal={Proceedings of the Fourteenth International Conference on Artificial Intelligence and Statistics},
-  year={2011}
-}
-
-@article{kelly2019hgdagger,
-  title={HG-DAgger: Interactive Imitation Learning with Human Experts},
-  author={Kelly, Michael and Sidrane, Chelsea and Driggs-Campbell, Katherine and Kochenderfer, Mykel J},
-  journal={arXiv preprint arXiv:1810.02890},
-  year={2019}
-}
-
-@article{hu2025rac,
-  title={RaC: Robot Learning for Long-Horizon Tasks by Scaling Recovery and Correction},
-  author={Hu, Zheyuan and Wu, Robyn and Enock, Naveen and Li, Jasmine and Kadakia, Riya and Erickson, Zackory and Kumar, Aviral},
-  journal={arXiv preprint arXiv:2509.07953},
-  year={2025}
-}
-
-@article{pi2025recap,
-  title={π0.6: a VLA That Learns From Experience},
-  author={Physical Intelligence},
-  year={2025}
-}
-```

docs/source/hope_jr.mdx

@@ -224,15 +224,12 @@ lerobot-record \
     --teleop.port=/dev/tty.usbmodem1201 \
     --teleop.id=right \
     --teleop.side=right \
-    --dataset.repo_id=<USER>/hand_record_test_with_video_data \
+    --dataset.repo_id=nepyope/hand_record_test_with_video_data \
     --dataset.single_task="Hand recording test with video data" \
     --dataset.num_episodes=1 \
     --dataset.episode_time_s=5 \
     --dataset.push_to_hub=true \
     --dataset.private=true \
-    --dataset.streaming_encoding=true \
-    --dataset.encoder_threads=2 \
-    # --dataset.vcodec=auto \
     --display_data=true
 ```
 
@@ -244,7 +241,7 @@ lerobot-replay \
     --robot.port=/dev/tty.usbmodem58760432281 \
     --robot.id=right \
     --robot.side=right \
-    --dataset.repo_id=<USER>/hand_record_test_with_camera \
+    --dataset.repo_id=nepyope/hand_record_test_with_camera \
     --dataset.episode=0
 ```
 
@@ -252,13 +249,13 @@ lerobot-replay \
 
 ```bash
 lerobot-train \
-  --dataset.repo_id=<USER>/hand_record_test_with_video_data \
+  --dataset.repo_id=nepyope/hand_record_test_with_video_data \
   --policy.type=act \
   --output_dir=outputs/train/hopejr_hand \
   --job_name=hopejr \
   --policy.device=mps \
   --wandb.enable=true \
-  --policy.repo_id=<USER>/hand_test_policy
+  --policy.repo_id=nepyope/hand_test_policy
 ```
 
 ### Evaluate
@@ -273,11 +270,8 @@ lerobot-record \
   --robot.side=right \
   --robot.cameras='{"main": {"type": "opencv", "index_or_path": 0, "width": 640, "height": 480, "fps": 30}}' \
   --display_data=false \
-  --dataset.repo_id=<USER>/eval_hopejr \
+  --dataset.repo_id=nepyope/eval_hopejr \
   --dataset.single_task="Evaluate hopejr hand policy" \
   --dataset.num_episodes=10 \
-  --dataset.streaming_encoding=true \
-  --dataset.encoder_threads=2 \
-  # --dataset.vcodec=auto \
   --policy.path=outputs/train/hopejr_hand/checkpoints/last/pretrained_model
 ```

docs/source/il_robots.mdx

@@ -58,8 +58,8 @@ lerobot-teleoperate \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.teleoperators.so_leader import SO101LeaderConfig, SO101Leader
-from lerobot.robots.so_follower import SO101FollowerConfig, SO101Follower
+from lerobot.teleoperators.so101_leader import SO101LeaderConfig, SO101Leader
+from lerobot.robots.so101_follower import SO101FollowerConfig, SO101Follower
 
 robot_config = SO101FollowerConfig(
     port="/dev/tty.usbmodem58760431541",
@@ -159,13 +159,13 @@ We use the Hugging Face hub features for uploading your dataset. If you haven't
 Add your token to the CLI by running this command:
 
 ```bash
-hf auth login --token ${HUGGINGFACE_TOKEN} --add-to-git-credential
+huggingface-cli login --token ${HUGGINGFACE_TOKEN} --add-to-git-credential
 ```
 
 Then store your Hugging Face repository name in a variable:
 
 ```bash
-HF_USER=$(NO_COLOR=1 hf auth whoami | awk -F': *' 'NR==1 {print $2}')
+HF_USER=$(hf auth whoami | head -n 1)
 echo $HF_USER
 ```
 
@@ -185,10 +185,7 @@ lerobot-record \
     --display_data=true \
     --dataset.repo_id=${HF_USER}/record-test \
     --dataset.num_episodes=5 \
-    --dataset.single_task="Grab the black cube" \
-    --dataset.streaming_encoding=true \
-    # --dataset.vcodec=auto \
-    --dataset.encoder_threads=2
+    --dataset.single_task="Grab the black cube"
 ```
 </hfoption>
 <hfoption id="API example">
@@ -198,9 +195,9 @@ lerobot-record \
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.utils import hw_to_dataset_features
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.teleoperators.so_leader.config_so100_leader import SO100LeaderConfig
-from lerobot.teleoperators.so_leader.so100_leader import SO100Leader
+from lerobot.robots.so100_follower import SO100Follower, SO100FollowerConfig
+from lerobot.teleoperators.so100_leader.config_so100_leader import SO100LeaderConfig
+from lerobot.teleoperators.so100_leader.so100_leader import SO100Leader
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say
 from lerobot.utils.visualization_utils import init_rerun
@@ -327,7 +324,7 @@ You can look for other LeRobot datasets on the hub by searching for `LeRobot` [t
 You can also push your local dataset to the Hub manually, running:
 
 ```bash
-hf upload ${HF_USER}/record-test ~/.cache/huggingface/lerobot/{repo-id} --repo-type dataset
+huggingface-cli upload ${HF_USER}/record-test ~/.cache/huggingface/lerobot/{repo-id} --repo-type dataset
 ```
 
 #### Record function
@@ -411,8 +408,8 @@ lerobot-replay \
 import time
 
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
-from lerobot.robots.so_follower.config_so100_follower import SO100FollowerConfig
-from lerobot.robots.so_follower.so100_follower import SO100Follower
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.utils import log_say
 
@@ -424,7 +421,7 @@ robot = SO100Follower(robot_config)
 robot.connect()
 
 dataset = LeRobotDataset("<hf_username>/<dataset_repo_id>", episodes=[episode_idx])
-actions = dataset.select_columns("action")
+actions = dataset.hf_dataset.select_columns("action")
 
 log_say(f"Replaying episode {episode_idx}")
 for idx in range(dataset.num_frames):
@@ -435,7 +432,7 @@ for idx in range(dataset.num_frames):
     }
     robot.send_action(action)
 
-    precise_sleep(max(1.0 / dataset.fps - (time.perf_counter() - t0), 0.0))
+    precise_sleep(1.0 / dataset.fps - (time.perf_counter() - t0))
 
 robot.disconnect()
 ```
@@ -491,7 +488,7 @@ If your local computer doesn't have a powerful GPU you could utilize Google Cola
 Once training is done, upload the latest checkpoint with:
 
 ```bash
-hf upload ${HF_USER}/act_so101_test \
+huggingface-cli upload ${HF_USER}/act_so101_test \
   outputs/train/act_so101_test/checkpoints/last/pretrained_model
 ```
 
@@ -499,7 +496,7 @@ You can also upload intermediate checkpoints with:
 
 ```bash
 CKPT=010000
-hf upload ${HF_USER}/act_so101_test${CKPT} \
+huggingface-cli upload ${HF_USER}/act_so101_test${CKPT} \
   outputs/train/act_so101_test/checkpoints/${CKPT}/pretrained_model
 ```
 
@@ -518,9 +515,6 @@ lerobot-record  \
   --display_data=false \
   --dataset.repo_id=${HF_USER}/eval_so100 \
   --dataset.single_task="Put lego brick into the transparent box" \
-  --dataset.streaming_encoding=true \
-  --dataset.encoder_threads=2 \
-  # --dataset.vcodec=auto \
   # <- Teleop optional if you want to teleoperate in between episodes \
   # --teleop.type=so100_leader \
   # --teleop.port=/dev/ttyACM0 \
@@ -537,8 +531,8 @@ from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.utils import hw_to_dataset_features
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors
-from lerobot.robots.so_follower.config_so100_follower import SO100FollowerConfig
-from lerobot.robots.so_follower.so100_follower import SO100Follower
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.scripts.lerobot_record import record_loop
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say

docs/source/installation.mdx

@@ -1,120 +1,44 @@
 # Installation
 
-This guide uses `conda` (via miniforge) to manage environments (recommended). If you prefer another environment manager (e.g. `uv`, `venv`), ensure you have Python >=3.12 and support PyTorch >= 2.10, then skip ahead to [Environment Setup](#step-2-environment-setup).
-
-## Step 1 (`conda` only): Install [`miniforge`](https://conda-forge.org/download/)
+## Install [`miniforge`](https://conda-forge.org/download/)
 
 ```bash
 wget "https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-$(uname)-$(uname -m).sh"
 bash Miniforge3-$(uname)-$(uname -m).sh
 ```
 
-## Step 2: Environment Setup
+## Environment Setup
 
-Create a virtual environment with Python 3.12:
+Create a virtual environment with Python 3.10, using conda:
 
-<!-- prettier-ignore-start -->
-<hfoptions id="create_venv">
-<hfoption id="conda">
 ```bash
-conda create -y -n lerobot python=3.12
+conda create -y -n lerobot python=3.10
 ```
-</hfoption>
-<hfoption id="uv (PyTorch >= 2.10 only)">
-```bash
-uv python install 3.12
-uv venv --python 3.12
-```
-</hfoption>
-</hfoptions>
-<!-- prettier-ignore-end -->
 
-Then activate your virtual environment, you have to do this each time you open a shell to use lerobot:
+Then activate your conda environment, you have to do this each time you open a shell to use lerobot:
 
-<!-- prettier-ignore-start -->
-
-<hfoptions id="activate_venv">
-<hfoption id="conda">
 ```bash
 conda activate lerobot
 ```
 
-> [!NOTE]
-> When installing LeRobot inside WSL (Windows Subsystem for Linux), make sure to also install `evdev`:
->
-> ```bash
-> conda install evdev -c conda-forge
-> ```
-
-</hfoption>
-<hfoption id="uv (PyTorch >= 2.10 only)">
-```bash
-# Linux/macOS
-source .venv/bin/activate
-# Windows PowerShell
-.venv\Scripts\activate
-```
-
-> [!NOTE]
-> When installing LeRobot inside WSL (Windows Subsystem for Linux), make sure to also install `evdev`:
->
-> ```bash
-> sudo apt install libevdev-dev
-> uv pip install evdev
-> ```
-
-</hfoption>
-</hfoptions>
-<!-- prettier-ignore-end -->
-
-### Install `ffmpeg` (for video decoding)
-
-LeRobot uses [TorchCodec](https://github.com/meta-pytorch/torchcodec) for video decoding by default, which requires `ffmpeg`.
-
-> [!NOTE]
-> **Platform support:** TorchCodec is **not available** on macOS Intel (x86_64), Linux ARM (aarch64, arm64, armv7l), or Windows with PyTorch < 2.8. On these platforms, LeRobot automatically falls back to `pyav` — so you do not need to install `ffmpeg` and can skip to Step 3.
-
-If your platform supports TorchCodec, install `ffmpeg` using one of the methods below:
-
-<!-- prettier-ignore-start -->
-
-<hfoptions id="install_ffmpeg">
-<hfoption id="conda (any PyTorch version)">
-
-Install `ffmpeg` in your conda environment. This works with **all PyTorch versions** and is **required for PyTorch < 2.10**:
+When using `conda`, install `ffmpeg` in your environment:
 
 ```bash
 conda install ffmpeg -c conda-forge
 ```
 
 > [!TIP]
-> This usually installs `ffmpeg 8.X` with the `libsvtav1` encoder. If you run into issues (e.g. `libsvtav1` missing — check with `ffmpeg -encoders` — or a version mismatch with `torchcodec`), you can explicitly install `ffmpeg 7.1.1` using:
+> This usually installs `ffmpeg 7.X` for your platform compiled with the `libsvtav1` encoder. If `libsvtav1` is not supported (check supported encoders with `ffmpeg -encoders`), you can:
+>
+> - _[On any platform]_ Explicitly install `ffmpeg 7.X` using:
 >
 > ```bash
 > conda install ffmpeg=7.1.1 -c conda-forge
 > ```
+>
+> - _[On Linux only]_ If you want to bring your own ffmpeg: Install [ffmpeg build dependencies](https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu#GettheDependencies) and [compile ffmpeg from source with libsvtav1](https://trac.ffmpeg.org/wiki/CompilationGuide/Ubuntu#libsvtav1), and make sure you use the corresponding ffmpeg binary to your install with `which ffmpeg`.
 
-</hfoption>
-<hfoption id="uv (PyTorch >= 2.10 only)">
-
-Starting with **PyTorch >= 2.10** (TorchCodec ≥ 0.10), TorchCodec can dynamically link to a system-wide `ffmpeg` installation. This is useful when using `uv` or other non-`conda` environment managers:
-
-```bash
-# Ubuntu/Debian
-sudo apt install ffmpeg
-
-# macOS (Apple Silicon)
-brew install ffmpeg
-```
-
-> [!IMPORTANT]
-> System-wide `ffmpeg` is **only supported with PyTorch >= 2.10** (TorchCodec ≥ 0.10). For older PyTorch versions, you **must** use `conda install ffmpeg -c conda-forge` instead.
-
-</hfoption>
-</hfoptions>
-<!-- prettier-ignore-end -->
-
-## Step 3: Install LeRobot 🤗
+## Install LeRobot 🤗
 
 ### From Source
 
@@ -127,45 +51,23 @@ cd lerobot
 
 Then, install the library in editable mode. This is useful if you plan to contribute to the code.
 
-<!-- prettier-ignore-start -->
-<hfoptions id="install_lerobot_src">
-<hfoption id="conda">
 ```bash
 pip install -e .
 ```
-</hfoption>
-<hfoption id="uv">
-```bash
-uv pip install -e .
-```
-</hfoption>
-</hfoptions>
-<!-- prettier-ignore-end -->
 
 ### Installation from PyPI
 
 **Core Library:**
 Install the base package with:
 
-<!-- prettier-ignore-start -->
-<hfoptions id="install_lerobot_pypi">
-<hfoption id="conda">
 ```bash
 pip install lerobot
 ```
-</hfoption>
-<hfoption id="uv">
-```bash
-uv pip install lerobot
-```
-</hfoption>
-</hfoptions>
-<!-- prettier-ignore-end -->
 
 _This installs only the default dependencies._
 
 **Extra Features:**
-To install additional functionality, use one of the following (If you are using `uv`, replace `pip install` with `uv pip install` in the commands below.):
+To install additional functionality, use one of the following:
 
 ```bash
 pip install 'lerobot[all]'          # All available features
@@ -179,10 +81,13 @@ _Replace `[...]` with your desired features._
 For a full list of optional dependencies, see:
 https://pypi.org/project/lerobot/
 
+> [!NOTE]
+> For lerobot 0.4.0, if you want to install pi, you will have to do: `pip install "lerobot[pi]@git+https://github.com/huggingface/lerobot.git"`
+
 ### Troubleshooting
 
 If you encounter build errors, you may need to install additional dependencies: `cmake`, `build-essential`, and `ffmpeg libs`.
-To install these for Linux run:
+To install these for linux run:
 
 ```bash
 sudo apt-get install cmake build-essential python3-dev pkg-config libavformat-dev libavcodec-dev libavdevice-dev libavutil-dev libswscale-dev libswresample-dev libavfilter-dev
@@ -192,7 +97,7 @@ For other systems, see: [Compiling PyAV](https://pyav.org/docs/develop/overview/
 
 ## Optional dependencies
 
-LeRobot provides optional extras for specific functionalities. Multiple extras can be combined (e.g., `.[aloha,feetech]`). For all available extras, refer to `pyproject.toml`. If you are using `uv`, replace `pip install` with `uv pip install` in the commands below.
+LeRobot provides optional extras for specific functionalities. Multiple extras can be combined (e.g., `.[aloha,feetech]`). For all available extras, refer to `pyproject.toml`.
 
 ### Simulations
 

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/so101_follower/so101_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

@@ -1,11 +1,5 @@
 # LeKiwi
 
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/1740517739083.jpeg"
-  alt="LeKiwi"
-  width="70%"
-/>
-
 In the steps below, we explain how to assemble the LeKiwi mobile robot.
 
 ## Source the parts
@@ -210,7 +204,7 @@ lerobot-calibrate \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.teleoperators.so_leader import SO100LeaderConfig, SO100Leader
+from lerobot.teleoperators.so100_leader import SO100LeaderConfig, SO100Leader
 
 config = SO100LeaderConfig(
     port="/dev/tty.usbmodem58760431551",
@@ -279,13 +273,13 @@ We use the Hugging Face hub features for uploading your dataset. If you haven't
 Add your token to the CLI by running this command:
 
 ```bash
-hf auth login --token ${HUGGINGFACE_TOKEN} --add-to-git-credential
+huggingface-cli login --token ${HUGGINGFACE_TOKEN} --add-to-git-credential
 ```
 
 Then store your Hugging Face repository name in a variable:
 
 ```bash
-HF_USER=$(hf auth whoami | awk -F': *' 'NR==1 {print $2}')
+HF_USER=$(huggingface-cli whoami | head -n 1)
 echo $HF_USER
 ```
 

docs/source/lerobot-dataset-v3.mdx

@@ -41,10 +41,7 @@ lerobot-record \
   --display_data=true \
   --dataset.repo_id=${HF_USER}/record-test \
   --dataset.num_episodes=5 \
-  --dataset.single_task="Grab the black cube" \
-  --dataset.streaming_encoding=true \
-  # --dataset.vcodec=auto \
-  --dataset.encoder_threads=2
+  --dataset.single_task="Grab the black cube"
 ```
 
 See the [recording guide](./il_robots#record-a-dataset) for more details.

docs/source/libero.mdx

@@ -1,61 +1,36 @@
 # LIBERO
 
-LIBERO is a benchmark designed to study **lifelong robot learning** — the idea that robots need to keep learning and adapting with their users over time, not just be pretrained once. It provides a set of standardized manipulation tasks that focus on **knowledge transfer**: how well a robot can apply what it has already learned to new situations. By evaluating on LIBERO, different algorithms can be compared fairly and researchers can build on each other's work.
+**LIBERO** is a benchmark designed to study **lifelong robot learning**. The idea is that robots won’t just be pretrained once in a factory, they’ll need to keep learning and adapting with their human users over time. This ongoing adaptation is called **lifelong learning in decision making (LLDM)**, and it’s a key step toward building robots that become truly personalized helpers.
 
-- Paper: [Benchmarking Knowledge Transfer for Lifelong Robot Learning](https://arxiv.org/abs/2306.03310)
-- GitHub: [Lifelong-Robot-Learning/LIBERO](https://github.com/Lifelong-Robot-Learning/LIBERO)
-- Project website: [libero-project.github.io](https://libero-project.github.io)
+- 📄 [LIBERO paper](https://arxiv.org/abs/2306.03310)
+- 💻 [Original LIBERO repo](https://github.com/Lifelong-Robot-Learning/LIBERO)
+
+To make progress on this challenge, LIBERO provides a set of standardized tasks that focus on **knowledge transfer**: how well a robot can apply what it has already learned to new situations. By evaluating on LIBERO, different algorithms can be compared fairly and researchers can build on each other’s work.
+
+LIBERO includes **five task suites**:
+
+- **LIBERO-Spatial (`libero_spatial`)** – tasks that require reasoning about spatial relations.
+- **LIBERO-Object (`libero_object`)** – tasks centered on manipulating different objects.
+- **LIBERO-Goal (`libero_goal`)** – goal-conditioned tasks where the robot must adapt to changing targets.
+- **LIBERO-90 (`libero_90`)** – 90 short-horizon tasks from the LIBERO-100 collection.
+- **LIBERO-Long (`libero_10`)** – 10 long-horizon tasks from the LIBERO-100 collection.
+
+Together, these suites cover **130 tasks**, ranging from simple object manipulations to complex multi-step scenarios. LIBERO is meant to grow over time, and to serve as a shared benchmark where the community can test and improve lifelong learning algorithms.
 
 ![An overview of the LIBERO benchmark](https://libero-project.github.io/assets/img/libero/fig1.png)
 
-## Available tasks
+## Evaluating with LIBERO
 
-LIBERO includes **five task suites** covering **130 tasks**, ranging from simple object manipulations to complex multi-step scenarios:
+At **LeRobot**, we ported [LIBERO](https://github.com/Lifelong-Robot-Learning/LIBERO) into our framework and used it mainly to **evaluate [SmolVLA](https://huggingface.co/docs/lerobot/en/smolvla)**, our lightweight Vision-Language-Action model.
 
-| Suite          | CLI name         | Tasks | Description                                        |
-| -------------- | ---------------- | ----- | -------------------------------------------------- |
-| LIBERO-Spatial | `libero_spatial` | 10    | Tasks requiring reasoning about spatial relations  |
-| LIBERO-Object  | `libero_object`  | 10    | Tasks centered on manipulating different objects   |
-| LIBERO-Goal    | `libero_goal`    | 10    | Goal-conditioned tasks with changing targets       |
-| LIBERO-90      | `libero_90`      | 90    | Short-horizon tasks from the LIBERO-100 collection |
-| LIBERO-Long    | `libero_10`      | 10    | Long-horizon tasks from the LIBERO-100 collection  |
+LIBERO is now part of our **multi-eval supported simulation**, meaning you can benchmark your policies either on a **single suite of tasks** or across **multiple suites at once** with just a flag.
 
-## Installation
-
-After following the LeRobot installation instructions:
-
-```bash
-pip install -e ".[libero]"
-```
-
-<Tip>
-LIBERO requires Linux (`sys_platform == 'linux'`). LeRobot uses MuJoCo for simulation — set the rendering backend before training or evaluation:
-
-```bash
-export MUJOCO_GL=egl  # for headless servers (HPC, cloud)
-```
-
-</Tip>
-
-## Evaluation
-
-### Default evaluation (recommended)
-
-Evaluate across the four standard suites (10 episodes per task):
-
-```bash
-lerobot-eval \
-  --policy.path="your-policy-id" \
-  --env.type=libero \
-  --env.task=libero_spatial,libero_object,libero_goal,libero_10 \
-  --eval.batch_size=1 \
-  --eval.n_episodes=10 \
-  --env.max_parallel_tasks=1
-```
+To Install LIBERO, after following LeRobot official instructions, just do:
+`pip install -e ".[libero]"`
 
 ### Single-suite evaluation
 
-Evaluate on one LIBERO suite:
+Evaluate a policy on one LIBERO suite:
 
 ```bash
 lerobot-eval \
@@ -67,13 +42,14 @@ lerobot-eval \
 ```
 
 - `--env.task` picks the suite (`libero_object`, `libero_spatial`, etc.).
-- `--env.task_ids` restricts to specific task indices (`[0]`, `[1,2,3]`, etc.). Omit to run all tasks in the suite.
 - `--eval.batch_size` controls how many environments run in parallel.
-- `--eval.n_episodes` sets how many episodes to run per task.
+- `--eval.n_episodes` sets how many episodes to run in total.
+
+---
 
 ### Multi-suite evaluation
 
-Benchmark a policy across multiple suites at once by passing a comma-separated list:
+Benchmark a policy across multiple suites at once:
 
 ```bash
 lerobot-eval \
@@ -84,49 +60,50 @@ lerobot-eval \
   --eval.n_episodes=2
 ```
 
-### Control mode
+- Pass a comma-separated list to `--env.task` for multi-suite evaluation.
 
-LIBERO supports two control modes — `relative` (default) and `absolute`. Different VLA checkpoints are trained with different action parameterizations, so make sure the mode matches your policy:
+### Control Mode
 
-```bash
---env.control_mode=relative   # or "absolute"
-```
+LIBERO now supports two control modes: relative and absolute. This matters because different VLA checkpoints are trained with different mode of action to output hence control parameterizations.
+You can switch them with: `env.control_mode = "relative"` and `env.control_mode = "absolute"`
 
 ### Policy inputs and outputs
 
-**Observations:**
+When using LIBERO through LeRobot, policies interact with the environment via **observations** and **actions**:
 
-- `observation.state` — 8-dim proprioceptive features (eef position, axis-angle orientation, gripper qpos)
-- `observation.images.image` — main camera view (`agentview_image`), HWC uint8
-- `observation.images.image2` — wrist camera view (`robot0_eye_in_hand_image`), HWC uint8
+- **Observations**
+  - `observation.state` – proprioceptive features (agent state).
+  - `observation.images.image` – main camera view (`agentview_image`).
+  - `observation.images.image2` – wrist camera view (`robot0_eye_in_hand_image`).
 
-<Tip warning={true}>
-  LeRobot enforces the `.images.*` prefix for visual features. Ensure your
-  policy config `input_features` use the same naming keys, and that your dataset
-  metadata keys follow this convention. If your data contains different keys,
-  you must rename the observations to match what the policy expects, since
-  naming keys are encoded inside the normalization statistics layer.
-</Tip>
+  ⚠️ **Note:** LeRobot enforces the `.images.*` prefix for any multi-modal visual features. Always ensure that your policy config `input_features` use the same naming keys, and that your dataset metadata keys follow this convention during evaluation.
+  If your data contains different keys, you must rename the observations to match what the policy expects, since naming keys are encoded inside the normalization statistics layer.
+  This will be fixed with the upcoming Pipeline PR.
 
-**Actions:**
+- **Actions**
+  - Continuous control values in a `Box(-1, 1, shape=(7,))` space.
 
-- Continuous control in `Box(-1, 1, shape=(7,))` — 6D end-effector delta + 1D gripper
+We also provide a notebook for quick testing:
+Training with LIBERO
 
-### Recommended evaluation episodes
+## Training with LIBERO
 
-For reproducible benchmarking, use **10 episodes per task** across all four standard suites (Spatial, Object, Goal, Long). This gives 400 total episodes and matches the protocol used for published results.
+When training on LIBERO tasks, make sure your dataset parquet and metadata keys follow the LeRobot convention.
 
-## Training
+The environment expects:
 
-### Dataset
+- `observation.state` → 8-dim agent state
+- `observation.images.image` → main camera (`agentview_image`)
+- `observation.images.image2` → wrist camera (`robot0_eye_in_hand_image`)
 
-We provide a preprocessed LIBERO dataset fully compatible with LeRobot:
+⚠️ Cleaning the dataset upfront is **cleaner and more efficient** than remapping keys inside the code.
+To avoid potential mismatches and key errors, we provide a **preprocessed LIBERO dataset** that is fully compatible with the current LeRobot codebase and requires no additional manipulation:
+👉 [HuggingFaceVLA/libero](https://huggingface.co/datasets/HuggingFaceVLA/libero)
 
-- [HuggingFaceVLA/libero](https://huggingface.co/datasets/HuggingFaceVLA/libero)
+For reference, here is the **original dataset** published by Physical Intelligence:
+👉 [physical-intelligence/libero](https://huggingface.co/datasets/physical-intelligence/libero)
 
-For reference, the original dataset published by Physical Intelligence:
-
-- [physical-intelligence/libero](https://huggingface.co/datasets/physical-intelligence/libero)
+---
 
 ### Example training command
 
@@ -143,39 +120,52 @@ lerobot-train \
   --batch_size=4 \
   --eval.batch_size=1 \
   --eval.n_episodes=1 \
-  --eval_freq=1000
+  --eval_freq=1000 \
 ```
 
-## Reproducing published results
+---
 
-We reproduce the results of Pi0.5 on the LIBERO benchmark. We take the Physical Intelligence LIBERO base model (`pi05_libero`) and finetune for an additional 6k steps in bfloat16, with batch size of 256 on 8 H100 GPUs using the [HuggingFace LIBERO dataset](https://huggingface.co/datasets/HuggingFaceVLA/libero).
+### Note on rendering
 
-The finetuned model: [lerobot/pi05_libero_finetuned](https://huggingface.co/lerobot/pi05_libero_finetuned)
+LeRobot uses MuJoCo for simulation. You need to set the rendering backend before training or evaluation:
 
-### Evaluation command
+- `export MUJOCO_GL=egl` → for headless servers (e.g. HPC, cloud)
+
+## Reproducing π₀.₅ results
+
+We reproduce the results of π₀.₅ on the LIBERO benchmark using the LeRobot implementation. We take the Physical Intelligence LIBERO base model (`pi05_libero`) and finetune for an additional 6k 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:
+
+- **π₀.₅ LIBERO**: [lerobot/pi05_libero_finetuned](https://huggingface.co/lerobot/pi05_libero_finetuned)
+
+We then evaluate the finetuned model using the LeRobot LIBERO implementation, by running the following command:
 
 ```bash
 lerobot-eval \
-  --output_dir=./eval_logs/ \
+  --output_dir=/logs/ \
   --env.type=libero \
   --env.task=libero_spatial,libero_object,libero_goal,libero_10 \
   --eval.batch_size=1 \
   --eval.n_episodes=10 \
   --policy.path=pi05_libero_finetuned \
   --policy.n_action_steps=10 \
+  --output_dir=./eval_logs/ \
   --env.max_parallel_tasks=1
 ```
 
-We set `n_action_steps=10`, matching the original OpenPI implementation.
+**Note:** We set `n_action_steps=10`, similar to the original OpenPI implementation.
 
 ### Results
 
-| Model               | LIBERO Spatial | LIBERO Object | LIBERO Goal | LIBERO 10 | Average  |
-| ------------------- | -------------- | ------------- | ----------- | --------- | -------- |
-| **Pi0.5 (LeRobot)** | 97.0           | 99.0          | 98.0        | 96.0      | **97.5** |
+We obtain the following results on the LIBERO benchmark:
 
-These results are consistent with the [original results](https://github.com/Physical-Intelligence/openpi/tree/main/examples/libero#results) reported by Physical Intelligence:
+| Model    | LIBERO Spatial | LIBERO Object | LIBERO Goal | LIBERO 10 | Average  |
+| -------- | -------------- | ------------- | ----------- | --------- | -------- |
+| **π₀.₅** | 97.0           | 99.0          | 98.0        | 96.0      | **97.5** |
 
-| Model              | LIBERO Spatial | LIBERO Object | LIBERO Goal | LIBERO 10 | Average   |
-| ------------------ | -------------- | ------------- | ----------- | --------- | --------- |
-| **Pi0.5 (OpenPI)** | 98.8           | 98.2          | 98.0        | 92.4      | **96.85** |
+These results are consistent with the original [results](https://github.com/Physical-Intelligence/openpi/tree/main/examples/libero#results) reported by Physical Intelligence:
+
+| Model    | LIBERO Spatial | LIBERO Object | LIBERO Goal | LIBERO 10 | Average   |
+| -------- | -------------- | ------------- | ----------- | --------- | --------- |
+| **π₀.₅** | 98.8           | 98.2          | 98.0        | 92.4      | **96.85** |

docs/source/metaworld.mdx

@@ -1,111 +1,32 @@
 # Meta-World
 
-Meta-World is an open-source simulation benchmark for **multi-task and meta reinforcement learning** in continuous-control robotic manipulation. It bundles 50 diverse manipulation tasks using everyday objects and a common tabletop Sawyer arm, providing a standardized playground to test whether algorithms can learn many different tasks and generalize quickly to new ones.
+Meta-World is a well-designed, open-source simulation benchmark for multi-task and meta reinforcement learning in continuous-control robotic manipulation. It gives researchers a shared, realistic playground to test whether algorithms can _learn many different tasks_ and _generalize quickly to new ones_ — two central challenges for real-world robotics.
 
-- Paper: [Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning](https://arxiv.org/abs/1910.10897)
-- GitHub: [Farama-Foundation/Metaworld](https://github.com/Farama-Foundation/Metaworld)
-- Project website: [metaworld.farama.org](https://metaworld.farama.org)
+- 📄 [MetaWorld paper](https://arxiv.org/pdf/1910.10897)
+- 💻 [Original MetaWorld repo](https://github.com/Farama-Foundation/Metaworld)
 
 ![MetaWorld MT10 demo](https://meta-world.github.io/figures/ml45.gif)
 
-## Available tasks
+## Why Meta-World matters
 
-Meta-World provides 50 tasks organized into difficulty groups. In LeRobot, you can evaluate on individual tasks, difficulty groups, or the full MT50 suite:
+- **Diverse, realistic tasks.** Meta-World bundles a large suite of simulated manipulation tasks (50 in the MT50 suite) using everyday objects and a common tabletop Sawyer arm. This diversity exposes algorithms to a wide variety of dynamics, contacts and goal specifications while keeping a consistent control and observation structure.
+- **Focus on generalization and multi-task learning.** By evaluating across task distributions that share structure but differ in goals and objects, Meta-World reveals whether an agent truly learns transferable skills rather than overfitting to a narrow task.
+- **Standardized evaluation protocol.** It provides clear evaluation modes and difficulty splits, so different methods can be compared fairly across easy, medium, hard and very-hard regimes.
+- **Empirical insight.** Past evaluations on Meta-World show impressive progress on some fronts, but also highlight that current multi-task and meta-RL methods still struggle with large, diverse task sets. That gap points to important research directions.
 
-| Group      | CLI name             | Tasks | Description                                            |
-| ---------- | -------------------- | ----- | ------------------------------------------------------ |
-| Easy       | `easy`               | 28    | Tasks with simple dynamics and single-step goals       |
-| Medium     | `medium`             | 11    | Tasks requiring multi-step reasoning                   |
-| Hard       | `hard`               | 6     | Tasks with complex contacts and precise manipulation   |
-| Very Hard  | `very_hard`          | 5     | The most challenging tasks in the suite                |
-| MT50 (all) | Comma-separated list | 50    | All 50 tasks — the most challenging multi-task setting |
+## What it enables in LeRobot
 
-You can also pass individual task names directly (e.g., `assembly-v3`, `dial-turn-v3`).
+In LeRobot, you can evaluate any policy or vision-language-action (VLA) model on Meta-World tasks and get a clear success-rate measure. The integration is designed to be straightforward:
 
-We provide a LeRobot-ready dataset for Meta-World MT50 on the HF Hub: [lerobot/metaworld_mt50](https://huggingface.co/datasets/lerobot/metaworld_mt50). This dataset is formatted for the MT50 evaluation that uses all 50 tasks with fixed object/goal positions and one-hot task vectors for consistency.
+- We provide a LeRobot-ready dataset for Meta-World (MT50) on the HF Hub: `https://huggingface.co/datasets/lerobot/metaworld_mt50`.
+  - This dataset is formatted for the MT50 evaluation that uses all 50 tasks (the most challenging multi-task setting).
+  - MT50 gives the policy a one-hot task vector and uses fixed object/goal positions for consistency.
 
-## Installation
+- Task descriptions and the exact keys required for evaluation are available in the repo/dataset — use these to ensure your policy outputs the right success signals.
 
-After following the LeRobot installation instructions:
+## Quick start, train a SmolVLA policy on Meta-World
 
-```bash
-pip install -e ".[metaworld]"
-```
-
-<Tip warning={true}>
-If you encounter an `AssertionError: ['human', 'rgb_array', 'depth_array']` when running Meta-World environments, this is a mismatch between Meta-World and your Gymnasium version. Fix it with:
-
-```bash
-pip install "gymnasium==1.1.0"
-```
-
-</Tip>
-
-## Evaluation
-
-### Default evaluation (recommended)
-
-Evaluate on the medium difficulty split (a good balance of coverage and compute):
-
-```bash
-lerobot-eval \
-  --policy.path="your-policy-id" \
-  --env.type=metaworld \
-  --env.task=medium \
-  --eval.batch_size=1 \
-  --eval.n_episodes=10
-```
-
-### Single-task evaluation
-
-Evaluate on a specific task:
-
-```bash
-lerobot-eval \
-  --policy.path="your-policy-id" \
-  --env.type=metaworld \
-  --env.task=assembly-v3 \
-  --eval.batch_size=1 \
-  --eval.n_episodes=10
-```
-
-### Multi-task evaluation
-
-Evaluate across multiple tasks or difficulty groups:
-
-```bash
-lerobot-eval \
-  --policy.path="your-policy-id" \
-  --env.type=metaworld \
-  --env.task=assembly-v3,dial-turn-v3,handle-press-side-v3 \
-  --eval.batch_size=1 \
-  --eval.n_episodes=10
-```
-
-- `--env.task` accepts explicit task lists (comma-separated) or difficulty groups (e.g., `easy`, `medium`, `hard`, `very_hard`).
-- `--eval.batch_size` controls how many environments run in parallel.
-- `--eval.n_episodes` sets how many episodes to run per task.
-
-### Policy inputs and outputs
-
-**Observations:**
-
-- `observation.image` — single camera view (`corner2`), 480x480 HWC uint8
-- `observation.state` — 4-dim proprioceptive state (end-effector position + gripper)
-
-**Actions:**
-
-- Continuous control in `Box(-1, 1, shape=(4,))` — 3D end-effector delta + 1D gripper
-
-### Recommended evaluation episodes
-
-For reproducible benchmarking, use **10 episodes per task**. For the full MT50 suite this gives 500 total episodes. If you care about generalization, run on the full MT50 — it is intentionally challenging and reveals strengths/weaknesses better than a few narrow tasks.
-
-## Training
-
-### Example training command
-
-Train a SmolVLA policy on a subset of Meta-World tasks:
+Example command to train a SmolVLA policy on a subset of tasks:
 
 ```bash
 lerobot-train \
@@ -123,8 +44,37 @@ lerobot-train \
   --eval_freq=1000
 ```
 
+Notes:
+
+- `--env.task` accepts explicit task lists (comma separated) or difficulty groups (e.g., `env.task="hard"`).
+- Adjust `batch_size`, `steps`, and `eval_freq` to match your compute budget.
+- **Gymnasium Assertion Error**: if you encounter an error like
+  `AssertionError: ['human', 'rgb_array', 'depth_array']` when running MetaWorld environments, this comes from a mismatch between MetaWorld and your Gymnasium version.
+  We recommend using:
+
+```bash
+  pip install "gymnasium==1.1.0"
+```
+
+to ensure proper compatibility.
+
+## Quick start — evaluate a trained policy
+
+To evaluate a trained policy on the Meta-World medium difficulty split:
+
+```bash
+lerobot-eval \
+  --policy.path="your-policy-id" \
+  --env.type=metaworld \
+  --env.task=medium \
+  --eval.batch_size=1 \
+  --eval.n_episodes=2
+```
+
+This will run episodes and return per-task success rates using the standard Meta-World evaluation keys.
+
 ## Practical tips
 
-- Use the one-hot task conditioning for multi-task training (MT10/MT50 conventions) so policies have explicit task context.
+- If you care about generalization, run on the full MT50 suite — it’s intentionally challenging and reveals strengths/weaknesses better than a few narrow tasks.
+- Use the one-hot task conditioning for multi-task training (MT10 / MT50 conventions) so policies have explicit task context.
 - Inspect the dataset task descriptions and the `info["is_success"]` keys when writing post-processing or logging so your success metrics line up with the benchmark.
-- Adjust `batch_size`, `steps`, and `eval_freq` to match your compute budget.

docs/source/multi_task_dit.mdx

@@ -1,388 +0,0 @@
-# Multitask DiT Policy
-
-Multitask Diffusion Transformer (DiT) Policy is an evolution of the original Diffusion Policy architecture, which leverages a large DiT with text and vision conditioning for multitask robot learning. This implementation supports both diffusion and flow matching objectives for action generation, enabling robots to perform diverse manipulation tasks conditioned on language instructions.
-
-## Model Overview
-
-The model uses:
-
-- **CLIP Vision Encoder**: Processes RGB images from multiple camera views
-- **CLIP Text Encoder**: Encodes language task instructions (frozen weights with learnable projection)
-- **Diffusion Transformer**: Predicts action sequences conditioned on observations and language
-- **Two Objectives**: Supports both diffusion (DDPM/DDIM) and flow matching for action generation
-
-This model is exciting because you can achieve extremely high dexterity, competitive with multi-billion parameter
-VLAs, with only ~450M parameters and significantly less training.
-
-## Installation Requirements
-
-Multitask DiT Policy has additional dependencies. Install it with:
-
-```bash
-pip install lerobot[multi_task_dit]
-```
-
-This will install all necessary dependencies including the HuggingFace Transformers library for CLIP models.
-
-## Usage
-
-To use Multitask DiT in your LeRobot configuration, specify the policy type as:
-
-```python
-policy.type=multi_task_dit
-```
-
-## Training
-
-### Basic Training Command
-
-Here's a complete training command for training Multitask DiT on your dataset:
-
-```bash
-lerobot-train \
-  --dataset.repo_id=YOUR_DATASET \
-  --output_dir=./outputs/multitask_dit_training \
-  --batch_size=32 \
-  --steps=5000 \
-  --save_freq=500 \
-  --log_freq=100 \
-  --policy.type=multi_task_dit \
-  --policy.device=cuda \
-  --policy.repo_id="HF_USER/multitask-dit-your-robot" \
-  --wandb.enable=true
-```
-
-### Recommended Hyperparameters and Dataset Details (30Hz Control Frequency)
-
-For reliable performance, start with these suggested default hyperparameters:
-
-```bash
-lerobot-train \
-  --dataset.repo_id=YOUR_DATASET \
-  --output_dir=./outputs/mutitask_dit_training \
-  --batch_size=320 \
-  --steps=30000 \
-  --policy.type=multi_task_dit \
-  --policy.device=cuda \
-  --policy.horizon=32 \
-  --policy.n_action_steps=24 \
-  --policy.objective=diffusion \
-  --policy.noise_scheduler_type=DDPM \
-  --policy.num_train_timesteps=100 \
-  --policy.repo_id="HF_USER/multitask-dit-your-robot" \
-  --wandb.enable=true
-```
-
-**Key Parameters:**
-
-- **Batch Size**: 192-320 - If you have access to a GPU that can support this, you will get the best training dynamics
-- **Horizon**: 32 - number of action steps to predict, ~1.0 sec at 30Hz
-- **n_action_steps**: 24 - ~0.8 seconds at 30Hz
-- **Objective**: `diffusion` - start with diffusion and experiment with flow matching if generation quality is poor
-- **Training Steps**: >30k steps recommended for a single task
-
-### Training Configuration Parameters
-
-#### Objective Selection
-
-Choose between diffusion and flow matching:
-
-```bash
-# Diffusion objective (default)
---policy.objective=diffusion \
---policy.noise_scheduler_type=DDPM \  # or "DDIM"
---policy.num_train_timesteps=100 \
---policy.num_inference_steps=10 \  # For faster inference
---policy.beta_schedule=squaredcos_cap_v2 \  # Noise schedule type
---policy.prediction_type=epsilon \  # "epsilon" (predict noise) or "sample" (predict clean)
---policy.clip_sample=true \  # Clip samples during denoising
---policy.clip_sample_range=1.0  # Clipping range [-x, x]
-
-# Flow matching objective
---policy.objective=flow_matching \
---policy.timestep_sampling_strategy=beta \  # or "uniform" | the beta sampling strategy performance appears much better in practice
---policy.num_integration_steps=100 \
---policy.integration_method=euler \  # or "rk4"
---policy.sigma_min=0.0  # Minimum noise in flow interpolation path
-```
-
-#### Transformer Architecture
-
-Adjust model capacity based on dataset size:
-
-```bash
-# Small datasets (< 100 examples)
---policy.num_layers=4 \
---policy.hidden_dim=512 \
---policy.num_heads=8  # should ideally be hidden_dim // 64
-
-# Medium datasets (100-5k examples) - default
---policy.num_layers=6 \
---policy.hidden_dim=512 \
---policy.num_heads=8  # should ideally be hidden_dim // 64
-
-# Large datasets (> 5k examples)
---policy.num_layers=8 \
---policy.hidden_dim=512 \
---policy.num_heads=8   # should ideally be hidden_dim // 64
-```
-
-**Positional Encoding Options:**
-
-The model supports two positional encoding methods for action sequences:
-
-```bash
-# Rotary Position Embedding (RoPE) - default, recommended
---policy.use_rope=true \
---policy.rope_base=10000.0  # Base frequency for RoPE
-
-# Absolute positional encoding
---policy.use_positional_encoding=true  # Disables RoPE when true
-```
-
-**Other Transformer Parameters:**
-
-```bash
---policy.dropout=0.1  # Dropout rate for DiT blocks (0.0-1.0)
---policy.timestep_embed_dim=256  # Timestep embedding dimension
-```
-
-#### Vision Encoder Configuration
-
-```bash
-# Use different CLIP model for more expressivity at the cost of inference time
-# experiment with larger or smaller models depending on the complexity of your tasks and size of dataset
---policy.vision_encoder_name=openai/clip-vit-large-patch14
-
-# Use separate vision encoder per camera
-# This may be useful when cameras have significantly different characteristics, but
-# be wary of increased VRAM footprint.
---policy.use_separate_rgb_encoder_per_camera=true
-
-# Image preprocessing
---policy.image_resize_shape=[XXX,YYY] \ # you may need to resize your images for inference speed ups
---policy.image_crop_shape=[224,224] \
---policy.image_crop_is_random=true  # Random during training, center at inference
-```
-
-#### Text Encoder Configuration
-
-```bash
-# Use different CLIP text encoder model
-# same as vision: experiment with larger or smaller models depending on the
-# complexity of your tasks and size of dataset
---policy.text_encoder_name=openai/clip-vit-large-patch14
-```
-
-#### Learning Rate Configuration
-
-The vision encoder uses a separate learning rate multiplier, where 1/10th is suggested to be the ideal staritng point:
-
-```bash
---policy.optimizer_lr=2e-5 \
---policy.vision_encoder_lr_multiplier=0.1  # Vision encoder LR = 0.1 * optimizer_lr
-```
-
-### Training Tuning Guidelines
-
-#### 1. Flow Matching with Beta Sampling
-
-The original diffusion implementation here is based on the work described in [TRI's LBM paper](https://arxiv.org/abs/2507.05331)
-
-Additionally, we have implemented a flow-matching objective, which is described at a high-level in [Boston Dynamics blog post](https://bostondynamics.com/blog/large-behavior-models-atlas-find-new-footing/).
-
-Consider testing the flow-matching objective and evaluating performance differences for your task:
-
-```bash
---policy.objective=flow_matching \
---policy.timestep_sampling_strategy=beta \
---policy.timestep_sampling_alpha=1.5 \
---policy.timestep_sampling_beta=1.0 \
---policy.timestep_sampling_s=0.999
-```
-
-This hasn't been shown to be a silver bullet across every user case, but it occasionally results in smoother and more consistent actions.
-
-#### 2. Number of Transformer Layers
-
-Match model capacity to your dataset size:
-
-- **Small datasets** (< 100 examples): Reduce to 4 layers
-- **Large datasets** (> 5k examples): Increase to 8 layers
-
-#### 3. `horizon` Tuning
-
-The model can be sensitive to the horizon you choose. Start with around a 1 second horizon based on your control frequency:
-
-- **30 Hz frequency**: `horizon=30`
-- **10 Hz frequency**: `horizon=10`
-
-Then experiment with increasing from there. The horizon determines how far into the future the model predicts actions.
-
-#### 4. `n_action_steps` Sensitivity
-
-The model can also be very sensitive to `n_action_steps`. Start with it being around 0.8 seconds based on your control frequency and tune from there:
-
-- **Lower values**: More reactive but potentially less stable for long-horizon tasks
-- **Higher values**: Better for long-horizon execution but open-loop failures are limited in their recovery
-
-### Inference Tuning
-
-For faster inference, use DDIM with fewer sampling steps:
-
-```bash
---policy.noise_scheduler_type=DDIM \
---policy.num_inference_steps=10
-```
-
-### Resuming Training
-
-To resume training from a checkpoint:
-
-```bash
-lerobot-train \
-  --config_path=./outputs/mutitask_dit_training/checkpoints/last/pretrained_model/train_config.json \
-  --resume=true
-```
-
-The checkpoint directory should contain `model.safetensors` and `config.json` files (saved automatically during training). When resuming, the configuration is loaded from the checkpoint, so you don't need to specify other parameters.
-
-## Common Failure Modes and Debugging
-
-Training these models can be finicky. Here are common failure modes and debugging approaches:
-
-### Idling / No Motion
-
-The model may "collapse" during inference, resulting in static or no motion. This can occur when:
-
-1. **Insufficient training data**: If you only have 20-50 examples, try to roughly double your dataset size. Once you have above 300 examples, if you're still seeing this, the task may be too complex.
-
-2. **Multiple similar tasks**: When your dataset contains multiple similar tasks (e.g., picking up 2 different objects), the model may rely too heavily on language conditioning which might not be rich enough.
-
-**Debugging tips:**
-
-- Increase dataset size (double until you get to over 300 examples)
-- Train for longer, up to 100k steps, even when the loss flatlines
-- Check if the model is receiving proper language instructions or increase diversity of instruction
-
-### Executing the Wrong Task
-
-Sometimes the robot will completely ignore your instruction and perform some other task. This generally only happens if you have trained on multiple tasks.
-
-**Potential causes:**
-
-- Language instruction ambiguity
-- Insufficient task-specific training data
-- Model confusion between similar tasks in the multitask dataset
-
-**Debugging tips:**
-
-- Verify language instruction specificity, especially if descriptions are similar between multiple tasks
-- Check task distribution in your training dataset and add weighting to the failing/ignored task
-- Consider task-specific fine-tuning
-
-### Training Instability
-
-If training loss is unstable or diverging:
-
-- Try adjusting learning rate between `1e-5` and `3e-4`
-- Increase batch size if possible
-- Check that your dataset normalization is correct
-- Verify image preprocessing is working correctly
-
-## Performance Considerations
-
-### GPU Requirements
-
-- **Inference**: At least an RTX 5070 Ti (or equivalent GPU) is recommended for reasonable speed performance
-- **Training**: A GPU with enough VRAM to load batch sizes of >64 is ideal, which will vary depending on the number of image observations, etc
-
-### Batch Size Recommendations
-
-- **Minimum**: 64 (less than this may result in unstable training)
-- **Recommended**: 256-320 (best performance, requires larger GPU)
-
-## Example: Training on Custom Dataset
-
-Here's a complete example training on a custom dataset:
-
-```bash
-lerobot-train \
-  --dataset.repo_id=YOUR_DATASET \
-  --output_dir=./outputs/mutitask_dit_training \
-  --batch_size=320 \
-  --steps=30000 \
-  --save_freq=1000 \
-  --log_freq=100 \
-  --eval_freq=1000 \
-  --policy.type=multi_task_dit \
-  --policy.device=cuda \
-  --policy.horizon=32 \
-  --policy.n_action_steps=24 \
-  --policy.objective=diffusion \
-  --policy.noise_scheduler_type=DDPM \
-  --policy.num_layers=6 \
-  --policy.hidden_dim=512 \
-  --policy.vision_encoder_name=openai/clip-vit-base-patch16 \
-  --policy.image_resize_shape=[320,240] \
-  --policy.image_crop_shape=[224,224] \
-  --policy.repo_id="HF_USER/multitask-dit-your-robot" \
-  --wandb.enable=true \
-  --wandb.project=multitask_dit
-```
-
-## Libero Results
-
-```
-python -m lerobot.scripts.lerobot_train \
-  --dataset.repo_id=HuggingFaceVLA/libero \
-  --policy.type=multi_task_dit \
-  --policy.push_to_hub=false \
-  --output_dir="./outputs/multitask_dit_libero" \
-  --job_name="multitask-dit-libero" \
-  --wandb.enable=true \
-  --wandb.project=multitask_dit_libero \
-  --dataset.image_transforms.enable=true \
-  --dataset.image_transforms.max_num_transforms=4 \
-  --dataset.image_transforms.tfs='{"brightness":{"type":"ColorJitter","kwargs":{"brightness":[0.75,1.25]}},"contrast":{"type":"ColorJitter","kwargs":{"contrast":[0.6,1.4]}},"saturation":{"type":"ColorJitter","kwargs":{"saturation":[0.8,1.2]}},"hue":{"type":"ColorJitter","kwargs":{"hue":[-0.05,0.05]}},"sharpness":{"type":"SharpnessJitter","kwargs":{"sharpness":[0.6,1.4]}},"rotation":{"type":"RandomRotation","kwargs":{"degrees":[-5,5]}},"translation":{"type":"RandomAffine","kwargs":{"degrees":0,"translate":[0.1,0.1]}}}' \
-  --dataset.video_backend=torchcodec \
-  --policy.use_amp=true \
-  --policy.horizon=48 \
-  --policy.n_obs_steps=2 \
-  --policy.use_rope=true \
-  --policy.use_positional_encoding=false \
-  --policy.hidden_dim=768 \
-  --policy.num_layers=8 \
-  --policy.num_heads=12 \
-  --policy.dropout=0.1 \
-  --policy.timestep_embed_dim=256 \
-  --policy.objective=diffusion \
-  --policy.optimizer_lr=3e-4 \
-  --policy.optimizer_weight_decay=0 \
-  --policy.scheduler_warmup_steps=0 \
-  --policy.vision_encoder_name=openai/clip-vit-base-patch16 \
-  --policy.image_resize_shape=[256,256] \
-  --policy.image_crop_is_random=true \
-  --policy.text_encoder_name=openai/clip-vit-base-patch16 \
-  --policy.vision_encoder_lr_multiplier=0.1 \
-  --policy.device=cuda \
-  --num_workers=8 \
-  --save_freq=4000 \
-  --log_freq=100 \
-  --steps=100000 \
-  --batch_size=320
-```
-
-Results:
-
-| LIBERO Spatial | LIBERO Object | LIBERO Goal | LIBERO 10 | Average |
-| -------------- | ------------- | ----------- | --------- | ------- |
-| 87.0           | 98.2          | 93.8        | 83.2      | 90.6    |
-
-## References
-
-For more details on the technical implementation and architecture, see:
-
-- [A Careful Examination of Large Behavior Models for Multitask Dexterous Manipulation](https://arxiv.org/abs/2507.05331)
-- [Large Behavior Models and Atlas Find New Footing](https://bostondynamics.com/blog/large-behavior-models-atlas-find-new-footing/)
-- [Dissecting and Open-Sourcing Multitask Diffusion Transformer Policy](https://brysonkjones.substack.com/p/dissecting-and-open-sourcing-multitask-diffusion-transformer-policy)

docs/source/omx.mdx

@@ -1,197 +0,0 @@
-## Order and Assemble the parts
-
-First, assemble the OMX hardware following the official assembly guide.
-
-OMX Assembly Guide: https://ai.robotis.com/omx/assembly_guide_omx.html
-
-OMX robots are shipped preconfigured from the factory. Motor IDs, communication parameters, and joint offsets are already set, so no additional motor setup or calibration is required before using LeRobot.
-
-## Install LeRobot 🤗
-
-To install LeRobot, follow our [Installation Guide](./installation)
-
-In addition to these instructions, you need to install the Dynamixel SDK:
-
-```bash
-pip install -e ".[dynamixel]"
-```
-
-## Connect the robot
-
-To find the port for each bus servo adapter, run this script:
-
-```bash
-lerobot-find-port
-```
-
-This command runs and when prompted, disconnect the USB cable from either the leader or follower arm and press Enter. The output will show 'The port of this MotorsBus is [port]'. This identifies the port for the disconnected arm. Repeat for the other arm to identify both ports.
-
-<hfoptions id="find_port">
-<hfoption id="Mac">
-
-Example output on macOS:
-
-```
-Finding all available ports for the MotorBus.
-['/dev/tty.usbmodem575E0032081', '/dev/tty.usbmodem575E0031751']
-Remove the USB cable from your MotorsBus and press Enter when done.
-
-[...Disconnect corresponding leader or follower arm and press Enter...]
-
-The port of this MotorsBus is /dev/tty.usbmodem575E0032081
-Reconnect the USB cable.
-```
-
-Where the found port is: `/dev/tty.usbmodem575E0032081` corresponding to your leader or follower arm.
-
-</hfoption>
-<hfoption id="Linux">
-
-On Linux, we strongly recommend using udev rules to assign persistent and human-readable device names to the OMX leader and follower arms. This avoids issues where device names such as ttyACM0 and ttyACM1 change when the robot is unplugged, replugged, or when the system is rebooted.
-
-#### 1. Find your device serial numbers
-
-You should have obtained the port numbers like ../../ttyACM? for the leader and follower using `lerobot-find-port`. You can match those results with the serial numbers using the `ls -l /dev/serial/by-id/` command.
-To create udev rules, you need the unique serial number for each OMX device. The easiest way is to list devices under:
-
-```bash
-ls -l /dev/serial/by-id/
-```
-
-You will see output similar to:
-
-```bash
-usb-ROBOTIS_OpenRB-150_228BDD7B503059384C2E3120FF0A2B19-if00 -> ../../ttyACM0
-usb-ROBOTIS_OpenRB-150_67E1ED68503059384C2E3120FF092234-if00 -> ../../ttyACM1
-```
-
-In each line, the serial number is the long string after `usb-ROBOTIS_OpenRB-150_` and before `-if00`.
-
-Follower serial: `228BDD7B503059384C2E3120FF0A2B19`
-
-Leader serial: `67E1ED68503059384C2E3120FF092234`
-
-#### 2. Create the udev rule
-
-Create a new udev rule file:
-
-```bash
-sudo nano /etc/udev/rules.d/99-omx.rules
-```
-
-Paste the following lines, replacing the serial numbers with the values you found above:
-
-```bash
-SUBSYSTEM=="tty", ATTRS{idVendor}=="0403", ATTRS{serial}=="228BDD7B503059384C2E3120FF0A2B19", SYMLINK+="omx_follower"
-SUBSYSTEM=="tty", ATTRS{idVendor}=="0403", ATTRS{serial}=="67E1ED68503059384C2E3120FF092234", SYMLINK+="omx_leader"
-```
-
-Save the file and reload udev rules:
-
-```bash
-sudo udevadm control --reload-rules
-sudo udevadm trigger
-```
-
-Now unplug and replug both devices once.
-
-#### 3. Verify the symlinks
-
-Check that the persistent device names exist:
-
-```bash
-ls -l /dev/omx_follower /dev/omx_leader
-```
-
-You should see them pointing to ttyACM\* devices:
-
-```bash
-/dev/omx_follower -> ttyACM*
-/dev/omx_leader   -> ttyACM*
-```
-
-These names remain stable across reboots and reconnections.
-
-</hfoption>
-</hfoptions>
-
-## Teleoperate
-
-After identifying the correct ports, you can directly teleoperate the follower arm using the leader arm.
-
-<hfoptions id="teleoperate">
-<hfoption id="Mac">
-
-### Teleoperate without camera
-
-```bash
-lerobot-teleoperate \
-  --robot.type=omx_follower \
-  --robot.port=<your_follower_port> \
-  --robot.id=omx_follower_arm \
-  --teleop.type=omx_leader \
-  --teleop.port=<your_leader_port> \
-  --teleop.id=omx_leader_arm
-```
-
-During teleoperation, motions of the leader arm are mirrored in real time by the follower arm. OMX is already preconfigured, teleoperation can begin immediately without any calibration steps.
-
-### Teleoperate with camera
-
-You can also enable camera input during teleoperation by providing a camera configuration for the follower arm.
-
-```bash
-lerobot-teleoperate \
-  --robot.type=omx_follower \
-  --robot.port=<your_follower_port> \
-  --robot.id=omx_follower_arm \
-  --robot.cameras="{front: {type: opencv, index_or_path: '/dev/video0', width: 640, height: 480, fps: 30}}" \
-  --teleop.type=omx_leader \
-  --teleop.port=<your_leader_port> \
-  --teleop.id=omx_leader_arm \
-  --display_data=true
-```
-
-When the camera is enabled, the camera stream is displayed in real time and synchronized with the robot state. This setup is useful for visual monitoring and can be reused later for demonstration recording and imitation learning.
-
-</hfoption>
-<hfoption id="Linux">
-
-### Teleoperate without camera
-
-```bash
-lerobot-teleoperate \
-  --robot.type=omx_follower \
-  --robot.port=/dev/omx_follower \
-  --robot.id=omx_follower_arm \
-  --teleop.type=omx_leader \
-  --teleop.port=/dev/omx_leader \
-  --teleop.id=omx_leader_arm
-```
-
-During teleoperation, motions of the leader arm are mirrored in real time by the follower arm. OMX is already preconfigured, teleoperation can begin immediately without any calibration steps.
-
-### Teleoperate with camera
-
-You can also enable camera input during teleoperation by providing a camera configuration for the follower arm.
-
-```bash
-lerobot-teleoperate \
-  --robot.type=omx_follower \
-  --robot.port=/dev/omx_follower \
-  --robot.id=omx_follower_arm \
-  --robot.cameras="{front: {type: opencv, index_or_path: '/dev/video0', width: 640, height: 480, fps: 30}}" \
-  --teleop.type=omx_leader \
-  --teleop.port=/dev/omx_leader \
-  --teleop.id=omx_leader_arm \
-  --display_data=true
-```
-
-When the camera is enabled, the camera stream is displayed in real time and synchronized with the robot state. This setup is useful for visual monitoring and can be reused later for demonstration recording and imitation learning.
-
-</hfoption>
-</hfoptions>
-
-Congrats 🎉, your robot is all set to learn a task on its own.
-
-> If you have any questions or need help, please reach out on [Discord](https://discord.com/invite/robotis).

docs/source/openarm.mdx

@@ -1,276 +0,0 @@
-# OpenArm
-
-[OpenArm](https://openarm.dev) is an open-source 7DOF humanoid arm designed for physical AI research and deployment.
-
-To get your OpenArm, assembled or DIY, and join the global community, browse verified and certified manufacturers worldwide at [openarm.dev](https://openarm.dev).
-
-## What's Unique?
-
-- **Human-Scale Design**: OpenArm is designed with human-like proportions, scaled for a person around 160-165cm tall. This provides an optimal balance between practical reach and manageable inertia for safe, responsive operation.
-
-- **Safety-First Architecture**: Built with QDD backdrivable motors and high compliance, OpenArm prioritizes safe human-robot interaction while maintaining practical payload capabilities (6.0kg peak / 4.1kg nominal) for real-world tasks.
-
-- **Built for Durability**: Critical structural components use aluminum and stainless steel construction, ensuring robust performance for repetitive data collection and continuous research use.
-
-- **Fully Accessible & Buildable**: Every component, from CNC parts and 3D-printed casings to electrical wiring is designed to be purchasable and buildable by individual researchers and labs, with complete fabrication data provided.
-
-- **Practical & Affordable**: At $6,500 USD for a complete bimanual system, OpenArm delivers research-grade capabilities at a fraction of traditional humanoid robot costs.
-
-## Platform Requirements
-
-<Tip warning={true}>
-  **Linux Only**: OpenArm currently only works on Linux. The CAN bus USB adapter
-  does not have macOS drivers and has not been tested on Windows.
-</Tip>
-
-## Safety Guide
-
-Before operating OpenArm, please read the [official safety guide](https://docs.openarm.dev/getting-started/safety-guide). Key points:
-
-- **Secure installation**: Fasten the arm to a flat, stable surface with screws or clamps
-- **Safe distance**: Keep body parts and objects outside the range of motion during operation
-- **Protective equipment**: Always wear safety goggles; use additional PPE as needed
-- **Payload limits**: Do not exceed specified payload limits (6.0kg peak / 4.1kg nominal per arm)
-- **Emergency stop**: Know the location and operation of the emergency stop device
-- **Regular inspection**: Check for loose screws, damaged mechanical limits, unusual noises, and wiring damage
-
-## Hardware Setup
-
-Follow the official [OpenArm hardware documentation](https://docs.openarm.dev) for:
-
-- Bill of materials and sourcing
-- 3D printing instructions
-- Mechanical assembly
-- Electrical wiring
-
-The hardware repositories are available at [github.com/enactic/openarm](https://github.com/enactic/openarm).
-
-## CAN Bus Setup
-
-OpenArm uses CAN bus communication with Damiao motors. Once you have the CAN bus USB adapter plugged into your Linux PC, follow the [Damiao Motors and CAN Bus guide](./damiao) to configure the interface.
-
-Quick setup:
-
-```bash
-# Setup CAN interfaces
-lerobot-setup-can --mode=setup --interfaces=can0,can1
-
-# Test motor communication
-lerobot-setup-can --mode=test --interfaces=can0,can1
-```
-
-## Install LeRobot 🤗
-
-Follow our [Installation Guide](./installation), then install the Damiao motor support:
-
-```bash
-pip install -e ".[damiao]"
-```
-
-## Usage
-
-### Follower Arm (Robot)
-
-<hfoptions id="follower">
-<hfoption id="Command">
-
-```bash
-lerobot-calibrate \
-    --robot.type=openarm_follower \
-    --robot.port=can0 \
-    --robot.side=right \
-    --robot.id=my_openarm_follower
-```
-
-</hfoption>
-<hfoption id="API example">
-
-```python
-from lerobot.robots.openarm_follower import OpenArmFollower, OpenArmFollowerConfig
-
-config = OpenArmFollowerConfig(
-    port="can0",
-    side="right",  # or "left" for left arm
-    id="my_openarm_follower",
-)
-
-follower = OpenArmFollower(config)
-follower.connect()
-
-# Read current state
-obs = follower.get_observation()
-print(obs)
-
-# Send action (position in degrees)
-action = {
-    "joint_1.pos": 0.0,
-    "joint_2.pos": 0.0,
-    "joint_3.pos": 0.0,
-    "joint_4.pos": 45.0,
-    "joint_5.pos": 0.0,
-    "joint_6.pos": 0.0,
-    "joint_7.pos": 0.0,
-    "gripper.pos": 0.0,
-}
-follower.send_action(action)
-
-follower.disconnect()
-```
-
-</hfoption>
-</hfoptions>
-
-### Leader Arm (Teleoperator)
-
-The leader arm is used for teleoperation - manually moving it to control the follower arm.
-
-<hfoptions id="leader">
-<hfoption id="Command">
-
-```bash
-lerobot-calibrate \
-    --teleop.type=openarm_leader \
-    --teleop.port=can1 \
-    --teleop.id=my_openarm_leader
-```
-
-</hfoption>
-<hfoption id="API example">
-
-```python
-from lerobot.teleoperators.openarm_leader import OpenArmLeader, OpenArmLeaderConfig
-
-config = OpenArmLeaderConfig(
-    port="can1",
-    id="my_openarm_leader",
-    manual_control=True,  # Disable torque for manual movement
-)
-
-leader = OpenArmLeader(config)
-leader.connect()
-
-# Read current position (as action to send to follower)
-action = leader.get_action()
-print(action)
-
-leader.disconnect()
-```
-
-</hfoption>
-</hfoptions>
-
-### Teleoperation
-
-To teleoperate OpenArm with leader-follower control:
-
-```bash
-lerobot-teleoperate \
-    --robot.type=openarm_follower \
-    --robot.port=can0 \
-    --robot.side=right \
-    --robot.id=my_follower \
-    --teleop.type=openarm_leader \
-    --teleop.port=can1 \
-    --teleop.id=my_leader
-```
-
-### Bimanual Teleoperation
-
-To teleoperate a bimanual OpenArm setup with two leader and two follower arms:
-
-```bash
-lerobot-teleoperate \
-    --robot.type=bi_openarm_follower \
-    --robot.left_arm_config.port=can0 \
-    --robot.left_arm_config.side=left \
-    --robot.right_arm_config.port=can1 \
-    --robot.right_arm_config.side=right \
-    --robot.id=my_bimanual_follower \
-    --teleop.type=bi_openarm_leader \
-    --teleop.left_arm_config.port=can2 \
-    --teleop.right_arm_config.port=can3 \
-    --teleop.id=my_bimanual_leader
-```
-
-### Recording Data
-
-To record a dataset during teleoperation:
-
-```bash
-lerobot-record \
-    --robot.type=openarm_follower \
-    --robot.port=can0 \
-    --robot.side=right \
-    --robot.id=my_follower \
-    --teleop.type=openarm_leader \
-    --teleop.port=can1 \
-    --teleop.id=my_leader \
-    --repo-id=my_hf_username/my_openarm_dataset \
-    --fps=30 \
-    --num-episodes=10
-```
-
-## Configuration Options
-
-### Follower Configuration
-
-| Parameter             | Default   | Description                                                |
-| --------------------- | --------- | ---------------------------------------------------------- |
-| `port`                | -         | CAN interface (e.g., `can0`)                               |
-| `side`                | `None`    | Arm side: `"left"`, `"right"`, or `None` for custom limits |
-| `use_can_fd`          | `True`    | Enable CAN FD for higher data rates                        |
-| `can_bitrate`         | `1000000` | Nominal bitrate (1 Mbps)                                   |
-| `can_data_bitrate`    | `5000000` | CAN FD data bitrate (5 Mbps)                               |
-| `max_relative_target` | `None`    | Safety limit for relative target positions                 |
-| `position_kp`         | Per-joint | Position control proportional gains                        |
-| `position_kd`         | Per-joint | Position control derivative gains                          |
-
-### Leader Configuration
-
-| Parameter          | Default   | Description                         |
-| ------------------ | --------- | ----------------------------------- |
-| `port`             | -         | CAN interface (e.g., `can1`)        |
-| `manual_control`   | `True`    | Disable torque for manual movement  |
-| `use_can_fd`       | `True`    | Enable CAN FD for higher data rates |
-| `can_bitrate`      | `1000000` | Nominal bitrate (1 Mbps)            |
-| `can_data_bitrate` | `5000000` | CAN FD data bitrate (5 Mbps)        |
-
-## Motor Configuration
-
-OpenArm uses Damiao motors with the following default configuration:
-
-| Joint                       | Motor Type | Send ID | Recv ID |
-| --------------------------- | ---------- | ------- | ------- |
-| joint_1 (Shoulder pan)      | DM8009     | 0x01    | 0x11    |
-| joint_2 (Shoulder lift)     | DM8009     | 0x02    | 0x12    |
-| joint_3 (Shoulder rotation) | DM4340     | 0x03    | 0x13    |
-| joint_4 (Elbow flex)        | DM4340     | 0x04    | 0x14    |
-| joint_5 (Wrist roll)        | DM4310     | 0x05    | 0x15    |
-| joint_6 (Wrist pitch)       | DM4310     | 0x06    | 0x16    |
-| joint_7 (Wrist rotation)    | DM4310     | 0x07    | 0x17    |
-| gripper                     | DM4310     | 0x08    | 0x18    |
-
-## Troubleshooting
-
-### No Response from Motors
-
-1. Check power supply connections
-2. Verify CAN wiring (CAN-H, CAN-L, GND)
-3. Run diagnostics: `lerobot-setup-can --mode=test --interfaces=can0`
-4. See the [Damiao troubleshooting guide](./damiao#troubleshooting) for more details
-
-### CAN Interface Not Found
-
-Ensure the CAN interface is configured:
-
-```bash
-ip link show can0
-```
-
-## Resources
-
-- [OpenArm Website](https://openarm.dev)
-- [OpenArm Documentation](https://docs.openarm.dev)
-- [OpenArm GitHub](https://github.com/enactic/openarm)
-- [Safety Guide](https://docs.openarm.dev/getting-started/safety-guide)
-- [Damiao Motors and CAN Bus](./damiao)

docs/source/phone_teleop.mdx

@@ -44,7 +44,7 @@ Modify the examples to use `PhoneOS.IOS` or `PhoneOS.ANDROID` in `PhoneConfig`.
 
 Teleoperation example:
 
-```python
+```36:43:examples/phone_so100_teleop.py
 from lerobot.teleoperators.phone.config_phone import PhoneConfig, PhoneOS
 
 teleop_config = PhoneConfig(phone_os=PhoneOS.IOS)  # or PhoneOS.ANDROID
@@ -66,13 +66,12 @@ Run on of the examples scripts to teleoperate, record a dataset, replay a datase
 
 All scripts assume you configured your robot (e.g., SO-100 follower) and set the correct serial port.
 
-Additionally you need to **copy the URDF of the robot into the examples folder**. For the examples in this tutorial (using SO100/SO101), copy the `SO101` folder from the [SO-ARM100 repo](https://github.com/TheRobotStudio/SO-ARM100/blob/main/Simulation/SO101) into the `examples/phone_to_so100/` directory, so that the URDF file path becomes `examples/phone_to_so100/SO101/so101_new_calib.urdf`.
+Additionally you need to **copy the urdf of the robot to the examples folder**. For the examples in this tutorial (Using SO100/SO101) it is highly recommended to use the urdf in the [SO-ARM100 repo](https://github.com/TheRobotStudio/SO-ARM100/blob/main/Simulation/SO101/so101_new_calib.urdf)
 
 - Run this example to teleoperate:
 
   ```bash
-  cd examples/phone_to_so100
-  python teleoperate.py
+  python examples/phone_to_so100/teleoperate.py
   ```
 
 After running the example:
@@ -85,29 +84,26 @@ Additionally you can customize mapping or safety limits by editing the processor
 - Run this example to record a dataset, which saves absolute end effector observations and actions:
 
   ```bash
-  cd examples/phone_to_so100
-  python record.py
+  python examples/phone_to_so100/record.py
   ```
 
 - Run this example to replay recorded episodes:
 
   ```bash
-  cd examples/phone_to_so100
-  python replay.py
+  python examples/phone_to_so100/replay.py
   ```
 
 - Run this example to evaluate a pretrained policy:
 
   ```bash
-  cd examples/phone_to_so100
-  python evaluate.py
+  python examples/phone_to_so100/evaluate.py
   ```
 
 ### Important pipeline steps and options
 
 - Kinematics are used in multiple steps. We use [Placo](https://github.com/Rhoban/placo) which is a wrapper around Pinocchio for handling our kinematics. We construct the kinematics object by passing the robot's URDF and target frame. We set `target_frame_name` to the gripper frame.
 
-  ```python
+  ```examples/phone_to_so100/teleoperate.py
   kinematics_solver = RobotKinematics(
     urdf_path="./SO101/so101_new_calib.urdf",
     target_frame_name="gripper_frame_link",
@@ -118,7 +114,7 @@ Additionally you can customize mapping or safety limits by editing the processor
 
 - The `MapPhoneActionToRobotAction` step converts the calibrated phone pose and inputs into target deltas and gripper commands, below is shown what the step outputs.
 
-  ```python
+  ```src/lerobot/teleoperators/phone/phone_processor.py
   action["enabled"] = enabled
         action["target_x"] = -pos[1] if enabled else 0.0
         action["target_y"] = pos[0] if enabled else 0.0
@@ -131,7 +127,7 @@ Additionally you can customize mapping or safety limits by editing the processor
 
 - The `EEReferenceAndDelta` step converts target deltas to an absolute desired EE pose, storing a reference on enable, the `end_effector_step_sizes` are the step sizes for the EE pose and can be modified to change the motion speed.
 
-  ```python
+  ```examples/phone_to_so100/teleoperate.py
   EEReferenceAndDelta(
       kinematics=kinematics_solver,
       end_effector_step_sizes={"x": 0.5, "y": 0.5, "z": 0.5},
@@ -142,7 +138,7 @@ Additionally you can customize mapping or safety limits by editing the processor
 
 - The `EEBoundsAndSafety` step clamps EE motion to a workspace and checks for large ee step jumps to ensure safety. The `end_effector_bounds` are the bounds for the EE pose and can be modified to change the workspace. The `max_ee_step_m` are the step limits for the EE pose and can be modified to change the safety limits.
 
-  ```python
+  ```examples/phone_to_so100/teleoperate.py
   EEBoundsAndSafety(
       end_effector_bounds={"min": [-1.0, -1.0, -1.0], "max": [1.0, 1.0, 1.0]},
       max_ee_step_m=0.10,
@@ -151,7 +147,7 @@ Additionally you can customize mapping or safety limits by editing the processor
 
 - The `GripperVelocityToJoint` step turns a velocity‑like gripper input into absolute gripper position using the current measured state. The `speed_factor` is the factor by which the velocity is multiplied.
 
-  ```python
+  ```examples/phone_to_so100/teleoperate.py
   GripperVelocityToJoint(speed_factor=20.0)
   ```
 
@@ -161,7 +157,7 @@ We use different IK initial guesses in the kinematic steps. As initial guess eit
 
 - Closed loop (used in record/eval): sets `initial_guess_current_joints=True` so IK starts from the measured joints each frame.
 
-  ```python
+  ```examples/phone_to_so100/record.py
   InverseKinematicsEEToJoints(
       kinematics=kinematics_solver,
       motor_names=list(robot.bus.motors.keys()),
@@ -171,7 +167,7 @@ We use different IK initial guesses in the kinematic steps. As initial guess eit
 
 - Open loop (used in replay): sets `initial_guess_current_joints=False` so IK continues from the previous IK solution rather than the measured state. This preserves action stability when we replay without feedback.
 
-  ```python
+  ```examples/phone_to_so100/replay.py
   InverseKinematicsEEToJoints(
       kinematics=kinematics_solver,
       motor_names=list(robot.bus.motors.keys()),

docs/source/pi0.mdx

@@ -6,12 +6,6 @@
 
 π₀ represents a breakthrough in robotics as the first general-purpose robot foundation model developed by [Physical Intelligence](https://www.physicalintelligence.company/blog/pi0). Unlike traditional robot programs that are narrow specialists programmed for repetitive motions, π₀ is designed to be a generalist policy that can understand visual inputs, interpret natural language instructions, and control a variety of different robots across diverse tasks.
 
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/lerobot-pi0%20(1).png"
-  alt="An overview of Pi0"
-  width="85%"
-/>
-
 ### The Vision for Physical Intelligence
 
 As described by Physical Intelligence, while AI has achieved remarkable success in digital domains, from chess-playing to drug discovery, human intelligence still dramatically outpaces AI in the physical world. To paraphrase Moravec's paradox, winning a game of chess represents an "easy" problem for AI, but folding a shirt or cleaning up a table requires solving some of the most difficult engineering problems ever conceived. π₀ represents a first step toward developing artificial physical intelligence that enables users to simply ask robots to perform any task they want, just like they can with large language models.
@@ -34,6 +28,11 @@ As described by Physical Intelligence, while AI has achieved remarkable success
    pip install -e ".[pi]"
    ```
 
+   > [!NOTE]
+   > For lerobot 0.4.0, if you want to install pi tag, you will have to do: `pip install "lerobot[pi]@git+https://github.com/huggingface/lerobot.git"`.
+   >
+   > This will be solved in the next patch release
+
 ## Training Data and Capabilities
 
 π₀ is trained on the largest robot interaction dataset to date, combining three key data sources:
@@ -55,7 +54,7 @@ policy.type=pi0
 For training π₀, you can use the standard LeRobot training script with the appropriate configuration:
 
 ```bash
-lerobot-train \
+python src/lerobot/scripts/lerobot_train.py \
     --dataset.repo_id=your_dataset \
     --policy.type=pi0 \
     --output_dir=./outputs/pi0_training \
@@ -65,8 +64,6 @@ lerobot-train \
     --policy.compile_model=true \
     --policy.gradient_checkpointing=true \
     --policy.dtype=bfloat16 \
-    --policy.freeze_vision_encoder=false \
-    --policy.train_expert_only=false \
     --steps=3000 \
     --policy.device=cuda \
     --batch_size=32
@@ -82,55 +79,6 @@ lerobot-train \
   - [lerobot/pi0_base](https://huggingface.co/lerobot/pi0_base)
   - [lerobot/pi0_libero](https://huggingface.co/lerobot/pi0_libero) (specifically trained on the Libero dataset)
 
-### Training Parameters Explained
-
-| Parameter               | Default | Description                                 |
-| ----------------------- | ------- | ------------------------------------------- |
-| `freeze_vision_encoder` | `false` | Do not freeze the vision encoder            |
-| `train_expert_only`     | `false` | Do not freeze the VLM, train all parameters |
-
-**💡 Tip**: Setting `train_expert_only=true` freezes the VLM and trains only the action expert and projections, allowing finetuning with reduced memory usage.
-
-## Relative Actions
-
-By default, π₀ predicts absolute actions. You can enable **relative actions** so the model predicts offsets relative to the current robot state. This can improve training stability for certain setups.
-
-To use relative actions, first recompute your dataset stats in relative space via the CLI:
-
-```bash
-lerobot-edit-dataset \
-    --repo_id your_dataset \
-    --operation.type recompute_stats \
-    --operation.relative_action true \
-    --operation.chunk_size 50 \
-    --operation.relative_exclude_joints "['gripper']" \
-    --push_to_hub true
-```
-
-Or equivalently in Python:
-
-```python
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
-from lerobot.datasets.dataset_tools import recompute_stats
-
-dataset = LeRobotDataset("your_dataset")
-recompute_stats(dataset, relative_action=True, chunk_size=50, relative_exclude_joints=["gripper"])
-dataset.push_to_hub()
-```
-
-The `chunk_size` should match your policy's `chunk_size` (default 50 for π₀). `relative_exclude_joints` lists joint names that should remain in absolute space (e.g. gripper commands). Use `--push_to_hub true` to upload the updated stats to the Hub.
-
-Then train with relative actions enabled:
-
-```bash
-lerobot-train \
-    --dataset.repo_id=your_dataset \
-    --policy.type=pi0 \
-    --policy.use_relative_actions=true \
-    --policy.relative_exclude_joints='["gripper"]' \
-    ...
-```
-
 ## License
 
 This model follows the **Apache 2.0 License**, consistent with the original [OpenPI repository](https://github.com/Physical-Intelligence/openpi).

docs/source/pi05.mdx

@@ -36,6 +36,11 @@ This diverse training mixture creates a "curriculum" that enables generalization
    pip install -e ".[pi]"
    ```
 
+   > [!NOTE]
+   > For lerobot 0.4.0, if you want to install pi tag, you will have to do: `pip install "lerobot[pi]@git+https://github.com/huggingface/lerobot.git"`.
+   >
+   > This will be solved in the next patch release
+
 ## Usage
 
 To use π₀.₅ in your LeRobot configuration, specify the policy type as:
@@ -51,7 +56,7 @@ policy.type=pi05
 Here's a complete training command for finetuning the base π₀.₅ model on your own dataset:
 
 ```bash
-lerobot-train \
+python src/lerobot/scripts/lerobot_train.py\
     --dataset.repo_id=your_dataset \
     --policy.type=pi05 \
     --output_dir=./outputs/pi05_training \
@@ -62,8 +67,6 @@ lerobot-train \
     --policy.gradient_checkpointing=true \
     --wandb.enable=true \
     --policy.dtype=bfloat16 \
-    --policy.freeze_vision_encoder=false \
-    --policy.train_expert_only=false \
     --steps=3000 \
     --policy.device=cuda \
     --batch_size=32
@@ -79,15 +82,6 @@ lerobot-train \
   - [lerobot/pi05_base](https://huggingface.co/lerobot/pi05_base)
   - [lerobot/pi05_libero](https://huggingface.co/lerobot/pi05_libero) (specifically trained on the Libero dataset)
 
-### Training Parameters Explained
-
-| Parameter               | Default | Description                                 |
-| ----------------------- | ------- | ------------------------------------------- |
-| `freeze_vision_encoder` | `false` | Do not freeze the vision encoder            |
-| `train_expert_only`     | `false` | Do not freeze the VLM, train all parameters |
-
-**💡 Tip**: Setting `train_expert_only=true` freezes the VLM and trains only the action expert and projections, allowing finetuning with reduced memory usage.
-
 If your dataset is not converted with `quantiles`, you can convert it with the following command:
 
 ```bash
@@ -97,46 +91,6 @@ python src/lerobot/datasets/v30/augment_dataset_quantile_stats.py \
 
 Or train pi05 with this normalization mapping: `--policy.normalization_mapping='{"ACTION": "MEAN_STD", "STATE": "MEAN_STD", "VISUAL": "IDENTITY"}'`
 
-## Relative Actions
-
-By default, π₀.₅ predicts absolute actions. You can enable **relative actions** so the model predicts offsets relative to the current robot state. This can improve training stability for certain setups.
-
-To use relative actions, first recompute your dataset stats in relative space via the CLI:
-
-```bash
-lerobot-edit-dataset \
-    --repo_id your_dataset \
-    --operation.type recompute_stats \
-    --operation.relative_action true \
-    --operation.chunk_size 50 \
-    --operation.relative_exclude_joints "['gripper']" \
-    --push_to_hub true
-```
-
-Or equivalently in Python:
-
-```python
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
-from lerobot.datasets.dataset_tools import recompute_stats
-
-dataset = LeRobotDataset("your_dataset")
-recompute_stats(dataset, relative_action=True, chunk_size=50, relative_exclude_joints=["gripper"])
-dataset.push_to_hub()
-```
-
-The `chunk_size` should match your policy's `chunk_size` (default 50 for π₀.₅). `relative_exclude_joints` lists joint names that should remain in absolute space (e.g. gripper commands). Use `--push_to_hub true` to upload the updated stats to the Hub.
-
-Then train with relative actions enabled:
-
-```bash
-lerobot-train \
-    --dataset.repo_id=your_dataset \
-    --policy.type=pi05 \
-    --policy.use_relative_actions=true \
-    --policy.relative_exclude_joints='["gripper"]' \
-    ...
-```
-
 ## Performance Results
 
 ### Libero Benchmark Results

docs/source/pi0fast.mdx

@@ -1,241 +0,0 @@
-# π₀-FAST (Pi0-FAST)
-
-π₀-FAST is a **Vision-Language-Action model for general robot control** that uses autoregressive next-token prediction to model continuous robot actions.
-
-## Model Overview
-
-π₀-FAST combines the power of Vision-Language Models with a novel action tokenization approach called **FAST (Frequency-space Action Sequence Tokenization)**. This enables training autoregressive VLAs on highly dexterous tasks that are impossible with standard binning-based discretization, while training **up to 5x faster** than diffusion-based approaches like π₀.
-
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/lerobot-pifast.png"
-  alt="An overview of Pi0-FAST"
-  width="85%"
-/>
-
-### Why FAST?
-
-Standard approaches for robot action tokenization use simple per-dimension, per-timestep binning schemes. While passable for simple behaviors, this rapidly breaks down for complex and dexterous skills that require precision and high-frequency control.
-
-FAST solves this by compressing action sequences using signal processing techniques, resulting in a dense sequence of action tokens that can be predicted autoregressively—just like language tokens.
-
-### How FAST Tokenization Works
-
-The FAST tokenizer compresses action sequences through the following steps:
-
-1. **Normalize**: Take a continuous action chunk of shape `(H, D)` where `H` is the horizon and `D` is the action dimension. Normalize using one of the supported normalization methods (Quantiles recommended to handle outliers).
-
-2. **Discrete Cosine Transform (DCT)**: Apply DCT (via scipy) to each action dimension separately. DCT is a compression algorithm commonly used in image and audio codecs (JPEG, MP3).
-
-3. **Quantization**: Round and remove insignificant coefficients for each action dimension, producing a sparse frequency matrix.
-
-4. **Flatten**: Flatten the matrix into a 1D vector, with low-frequency components first.
-
-5. **Byte Pair Encoding (BPE)**: Train a BPE tokenizer to compress the DCT coefficients into dense action tokens, typically achieving **10x compression** over prior tokenization approaches.
-
-This approach can transform **any existing VLM** into a VLA by training it to predict these FAST tokens.
-
-## Installation Requirements
-
-1. Install LeRobot by following our [Installation Guide](./installation).
-2. Install π₀-FAST dependencies by running:
-
-   ```bash
-   pip install -e ".[pi]"
-   ```
-
-## Training a Custom FAST Tokenizer
-
-You have two options for the FAST tokenizer:
-
-1. **Use the pre-trained tokenizer**: The `lerobot/fast-action-tokenizer` tokenizer was trained on 1M+ real robot action sequences and works as a general-purpose tokenizer.
-
-2. **Train your own tokenizer**: For maximum performance on your specific dataset, you can finetune the tokenizer on your own data.
-
-### Training Your Own Tokenizer
-
-```bash
-lerobot-train-tokenizer \
-    --repo_id "user/my-lerobot-dataset" \
-    --action_horizon 10 \
-    --encoded_dims "0:6" \
-    --vocab_size 1024 \
-    --scale 10.0 \
-    --normalization_mode QUANTILES \
-    --output_dir "./my_fast_tokenizer" \
-    --push_to_hub \
-    --hub_repo_id "username/my-action-tokenizer"
-```
-
-### Key Tokenizer Parameters
-
-| Parameter              | Description                                                                       | Default      |
-| ---------------------- | --------------------------------------------------------------------------------- | ------------ |
-| `--repo_id`            | LeRobot dataset repository ID                                                     | Required     |
-| `--action_horizon`     | Number of future actions in each chunk                                            | `10`         |
-| `--encoded_dims`       | Comma-separated dimension ranges to encode (e.g., `"0:6,7:23"`)                   | `"0:6,7:23"` |
-| `--vocab_size`         | BPE vocabulary size                                                               | `1024`       |
-| `--scale`              | DCT scaling factor for quantization                                               | `10.0`       |
-| `--normalization_mode` | Normalization mode (`MEAN_STD`, `MIN_MAX`, `QUANTILES`, `QUANTILE10`, `IDENTITY`) | `QUANTILES`  |
-| `--sample_fraction`    | Fraction of chunks to sample per episode                                          | `0.1`        |
-
-## Usage
-
-To use π₀-FAST in LeRobot, specify the policy type as:
-
-```python
-policy.type=pi0_fast
-```
-
-## Training
-
-For training π₀-FAST, you can use the LeRobot training script:
-
-```bash
-lerobot-train \
-    --dataset.repo_id=your_dataset \
-    --policy.type=pi0_fast \
-    --output_dir=./outputs/pi0fast_training \
-    --job_name=pi0fast_training \
-    --policy.pretrained_path=lerobot/pi0_fast_base \
-    --policy.dtype=bfloat16 \
-    --policy.gradient_checkpointing=true \
-    --policy.chunk_size=10 \
-    --policy.n_action_steps=10 \
-    --policy.max_action_tokens=256 \
-    --steps=100000 \
-    --batch_size=4 \
-    --policy.device=cuda
-```
-
-### Key Training Parameters
-
-| Parameter                              | Description                                        | Default                         |
-| -------------------------------------- | -------------------------------------------------- | ------------------------------- |
-| `--policy.gradient_checkpointing=true` | Reduces memory usage significantly during training | `false`                         |
-| `--policy.dtype=bfloat16`              | Use mixed precision training for efficiency        | `float32`                       |
-| `--policy.chunk_size`                  | Number of action steps to predict (action horizon) | `50`                            |
-| `--policy.n_action_steps`              | Number of action steps to execute                  | `50`                            |
-| `--policy.max_action_tokens`           | Maximum number of FAST tokens per action chunk     | `256`                           |
-| `--policy.action_tokenizer_name`       | FAST tokenizer to use                              | `lerobot/fast-action-tokenizer` |
-| `--policy.compile_model=true`          | Enable torch.compile for faster training           | `false`                         |
-
-## Inference
-
-### KV-Caching for Fast Inference
-
-π₀-FAST supports **KV-caching**, a widely used optimization in LLM inference. This caches the key-value pairs from the attention mechanism, avoiding redundant computation during autoregressive decoding.
-
-```python
-# KV-caching is enabled by default
-policy.use_kv_cache=true
-```
-
-### Inference Example
-
-```python
-from lerobot.policies.pi0_fast import PI0FastPolicy, PI0FastConfig
-
-# Load the policy
-policy = PI0FastPolicy.from_pretrained("your-model-path")
-
-# During inference
-actions = policy.predict_action_chunk(batch)
-```
-
-## Model Architecture
-
-π₀-FAST uses a PaliGemma-based architecture:
-
-- **Vision Encoder**: SigLIP vision tower for image understanding
-- **Language Model**: Gemma 2B for processing language instructions and predicting action tokens
-
-The model takes images, text instructions, and robot state as input, and outputs discrete FAST tokens that are decoded back to continuous actions.
-
-## Configuration Options
-
-| Parameter            | Description                                     | Default    |
-| -------------------- | ----------------------------------------------- | ---------- |
-| `paligemma_variant`  | VLM backbone variant (`gemma_300m`, `gemma_2b`) | `gemma_2b` |
-| `max_state_dim`      | Maximum state vector dimension (padded)         | `32`       |
-| `max_action_dim`     | Maximum action vector dimension (padded)        | `32`       |
-| `temperature`        | Sampling temperature (0.0 for greedy)           | `0.0`      |
-| `max_decoding_steps` | Maximum decoding steps                          | `256`      |
-| `use_kv_cache`       | Enable KV caching for faster inference          | `true`     |
-
-## Comparison with π₀
-
-| Feature               | π₀                        | π₀-FAST                      |
-| --------------------- | ------------------------- | ---------------------------- |
-| Action Representation | Flow Matching (Diffusion) | Autoregressive Tokens (FAST) |
-| Training Speed        | 1x                        | **5x faster**                |
-| Dexterity             | High                      | High                         |
-| Inference Method      | Iterative Denoising       | Autoregressive Decoding      |
-| KV-Caching            | N/A                       | Supported                    |
-
-## 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).
-
-The finetuned model can be found here:
-
-- **π₀Fast LIBERO**: [lerobot/pi0fast-libero](https://huggingface.co/lerobot/pi0fast-libero)
-
-With the following training command:
-
-```bash
-lerobot-train \
-  --dataset.repo_id=lerobot/libero \
-  --output_dir=outputs/libero_pi0fast \
-  --job_name=libero_pi0fast \
-  --policy.path=lerobot/pi0fast_base \
-  --policy.dtype=bfloat16 \
-  --steps=100000 \
-  --save_freq=20000 \
-  --batch_size=4 \
-  --policy.device=cuda \
-  --policy.scheduler_warmup_steps=4000 \
-  --policy.scheduler_decay_steps=100000 \
-  --policy.scheduler_decay_lr=1e-5 \
-  --policy.gradient_checkpointing=true \
-  --policy.chunk_size=10 \
-  --policy.n_action_steps=10 \
-  --policy.max_action_tokens=256 \
-  --policy.empty_cameras=1 \
-```
-
-We then evaluate the finetuned model using the LeRobot LIBERO implementation, by running the following command:
-
-```bash
-tasks="libero_object,libero_spatial,libero_goal,libero_10"
-lerobot-eval \
-  --policy.path=lerobot/pi0fast-libero \
-  --policy.max_action_tokens=256 \
-  --env.type=libero \
-  --policy.gradient_checkpointing=false \
-  --env.task=${tasks} \
-  --eval.batch_size=1 \
-  --eval.n_episodes=1 \
-  --rename_map='{"observation.images.image":"observation.images.base_0_rgb","observation.images.image2":"observation.images.left_wrist_0_rgb"}'
-```
-
-**Note:** We set `n_action_steps=10`, similar to the original OpenPI implementation.
-
-### Results
-
-We obtain the following results on the LIBERO benchmark:
-
-| Model       | LIBERO Spatial | LIBERO Object | LIBERO Goal | LIBERO 10 | Average  |
-| ----------- | -------------- | ------------- | ----------- | --------- | -------- |
-| **π₀-fast** | 70.0           | 100.0         | 100.0       | 60.0      | **82.5** |
-
-The full evaluation output folder, including videos, is available [here](https://drive.google.com/drive/folders/1HXpwPTRm4hx6g1sF2P7OOqGG0TwPU7LQ?usp=sharing)
-
-## License
-
-This model follows the **Apache 2.0 License**, consistent with the original [OpenPI repository](https://github.com/Physical-Intelligence/openpi).
-
-## References
-
-- [FAST: Efficient Robot Action Tokenization](https://www.physicalintelligence.company/research/fast) - Physical Intelligence Blog
-- [OpenPI Repository](https://github.com/Physical-Intelligence/openpi) - Original implementation
-- [FAST Tokenizer on Hugging Face](https://huggingface.co/physical-intelligence/fast) - Pre-trained tokenizer

docs/source/policy_multi_task_dit_README.md

@@ -1,37 +0,0 @@
-# Multitask DiT Policy
-
-## Citation
-
-If you use this work, please cite the following works:
-
-```bibtex
-@misc{jones2025multitaskditpolicy,
-  author = {Bryson Jones},
-  title = {Dissecting and Open-Sourcing Multitask Diffusion Transformer Policy},
-  year = {2025},
-  url = {https://brysonkjones.substack.com/p/dissecting-and-open-sourcing-multitask-diffusion-transformer-policy},
-  note = {Blog post}
-}
-```
-
-```bibtex
-@misc{trilbmteam2025carefulexaminationlargebehaviormodels,
-  author       = {TRI LBM Team},
-  title        = {A Careful Examination of Large Behavior Models for Multitask Dexterous Manipulation},
-  year         = {2025},
-  eprint       = {arXiv:2507.05331},
-  archivePrefix = {arXiv},
-  primaryClass = {cs.RO},
-  url          = {https://arxiv.org/abs/2507.05331}
-}
-```
-
-```bibtex
-@misc{bostondynamics2025largebehaviormodelsatlas,
-  author       = {Boston Dynamics and TRI Research Team},
-  title        = {Large Behavior Models and Atlas Find New Footing},
-  year         = {2025},
-  url          = {https://bostondynamics.com/blog/large-behavior-models-atlas-find-new-footing/},
-  note         = {Blog post}
-}
-```

docs/source/policy_pi05_README.md

@@ -1,91 +0,0 @@
-# π₀.₅ (pi05)
-
-This repository contains the Hugging Face port of **π₀.₅**, adapted from [OpenPI](https://github.com/Physical-Intelligence/openpi) by the Physical Intelligence.
-It is designed as a **Vision-Language-Action model with open-world generalization**.
-
----
-
-## Model Overview
-
-| Feature              | π₀                                                     | π₀.₅                                      |
-| -------------------- | ------------------------------------------------------ | ----------------------------------------- |
-| Time Conditioning    | Concatenates time with actions via `action_time_mlp_*` | Uses `time_mlp_*` for AdaRMS conditioning |
-| AdaRMS               | Not used                                               | Used in action expert                     |
-| Tokenizer Length     | 48 tokens                                              | 200 tokens                                |
-| Discrete State Input | False (Uses `state_proj` layer)                        | True                                      |
-| Parameter Count      | Higher (includes state embedding)                      | Lower (no state embedding)                |
-
----
-
-## Relative Actions
-
-π₀.₅ supports training with **relative actions**, where the model learns relative offsets
-from the current robot state instead of absolute joint positions. This mirrors the
-relative-action transform in OpenPI (`DeltaActions`) and can improve performance.
-
-### How it works
-
-1. **During preprocessing**, absolute actions are converted to relative offsets:
-   `relative = action - state` (for selected joints).
-2. The relative actions are normalized using statistics computed from the relative distribution.
-3. **During postprocessing**, predicted relative actions are converted back to absolute:
-   `absolute = relative + state`.
-
-Joints listed in `relative_exclude_joints` (e.g., gripper) are kept absolute.
-
-### Configuration
-
-| Parameter                 | Type        | Default       | Description                                                      |
-| ------------------------- | ----------- | ------------- | ---------------------------------------------------------------- |
-| `use_relative_actions`    | `bool`      | `False`       | Enable relative-action training                                  |
-| `relative_exclude_joints` | `list[str]` | `["gripper"]` | Joint names to keep absolute (matched by substring)              |
-| `action_feature_names`    | `list[str]` | `None`        | Auto-populated from dataset metadata at runtime by `make_policy` |
-
-### Training example
-
-```bash
-python -m lerobot.scripts.lerobot_train \
-  --policy.type=pi05 \
-  --dataset.repo_id=your_org/your_dataset \
-  --policy.use_relative_actions=true \
-  --policy.relative_exclude_joints='["gripper"]'
-```
-
-When `use_relative_actions=true`, the training script automatically:
-
-- Computes relative action statistics from the dataset (sampled chunk-level relative actions)
-- Replaces the standard action stats with relative stats for normalization
-- Broadcasts these stats across all ranks in distributed training
-
----
-
-## Citation
-
-If you use this work, please cite both **OpenPI** and the π₀.₅ paper:
-
-```bibtex
-@misc{openpi2024,
-  author       = {Physical Intelligence Lab},
-  title        = {OpenPI: PyTorch Implementation of π0 and π0.5 Policies},
-  year         = {2024},
-  publisher    = {GitHub},
-  howpublished = {\url{https://github.com/Physical-Intelligence/openpi}},
-  license      = {Apache-2.0}
-}
-
-@misc{intelligence2025pi05visionlanguageactionmodelopenworld,
-  title        = {π₀.₅: a Vision-Language-Action Model with Open-World Generalization},
-  author       = {Physical Intelligence and Kevin Black and Noah Brown and James Darpinian and Karan Dhabalia and Danny Driess and Adnan Esmail and Michael Equi and Chelsea Finn and Niccolo Fusai and Manuel Y. Galliker and Dibya Ghosh and Lachy Groom and Karol Hausman and Brian Ichter and Szymon Jakubczak and Tim Jones and Liyiming Ke and Devin LeBlanc and Sergey Levine and Adrian Li-Bell and Mohith Mothukuri and Suraj Nair and Karl Pertsch and Allen Z. Ren and Lucy Xiaoyang Shi and Laura Smith and Jost Tobias Springenberg and Kyle Stachowicz and James Tanner and Quan Vuong and Homer Walke and Anna Walling and Haohuan Wang and Lili Yu and Ury Zhilinsky},
-  year         = {2025},
-  eprint       = {2504.16054},
-  archivePrefix= {arXiv},
-  primaryClass = {cs.LG},
-  url          = {https://arxiv.org/abs/2504.16054},
-}
-```
-
----
-
-## License
-
-This port follows the **Apache 2.0 License**, consistent with the original [OpenPI repository](https://github.com/Physical-Intelligence/openpi).

docs/source/policy_pi0_README.md

@@ -1,108 +0,0 @@
-# π₀ (pi0)
-
-This repository contains the Hugging Face port of **π₀**, adapted from [OpenPI](https://github.com/Physical-Intelligence/openpi) by the Physical Intelligence.
-It is designed as a **Vision-Language-Action model for general robot control**.
-
----
-
-## Model Overview
-
-| Feature              | π₀                                                     | π₀.₅                                      |
-| -------------------- | ------------------------------------------------------ | ----------------------------------------- |
-| Time Conditioning    | Concatenates time with actions via `action_time_mlp_*` | Uses `time_mlp_*` for AdaRMS conditioning |
-| AdaRMS               | Not used                                               | Used in action expert                     |
-| Tokenizer Length     | 48 tokens                                              | 200 tokens                                |
-| Discrete State Input | False (Uses `state_proj` layer)                        | True                                      |
-| Parameter Count      | Higher (includes state embedding)                      | Lower (no state embedding)                |
-
----
-
-## Relative Actions
-
-π₀ supports training with **relative actions**, where the model learns relative offsets
-from the current robot state instead of absolute joint positions. This mirrors the
-relative-action transform in OpenPI (`DeltaActions`) and can improve performance.
-
-### How it works
-
-1. **During preprocessing**, absolute actions are converted to relative offsets:
-   `relative = action - state` (for selected joints).
-2. The relative actions are normalized using statistics computed from the relative distribution.
-3. **During postprocessing**, predicted relative actions are converted back to absolute:
-   `absolute = relative + state`.
-
-Joints listed in `relative_exclude_joints` (e.g., gripper) are kept absolute.
-
-### Configuration
-
-| Parameter                 | Type        | Default       | Description                                                      |
-| ------------------------- | ----------- | ------------- | ---------------------------------------------------------------- |
-| `use_relative_actions`    | `bool`      | `False`       | Enable relative-action training                                  |
-| `relative_exclude_joints` | `list[str]` | `["gripper"]` | Joint names to keep absolute (matched by substring)              |
-| `action_feature_names`    | `list[str]` | `None`        | Auto-populated from dataset metadata at runtime by `make_policy` |
-
-### Training example
-
-```bash
-python -m lerobot.scripts.lerobot_train \
-  --policy.type=pi0 \
-  --dataset.repo_id=your_org/your_dataset \
-  --policy.use_relative_actions=true \
-  --policy.relative_exclude_joints='["gripper"]'
-```
-
-When `use_relative_actions=true`, the training script automatically:
-
-- Computes relative action statistics from the dataset (sampled chunk-level relative actions)
-- Replaces the standard action stats with relative stats for normalization
-- Broadcasts these stats across all ranks in distributed training
-
-### Recomputing stats for an existing dataset
-
-If you want to precompute relative action stats offline, use `recompute_stats` from
-`lerobot.datasets.dataset_tools`:
-
-```python
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
-from lerobot.datasets.dataset_tools import recompute_stats
-
-dataset = LeRobotDataset("your_org/your_dataset")
-dataset = recompute_stats(
-    dataset,
-    relative_action=True,
-    relative_exclude_joints=["gripper"],
-)
-```
-
----
-
-## Citation
-
-If you use this work, please cite both **OpenPI** and the π₀ paper:
-
-```bibtex
-@misc{openpi2024,
-  author       = {Physical Intelligence Lab},
-  title        = {OpenPI: PyTorch Implementation of π0 and π0.5 Policies},
-  year         = {2024},
-  publisher    = {GitHub},
-  howpublished = {\url{https://github.com/Physical-Intelligence/openpi}},
-  license      = {Apache-2.0}
-}
-
-@misc{black2024pi0visionlanguageactionflowmodel,
-  title        = {π₀: A Vision-Language-Action Flow Model for General Robot Control},
-  author       = {Kevin Black and Noah Brown and Danny Driess and Adnan Esmail and Michael Equi and Chelsea Finn and Niccolo Fusai and Lachy Groom and Karol Hausman and Brian Ichter and Szymon Jakubczak and Tim Jones and Liyiming Ke and Sergey Levine and Adrian Li-Bell and Mohith Mothukuri and Suraj Nair and Karl Pertsch and Lucy Xiaoyang Shi and James Tanner and Quan Vuong and Anna Walling and Haohuan Wang and Ury Zhilinsky},
-  year         = {2024},
-  eprint       = {2410.24164},
-  archivePrefix= {arXiv},
-  primaryClass = {cs.LG},
-  url          = {https://arxiv.org/abs/2410.24164},
-}
-```
-
----
-
-## License
-
-This port follows the **Apache 2.0 License**, consistent with the original [OpenPI repository](https://github.com/Physical-Intelligence/openpi).

docs/source/policy_rtc_README.md

@@ -1,38 +0,0 @@
-# Real-Time Chunking (RTC)
-
-This module contains the LeRobot implementation of **Real-Time Chunking (RTC)**, an inference-time technique for flow-matching based policies.
-
-**Note**: RTC is not a policy itself, but rather an inference enhancement that works with flow-matching based policies including [π₀](../pi0/), [π₀.₅](../pi05/), and [SmolVLA](../smolvla/).
-
----
-
-## Citation
-
-If you use Real-Time Chunking in your work, please cite:
-
-```bibtex
-@misc{openpi2024,
-  author       = {Physical Intelligence Lab},
-  title        = {OpenPI: PyTorch Implementation of π0 and π0.5 Policies},
-  year         = {2024},
-  publisher    = {GitHub},
-  howpublished = {\url{https://github.com/Physical-Intelligence/openpi}},
-  license      = {Apache-2.0}
-}
-
-@misc{black2025realtimeexecutionactionchunking,
-      title={Real-Time Execution of Action Chunking Flow Policies},
-      author={Kevin Black and Manuel Y. Galliker and Sergey Levine},
-      year={2025},
-      eprint={2506.07339},
-      archivePrefix={arXiv},
-      primaryClass={cs.RO},
-      url={https://arxiv.org/abs/2506.07339},
-}
-```
-
----
-
-## License
-
-This implementation follows the **Apache 2.0 License**, consistent with the LeRobot project.

docs/source/policy_sarm_README.md

@@ -1,14 +0,0 @@
-## Paper
-
-https://arxiv.org/abs/2509.25358
-
-## Citation
-
-```bibtex
-@article{chen2025sarm,
-  title={SARM: Stage-Aware Reward Modeling for Long Horizon Robot Manipulation},
-  author={Chen, Qianzhong and Yu, Justin and Schwager, Mac and Abbeel, Pieter and Shentu, Yide and Wu, Philipp},
-  journal={arXiv preprint arXiv:2509.25358},
-  year={2025}
-}
-```

docs/source/policy_walloss_README.md

@@ -1,45 +0,0 @@
-# WALL-OSS
-
-This repository contains the Hugging Face port of [**WALL-OSS**](https://x2robot.com/en/research/68bc2cde8497d7f238dde690), a Vision-Language-Action model for cross-embodiment robotic control based on Qwen2.5-VL with flow matching/FAST action prediction.
-
----
-
-## Model Overview
-
-| Feature            | Description                                           |
-| ------------------ | ----------------------------------------------------- |
-| Base Model         | Qwen2.5-VL (Vision-Language Model)                    |
-| Action Prediction  | Flow Matching (diffusion) or FAST (discrete tokens)   |
-| Architecture       | Mixture of Experts (MoE) with action-specific routing |
-| Multi-Modal Inputs | Vision (images/videos), Language, Proprioception      |
-
----
-
-## Additional Resources
-
-Paper: https://arxiv.org/pdf/2509.11766
-
-Official Repository: https://github.com/X-Square-Robot/wall-x
-
-Hugging Face: https://huggingface.co/x-square-robot
-
----
-
-## Citation
-
-If you use this work, please cite:
-
-```bibtex
-@article{zhai2025igniting,
-    title   = {Igniting VLMs Toward the Embodied Space},
-    author  = {Zhai, Andy and Liu, Brae and Fang, Bruno and Cai, Chalse and Ma, Ellie and Yin, Ethan and Wang, Hao and Zhou, Hugo and Wang, James and Shi, Lights and Liang, Lucy and Wang, Make and Wang, Qian and Gan, Roy and Yu, Ryan and Li, Shalfun and Liu, Starrick and Chen, Sylas and Chen, Vincent and Xu, Zach},
-    journal = {arXiv preprint arXiv:2509.11766},
-    year    = {2025}
-}
-```
-
----
-
-## License
-
-This model follows the **Apache 2.0 License**, consistent with the original [WallX repository](https://github.com/X-Square-Robot/wall-x).

docs/source/processors_robots_teleop.mdx

@@ -30,7 +30,7 @@ Each of these pipelines handle different conversions between different action an
 
 Below is an example of the three pipelines that we use in the phone to SO-100 follower examples:
 
-```python
+```69:90:examples/phone_so100_record.py
 phone_to_robot_ee_pose_processor = RobotProcessorPipeline[RobotAction, RobotAction]( # teleop -> dataset action
     steps=[
         MapPhoneActionToRobotAction(platform=teleop_config.phone_os),
@@ -84,7 +84,7 @@ Dataset features are determined by the keys saved in the dataset. Each step can
 
 Below is and example of how we declare features with the `transform_features` method in the phone to SO-100 follower examples:
 
-```python
+```src/lerobot/robots/so100_follower/robot_kinematic_processor.py
     def transform_features(
         self, features: dict[PipelineFeatureType, dict[str, PolicyFeature]]
     ) -> dict[PipelineFeatureType, dict[str, PolicyFeature]]:
@@ -103,7 +103,7 @@ Here we declare what PolicyFeatures we modify in this step, so we know what feat
 
 Below is an example of how we aggregate and merge features in the phone to SO-100 record example:
 
-```python
+```121:145:examples/phone_so100_record.py
 features=combine_feature_dicts(
         # Run the feature contract of the pipelines
         # This tells you how the features would look like after the pipeline steps

docs/source/reachy2.mdx

@@ -38,7 +38,6 @@ docker run --rm -it \
   start_rviz:=true start_sdk_server:=true mujoco:=true
 ```
 
-> [!NOTE]
 > If MuJoCo runs slowly (low simulation frequency), append `-e LD_LIBRARY_PATH="/opt/host-libs:$LD_LIBRARY_PATH" \` to the previous command to improve performance:
 >
 > ```
@@ -142,7 +141,7 @@ If you choose this option but still want to use the VR teleoperation application
 First add reachy2 and reachy2_teleoperator to the imports of the record script. Then you can use the following command:
 
 ```bash
-lerobot-record \
+python -m lerobot.record \
     --robot.type=reachy2 \
     --robot.ip_address=192.168.0.200 \
     --robot.id=r2-0000 \
@@ -151,7 +150,6 @@ lerobot-record \
     --teleop.type=reachy2_teleoperator \
     --teleop.ip_address=192.168.0.200 \
     --teleop.with_mobile_base=false \
-    --robot.with_torso_camera=true \
     --dataset.repo_id=pollen_robotics/record_test \
     --dataset.single_task="Reachy 2 recording test" \
     --dataset.num_episodes=1 \
@@ -159,9 +157,6 @@ lerobot-record \
     --dataset.fps=15 \
     --dataset.push_to_hub=true \
     --dataset.private=true \
-    --dataset.streaming_encoding=true \
-    --dataset.encoder_threads=2 \
-    # --dataset.vcodec=auto \
     --display_data=true
 ```
 
@@ -170,7 +165,7 @@ lerobot-record \
 **Extended setup overview (all options included):**
 
 ```bash
-lerobot-record \
+python -m lerobot.record \
     --robot.type=reachy2 \
     --robot.ip_address=192.168.0.200 \
     --robot.use_external_commands=true \
@@ -182,8 +177,6 @@ lerobot-record \
     --robot.with_left_teleop_camera=true \
     --robot.with_right_teleop_camera=true \
     --robot.with_torso_camera=false \
-    --robot.camera_width=640 \
-    --robot.camera_height=480 \
     --robot.disable_torque_on_disconnect=false \
     --robot.max_relative_target=5.0 \
     --teleop.type=reachy2_teleoperator \
@@ -201,9 +194,6 @@ lerobot-record \
     --dataset.fps=15 \
     --dataset.push_to_hub=true \
     --dataset.private=true \
-    --dataset.streaming_encoding=true \
-    --dataset.encoder_threads=2 \
-    # --dataset.vcodec=auto \
     --display_data=true
 ```
 
@@ -222,10 +212,9 @@ Must be set to true if a compliant Reachy 2 is used to control another one.
 From our initial tests, recording **all** joints when only some are moving can reduce model quality with certain policies.
 To avoid this, you can exclude specific parts from recording and replay using:
 
-```bash
+````
 --robot.with_<part>=false
-```
-
+```,
 with `<part>` being one of : `mobile_base`, `l_arm`, `r_arm", `neck`, `antennas`.
 It determine whether the corresponding part is recorded in the observations. True if not set.
 
@@ -233,60 +222,49 @@ By default, **all parts are recorded**.
 
 The same per-part mechanism is available in `reachy2_teleoperator` as well.
 
-```bash
---teleop.with\_<part>
-```
+````
 
+--teleop.with\_<part>
+
+```
 with `<part>` being one of : `mobile_base`, `l_arm`, `r_arm", `neck`, `antennas`.
 Determine whether the corresponding part is recorded in the actions. True if not set.
 
 > **Important:** In a given session, the **enabled parts must match** on both the robot and the teleoperator.
-> For example, if the robot runs with `--robot.with_mobile_base=false`, the teleoperator must disable the same part `--teleoperator.with_mobile_base=false`.
+For example, if the robot runs with `--robot.with_mobile_base=false`, the teleoperator must disable the same part `--teleoperator.with_mobile_base=false`.
 
 ##### Use the relevant cameras
 
-You can do the same for **cameras**. Enable or disable each camera with default parameters using:
+You can do the same for **cameras**. By default, only the **teleoperation cameras** are recorded (both `left_teleop_camera` and `right_teleop_camera`). Enable or disable each camera with:
 
-```bash
---robot.with_left_teleop_camera=<true|false> \
---robot.with_right_teleop_camera=<true|false> \
+```
+
+--robot.with_left_teleop_camera=<true|false>
+--robot.with_right_teleop_camera=<true|false>
 --robot.with_torso_camera=<true|false>
-```
 
-By default, no camera is recorded, all camera arguments are set to `false`.
-If you want to, you can use custom `width` and `height` parameters for Reachy 2's cameras using the `--robot.camera_width` & `--robot.camera_height` argument:
+````
 
-```bash
---robot.camera_width=1920 \
---robot.camera_height=1080
-```
-
-This will change the resolution of all 3 default robot cameras (enabled by the above bool arguments).
-
-If you want, you can add additional cameras other than the ones in the robot as usual with:
-
-```bash
---robot.cameras="{ extra: {type: opencv, index_or_path: 42, width: 640, height: 480, fps: 30}}" \
-```
 
 ## Step 2: Replay
 
 Make sure the robot is configured with the same parts as the dataset:
 
 ```bash
-lerobot-replay \
+python -m lerobot.replay \
     --robot.type=reachy2 \
     --robot.ip_address=192.168.0.200 \
     --robot.use_external_commands=false \
     --robot.with_mobile_base=false \
     --dataset.repo_id=pollen_robotics/record_test \
     --dataset.episode=0
-```
+    --display_data=true
+````
 
 ## Step 3: Train
 
 ```bash
-lerobot-train \
+python -m lerobot.scripts.train \
   --dataset.repo_id=pollen_robotics/record_test \
   --policy.type=act \
   --output_dir=outputs/train/reachy2_test \
@@ -299,9 +277,10 @@ lerobot-train \
 ## Step 4: Evaluate
 
 ```bash
-lerobot-eval \
+python -m lerobot.record \
   --robot.type=reachy2 \
   --robot.ip_address=192.168.0.200 \
+  --display_data=false \
   --dataset.repo_id=pollen_robotics/eval_record_test \
   --dataset.single_task="Evaluate reachy2 policy" \
   --dataset.num_episodes=10 \

docs/source/rename_map.mdx

@@ -1,114 +0,0 @@
-# Rename Map and Empty Cameras
-
-When you train, evaluate, or record with a robot policy, your **dataset** or **environment** provides observations under one set of keys (e.g. `observation.images.front`, `observation.images.eagle`), while your **policy** expects another (e.g. `observation.images.image`, `observation.images.image2`). The **rename map** bridges that gap without changing the policy or data source.
-
-> **Scope:** The rename map only renames **observation** keys (images and state). Action keys are not affected.
-
-## Why observation keys don't always match
-
-Policies have a fixed set of **input feature names** baked into their pretrained config. For example:
-
-- [pi0fast-libero](https://huggingface.co/lerobot/pi0fast-libero) expects `observation.images.base_0_rgb` and `observation.images.left_wrist_0_rgb`.
-- [xvla-base](https://huggingface.co/lerobot/xvla-base) expects `observation.images.image`, `observation.images.image2`, and `observation.images.image3`.
-
-Your dataset might use different names entirely (e.g. `observation.images.front`, `observation.images.eagle`, `observation.images.glove`), and your eval environment might use yet another set. Rather than editing the policy config or renaming columns in the dataset, you pass a **rename map**: a JSON dictionary that maps source keys to the keys the policy expects. Renaming happens inside the preprocessor pipeline, so the policy always sees its expected keys.
-
-## Using the rename map
-
-Pass the mapping as a JSON string on the command line. The convention is always:
-
-```
---rename_map='{"source_key": "policy_key", ...}'
-```
-
-where **source_key** is what the dataset or environment provides, and **policy_key** is what the policy expects.
-
-Only listed keys are renamed; everything else passes through unchanged. Order of entries doesn't matter.
-
-Supported policies: **PI0**, **PI05**, **PI0Fast**, **SmolVLA**, and **XVLA**.
-
-### Training
-
-Suppose you fine-tune [lerobot/xvla-base](https://huggingface.co/lerobot/xvla-base) on a dataset with images under `observation.images.front`, `observation.images.eagle`, and `observation.images.glove`. XVLA expects `observation.images.image`, `observation.images.image2`, and `observation.images.image3`:
-
-```bash
-lerobot-train \
-  --dataset.repo_id=YOUR_DATASET \
-  --output_dir=./outputs/xvla_training \
-  --job_name=xvla_training \
-  --policy.path="lerobot/xvla-base" \
-  --policy.repo_id="HF_USER/xvla-your-robot" \
-  --policy.dtype=bfloat16 \
-  --policy.action_mode=auto \
-  --steps=20000 \
-  --policy.device=cuda \
-  --policy.freeze_vision_encoder=false \
-  --policy.freeze_language_encoder=false \
-  --policy.train_policy_transformer=true \
-  --policy.train_soft_prompts=true \
-  --rename_map='{"observation.images.front": "observation.images.image", "observation.images.eagle": "observation.images.image2", "observation.images.glove": "observation.images.image3"}'
-```
-
-### Evaluation
-
-A policy that expects `observation.images.base_0_rgb` and `observation.images.left_wrist_0_rgb` (e.g. [pi0fast-libero](https://huggingface.co/lerobot/pi0fast-libero)), but the LIBERO environment returns `observation.images.image` and `observation.images.image2`:
-
-```bash
-lerobot-eval \
-  --policy.path=lerobot/pi0fast-libero \
-  --env.type=libero \
-  ... \
-  --rename_map='{"observation.images.image": "observation.images.base_0_rgb", "observation.images.image2": "observation.images.left_wrist_0_rgb"}'
-```
-
-### Recording
-
-`lerobot-record` also supports rename maps, nested under the dataset config:
-
-```bash
-lerobot-record \ # When running inference
-  --policy.path="<user>/smolVLA_finetuned" \
-  ... \
-  --dataset.rename_map='{"observation.images.glove2": "observation.images.image"}'
-```
-
-## Alternative: edit the policy config directly
-
-If you always use the same dataset or environment, you can **edit the policy's `config.json`** so its observation keys match your data source. Then no rename map is needed.
-
-The tradeoff: modifying the policy config ties it to one data source. A rename map keeps one policy usable across many datasets and environments.
-
-## Empty cameras: fewer views than the policy expects
-
-Some policies are built for a fixed number of image inputs. If your dataset has fewer cameras, you can set **`empty_cameras`** in the policy config instead of modifying the model architecture.
-
-### How it works
-
-Setting `empty_cameras=N` adds N placeholder image features to the policy config, named:
-
-```
-observation.images.empty_camera_0
-observation.images.empty_camera_1
-...
-```
-
-At runtime, these keys have no corresponding data in the batch. The policy fills them with masked dummy tensors (padded with `-1` for SigLIP-based vision encoders, with a zero attention mask), so the extra image slots are effectively ignored during training and inference.
-
-### Example
-
-XVLA-base has three visual inputs and `empty_cameras=0` by default. Your dataset only has two cameras:
-
-1. Set `--policy.empty_cameras=1`.
-2. The config adds a third key: `observation.images.empty_camera_0`.
-3. Use the rename map for your two real cameras as usual.
-4. The third slot is masked out — no fake images needed in your dataset.
-
-## Quick reference
-
-| Goal                                      | What to do                                                                  |
-| ----------------------------------------- | --------------------------------------------------------------------------- |
-| Dataset keys ≠ policy keys                | `--rename_map='{"dataset_key": "policy_key", ...}'`                         |
-| Env keys ≠ policy keys (eval)             | `--rename_map='{"env_key": "policy_key", ...}'`                             |
-| Recording with different keys (inference) | `--dataset.rename_map='{"source_key": "policy_key", ...}'`.                 |
-| Fewer cameras than policy expects         | `--policy.empty_cameras=N` (supported by PI0, PI05, PI0Fast, SmolVLA, XVLA) |
-| Avoid passing a rename map                | Edit the policy's `config.json` so its keys match your data source          |

docs/source/sarm.mdx

@@ -4,12 +4,6 @@ SARM (Stage-Aware Reward Modeling) is a video-based reward modeling framework fo
 
 **Paper**: [SARM: Stage-Aware Reward Modeling for Long Horizon Robot Manipulation](https://arxiv.org/abs/2509.25358)
 
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/lerobot-sarm.png"
-  alt="An overview of SARM"
-  width="80%"
-/>
-
 ## Why Reward Models?
 
 Standard behavior cloning treats all demonstration frames equally, but real-world robot datasets are messy. They contain hesitations, corrections, and variable-quality trajectories. Reward models solve this by learning a generalizable notion of **task progress** from demonstrations: given video frames and a task description, they predict how close the robot is to completing the task (0→1). This learned "progress signal" can be used in multiple ways, two promising applications are: (1) **weighted imitation learning** (RA-BC), where high-progress frames receive more weight during policy training, and (2) **reinforcement learning**, where the reward model provides dense rewards for online or offline policy improvement.
@@ -269,7 +263,7 @@ This generates visualizations showing video frames with subtask boundaries overl
 Train with **no annotations** - uses linear progress from 0 to 1:
 
 ```bash
-lerobot-train \
+python src/lerobot/scripts/lerobot_train.py \
   --dataset.repo_id=your-username/your-dataset \
   --policy.type=sarm \
   --policy.annotation_mode=single_stage \
@@ -288,7 +282,7 @@ lerobot-train \
 Train with **dense annotations only** (sparse auto-generated):
 
 ```bash
-lerobot-train \
+python src/lerobot/scripts/lerobot_train.py \
   --dataset.repo_id=your-username/your-dataset \
   --policy.type=sarm \
   --policy.annotation_mode=dense_only \
@@ -307,7 +301,7 @@ lerobot-train \
 Train with **both sparse and dense annotations**:
 
 ```bash
-lerobot-train \
+python src/lerobot/scripts/lerobot_train.py \
   --dataset.repo_id=your-username/your-dataset \
   --policy.type=sarm \
   --policy.annotation_mode=dual \
@@ -468,7 +462,7 @@ This script:
 Once you have the progress file, train your policy with RA-BC weighting. The progress file is auto-detected from the dataset path (`sarm_progress.parquet`). Currently PI0, PI0.5 and SmolVLA are supported with RA-BC:
 
 ```bash
-lerobot-train \
+python src/lerobot/scripts/lerobot_train.py \
   --dataset.repo_id=your-username/your-dataset \
   --policy.type=pi0 \
   --use_rabc=true \

docs/source/smolvla.mdx

@@ -106,9 +106,6 @@ lerobot-record \
   --dataset.repo_id=${HF_USER}/eval_DATASET_NAME_test \  # <- This will be the dataset name on HF Hub
   --dataset.episode_time_s=50 \
   --dataset.num_episodes=10 \
-  --dataset.streaming_encoding=true \
-  --dataset.encoder_threads=2 \
-  # --dataset.vcodec=auto \
   # <- Teleop optional if you want to teleoperate in between episodes \
   # --teleop.type=so100_leader \
   # --teleop.port=/dev/ttyACM0 \

docs/source/so100.mdx

@@ -103,7 +103,7 @@ lerobot-setup-motors \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
+from lerobot.robots.so100_follower import SO100Follower, SO100FollowerConfig
 
 config = SO100FollowerConfig(
     port="/dev/tty.usbmodem585A0076841",
@@ -177,7 +177,7 @@ lerobot-setup-motors \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.teleoperators.so_leader import SO100Leader, SO100LeaderConfig
+from lerobot.teleoperators.so100_leader import SO100Leader, SO100LeaderConfig
 
 config = SO100LeaderConfig(
     port="/dev/tty.usbmodem585A0076841",
@@ -579,7 +579,7 @@ lerobot-calibrate \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.robots.so_follower import SO100FollowerConfig, SO100Follower
+from lerobot.robots.so100_follower import SO100FollowerConfig, SO100Follower
 
 config = SO100FollowerConfig(
     port="/dev/tty.usbmodem585A0076891",
@@ -617,7 +617,7 @@ lerobot-calibrate \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.teleoperators.so_leader import SO100LeaderConfig, SO100Leader
+from lerobot.teleoperators.so100_leader import SO100LeaderConfig, SO100Leader
 
 config = SO100LeaderConfig(
     port="/dev/tty.usbmodem58760431551",

docs/source/so101.mdx

@@ -1,18 +1,5 @@
 # SO-101
 
-<div style="display: flex; align-items: center; gap: 10px;">
-  <img
-    src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/SO101_Follower.webp"
-    alt="SO-101"
-    width="60%"
-  />
-  <img
-    src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/SO101_Leader.webp"
-    alt="SO-101"
-    width="60%"
-  />
-</div>
-
 In the steps below, we explain how to assemble our flagship robot, the SO-101.
 
 ## Source the parts
@@ -138,7 +125,7 @@ lerobot-setup-motors \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.robots.so_follower import SO101Follower, SO101FollowerConfig
+from lerobot.robots.so101_follower import SO101Follower, SO101FollowerConfig
 
 config = SO101FollowerConfig(
     port="/dev/tty.usbmodem585A0076841",
@@ -214,7 +201,7 @@ lerobot-setup-motors \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.teleoperators.so_leader import SO101Leader, SO101LeaderConfig
+from lerobot.teleoperators.so101_leader import SO101Leader, SO101LeaderConfig
 
 config = SO101LeaderConfig(
     port="/dev/tty.usbmodem585A0076841",
@@ -236,10 +223,10 @@ It is advisable to install one 3-pin cable in the motor after placing them befor
 
 ### Joint 1
 
-- Install both motor horns. Secure the top horn with a M3x6mm screw. No screws are required for the bottom horn.
 - Place the first motor into the base.
 - Fasten the motor with 4 M2x6mm screws (smallest screws). Two from the top and two from the bottom.
 - Slide over the first motor holder and fasten it using two M2x6mm screws (one on each side).
+- Install both motor horns, securing the top horn with a M3x6mm screw.
 - Attach the shoulder part.
 - Tighten the shoulder part with 4 M3x6mm screws on top and 4 M3x6mm screws on the bottom
 - Add the shoulder motor holder.
@@ -255,9 +242,9 @@ It is advisable to install one 3-pin cable in the motor after placing them befor
 
 ### Joint 2
 
-- Install both motor horns. Secure the top horn with a M3x6mm screw. No screws are required for the bottom horn.
 - Slide the second motor in from the top.
 - Fasten the second motor with 4 M2x6mm screws.
+- Attach both motor horns to motor 2, again use the M3x6mm horn screw.
 - Attach the upper arm with 4 M3x6mm screws on each side.
 
 <div class="video-container">
@@ -271,8 +258,8 @@ It is advisable to install one 3-pin cable in the motor after placing them befor
 
 ### Joint 3
 
-- Install both motor horns. Secure the top horn with a M3x6mm screw. No screws are required for the bottom horn.
-- Insert motor 3 and fasten using 4 M2x6mm screws.
+- Insert motor 3 and fasten using 4 M2x6mm screws
+- Attach both motor horns to motor 3 and secure one again with a M3x6mm horn screw.
 - Connect the forearm to motor 3 using 4 M3x6mm screws on each side.
 
 <div class="video-container">
@@ -286,10 +273,9 @@ It is advisable to install one 3-pin cable in the motor after placing them befor
 
 ### Joint 4
 
-- Install both motor horns. Secure the top horn with a M3x6mm screw. No screws are required for the bottom horn.
 - Slide over motor holder 4.
 - Slide in motor 4.
-- Fasten motor 4 with 4 M2x6mm screws.
+- Fasten motor 4 with 4 M2x6mm screws and attach its motor horns, use a M3x6mm horn screw.
 
 <div class="video-container">
   <video controls width="600">
@@ -322,7 +308,7 @@ It is advisable to install one 3-pin cable in the motor after placing them befor
 
 - Attach the gripper to motor 5, attach it to the motor horn on the wrist using 4 M3x6mm screws.
 - Insert the gripper motor and secure it with 2 M2x6mm screws on each side.
-- Install both motor horns on the gripper motor. Secure the top horn with a M3x6mm screw; no screws are required for the bottom horn.
+- Attach the motor horns and again use a M3x6mm horn screw.
 - Install the gripper claw and secure it with 4 M3x6mm screws on both sides.
 
 <div class="video-container">
@@ -378,7 +364,7 @@ lerobot-calibrate \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.robots.so_follower import SO101FollowerConfig, SO101Follower
+from lerobot.robots.so101_follower import SO101FollowerConfig, SO101Follower
 
 config = SO101FollowerConfig(
     port="/dev/tty.usbmodem585A0076891",
@@ -427,7 +413,7 @@ lerobot-calibrate \
 
 <!-- prettier-ignore-start -->
 ```python
-from lerobot.teleoperators.so_leader import SO101LeaderConfig, SO101Leader
+from lerobot.teleoperators.so101_leader import SO101LeaderConfig, SO101Leader
 
 config = SO101LeaderConfig(
     port="/dev/tty.usbmodem58760431551",

docs/source/streaming_video_encoding.mdx

@@ -1,155 +0,0 @@
-# Streaming Video Encoding Guide
-
-## 1. Overview
-
-Streaming video encoding eliminates the traditional PNG round-trip during video dataset recording. Instead of:
-
-1. Capture frame -> write PNG to disk -> (at episode end) read PNG's -> encode to MP4 -> delete PNG's
-
-Frames can be encoded in real-time during capture:
-
-1. Capture frame -> queue to encoder thread -> encode to MP4 directly
-
-This makes `save_episode()` near-instant (the video is already encoded by the time the episode ends) and removes the blocking wait that previously occurred between episodes, especially with multiple cameras in long episodes.
-
-## 2. Tuning Parameters
-
-| Parameter               | CLI Flag                          | Type          | Default       | Description                                                       |
-| ----------------------- | --------------------------------- | ------------- | ------------- | ----------------------------------------------------------------- |
-| `streaming_encoding`    | `--dataset.streaming_encoding`    | `bool`        | `True`        | Enable real-time encoding during capture                          |
-| `vcodec`                | `--dataset.vcodec`                | `str`         | `"libsvtav1"` | Video codec. `"auto"` detects best HW encoder                     |
-| `encoder_threads`       | `--dataset.encoder_threads`       | `int \| None` | `None` (auto) | Threads per encoder instance. `None` will leave the vcoded decide |
-| `encoder_queue_maxsize` | `--dataset.encoder_queue_maxsize` | `int`         | `60`          | Max buffered frames per camera (~2s at 30fps). Consumes RAM       |
-
-## 3. Performance Considerations
-
-Streaming encoding means the CPU is encoding video **during** the capture loop, not after. This creates a CPU budget that must be shared between:
-
-- **Control loop** (reading cameras, control the robot, writing non-video data)
-- **Encoder threads** (one pool per camera)
-- **Rerun visualization** (if enabled)
-- **OS and other processes**
-
-### Resolution & Number of Cameras Impact
-
-| Setup                     | Throughput (px/sec) | CPU Encoding Load | Notes                          |
-| ------------------------- | ------------------- | ----------------- | ------------------------------ |
-| 2camsx 640x480x3 @30fps   | 55M                 | Low               | Works on most systems          |
-| 2camsx 1280x720x3 @30fps  | 165M                | Moderate          | Comfortable on modern systems  |
-| 2camsx 1920x1080x3 @30fps | 373M                | High              | Requires powerful high-end CPU |
-
-### `encoder_threads` Tuning
-
-This parameter controls how many threads each encoder instance uses internally:
-
-- **Higher values** (e.g., 4-5): Faster encoding, but uses more CPU cores per camera. Good for high-end systems with many cores.
-- **Lower values** (e.g., 1-2): Less CPU per camera, freeing cores for capture and visualization. Good for low-res images and capable CPUs.
-- **`None` (default)**: Lets the codec decide. Information available in the codec logs.
-
-### Backpressure and Frame Dropping
-
-Each camera has a bounded queue (`encoder_queue_maxsize`, default 60 frames). When the encoder can't keep up:
-
-1. The queue fills up (consuming RAM)
-2. New frames are **dropped** (not blocked) — the capture loop continues uninterrupted
-3. A warning is logged: `"Encoder queue full for {camera}, dropped N frame(s)"`
-4. At episode end, total dropped frames per camera are reported
-
-### Symptoms of Encoder Falling Behind
-
-- **System feels laggy and freezes**: all CPUs are at 100%
-- **Dropped frame warnings** in the log or lower frames/FPS than expected in the recorded dataset
-- **Choppy robot movement**: If CPU is severely overloaded, even the capture loop may be affected
-- **Accumulated rerun lag**: Visualization falls behind real-time
-
-## 4. Hardware-Accelerated Encoding
-
-### When to Use
-
-Use HW encoding when:
-
-- CPU is the bottleneck (dropped frames, choppy robot, rerun lag)
-- You have compatible hardware (GPU or dedicated encoder)
-- You're recording at high throughput (high resolution or with many cameras)
-
-### Choosing a Codec
-
-| Codec                 | CPU Usage | File Size      | Quality | Notes                                                            |
-| --------------------- | --------- | -------------- | ------- | ---------------------------------------------------------------- |
-| `libsvtav1` (default) | High      | Smallest       | Best    | Default. Best compression but most CPU-intensive                 |
-| `h264`                | Medium    | ~30-50% larger | Good    | Software H.264. Lower CPU                                        |
-| HW encoders           | Very Low  | Largest        | Good    | Offloads to dedicated hardware. Best for CPU-constrained systems |
-
-### Available HW Encoders
-
-| Encoder             | Platform      | Hardware                                                                                         | CLI Value                            |
-| ------------------- | ------------- | ------------------------------------------------------------------------------------------------ | ------------------------------------ |
-| `h264_videotoolbox` | macOS         | Apple Silicon / Intel                                                                            | `--dataset.vcodec=h264_videotoolbox` |
-| `hevc_videotoolbox` | macOS         | Apple Silicon / Intel                                                                            | `--dataset.vcodec=hevc_videotoolbox` |
-| `h264_nvenc`        | Linux/Windows | NVIDIA GPU                                                                                       | `--dataset.vcodec=h264_nvenc`        |
-| `hevc_nvenc`        | Linux/Windows | NVIDIA GPU                                                                                       | `--dataset.vcodec=hevc_nvenc`        |
-| `h264_vaapi`        | Linux         | Intel/AMD GPU                                                                                    | `--dataset.vcodec=h264_vaapi`        |
-| `h264_qsv`          | Linux/Windows | Intel Quick Sync                                                                                 | `--dataset.vcodec=h264_qsv`          |
-| `auto`              | Any           | Probes the system for available HW encoders. Falls back to `libsvtav1` if no HW encoder is found | `--dataset.vcodec=auto`              |
-
-> [!NOTE]
-> In order to use the HW accelerated encoders you might need to upgrade your GPU drivers.
-
-> [!NOTE]
-> `libsvtav1` is the default because it provides the best training performance; other vcodecs can reduce CPU usage and be faster, but they typically produce larger files and may affect training time.
-
-## 5. Troubleshooting
-
-| Symptom                                                            | Likely Cause                                 | Fix                                                                                                                                                                                                                                                                                  |
-| ------------------------------------------------------------------ | -------------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
-| System freezes or choppy robot movement or Rerun visualization lag | CPU starved (100% load usage)                | Close other apps, reduce encoding throughput, lower `encoder_threads`, use `h264`, use `display_data=False`. If the CPU continues to be at 100% then it might be insufficient for your setup, consider `--dataset.streaming_encoding=false` or HW encoding (`--dataset.vcodec=auto`) |
-| "Encoder queue full" warnings or dropped frames in dataset         | Encoder can't keep up (Queue overflow)       | If CPU is not at 100%: Increase `encoder_threads`, increase `encoder_queue_maxsize` or use HW encoding (`--dataset.vcodec=auto`).                                                                                                                                                    |
-| High RAM usage                                                     | Queue filling faster than encoding           | `encoder_threads` too low or CPU insufficient. Reduce `encoder_queue_maxsize` or use HW encoding                                                                                                                                                                                     |
-| Large video files                                                  | Using HW encoder or H.264                    | Expected trade-off. Switch to `libsvtav1` if CPU allows                                                                                                                                                                                                                              |
-| `save_episode()` still slow                                        | `streaming_encoding` is `False`              | Set `--dataset.streaming_encoding=true`                                                                                                                                                                                                                                              |
-| Encoder thread crash                                               | Codec not available or invalid settings      | Check `vcodec` is installed, try `--dataset.vcodec=auto`                                                                                                                                                                                                                             |
-| Recorded dataset is missing frames                                 | CPU/GPU starvation or occasional load spikes | If ~5% of frames are missing, your system is likely overloaded — follow the recommendations above. If fewer frames are missing (~2%), they are probably due to occasional transient load spikes (often at startup) and can be considered expected.                                   |
-
-## 6. Recommended Configurations
-
-These estimates are conservative; we recommend testing them on your setup—start with a low load and increase it gradually.
-
-### High-End Systems: modern 12+ cores (24+ threads)
-
-A throughput between ~250-500M px/sec should be comfortable in CPU. For even better results try HW encoding if available.
-
-```bash
-# 3camsx 1280x720x3 @30fps: Defaults work well. Optionally increase encoder parallelism.
-# 2camsx 1920x1080x3 @30fps: Defaults work well. Optionally increase encoder parallelism.
-lerobot-record --dataset.encoder_threads=5 ...
-
-# 3camsx 1920x1080x3 @30fps: Might require some tuning.
-```
-
-### Mid-Range Systems: modern 8+ cores (16+ threads) or Apple Silicon
-
-A throughput between ~80-300M px/sec should be possible in CPU.
-
-```bash
-# 3camsx 640x480x3 @30fps: Defaults work well. Optionally decrease encoder parallelism.
-# 2camsx 1280x720x3 @30fps: Defaults work well. Optionally decrease encoder parallelism.
-lerobot-record --dataset.encoder_threads=2 ...
-
-# 2camsx 1920x1080x3 @30fps: Might require some tuning.
-```
-
-### Low-Resource Systems: modern 4+ cores (8+ threads) or Raspberry Pi 5
-
-On very constrained systems, streaming encoding may compete too heavily with the capture loop. Disabling it falls back to the PNG-based approach where encoding happens between episodes (blocking, but doesn't interfere with capture). Alternatively, record at a lower throughput to reduce both capture and encoding load. Consider also changing codec to `h264` and using batch encoding.
-
-```bash
-# 2camsx 640x480x3 @30fps: Requires some tuning.
-
-# Use H.264, disable streaming, consider batching encoding
-lerobot-record --dataset.vcodec=h264 --dataset.streaming_encoding=false ...
-```
-
-## 7. Closing note
-
-Performance ultimately depends on your exact setup — frames-per-second, resolution, CPU cores and load, available memory, episode length, and the encoder you choose. Always test with your target workload, be mindful about your CPU & system capabilities and tune `encoder_threads`, `encoder_queue_maxsize`, and
-`vcodec` reasonably. That said, a common practical configuration (for many applications) is three cameras at 640×480x3 @30fps; this usually runs fine with the default streaming video encoding settings in modern systems. Always verify your recorded dataset is healthy by comparing the video duration to the CLI episode duration and confirming the row count equals FPS × CLI duration.

docs/source/unitree_g1.mdx

@@ -1,72 +1,23 @@
-# Unitree G1
+# Unitree G1 Robot Setup and Control
 
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/unitree_thumbnail.jpg"
-  alt="Unitree G1 locomanipulation demo"
-  style={{ width: "100%" }}
-/>
+This guide covers the complete setup process for the Unitree G1 humanoid, from initial connection to running gr00t_wbc locomotion.
 
-The Unitree G1 humanoid is now supported in LeRobot! You can teleoperate, train locomanipulation policies, test in sim, and more. Both 29 and 23 DoF variants are supported.
+## About the Unitree G1
+
+We offer support for both 29 and 23 DOF G1. We introduce:
+
+- **`unitree g1` robot class, handling low level communication with the humanoid**
+- **ZMQ socket bridge** for remote communication over WiFi, allowing one to deploy policies remotely instead of over ethernet or directly on the Orin
+- **GR00T locomotion policy** for bipedal walking and balance
+- **MuJoCo simulation mode** for testing policies without the physical robot
 
 ---
 
-## Part 1: Getting Started
+## Part 1: Connect to Robot over Ethernet
 
-### Install the Unitree SDK
+### Step 1: Configure Your Computer's Ethernet Interface
 
-Follow the [unitree_sdk2_python installation guide](https://github.com/unitreerobotics/unitree_sdk2_python#installation). Tested with `unitree_sdk2py==1.0.1` and `cyclonedds==0.10.2`:
-
-```bash
-conda create -y -n lerobot python=3.12
-conda activate lerobot
-git clone https://github.com/unitreerobotics/unitree_sdk2_python.git
-cd unitree_sdk2_python
-pip install -e .
-cd ..
-```
-
-### Install LeRobot
-
-```bash
-conda install ffmpeg -c conda-forge
-conda install -c conda-forge "pinocchio>=3.0.0,<4.0.0"
-git clone https://github.com/huggingface/lerobot.git
-cd lerobot
-pip install -e '.[unitree_g1]'
-```
-
-<Tip>
-  For now, pinocchio must be installed from conda-forge (not pip) to include the
-  CasADi bindings needed for arm IK.
-</Tip>
-
-### Test the Installation (Simulation)
-
-The simulation environment has its own dependencies. Check the Simulation environment dependencies: [Unitree G1 Mujoco EnvHub](https://huggingface.co/lerobot/unitree-g1-mujoco/tree/main).
-
-```bash
-pip install mujoco loguru msgpack msgpack-numpy
-```
-
-```bash
-lerobot-teleoperate \
-  --robot.type=unitree_g1 \
-  --robot.is_simulation=true \
-  --teleop.type=unitree_g1 \
-  --teleop.id=wbc_unitree \
-  --robot.cameras='{"global_view": {"type": "zmq", "server_address": "localhost", "port": 5555, "camera_name": "head_camera", "width": 640, "height": 480, "fps": 30, "warmup_s": 5}}' \
-  --display_data=true \
-  --robot.controller=GrootLocomotionController
-```
-
-This will launch a [MuJoCo sim instance](https://huggingface.co/lerobot/unitree-g1-mujoco/tree/main) for the G1. You can connect a gamepad to your machine before launching in order to control the robot's locomotion in sim. We support both [HolosomaLocomotionController](https://github.com/amazon-far/holosoma) and [GrootLocomotionController](https://github.com/NVlabs/GR00T-WholeBodyControl) via `--robot.controller`.
-
-- Press `9` to release the robot
-- Press `7` / `8` to increase / decrease waist height
-
-### Connect to the Physical Robot
-
-The G1's Ethernet IP is fixed at `192.168.123.164`. Your machine must have a static IP on the same subnet: `192.168.123.x` where `x ≠ 164`.
+Set a static IP on the same subnet as the robot:
 
 ```bash
 # Replace 'enp131s0' with your ethernet interface name (check with `ip a`)
@@ -75,228 +26,183 @@ sudo ip addr add 192.168.123.200/24 dev enp131s0
 sudo ip link set enp131s0 up
 ```
 
-### SSH into the Robot
+**Note**: The robot's Ethernet IP is fixed at `192.168.123.164`. Your computer must use `192.168.123.x` where x ≠ 164.
+
+### Step 2: SSH into the Robot
 
 ```bash
 ssh unitree@192.168.123.164
 # Password: 123
 ```
 
-### Share Internet via Ethernet
+You should now be connected to the robot's onboard computer.
 
-The G1 needs internet access to clone repos and install packages. Share your laptop's connection over Ethernet:
+---
 
-**On your laptop:**
-
-```bash
-sudo sysctl -w net.ipv4.ip_forward=1
-
-# Replace wlp132s0f0 with your WiFi interface name
-sudo iptables -t nat -A POSTROUTING -o wlp132s0f0 -s 192.168.123.0/24 -j MASQUERADE
-sudo iptables -A FORWARD -i wlp132s0f0 -o enp131s0 -m state --state RELATED,ESTABLISHED -j ACCEPT
-sudo iptables -A FORWARD -i enp131s0 -o wlp132s0f0 -j ACCEPT
-```
-
-**On the G1:**
-
-```bash
-sudo ip route del default 2>/dev/null || true
-sudo ip route add default via 192.168.123.200 dev eth0
-echo "nameserver 8.8.8.8" | sudo tee /etc/resolv.conf
-
-# Verify
-ping -c 3 8.8.8.8
-```
-
-### Install the Unitree SDK on the G1
-
-Follow the [unitree_sdk2_python installation guide](https://github.com/unitreerobotics/unitree_sdk2_python#installation):
-
-```bash
-conda create -y -n lerobot python=3.12
-conda activate lerobot
-git clone https://github.com/unitreerobotics/unitree_sdk2_python.git
-cd unitree_sdk2_python
-python -m pip install -e .
-cd ..
-```
-
-### Install LeRobot on the G1
-
-```bash
-git clone https://github.com/huggingface/lerobot.git
-cd lerobot
-conda install -c conda-forge "pinocchio>=3.0.0,<4.0.0"
-python -m pip install -e '.[unitree_g1]'
-```
-
-<Tip>
-  For now, pinocchio must be installed from conda-forge (not pip) to include the
-  CasADi bindings needed for arm IK.
-</Tip>
-
-### (Optional) Enable WiFi on the Robot
-
-For wireless SSH access, you can enable WiFi on the G1 (it's blocked by default):
+## Part 2: Enable WiFi on the Robot
+
+Once connected via Ethernet, follow these steps to enable WiFi:
+
+### Step 1: Enable WiFi Hardware
 
 ```bash
+# Unblock WiFi radio
+sudo rfkill unblock wifi
 sudo rfkill unblock all
+
+# Bring up WiFi interface
 sudo ip link set wlan0 up
+
+# Enable NetworkManager control
 sudo nmcli radio wifi on
 sudo nmcli device set wlan0 managed yes
 sudo systemctl restart NetworkManager
 ```
 
-**Connect to a WiFi network:**
+### Step 2: Enable Internet Forwarding
+
+**On your laptop:**
 
 ```bash
+# Enable IP forwarding
+sudo sysctl -w net.ipv4.ip_forward=1
+
+# Set up NAT (replace wlp132s0f0 with your WiFi interface)
+sudo iptables -t nat -A POSTROUTING -o wlp132s0f0 -s 192.168.123.0/24 -j MASQUERADE
+sudo iptables -A FORWARD -i wlp132s0f0 -o enp131s0 -m state --state RELATED,ESTABLISHED -j ACCEPT
+sudo iptables -A FORWARD -i enp131s0 -o wlp132s0f0 -j ACCEPT
+```
+
+**On the robot:**
+
+```bash
+# Add laptop as default gateway
+sudo ip route del default 2>/dev/null || true
+sudo ip route add default via 192.168.123.200 dev eth0
+echo "nameserver 8.8.8.8" | sudo tee /etc/resolv.conf
+
+# Test connection
+ping -c 3 8.8.8.8
+```
+
+### Step 3: Connect to WiFi Network
+
+```bash
+# List available networks
 nmcli device wifi list
 
+# Connect to your WiFi (example)
 sudo nmcli connection add type wifi ifname wlan0 con-name "YourNetwork" ssid "YourNetwork"
 sudo nmcli connection modify "YourNetwork" wifi-sec.key-mgmt wpa-psk
 sudo nmcli connection modify "YourNetwork" wifi-sec.psk "YourPassword"
 sudo nmcli connection modify "YourNetwork" connection.autoconnect yes
 sudo nmcli connection up "YourNetwork"
 
+# Check WiFi IP address
 ip a show wlan0
 ```
 
-You can then SSH over WiFi instead of Ethernet:
+### Step 4: SSH Over WiFi
+
+Once connected to WiFi, note the robot's IP address and disconnect the Ethernet cable. You can now SSH over WiFi:
 
 ```bash
-ssh unitree@<ROBOT_WIFI_IP>
+ssh unitree@<YOUR_ROBOT_IP>
 # Password: 123
 ```
 
----
-
-## Part 2: Teleoperation & Locomotion
-
-### Run the Robot Server
-
-On the robot (from `~/lerobot`):
-
-```bash
-cd ~/lerobot
-python src/lerobot/robots/unitree_g1/run_g1_server.py --camera
-```
-
-### Run the Locomotion Policy
-
-You can run the teleoperation client from your laptop over Ethernet, over WiFi (experimental), or directly on the robot itself. Mind potential latency introduced by your network.
-
-**From your laptop:**
-
-```bash
-lerobot-teleoperate \
-  --robot.type=unitree_g1 \
-  --robot.is_simulation=false \
-  --robot.robot_ip=<ROBOT_IP> \
-  --teleop.type=unitree_g1 \
-  --teleop.id=wbc_unitree \
-  --robot.cameras='{"global_view": {"type": "zmq", "server_address": "<ROBOT_IP>", "port": 5555, "camera_name": "head_camera", "width": 640, "height": 480, "fps": 30}}' \
-  --display_data=true \
-  --robot.controller=HolosomaLocomotionController
-```
-
-We support both [GrootLocomotionController](https://github.com/NVlabs/GR00T-WholeBodyControl) and [HolosomaLocomotionController](https://github.com/amazon-far/holosoma) via `--robot.controller`.
+Replace `<YOUR_ROBOT_IP>` with your robot's actual WiFi IP address (e.g., `172.18.129.215`).
 
 ---
 
-## Part 3: Loco-Manipulation with the Homunculus Exoskeleton
+## Part 3: Robot Server Setup
 
-We provide a loco-manipulation solution via the Homunculus Exoskeleton — an open-source 7 DoF exoskeleton for whole-body control. Check it out [here](https://github.com/nepyope/hmc_exo).
+### Step 1: Install LeRobot on the Orin
 
-### Calibrate
+SSH into the robot and install LeRobot:
 
 ```bash
-lerobot-calibrate \
-  --teleop.type=unitree_g1 \
-  --teleop.left_arm_config.port=/dev/ttyACM1 \
-  --teleop.right_arm_config.port=/dev/ttyACM0 \
-  --teleop.id=exo
+ssh unitree@<YOUR_ROBOT_IP>
+
+conda create -y -n lerobot python=3.10
+conda activate lerobot
+git clone https://github.com/huggingface/lerobot.git
+cd lerobot
+pip install -e '.[unitree_g1]'
+git clone https://github.com/unitreerobotics/unitree_sdk2_python.git
+cd unitree_sdk2_python  && pip install -e .
 ```
 
-During calibration move each joint through its entire range. After fitting, move the joint in a neutral position and press `n` to advance.
+**Note**: The Unitree SDK requires CycloneDDS v0.10.2 to be installed. See the [Unitree SDK documentation](https://github.com/unitreerobotics/unitree_sdk2_python) for details.
 
-### Record a Dataset
+### Step 2: Run the Robot Server
+
+On the robot:
 
 ```bash
-lerobot-record \
-  --robot.type=unitree_g1 \
-  --robot.is_simulation=true \
-  --robot.cameras='{"global_view": {"type": "zmq", "server_address": "localhost", "port": 5555, "camera_name": "head_camera", "width": 640, "height": 480, "fps": 30}}' \
-  --teleop.type=unitree_g1 \
-  --teleop.left_arm_config.port=/dev/ttyACM1 \
-  --teleop.right_arm_config.port=/dev/ttyACM0 \
-  --teleop.id=exo \
-  --dataset.repo_id=your-username/dataset-name \
-  --dataset.single_task="Test" \
-  --dataset.num_episodes=2 \
-  --dataset.episode_time_s=5 \
-  --dataset.reset_time_s=5 \
-  --dataset.push_to_hub=true \
-  --dataset.streaming_encoding=true \
-  --dataset.encoder_threads=2
+python src/lerobot/robots/unitree_g1/run_g1_server.py
 ```
 
-> **Note:** Omit `--teleop.left_arm_config.port` and `--teleop.right_arm_config.port` if you're only using the joystick.
-
-Example dataset: [nepyope/unitree_box_move_blue_full](https://huggingface.co/datasets/nepyope/unitree_box_move_blue_full)
+**Important**: Keep this terminal running. The server must be active for remote control.
 
 ---
 
-## Part 4: Training & Inference
+## Part 4: Running GR00T Locomotion
 
-### Train
+With the robot server running, you can now control the robot from your laptop.
+
+### Step 1: Install LeRobot on your machine
 
 ```bash
-python src/lerobot/scripts/lerobot_train.py \
-  --dataset.repo_id=your-username/dataset-name  \
-  --policy.type=pi05 \
-  --output_dir=./outputs/pi05_training \
-  --job_name=pi05_training \
-  --policy.repo_id=your-username/your-repo-id \
-  --policy.pretrained_path=lerobot/pi05_base \
-  --policy.compile_model=true \
-  --policy.gradient_checkpointing=true \
-  --wandb.enable=true \
-  --policy.dtype=bfloat16 \
-  --policy.freeze_vision_encoder=false \
-  --policy.train_expert_only=false \
-  --steps=3000 \
-  --policy.device=cuda \
-  --batch_size=32
+conda create -y -n lerobot python=3.10
+conda activate lerobot
+git clone https://github.com/huggingface/lerobot.git
+cd lerobot
+pip install -e '.[unitree_g1]'
+git clone https://github.com/unitreerobotics/unitree_sdk2_python.git
+cd unitree_sdk2_python  && pip install -e .
 ```
 
-### Inference with RTC
+### Step 2: Update Robot IP in Config
 
-Once trained, we recommend deploying policies using inference-time RTC:
+Edit the config file to match your robot's WiFi IP:
+
+```python
+# In src/lerobot/robots/unitree_g1/config_unitree_g1.py
+robot_ip: str = "<YOUR_ROBOT_IP>"  # Replace with your robot's WiFi IP.
+```
+
+**Note**: When running directly on the G1 (not remotely), set `robot_ip: str = "127.0.0.1"` instead.
+
+### Step 3: Run the Locomotion Policy
 
 ```bash
-python examples/rtc/eval_with_real_robot.py \
-  --policy.path=your-username/your-repo-id \
-  --policy.device=cuda \
-  --robot.type=unitree_g1 \
-  --robot.is_simulation=false \
-  --robot.controller=HolosomaLocomotionController \
-  --robot.cameras='{"global_view": {"type": "zmq", "server_address": "<ROBOT_IP>", "port": 5555, "camera_name": "head_camera", "width": 640, "height": 480, "fps": 30}}' \
-  --task="task_description" \
-  --duration=1000 \
-  --fps=30 \
-  --rtc.enabled=true
+# Run GR00T locomotion controller
+python examples/unitree_g1/gr00t_locomotion.py --repo-id "nepyope/GR00T-WholeBodyControl_g1"
 ```
 
+### Step 4: Control with Remote
+
+- **Left stick**: Forward/backward and left/right movement
+- **Right stick**: Rotation
+- **R1 button**: Raise waist height
+- **R2 button**: Lower waist height
+
+Press `Ctrl+C` to stop the policy.
+
 ---
 
+## Extra: Running in Simulation Mode (MuJoCo)
+
+You can now test and develop policies without a physical robot using MuJoCo. to do so set `is_simulation=True` in config.
+
 ## Additional Resources
 
 - [Unitree SDK Documentation](https://github.com/unitreerobotics/unitree_sdk2_python)
-- [GR00T-WholeBodyControl](https://github.com/NVlabs/GR00T-WholeBodyControl)
-- [Holosoma](https://github.com/amazon-far/holosoma)
+- [GR00T Policy Repository](https://huggingface.co/nepyope/GR00T-WholeBodyControl_g1)
 - [LeRobot Documentation](https://github.com/huggingface/lerobot)
-- [Unitree IL LeRobot](https://github.com/unitreerobotics/unitree_IL_lerobot)
+- [Unitree_IL_Lerobot](https://github.com/unitreerobotics/unitree_IL_lerobot)
 
 ---
 
-_Last updated: March 2026_
+_Last updated: December 2025_

docs/source/using_dataset_tools.mdx

@@ -12,7 +12,6 @@ LeRobot provides several utilities for manipulating datasets:
 4. **Add Features** - Add new features to a dataset
 5. **Remove Features** - Remove features from a dataset
 6. **Convert to Video** - Convert image-based datasets to video format for efficient storage
-7. **Show the Info of Datasets** - Show the summary of datasets information such as number of episode etc.
 
 The core implementation is in `lerobot.datasets.dataset_tools`.
 An example script detailing how to use the tools API is available in `examples/dataset/use_dataset_tools.py`.
@@ -96,26 +95,26 @@ Convert an image-based dataset to video format, creating a new LeRobotDataset wh
 # Local-only: Save to a custom output directory (no hub push)
 lerobot-edit-dataset \
     --repo_id lerobot/pusht_image \
-    --operation.type convert_image_to_video \
+    --operation.type convert_to_video \
     --operation.output_dir /path/to/output/pusht_video
 
 # Save with new repo_id (local storage)
 lerobot-edit-dataset \
     --repo_id lerobot/pusht_image \
     --new_repo_id lerobot/pusht_video \
-    --operation.type convert_image_to_video
+    --operation.type convert_to_video
 
 # Convert and push to Hugging Face Hub
 lerobot-edit-dataset \
     --repo_id lerobot/pusht_image \
     --new_repo_id lerobot/pusht_video \
-    --operation.type convert_image_to_video \
+    --operation.type convert_to_video \
     --push_to_hub true
 
 # Convert with custom video codec and quality settings
 lerobot-edit-dataset \
     --repo_id lerobot/pusht_image \
-    --operation.type convert_image_to_video \
+    --operation.type convert_to_video \
     --operation.output_dir outputs/pusht_video \
     --operation.vcodec libsvtav1 \
     --operation.pix_fmt yuv420p \
@@ -125,23 +124,16 @@ lerobot-edit-dataset \
 # Convert only specific episodes
 lerobot-edit-dataset \
     --repo_id lerobot/pusht_image \
-    --operation.type convert_image_to_video \
+    --operation.type convert_to_video \
     --operation.output_dir outputs/pusht_video \
     --operation.episode_indices "[0, 1, 2, 5, 10]"
 
 # Convert with multiple workers for parallel processing
 lerobot-edit-dataset \
     --repo_id lerobot/pusht_image \
-    --operation.type convert_image_to_video \
+    --operation.type convert_to_video \
     --operation.output_dir outputs/pusht_video \
     --operation.num_workers 8
-
-# For memory-constrained systems, users can now specify limits:
-lerobot-edit-dataset \
-    --repo_id lerobot/pusht_image \
-    --operation.type convert_to_video \
-    --operation.max_episodes_per_batch 50 \
-    --operation.max_frames_per_batch 10000
 ```
 
 **Parameters:**
@@ -157,30 +149,6 @@ lerobot-edit-dataset \
 
 **Note:** The resulting dataset will be a proper LeRobotDataset with all cameras encoded as videos in the `videos/` directory, with parquet files containing only metadata (no raw image data). All episodes, stats, and tasks are preserved.
 
-### Show the information of datasets
-
-Show the information of datasets such as number of episode, number of frame, File size and so on.
-No change will be made to the dataset
-
-```bash
-
-# Show dataset information without feature details
-lerobot-edit-dataset \
-    --repo_id lerobot/pusht_image \
-    --operation.type info \
-
-# Show dataset information with feature details
-lerobot-edit-dataset \
-    --repo_id lerobot/pusht_image \
-    --operation.type info \
-    --operation.show_features true
-
-```
-
-**Parameters:**
-
-- `parameters`: The flag to control show or no show dataset information with feature details.(default=false)
-
 ### Push to Hub
 
 Add the `--push_to_hub true` flag to any command to automatically upload the resulting dataset to the Hugging Face Hub:

docs/source/walloss.mdx

@@ -1,80 +0,0 @@
-# WALL-OSS
-
-WALL-OSS is an open-source foundation model for embodied intelligence, proposed by the [XSquare Robot](https://x2robot.com/en/research/68bc2cde8497d7f238dde690) team in 2025. The LeRobot implementation is adapted from their open-source [WallX](https://github.com/X-Square-Robot/wall-x) repository.
-
-X Square Robot’s WALL-OSS is now integrated into Hugging Face’s LeRobot ecosystem. This is an exciting collaborative project between the LeRobot and X Square Robot teams. You can now post-train, evaluate, and deploy WALL-OSS directly through LeRobot. With this, we’re aiming to make it easier for the open-source robotics community to customize and deploy WALL-OSS foundation models. Read and explore WALL-OSS [paper](https://arxiv.org/pdf/2509.11766) and [code](https://github.com/X-Square-Robot/wall-x).
-
-## Model Overview
-
-The WALL-OSS team is building the embodied foundation model to capture and compress the world's most valuable data: the continuous, high-fidelity stream of physical interaction. By creating a direct feedback loop between the model's decisions and the body's lived experience, the emergence of a truly generalizable intelligence is enabled—one that understands not just how the world works, but how to act effectively within it.
-
-<img
-  src="https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/lerobot/walloss-lerobot-paper.png"
-  alt="An overview of WALL-OSS"
-  width="85%"
-/>
-
-Technically, WALL-OSS introduces a tightly coupled multimodal architecture (tightly-coupled MoE structure) that integrates both discrete and continuous action modeling strategies. Through a two-stage training pipeline (Inspiration → Integration), the model gradually unifies semantic reasoning and high-frequency action generation. Its core innovations include:
-
-- **Embodied perception–enhanced multimodal pretraining**: Large-scale training on unified vision–language–action data to strengthen spatial, causal, and manipulation understanding.
-- **Unified Cross-Level Chain-of-Thought (Uni-CoT)**: A single differentiable framework that unifies high-level instruction reasoning, sub-task decomposition, and fine-grained action synthesis, forming a continuous chain from “understanding” to “execution.”
-- **Mixture-of-Experts (MoE) action heads**: Dynamically activating experts depending on the task phase and modeling actions in discrete or continuous space to maintain stable VLM priors.
-- **Two-stage training paradigm**:
-  - **Inspiration stage**: Injecting discrete action priors to strengthen spatial understanding and semantic-action alignment.
-  - **Integration stage**: Using flow matching to achieve high-frequency continuous control.
-
-## Installation Requirements
-
-1. Install LeRobot by following our [Installation Guide](./installation).
-2. Install WallX dependencies by running:
-
-   ```bash
-   pip install -e ".[wallx]"
-   ```
-
-## Usage
-
-To use WallX in LeRobot, specify the policy type as:
-
-```python
-policy.type=wall_x
-```
-
-## Training
-
-For training WallX, you can use the standard LeRobot training script with the appropriate configuration:
-
-```bash
-lerobot-train \
-    --dataset.repo_id=your_dataset \
-    --policy.type=wall_x \
-    --output_dir=./outputs/wallx_training \
-    --job_name=wallx_training \
-    --policy.repo_id=your_repo_id \
-    --policy.pretrained_name_or_path=x-square-robot/wall-oss-flow \
-    --policy.prediction_mode=diffusion \
-    --policy.attn_implementation=eager \
-    --steps=3000 \
-    --policy.device=cuda \
-    --batch_size=32
-```
-
-### Training Arguments
-
-| Argument                       | Description                                                                                                                                                   |
-| ------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------- |
-| `--dataset.repo_id`            | The Hugging Face Hub repository ID for your training dataset (e.g., `lerobot/aloha_sim_insertion_human`)                                                      |
-| `--policy.type`                | Specifies using the WallX policy architecture                                                                                                                 |
-| `--output_dir`                 | Local directory where training checkpoints and logs will be saved                                                                                             |
-| `--job_name`                   | A name identifier for this training run (used in logging/tracking)                                                                                            |
-| `--policy.repo_id`             | Your Hugging Face Hub repo ID where the trained model will be pushed                                                                                          |
-| `--policy.pretrained_path`     | Path to pretrained WallX weights to initialize from (the official WALL-OSS checkpoint)                                                                        |
-| `--policy.prediction_mode`     | The action prediction strategy: `diffusion` or `fast` - `diffusion` uses iterative denoising for action generation, `fast` uses next token prediction instead |
-| `--policy.attn_implementation` | Attention implementation backend - `eager` uses standard PyTorch attention (alternatives include `flash_attention_2` or `sdpa`)                               |
-| `--steps`                      | Total number of training steps to run                                                                                                                         |
-| `--policy.device`              | Device to train on (`cuda` for GPU, `cpu` for CPU)                                                                                                            |
-| `--batch_size`                 | Number of samples per training batch                                                                                                                          |
-
-## License
-
-This model follows the **Apache 2.0 License**, consistent with the original [WallX repository](https://github.com/X-Square-Robot/wall-x).

docs/source/xvla.mdx

@@ -154,7 +154,7 @@ lerobot-train \
 
 ```bash
 lerobot-train \
-  --dataset.repo_id=<USER>/bimanual-so100-handover-cube \
+  --dataset.repo_id=pepijn223/bimanual-so100-handover-cube \
   --output_dir=./outputs/xvla_bimanual \
   --job_name=xvla_so101_training \
   --policy.path="lerobot/xvla-base" \

examples/backward_compatibility/replay.py

@@ -22,7 +22,7 @@ lerobot-replay \
     --robot.type=so100_follower \
     --robot.port=/dev/tty.usbmodem58760431541 \
     --robot.id=black \
-    --dataset.repo_id=<USER>/record-test \
+    --dataset.repo_id=aliberts/record-test \
     --dataset.episode=2
 ```
 """
@@ -41,7 +41,8 @@ from lerobot.robots import (  # noqa: F401
     RobotConfig,
     koch_follower,
     make_robot_from_config,
-    so_follower,
+    so100_follower,
+    so101_follower,
 )
 from lerobot.utils.constants import ACTION
 from lerobot.utils.robot_utils import precise_sleep
@@ -57,7 +58,7 @@ class DatasetReplayConfig:
     repo_id: str
     # Episode to replay.
     episode: int
-    # Root directory where the dataset will be stored (e.g. 'dataset/path'). If None, defaults to $HF_LEROBOT_HOME/repo_id.
+    # Root directory where the dataset will be stored (e.g. 'dataset/path').
     root: str | Path | None = None
     # Limit the frames per second. By default, uses the policy fps.
     fps: int = 30
@@ -78,28 +79,27 @@ def replay(cfg: ReplayConfig):
 
     robot = make_robot_from_config(cfg.robot)
     dataset = LeRobotDataset(cfg.dataset.repo_id, root=cfg.dataset.root, episodes=[cfg.dataset.episode])
-    actions = dataset.select_columns(ACTION)
+    actions = dataset.hf_dataset.select_columns(ACTION)
     robot.connect()
 
-    try:
-        log_say("Replaying episode", cfg.play_sounds, blocking=True)
-        for idx in range(dataset.num_frames):
-            start_episode_t = time.perf_counter()
+    log_say("Replaying episode", cfg.play_sounds, blocking=True)
+    for idx in range(dataset.num_frames):
+        start_episode_t = time.perf_counter()
 
-            action_array = actions[idx][ACTION]
-            action = {}
-            for i, name in enumerate(dataset.features[ACTION]["names"]):
-                key = f"{name.removeprefix('main_')}.pos"
-                action[key] = action_array[i].item()
+        action_array = actions[idx][ACTION]
+        action = {}
+        for i, name in enumerate(dataset.features[ACTION]["names"]):
+            key = f"{name.removeprefix('main_')}.pos"
+            action[key] = action_array[i].item()
 
-            action["shoulder_lift.pos"] = -(action["shoulder_lift.pos"] - 90)
-            action["elbow_flex.pos"] -= 90
-            robot.send_action(action)
+        action["shoulder_lift.pos"] = -(action["shoulder_lift.pos"] - 90)
+        action["elbow_flex.pos"] -= 90
+        robot.send_action(action)
 
-            dt_s = time.perf_counter() - start_episode_t
-            precise_sleep(max(1 / dataset.fps - dt_s, 0.0))
-    finally:
-        robot.disconnect()
+        dt_s = time.perf_counter() - start_episode_t
+        precise_sleep(1 / dataset.fps - dt_s)
+
+    robot.disconnect()
 
 
 if __name__ == "__main__":

examples/dataset/create_progress_videos.py

@@ -1,680 +0,0 @@
-#!/usr/bin/env python
-
-# Copyright 2025 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.
-
-"""
-Create MP4 (or GIF) videos with sarm_progress overlay for specified episodes.
-
-Downloads datasets from HuggingFace, seeks directly into the episode segment
-of the source video, draws a progress line on each frame, and writes the result.
-
-Usage:
-    python examples/dataset/create_progress_videos.py \
-        --repo-id lerobot-data-collection/level2_final_quality3 \
-        --episode 1100
-
-    python examples/dataset/create_progress_videos.py \
-        --repo-id lerobot-data-collection/level2_final_quality3 \
-        --episode 1100 \
-        --camera-key observation.images.top \
-        --output-dir ./my_videos \
-        --gif
-"""
-
-from __future__ import annotations
-
-import argparse
-import json
-import logging
-import subprocess
-from pathlib import Path
-
-import cv2
-import numpy as np
-import pandas as pd
-from huggingface_hub import snapshot_download
-
-GRAPH_Y_TOP_FRAC = 0.01
-GRAPH_Y_BOT_FRAC = 0.99
-LINE_THICKNESS = 3
-SHADOW_THICKNESS = 6
-REF_ALPHA = 0.45
-FILL_ALPHA = 0.55
-SCORE_FONT_SCALE = 0.8
-TASK_FONT_SCALE = 0.55
-
-
-def download_episode_metadata(repo_id: str, episode: int) -> Path:
-    """Download only the metadata and sarm_progress files for a dataset.
-
-    Args:
-        repo_id: HuggingFace dataset repository ID.
-        episode: Episode index (used for logging only; all meta is fetched).
-
-    Returns:
-        Local cache path for the downloaded snapshot.
-    """
-    logging.info("[1/4] Downloading metadata for %s (episode %d) ...", repo_id, episode)
-    local_path = Path(
-        snapshot_download(
-            repo_id=repo_id,
-            repo_type="dataset",
-            allow_patterns=["meta/**", "sarm_progress.parquet"],
-            ignore_patterns=["*.mp4"],
-        )
-    )
-    return local_path
-
-
-def load_episode_meta(local_path: Path, episode: int, camera_key: str | None) -> dict:
-    """Read info.json and episode parquet to resolve fps, video path, and timestamps.
-
-    Args:
-        local_path: Local cache directory containing meta/.
-        episode: Episode index to look up.
-        camera_key: Camera observation key (e.g. "observation.images.base").
-            If None, the first available video key is used.
-
-    Returns:
-        Dict with keys: fps, camera, video_rel, chunk_index, file_index,
-        from_ts, to_ts, task_name.
-    """
-    info = json.loads((local_path / "meta" / "info.json").read_text())
-    fps = info["fps"]
-    features = info["features"]
-
-    video_keys = [k for k, v in features.items() if v.get("dtype") == "video"]
-    if not video_keys:
-        raise RuntimeError("No video keys found in dataset features")
-
-    if camera_key is not None:
-        if camera_key not in video_keys:
-            raise RuntimeError(f"camera_key='{camera_key}' not found. Available: {video_keys}")
-        selected_camera = camera_key
-    else:
-        selected_camera = video_keys[0]
-    logging.info("   fps=%d  camera='%s'  all_cams=%s", fps, selected_camera, video_keys)
-
-    episode_rows = []
-    for parquet_file in sorted((local_path / "meta" / "episodes").glob("**/*.parquet")):
-        episode_rows.append(pd.read_parquet(parquet_file))
-    episode_df = pd.concat(episode_rows, ignore_index=True)
-    row = episode_df[episode_df["episode_index"] == episode]
-    if row.empty:
-        raise RuntimeError(f"Episode {episode} not found in episode metadata")
-    row = row.iloc[0]
-
-    chunk_col = f"videos/{selected_camera}/chunk_index"
-    file_col = f"videos/{selected_camera}/file_index"
-    ts_from_col = f"videos/{selected_camera}/from_timestamp"
-    ts_to_col = f"videos/{selected_camera}/to_timestamp"
-
-    if chunk_col not in row.index:
-        chunk_col = f"{selected_camera}/chunk_index"
-        file_col = f"{selected_camera}/file_index"
-        ts_from_col = f"{selected_camera}/from_timestamp"
-        ts_to_col = f"{selected_camera}/to_timestamp"
-    if chunk_col not in row.index:
-        raise RuntimeError(
-            f"Cannot find video metadata columns for {selected_camera}.\nAvailable: {list(row.index)}"
-        )
-
-    chunk_index = int(row[chunk_col])
-    file_index = int(row[file_col])
-    from_timestamp = float(row[ts_from_col])
-    to_timestamp = float(row[ts_to_col])
-
-    video_template = info.get(
-        "video_path", "videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4"
-    )
-    video_rel = video_template.format(
-        video_key=selected_camera,
-        chunk_index=chunk_index,
-        file_index=file_index,
-    )
-
-    task_name = _resolve_task_name(row, local_path)
-
-    return {
-        "fps": fps,
-        "camera": selected_camera,
-        "video_rel": video_rel,
-        "chunk_index": chunk_index,
-        "file_index": file_index,
-        "from_ts": from_timestamp,
-        "to_ts": to_timestamp,
-        "task_name": task_name,
-    }
-
-
-def _resolve_task_name(row: pd.Series, local_path: Path) -> str:
-    """Best-effort extraction of the task name for an episode row.
-
-    Args:
-        row: Single-episode row from the episodes parquet.
-        local_path: Dataset cache root.
-
-    Returns:
-        Task name string, or empty string if unavailable.
-    """
-    try:
-        if "tasks" in row.index and row["tasks"] is not None:
-            tasks_val = row["tasks"]
-            if isinstance(tasks_val, (list, tuple, np.ndarray)) and len(tasks_val) > 0:
-                return str(tasks_val[0])
-            return str(tasks_val).strip("[]'")
-
-        tasks_parquet = local_path / "meta" / "tasks.parquet"
-        if tasks_parquet.exists():
-            tasks_df = pd.read_parquet(tasks_parquet)
-            task_idx = int(row.get("task_index", 0)) if "task_index" in row.index else 0
-            match = tasks_df[tasks_df["task_index"] == task_idx]
-            if not match.empty:
-                return str(match.index[0])
-    except Exception as exc:
-        logging.warning("Could not load task name: %s", exc)
-    return ""
-
-
-def download_video_file(repo_id: str, local_path: Path, video_rel: str) -> Path:
-    """Download the specific video file if not already cached.
-
-    Args:
-        repo_id: HuggingFace dataset repository ID.
-        local_path: Local cache directory.
-        video_rel: Relative path to the video file within the dataset.
-
-    Returns:
-        Absolute path to the downloaded video file.
-    """
-    video_path = local_path / video_rel
-    if video_path.exists():
-        logging.info("   Video already cached: %s", video_path)
-        return video_path
-    logging.info("[2/4] Downloading video file %s ...", video_rel)
-    snapshot_download(
-        repo_id=repo_id,
-        repo_type="dataset",
-        local_dir=str(local_path),
-        allow_patterns=[video_rel],
-    )
-    if not video_path.exists():
-        raise RuntimeError(f"Video not found after download: {video_path}")
-    return video_path
-
-
-def load_progress_data(local_path: Path, episode: int) -> np.ndarray | None:
-    """Load sarm_progress values for an episode.
-
-    Args:
-        local_path: Dataset cache root.
-        episode: Episode index.
-
-    Returns:
-        Sorted (N, 2) array of (frame_index, progress), or None if unavailable.
-    """
-    parquet_path = local_path / "sarm_progress.parquet"
-    if not parquet_path.exists():
-        logging.warning("sarm_progress.parquet not found")
-        return None
-    df = pd.read_parquet(parquet_path)
-    logging.info("   sarm_progress.parquet columns: %s", list(df.columns))
-    episode_df = df[df["episode_index"] == episode].copy()
-    if episode_df.empty:
-        logging.warning("No sarm_progress rows for episode %d", episode)
-        return None
-    episode_df = episode_df.sort_values("frame_index")
-
-    if "progress_dense" in episode_df.columns and episode_df["progress_dense"].notna().any():
-        progress_column = "progress_dense"
-    elif "progress_sparse" in episode_df.columns:
-        progress_column = "progress_sparse"
-    else:
-        progress_columns = [c for c in episode_df.columns if "progress" in c.lower()]
-        if not progress_columns:
-            return None
-        progress_column = progress_columns[0]
-
-    logging.info("   Using progress column: '%s'", progress_column)
-    return episode_df[["frame_index", progress_column]].rename(columns={progress_column: "progress"}).values
-
-
-def _precompute_pixel_coords(
-    progress_data: np.ndarray,
-    num_frames: int,
-    frame_width: int,
-    frame_height: int,
-) -> np.ndarray:
-    """Map progress samples to pixel coordinates for overlay drawing.
-
-    Args:
-        progress_data: (N, 2) array of (frame_index, progress).
-        num_frames: Total number of video frames.
-        frame_width: Video width in pixels.
-        frame_height: Video height in pixels.
-
-    Returns:
-        (N, 2) array of (x, y) pixel coordinates.
-    """
-    frame_indices = progress_data[:, 0].astype(float)
-    progress_values = np.clip(progress_data[:, 1].astype(float), 0.0, 1.0)
-
-    y_top = int(frame_height * GRAPH_Y_TOP_FRAC)
-    y_bot = int(frame_height * GRAPH_Y_BOT_FRAC)
-    graph_height = y_bot - y_top
-
-    x_coords = (frame_indices / (num_frames - 1) * (frame_width - 1)).astype(int)
-    y_coords = (y_bot - progress_values * graph_height).astype(int)
-
-    return np.stack([x_coords, y_coords], axis=1)
-
-
-def _progress_color(normalized_position: float) -> tuple[int, int, int]:
-    """Interpolate BGR color from red to green based on position in [0, 1].
-
-    Args:
-        normalized_position: Value in [0, 1] indicating how far along the episode.
-
-    Returns:
-        BGR color tuple.
-    """
-    red = int(255 * (1.0 - normalized_position))
-    green = int(255 * normalized_position)
-    return (0, green, red)
-
-
-def _prerender_fill_polygon(
-    pixel_coords: np.ndarray,
-    frame_width: int,
-    frame_height: int,
-) -> np.ndarray:
-    """Pre-render the grey fill polygon under the progress curve as a BGRA image.
-
-    Args:
-        pixel_coords: (N, 2) array of (x, y) pixel coordinates.
-        frame_width: Video width in pixels.
-        frame_height: Video height in pixels.
-
-    Returns:
-        BGRA image array of shape (frame_height, frame_width, 4).
-    """
-    y_bot = int(frame_height * GRAPH_Y_BOT_FRAC)
-    fill_image = np.zeros((frame_height, frame_width, 4), dtype=np.uint8)
-    polygon = np.concatenate(
-        [
-            pixel_coords,
-            [[pixel_coords[-1][0], y_bot], [pixel_coords[0][0], y_bot]],
-        ],
-        axis=0,
-    ).astype(np.int32)
-    cv2.fillPoly(fill_image, [polygon], color=(128, 128, 128, int(255 * FILL_ALPHA)))
-    return fill_image
-
-
-def _alpha_composite_region(base: np.ndarray, overlay_bgra: np.ndarray, x_limit: int) -> None:
-    """Blend BGRA overlay onto BGR base in-place, up to x_limit columns.
-
-    Args:
-        base: BGR frame to draw on (modified in-place).
-        overlay_bgra: BGRA overlay image.
-        x_limit: Only blend columns [0, x_limit).
-    """
-    if x_limit <= 0:
-        return
-    region_base = base[:, :x_limit]
-    region_overlay = overlay_bgra[:, :x_limit]
-    alpha = region_overlay[:, :, 3:4].astype(np.float32) / 255.0
-    region_base[:] = np.clip(
-        region_overlay[:, :, :3].astype(np.float32) * alpha + region_base.astype(np.float32) * (1.0 - alpha),
-        0,
-        255,
-    ).astype(np.uint8)
-
-
-def _draw_text_outlined(
-    frame: np.ndarray,
-    text: str,
-    position: tuple[int, int],
-    font_scale: float,
-    thickness: int = 1,
-) -> None:
-    """Draw white text with a dark outline for readability on any background.
-
-    Args:
-        frame: BGR image to draw on (modified in-place).
-        text: String to render.
-        position: (x, y) bottom-left corner of the text.
-        font_scale: OpenCV font scale.
-        thickness: Text stroke thickness.
-    """
-    font = cv2.FONT_HERSHEY_SIMPLEX
-    cv2.putText(frame, text, position, font, font_scale, (0, 0, 0), thickness + 2, cv2.LINE_AA)
-    cv2.putText(frame, text, position, font, font_scale, (255, 255, 255), thickness, cv2.LINE_AA)
-
-
-def composite_progress_video(
-    video_path: Path,
-    from_timestamp: float,
-    to_timestamp: float,
-    progress_data: np.ndarray,
-    output_path: Path,
-    fps: float,
-    task_name: str = "",
-) -> Path:
-    """Read episode frames by seeking into the source video, draw progress overlay, write output.
-
-    Uses cv2.CAP_PROP_POS_MSEC to seek directly into the source video,
-    eliminating the need for an intermediate clip file.
-
-    Args:
-        video_path: Path to the full source video file.
-        from_timestamp: Start timestamp of the episode in seconds.
-        to_timestamp: End timestamp of the episode in seconds.
-        progress_data: (N, 2) array of (frame_index, progress).
-        output_path: Path to write the output MP4.
-        fps: Frames per second for the output video.
-        task_name: Optional task name to display at the top of the video.
-
-    Returns:
-        Path to the written output file (MP4).
-    """
-    capture = cv2.VideoCapture(str(video_path))
-    try:
-        capture.set(cv2.CAP_PROP_POS_MSEC, from_timestamp * 1000)
-
-        frame_width = int(capture.get(cv2.CAP_PROP_FRAME_WIDTH))
-        frame_height = int(capture.get(cv2.CAP_PROP_FRAME_HEIGHT))
-        duration_seconds = to_timestamp - from_timestamp
-        num_frames = int(round(duration_seconds * fps))
-
-        logging.info(
-            "   Video: %dx%d, %d frames @ %.1f fps (%.2fs)",
-            frame_width,
-            frame_height,
-            num_frames,
-            fps,
-            duration_seconds,
-        )
-
-        pixel_coords = _precompute_pixel_coords(progress_data, num_frames, frame_width, frame_height)
-        y_ref = int(frame_height * GRAPH_Y_TOP_FRAC)
-
-        fill_image = _prerender_fill_polygon(pixel_coords, frame_width, frame_height)
-
-        ref_line_image = np.zeros((frame_height, frame_width, 4), dtype=np.uint8)
-        cv2.line(
-            ref_line_image,
-            (0, y_ref),
-            (frame_width - 1, y_ref),
-            (200, 200, 200, int(255 * REF_ALPHA)),
-            1,
-            cv2.LINE_AA,
-        )
-
-        frame_indices = progress_data[:, 0].astype(int)
-        progress_values = progress_data[:, 1].astype(float)
-
-        logging.info("[3/4] Compositing %d frames ...", num_frames)
-        fourcc = cv2.VideoWriter_fourcc(*"mp4v")
-        writer = cv2.VideoWriter(str(output_path), fourcc, fps, (frame_width, frame_height))
-
-        for frame_idx in range(num_frames):
-            ret, frame = capture.read()
-            if not ret:
-                break
-
-            drawn_count = int(np.searchsorted(frame_indices, frame_idx, side="right"))
-            x_current = (
-                int(pixel_coords[min(drawn_count, len(pixel_coords)) - 1][0]) + 1 if drawn_count > 0 else 0
-            )
-
-            _alpha_composite_region(frame, ref_line_image, frame_width)
-            _alpha_composite_region(frame, fill_image, x_current)
-
-            if drawn_count >= 2:
-                time_position = (drawn_count - 1) / max(len(progress_values) - 1, 1)
-                line_color = _progress_color(time_position)
-                points = pixel_coords[:drawn_count].reshape(-1, 1, 2).astype(np.int32)
-                cv2.polylines(
-                    frame,
-                    [points],
-                    isClosed=False,
-                    color=(255, 255, 255),
-                    thickness=SHADOW_THICKNESS,
-                    lineType=cv2.LINE_AA,
-                )
-                cv2.polylines(
-                    frame,
-                    [points],
-                    isClosed=False,
-                    color=line_color,
-                    thickness=LINE_THICKNESS,
-                    lineType=cv2.LINE_AA,
-                )
-
-            if drawn_count > 0:
-                score = float(progress_values[min(drawn_count, len(progress_values)) - 1])
-                score_text = f"{score:.2f}"
-                (text_width, _), _ = cv2.getTextSize(
-                    score_text, cv2.FONT_HERSHEY_SIMPLEX, SCORE_FONT_SCALE, 2
-                )
-                score_x = frame_width - text_width - 12
-                score_y = frame_height - 12
-                time_position = (drawn_count - 1) / max(len(progress_values) - 1, 1)
-                score_color = _progress_color(time_position)
-                cv2.putText(
-                    frame,
-                    score_text,
-                    (score_x, score_y),
-                    cv2.FONT_HERSHEY_SIMPLEX,
-                    SCORE_FONT_SCALE,
-                    (0, 0, 0),
-                    4,
-                    cv2.LINE_AA,
-                )
-                cv2.putText(
-                    frame,
-                    score_text,
-                    (score_x, score_y),
-                    cv2.FONT_HERSHEY_SIMPLEX,
-                    SCORE_FONT_SCALE,
-                    score_color,
-                    2,
-                    cv2.LINE_AA,
-                )
-
-            if task_name:
-                (text_width, _), _ = cv2.getTextSize(task_name, cv2.FONT_HERSHEY_SIMPLEX, TASK_FONT_SCALE, 1)
-                task_x = max((frame_width - text_width) // 2, 4)
-                _draw_text_outlined(frame, task_name, (task_x, 22), TASK_FONT_SCALE)
-
-            writer.write(frame)
-            if frame_idx % 100 == 0:
-                logging.info("   Frame %d/%d ...", frame_idx, num_frames)
-
-        writer.release()
-    finally:
-        capture.release()
-
-    logging.info("   MP4 written: %s", output_path)
-    return output_path
-
-
-def convert_mp4_to_gif(mp4_path: Path) -> Path:
-    """Convert an MP4 to an optimized GIF using ffmpeg palette generation.
-
-    Args:
-        mp4_path: Path to the source MP4 file.
-
-    Returns:
-        Path to the generated GIF file.
-    """
-    capture = cv2.VideoCapture(str(mp4_path))
-    frame_width = int(capture.get(cv2.CAP_PROP_FRAME_WIDTH))
-    capture.release()
-
-    gif_path = mp4_path.with_suffix(".gif")
-    palette_path = mp4_path.parent / "_palette.png"
-
-    logging.info("[4/4] Converting to GIF ...")
-    result_palette = subprocess.run(  # nosec B607
-        [
-            "ffmpeg",
-            "-y",
-            "-i",
-            str(mp4_path),
-            "-vf",
-            f"fps=10,scale={frame_width}:-1:flags=lanczos,palettegen=max_colors=128:stats_mode=diff",
-            "-update",
-            "1",
-            str(palette_path),
-        ],
-        capture_output=True,
-        text=True,
-    )
-    if result_palette.returncode != 0:
-        logging.warning("palettegen failed:\n%s", result_palette.stderr[-500:])
-
-    result_gif = subprocess.run(  # nosec B607
-        [
-            "ffmpeg",
-            "-y",
-            "-i",
-            str(mp4_path),
-            "-i",
-            str(palette_path),
-            "-filter_complex",
-            f"fps=10,scale={frame_width}:-1:flags=lanczos[v];[v][1:v]paletteuse=dither=bayer:bayer_scale=3",
-            str(gif_path),
-        ],
-        capture_output=True,
-        text=True,
-    )
-    if result_gif.returncode != 0:
-        logging.warning("GIF encode failed:\n%s", result_gif.stderr[-500:])
-
-    palette_path.unlink(missing_ok=True)
-    logging.info("   GIF written: %s", gif_path)
-    return gif_path
-
-
-def process_dataset(
-    repo_id: str,
-    episode: int,
-    camera_key: str | None,
-    output_dir: Path,
-    create_gif: bool = False,
-) -> Path | None:
-    """Full pipeline: download, extract metadata, composite progress, write output.
-
-    Args:
-        repo_id: HuggingFace dataset repository ID.
-        episode: Episode index.
-        camera_key: Camera key to use, or None for auto-selection.
-        output_dir: Directory to write output files.
-        create_gif: If True, also generate a GIF from the MP4.
-
-    Returns:
-        Path to the final output file, or None on failure.
-    """
-    safe_name = repo_id.replace("/", "_")
-    logging.info("Processing: %s  |  episode %d", repo_id, episode)
-
-    local_path = download_episode_metadata(repo_id, episode)
-    logging.info("   Local cache: %s", local_path)
-
-    episode_meta = load_episode_meta(local_path, episode, camera_key)
-    logging.info("   Episode meta: %s", episode_meta)
-
-    video_path = download_video_file(repo_id, local_path, episode_meta["video_rel"])
-
-    progress_data = load_progress_data(local_path, episode)
-    if progress_data is None:
-        logging.error("Could not load sarm_progress data. Skipping overlay.")
-        return None
-
-    logging.info("   Progress frames: %d", len(progress_data))
-
-    output_path = output_dir / f"{safe_name}_ep{episode}_progress.mp4"
-    final_path = composite_progress_video(
-        video_path=video_path,
-        from_timestamp=episode_meta["from_ts"],
-        to_timestamp=episode_meta["to_ts"],
-        progress_data=progress_data,
-        output_path=output_path,
-        fps=episode_meta["fps"],
-        task_name=episode_meta.get("task_name", ""),
-    )
-
-    if create_gif:
-        final_path = convert_mp4_to_gif(final_path)
-
-    logging.info("Done: %s", final_path)
-    return final_path
-
-
-def main() -> None:
-    parser = argparse.ArgumentParser(
-        description="Create MP4/GIF videos with sarm_progress overlay for dataset episodes."
-    )
-    parser.add_argument(
-        "--repo-id",
-        type=str,
-        required=True,
-        help="HuggingFace dataset repository ID (e.g. 'lerobot-data-collection/level2_final_quality3').",
-    )
-    parser.add_argument(
-        "--episode",
-        type=int,
-        required=True,
-        help="Episode index to visualize.",
-    )
-    parser.add_argument(
-        "--camera-key",
-        type=str,
-        default=None,
-        help="Camera observation key (e.g. 'observation.images.base'). Auto-selects first camera if omitted.",
-    )
-    parser.add_argument(
-        "--output-dir",
-        type=Path,
-        default=Path("progress_videos"),
-        help="Directory to write output files (default: ./progress_videos).",
-    )
-    parser.add_argument(
-        "--gif",
-        action="store_true",
-        help="Also generate a GIF from the MP4 output.",
-    )
-    args = parser.parse_args()
-
-    logging.basicConfig(level=logging.INFO, format="%(levelname)s: %(message)s")
-
-    args.output_dir.mkdir(parents=True, exist_ok=True)
-
-    result = process_dataset(
-        repo_id=args.repo_id,
-        episode=args.episode,
-        camera_key=args.camera_key,
-        output_dir=args.output_dir,
-        create_gif=args.gif,
-    )
-
-    if result:
-        logging.info("Output: %s", result)
-
-
-if __name__ == "__main__":
-    main()

examples/dataset/load_lerobot_dataset.py

@@ -32,8 +32,7 @@ import torch
 from huggingface_hub import HfApi
 
 import lerobot
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.lerobot_dataset import LeRobotDataset, LeRobotDatasetMetadata
 
 
 def main():
@@ -88,8 +87,9 @@ def main():
     # The previous metadata class is contained in the 'meta' attribute of the dataset:
     print(dataset.meta)
 
-    # You can inspect the dataset using its repr:
-    print(dataset)
+    # LeRobotDataset actually wraps an underlying Hugging Face dataset
+    # (see https://huggingface.co/docs/datasets for more information).
+    print(dataset.hf_dataset)
 
     # LeRobot datasets also subclasses PyTorch datasets so you can do everything you know and love from working
     # with the latter, like iterating through the dataset.

examples/dataset/slurm_compute_rabc.py

@@ -1,490 +0,0 @@
-#!/usr/bin/env python
-
-# Copyright 2025 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.
-
-"""
-SLURM-distributed SARM RA-BC annotation pipeline.
-
-Computes SARM progress values for all frames in a dataset, distributed across
-SLURM workers, then merges the shards into a single sarm_progress.parquet.
-
-Two subcommands, each a separate SLURM submission:
-
-  compute    – N workers, each computes progress for a subset of episodes
-  aggregate  – 1 worker, merges N shards into sarm_progress.parquet, pushes to hub
-
-Usage:
-    python slurm_compute_rabc.py compute \\
-        --repo-id user/dataset --reward-model-path user/sarm_model \\
-        --stride 10 --device cpu --workers 50 --partition cpu
-
-    python slurm_compute_rabc.py aggregate \\
-        --repo-id user/dataset --reward-model-path user/sarm_model \\
-        --partition cpu --push-to-hub
-"""
-
-import argparse
-from pathlib import Path
-
-from datatrove.executor import LocalPipelineExecutor
-from datatrove.executor.slurm import SlurmPipelineExecutor
-from datatrove.pipeline.base import PipelineStep
-
-
-class ComputeProgressShards(PipelineStep):
-    """Each worker computes SARM progress for its assigned episodes."""
-
-    def __init__(
-        self, repo_id, reward_model_path, stride=1, head_mode="sparse", device="cpu", shard_dir="rabc_shards"
-    ):
-        super().__init__()
-        if stride < 1:
-            raise ValueError(f"stride must be >= 1, got {stride}")
-        self.repo_id = repo_id
-        self.reward_model_path = reward_model_path
-        self.stride = stride
-        self.head_mode = head_mode
-        self.device = device
-        self.shard_dir = shard_dir
-
-    def run(self, data=None, rank: int = 0, world_size: int = 1):
-        import logging
-        from pathlib import Path
-
-        import numpy as np
-        import pyarrow as pa
-        import pyarrow.parquet as pq
-        import torch
-        from tqdm import tqdm
-
-        from lerobot.policies.sarm.compute_rabc_weights import (
-            generate_all_frame_indices,
-            interpolate_progress,
-            load_sarm_resources,
-        )
-        from lerobot.utils.utils import init_logging
-
-        init_logging()
-
-        dataset, reward_model, preprocess = load_sarm_resources(
-            self.repo_id,
-            self.reward_model_path,
-            self.device,
-        )
-
-        if hasattr(preprocess, "eval"):
-            preprocess.eval()
-        for step in preprocess.steps:
-            if hasattr(step, "eval"):
-                step.eval()
-
-        image_key = reward_model.config.image_key
-        state_key = reward_model.config.state_key
-        frame_gap = reward_model.config.frame_gap
-        center_idx = reward_model.config.n_obs_steps // 2
-
-        dual_mode = reward_model.config.uses_dual_heads
-        compute_sparse = self.head_mode in ("sparse", "both") or not dual_mode
-        compute_dense = self.head_mode in ("dense", "both") and dual_mode
-
-        my_episodes = list(range(dataset.num_episodes))[rank::world_size]
-        if not my_episodes:
-            logging.info(f"Rank {rank}: no episodes assigned")
-            return
-        logging.info(f"Rank {rank}: {len(my_episodes)} / {dataset.num_episodes} episodes")
-
-        all_rows = []
-
-        for ep_idx in tqdm(my_episodes, desc=f"Rank {rank}"):
-            ep = dataset.meta.episodes[ep_idx]
-            ep_start, ep_end = ep["dataset_from_index"], ep["dataset_to_index"]
-            task = dataset[ep_start].get("task", "perform the task")
-
-            all_ep_indices = generate_all_frame_indices(ep_start, ep_end, frame_gap)
-            if self.stride > 1:
-                compute_indices = [i for i in all_ep_indices if (i - ep_start) % self.stride == 0]
-                if (ep_end - 1) not in compute_indices:
-                    compute_indices.append(ep_end - 1)
-                compute_indices = sorted(set(compute_indices))
-            else:
-                compute_indices = all_ep_indices
-
-            frame_results = {}
-            for qi in tqdm(compute_indices, desc=f"  Ep {ep_idx}", leave=False):
-                try:
-                    sample = dataset[qi]
-                    batch = {
-                        image_key: sample[image_key],
-                        "task": task,
-                        "index": qi,
-                        "episode_index": ep_idx,
-                    }
-                    if state_key in sample:
-                        batch[state_key] = sample[state_key]
-
-                    with torch.no_grad():
-                        processed = preprocess(batch)
-                        vf = processed["video_features"].to(self.device)
-                        tf = processed["text_features"].to(self.device)
-                        sf = processed.get("state_features")
-                        if sf is not None:
-                            sf = sf.to(self.device)
-                        lengths = processed.get("lengths")
-
-                        sparse_val = dense_val = np.nan
-                        if compute_sparse:
-                            r = reward_model.calculate_rewards(
-                                text_embeddings=tf,
-                                video_embeddings=vf,
-                                state_features=sf,
-                                lengths=lengths,
-                                return_all_frames=True,
-                                head_mode="sparse",
-                            )
-                            sparse_val = float(r[0, center_idx] if r.ndim == 2 else r[center_idx])
-                        if compute_dense:
-                            r = reward_model.calculate_rewards(
-                                text_embeddings=tf,
-                                video_embeddings=vf,
-                                state_features=sf,
-                                lengths=lengths,
-                                return_all_frames=True,
-                                head_mode="dense",
-                            )
-                            dense_val = float(r[0, center_idx] if r.ndim == 2 else r[center_idx])
-
-                        frame_results[qi] = (sparse_val, dense_val)
-                except Exception as e:
-                    logging.warning(f"Failed frame {qi}: {e}")
-
-            if not frame_results:
-                logging.warning(f"Episode {ep_idx}: all frames failed, skipping")
-                continue
-
-            # Interpolate to all frames in this episode
-            computed_idx = np.array(sorted(frame_results.keys()))
-            all_frame_arr = np.arange(ep_start, ep_end)
-
-            sparse_vals = np.array([frame_results[i][0] for i in computed_idx]) if compute_sparse else None
-            dense_vals = np.array([frame_results[i][1] for i in computed_idx]) if compute_dense else None
-
-            if self.stride > 1 and len(computed_idx) > 1:
-                if compute_sparse:
-                    sparse_vals = interpolate_progress(computed_idx, sparse_vals, all_frame_arr)
-                if compute_dense:
-                    dense_vals = interpolate_progress(computed_idx, dense_vals, all_frame_arr)
-                output_frames = all_frame_arr
-            else:
-                # Use only successfully computed frames to avoid indexing mismatch on failures
-                output_frames = computed_idx
-
-            for i, fi in enumerate(output_frames):
-                row = {"index": int(fi), "episode_index": ep_idx, "frame_index": int(fi - ep_start)}
-                if compute_sparse:
-                    row["progress_sparse"] = float(sparse_vals[i])
-                if compute_dense:
-                    row["progress_dense"] = float(dense_vals[i])
-                all_rows.append(row)
-
-        if all_rows:
-            import pandas as pd
-
-            df = pd.DataFrame(all_rows).sort_values("index").reset_index(drop=True)
-            table = pa.Table.from_pandas(df, preserve_index=False)
-            table = table.replace_schema_metadata({b"reward_model_path": self.reward_model_path.encode()})
-            shard_dir = Path(self.shard_dir)
-            shard_dir.mkdir(parents=True, exist_ok=True)
-            out = shard_dir / f"shard_{rank:05d}.parquet"
-            pq.write_table(table, out)
-            logging.info(f"Rank {rank}: saved {len(df)} rows to {out}")
-
-
-class AggregateProgress(PipelineStep):
-    """Merge all shard parquets into final sarm_progress.parquet."""
-
-    def __init__(self, repo_id, reward_model_path, shard_dir="rabc_shards", push_to_hub=False):
-        super().__init__()
-        self.repo_id = repo_id
-        self.reward_model_path = reward_model_path
-        self.shard_dir = shard_dir
-        self.push_to_hub = push_to_hub
-
-    def run(self, data=None, rank: int = 0, world_size: int = 1):
-        import datetime
-        import logging
-        import os
-        from pathlib import Path
-
-        import pandas as pd
-        import pyarrow as pa
-        import pyarrow.parquet as pq
-
-        from lerobot.datasets.lerobot_dataset import LeRobotDataset
-        from lerobot.utils.utils import init_logging
-
-        init_logging()
-        if rank != 0:
-            return
-
-        shard_dir = Path(self.shard_dir)
-        shards = sorted(shard_dir.glob("shard_*.parquet"))
-        if not shards:
-            raise FileNotFoundError(f"No shards found in {shard_dir}")
-
-        # Log shard modification time range to help detect stale files
-        mtimes = [os.path.getmtime(s) for s in shards]
-        oldest = datetime.datetime.fromtimestamp(min(mtimes)).isoformat(timespec="seconds")
-        newest = datetime.datetime.fromtimestamp(max(mtimes)).isoformat(timespec="seconds")
-        logging.info(f"Aggregating {len(shards)} shards (oldest: {oldest}, newest: {newest})")
-
-        df = pd.concat([pd.read_parquet(s) for s in shards], ignore_index=True)
-        df = df.sort_values("index").reset_index(drop=True)
-
-        table = pa.Table.from_pandas(df, preserve_index=False)
-        table = table.replace_schema_metadata({b"reward_model_path": self.reward_model_path.encode()})
-
-        temp_ds = LeRobotDataset(self.repo_id, download_videos=False)
-        out_path = Path(temp_ds.root) / "sarm_progress.parquet"
-        out_path.parent.mkdir(parents=True, exist_ok=True)
-        pq.write_table(table, out_path)
-        logging.info(f"Saved {len(df)} rows to {out_path}")
-
-        for col in ["progress_sparse", "progress_dense"]:
-            if col in df.columns:
-                v = df[col].dropna()
-                logging.info(
-                    f"{col}: mean={v.mean():.4f} std={v.std():.4f} min={v.min():.4f} max={v.max():.4f}"
-                )
-
-        if self.push_to_hub:
-            from huggingface_hub import HfApi
-
-            api = HfApi()
-            hub_path = "sarm_progress.parquet"
-            logging.info(f"Uploading to {self.repo_id}/{hub_path}")
-            api.upload_file(
-                path_or_fileobj=str(out_path),
-                path_in_repo=hub_path,
-                repo_id=self.repo_id,
-                repo_type="dataset",
-            )
-            logging.info(f"Uploaded: https://huggingface.co/datasets/{self.repo_id}/blob/main/{hub_path}")
-
-
-def make_compute_executor(
-    repo_id,
-    reward_model_path,
-    stride,
-    head_mode,
-    device,
-    shard_dir,
-    logs_dir,
-    job_name,
-    slurm,
-    workers,
-    partition,
-    cpus_per_task,
-    mem_per_cpu,
-):
-    kwargs = {
-        "pipeline": [
-            ComputeProgressShards(repo_id, reward_model_path, stride, head_mode, device, str(shard_dir)),
-        ],
-        "logging_dir": str(logs_dir / job_name),
-    }
-
-    if slurm:
-        kwargs.update(
-            {
-                "job_name": job_name,
-                "tasks": workers,
-                "workers": workers,
-                "time": "24:00:00",
-                "partition": partition,
-                "cpus_per_task": cpus_per_task,
-                "sbatch_args": {"mem-per-cpu": mem_per_cpu},
-            }
-        )
-        return SlurmPipelineExecutor(**kwargs)
-
-    kwargs.update({"tasks": workers, "workers": 1})
-    return LocalPipelineExecutor(**kwargs)
-
-
-def make_aggregate_executor(
-    repo_id,
-    reward_model_path,
-    shard_dir,
-    logs_dir,
-    job_name,
-    slurm,
-    partition,
-    cpus_per_task,
-    mem_per_cpu,
-    push_to_hub,
-):
-    kwargs = {
-        "pipeline": [
-            AggregateProgress(repo_id, reward_model_path, str(shard_dir), push_to_hub),
-        ],
-        "logging_dir": str(logs_dir / job_name),
-    }
-
-    if slurm:
-        kwargs.update(
-            {
-                "job_name": job_name,
-                "tasks": 1,
-                "workers": 1,
-                "time": "02:00:00",
-                "partition": partition,
-                "cpus_per_task": cpus_per_task,
-                "sbatch_args": {"mem-per-cpu": mem_per_cpu},
-            }
-        )
-        return SlurmPipelineExecutor(**kwargs)
-
-    kwargs.update({"tasks": 1, "workers": 1})
-    return LocalPipelineExecutor(**kwargs)
-
-
-def _add_shared_args(p):
-    p.add_argument(
-        "--repo-id",
-        type=str,
-        required=True,
-        help="Hugging Face repository identifier, e.g. 'user/dataset'.",
-    )
-    p.add_argument(
-        "--shard-dir",
-        type=Path,
-        default=Path("rabc_shards"),
-        help="Directory to read/write per-rank parquet shards.",
-    )
-    p.add_argument(
-        "--logs-dir",
-        type=Path,
-        default=Path("logs"),
-        help="Directory for datatrove logs.",
-    )
-    p.add_argument(
-        "--job-name",
-        type=str,
-        default=None,
-        help="SLURM job name (defaults to rabc_<subcommand>).",
-    )
-    p.add_argument(
-        "--slurm",
-        type=int,
-        default=1,
-        help="1 = submit via SLURM; 0 = run locally (useful for debugging).",
-    )
-    p.add_argument(
-        "--partition",
-        type=str,
-        default=None,
-        help="SLURM partition to submit to.",
-    )
-    p.add_argument(
-        "--cpus-per-task",
-        type=int,
-        default=4,
-        help="Number of CPUs per SLURM task.",
-    )
-    p.add_argument(
-        "--mem-per-cpu",
-        type=str,
-        default="4G",
-        help="Memory per CPU, e.g. '4G' or '1950M'.",
-    )
-
-
-def main():
-    parser = argparse.ArgumentParser(
-        description="SLURM-distributed SARM RA-BC annotation pipeline",
-        formatter_class=argparse.RawDescriptionHelpFormatter,
-    )
-    sub = parser.add_subparsers(dest="command", required=True)
-
-    # compute subcommand
-    cp = sub.add_parser(
-        "compute",
-        help="Distribute progress computation across SLURM workers.",
-    )
-    _add_shared_args(cp)
-    cp.add_argument(
-        "--reward-model-path",
-        type=str,
-        required=True,
-        help="Path or HF repo id of the SARM reward model.",
-    )
-    cp.add_argument(
-        "--stride",
-        type=int,
-        default=1,
-        help="Compute every Nth frame; intermediate frames are interpolated (must be >= 1).",
-    )
-    cp.add_argument(
-        "--head-mode",
-        type=str,
-        default="sparse",
-        choices=["sparse", "dense", "both"],
-        help="Which reward head(s) to compute.",
-    )
-    cp.add_argument(
-        "--device",
-        type=str,
-        default="cpu",
-        help="Device for reward model inference, e.g. 'cpu' or 'cuda'.",
-    )
-    cp.add_argument(
-        "--workers",
-        type=int,
-        default=50,
-        help="Number of parallel SLURM tasks (one shard per worker).",
-    )
-
-    # aggregate subcommand
-    ap = sub.add_parser(
-        "aggregate",
-        help="Merge per-rank shards into a single sarm_progress.parquet.",
-    )
-    _add_shared_args(ap)
-    ap.add_argument(
-        "--reward-model-path",
-        type=str,
-        required=True,
-        help="Path or HF repo id of the SARM reward model (stored in parquet metadata).",
-    )
-    ap.add_argument(
-        "--push-to-hub",
-        action="store_true",
-        help="Upload sarm_progress.parquet to the Hugging Face Hub after aggregation.",
-    )
-
-    args = parser.parse_args()
-    job_name = args.job_name or f"rabc_{args.command}"
-    kwargs = vars(args)
-    kwargs["slurm"] = kwargs.pop("slurm") == 1
-    kwargs["job_name"] = job_name
-    command = kwargs.pop("command")
-
-    executor = make_compute_executor(**kwargs) if command == "compute" else make_aggregate_executor(**kwargs)
-
-    executor.run()
-
-
-if __name__ == "__main__":
-    main()

examples/hil/hil_utils.py

@@ -1,228 +0,0 @@
-# Copyright 2025 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 utilities for Human-in-the-Loop data collection scripts."""
-
-import logging
-import time
-from dataclasses import dataclass, field
-from pathlib import Path
-
-from lerobot.processor import (
-    IdentityProcessorStep,
-    RobotAction,
-    RobotObservation,
-    RobotProcessorPipeline,
-)
-from lerobot.processor.converters import (
-    observation_to_transition,
-    robot_action_observation_to_transition,
-    transition_to_observation,
-    transition_to_robot_action,
-)
-from lerobot.robots import Robot
-from lerobot.teleoperators import Teleoperator
-from lerobot.utils.control_utils import is_headless
-from lerobot.utils.robot_utils import precise_sleep
-
-logger = logging.getLogger(__name__)
-
-
-@dataclass
-class HILDatasetConfig:
-    repo_id: str
-    single_task: str
-    root: str | Path | None = None
-    fps: int = 30
-    episode_time_s: float = 120
-    num_episodes: int = 50
-    video: bool = True
-    push_to_hub: bool = True
-    private: bool = False
-    tags: list[str] | None = None
-    num_image_writer_processes: int = 0
-    num_image_writer_threads_per_camera: int = 4
-    video_encoding_batch_size: int = 1
-    vcodec: str = "auto"
-    streaming_encoding: bool = True
-    encoder_queue_maxsize: int = 30
-    encoder_threads: int | None = None
-    rename_map: dict[str, str] = field(default_factory=dict)
-
-
-def teleop_has_motor_control(teleop: Teleoperator) -> bool:
-    """Check if teleoperator has motor control capabilities."""
-    return all(hasattr(teleop, attr) for attr in ("enable_torque", "disable_torque", "write_goal_positions"))
-
-
-def teleop_disable_torque(teleop: Teleoperator) -> None:
-    """Disable teleop torque if supported."""
-    if hasattr(teleop, "disable_torque"):
-        teleop.disable_torque()
-
-
-def teleop_enable_torque(teleop: Teleoperator) -> None:
-    """Enable teleop torque if supported."""
-    if hasattr(teleop, "enable_torque"):
-        teleop.enable_torque()
-
-
-def teleop_smooth_move_to(teleop: Teleoperator, target_pos: dict, duration_s: float = 2.0, fps: int = 50):
-    """Smoothly move teleop to target position if motor control is available."""
-    if not teleop_has_motor_control(teleop):
-        logger.warning("Teleop does not support motor control - cannot mirror robot position")
-        return
-
-    teleop_enable_torque(teleop)
-    current = teleop.get_action()
-    steps = max(int(duration_s * fps), 1)
-
-    for step in range(steps + 1):
-        t = step / steps
-        interp = {}
-        for k in current:
-            if k in target_pos:
-                interp[k] = current[k] * (1 - t) + target_pos[k] * t
-            else:
-                interp[k] = current[k]
-        teleop.write_goal_positions(interp)
-        time.sleep(1 / fps)
-
-
-def init_keyboard_listener():
-    """Initialize keyboard listener with HIL controls."""
-    events = {
-        "exit_early": False,
-        "rerecord_episode": False,
-        "stop_recording": False,
-        "policy_paused": False,
-        "correction_active": False,
-        "resume_policy": False,
-        "in_reset": False,
-        "start_next_episode": False,
-    }
-
-    if is_headless():
-        logger.warning("Headless environment - keyboard controls unavailable")
-        return None, events
-
-    from pynput import keyboard
-
-    def on_press(key):
-        try:
-            if events["in_reset"]:
-                if key in [keyboard.Key.space, keyboard.Key.right]:
-                    logger.info("[HIL] Starting next episode...")
-                    events["start_next_episode"] = True
-                elif hasattr(key, "char") and key.char == "c":
-                    events["start_next_episode"] = True
-                elif key == keyboard.Key.esc:
-                    logger.info("[HIL] ESC - Stop recording, pushing to hub...")
-                    events["stop_recording"] = True
-                    events["start_next_episode"] = True
-            else:
-                if key == keyboard.Key.space:
-                    if not events["policy_paused"] and not events["correction_active"]:
-                        logger.info("[HIL] PAUSED - Press 'c' to take control or 'p' to resume policy")
-                        events["policy_paused"] = True
-                elif hasattr(key, "char") and key.char == "c":
-                    if events["policy_paused"] and not events["correction_active"]:
-                        logger.info("[HIL] Taking control...")
-                        events["start_next_episode"] = True
-                elif hasattr(key, "char") and key.char == "p":
-                    if events["policy_paused"] or events["correction_active"]:
-                        logger.info("[HIL] Resuming policy...")
-                        events["resume_policy"] = True
-                elif key == keyboard.Key.right:
-                    logger.info("[HIL] End episode")
-                    events["exit_early"] = True
-                elif key == keyboard.Key.left:
-                    logger.info("[HIL] Re-record episode")
-                    events["rerecord_episode"] = True
-                    events["exit_early"] = True
-                elif key == keyboard.Key.esc:
-                    logger.info("[HIL] ESC - Stop recording...")
-                    events["stop_recording"] = True
-                    events["exit_early"] = True
-        except Exception as e:
-            logger.info(f"Key error: {e}")
-
-    listener = keyboard.Listener(on_press=on_press)
-    listener.start()
-    return listener, events
-
-
-def make_identity_processors():
-    """Create identity processors for recording."""
-    teleop_proc = RobotProcessorPipeline[tuple[RobotAction, RobotObservation], RobotAction](
-        steps=[IdentityProcessorStep()],
-        to_transition=robot_action_observation_to_transition,
-        to_output=transition_to_robot_action,
-    )
-    obs_proc = RobotProcessorPipeline[RobotObservation, RobotObservation](
-        steps=[IdentityProcessorStep()],
-        to_transition=observation_to_transition,
-        to_output=transition_to_observation,
-    )
-    return teleop_proc, obs_proc
-
-
-def reset_loop(robot: Robot, teleop: Teleoperator, events: dict, fps: int):
-    """Reset period where human repositions environment."""
-    logger.info("[HIL] RESET")
-
-    events["in_reset"] = True
-    events["start_next_episode"] = False
-
-    obs = robot.get_observation()
-    robot_pos = {k: v for k, v in obs.items() if k.endswith(".pos") and k in robot.observation_features}
-    teleop_smooth_move_to(teleop, robot_pos, duration_s=2.0, fps=50)
-
-    logger.info("Press any key to enable teleoperation")
-    while not events["start_next_episode"] and not events["stop_recording"]:
-        precise_sleep(0.05)
-
-    if events["stop_recording"]:
-        return
-
-    events["start_next_episode"] = False
-    teleop_disable_torque(teleop)
-    logger.info("Teleop enabled - press any key to start episode")
-
-    while not events["start_next_episode"] and not events["stop_recording"]:
-        loop_start = time.perf_counter()
-        action = teleop.get_action()
-        robot.send_action(action)
-        precise_sleep(1 / fps - (time.perf_counter() - loop_start))
-
-    events["in_reset"] = False
-    events["start_next_episode"] = False
-    events["exit_early"] = False
-    events["policy_paused"] = False
-    events["correction_active"] = False
-    events["resume_policy"] = False
-
-
-def print_controls(rtc: bool = False):
-    """Print control instructions."""
-    mode = "Human-in-the-Loop Data Collection" + (" (RTC)" if rtc else "")
-    logger.info(
-        "%s\n  Controls:\n"
-        "    SPACE  - Pause policy\n"
-        "    c      - Take control\n"
-        "    p      - Resume policy after pause/correction\n"
-        "    →      - End episode\n"
-        "    ESC    - Stop and push to hub",
-        mode,
-    )

examples/lekiwi/evaluate.py

@@ -14,8 +14,8 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from lerobot.datasets.feature_utils import hw_to_dataset_features
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.utils import hw_to_dataset_features
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors
 from lerobot.processor import make_default_processors
@@ -78,24 +78,40 @@ def main():
     listener, events = init_keyboard_listener()
     init_rerun(session_name="lekiwi_evaluate")
 
-    try:
-        if not robot.is_connected:
-            raise ValueError("Robot is not connected!")
+    if not robot.is_connected:
+        raise ValueError("Robot is not connected!")
 
-        print("Starting evaluate loop...")
-        recorded_episodes = 0
-        while recorded_episodes < NUM_EPISODES and not events["stop_recording"]:
-            log_say(f"Running inference, recording eval episode {recorded_episodes} of {NUM_EPISODES}")
+    print("Starting evaluate loop...")
+    recorded_episodes = 0
+    while recorded_episodes < NUM_EPISODES and not events["stop_recording"]:
+        log_say(f"Running inference, recording eval episode {recorded_episodes} of {NUM_EPISODES}")
 
-            # Main record loop
+        # Main record loop
+        record_loop(
+            robot=robot,
+            events=events,
+            fps=FPS,
+            policy=policy,
+            preprocessor=preprocessor,  # Pass the pre and post policy processors
+            postprocessor=postprocessor,
+            dataset=dataset,
+            control_time_s=EPISODE_TIME_SEC,
+            single_task=TASK_DESCRIPTION,
+            display_data=True,
+            teleop_action_processor=teleop_action_processor,
+            robot_action_processor=robot_action_processor,
+            robot_observation_processor=robot_observation_processor,
+        )
+
+        # Reset the environment if not stopping or re-recording
+        if not events["stop_recording"] and (
+            (recorded_episodes < NUM_EPISODES - 1) or events["rerecord_episode"]
+        ):
+            log_say("Reset the environment")
             record_loop(
                 robot=robot,
                 events=events,
                 fps=FPS,
-                policy=policy,
-                preprocessor=preprocessor,  # Pass the pre and post policy processors
-                postprocessor=postprocessor,
-                dataset=dataset,
                 control_time_s=EPISODE_TIME_SEC,
                 single_task=TASK_DESCRIPTION,
                 display_data=True,
@@ -104,42 +120,24 @@ def main():
                 robot_observation_processor=robot_observation_processor,
             )
 
-            # Reset the environment if not stopping or re-recording
-            if not events["stop_recording"] and (
-                (recorded_episodes < NUM_EPISODES - 1) or events["rerecord_episode"]
-            ):
-                log_say("Reset the environment")
-                record_loop(
-                    robot=robot,
-                    events=events,
-                    fps=FPS,
-                    control_time_s=EPISODE_TIME_SEC,
-                    single_task=TASK_DESCRIPTION,
-                    display_data=True,
-                    teleop_action_processor=teleop_action_processor,
-                    robot_action_processor=robot_action_processor,
-                    robot_observation_processor=robot_observation_processor,
-                )
+        if events["rerecord_episode"]:
+            log_say("Re-record episode")
+            events["rerecord_episode"] = False
+            events["exit_early"] = False
+            dataset.clear_episode_buffer()
+            continue
 
-            if events["rerecord_episode"]:
-                log_say("Re-record episode")
-                events["rerecord_episode"] = False
-                events["exit_early"] = False
-                dataset.clear_episode_buffer()
-                continue
+        # Save episode
+        dataset.save_episode()
+        recorded_episodes += 1
 
-            # Save episode
-            dataset.save_episode()
-            recorded_episodes += 1
+    # Clean up
+    log_say("Stop recording")
+    robot.disconnect()
+    listener.stop()
 
-    finally:
-        # Clean up
-        log_say("Stop recording")
-        robot.disconnect()
-        listener.stop()
-
-        dataset.finalize()
-        dataset.push_to_hub()
+    dataset.finalize()
+    dataset.push_to_hub()
 
 
 if __name__ == "__main__":

examples/lekiwi/record.py

@@ -14,14 +14,14 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from lerobot.datasets.feature_utils import hw_to_dataset_features
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.utils import hw_to_dataset_features
 from lerobot.processor import make_default_processors
 from lerobot.robots.lekiwi.config_lekiwi import LeKiwiClientConfig
 from lerobot.robots.lekiwi.lekiwi_client import LeKiwiClient
 from lerobot.scripts.lerobot_record import record_loop
 from lerobot.teleoperators.keyboard import KeyboardTeleop, KeyboardTeleopConfig
-from lerobot.teleoperators.so_leader import SO100Leader, SO100LeaderConfig
+from lerobot.teleoperators.so100_leader import SO100Leader, SO100LeaderConfig
 from lerobot.utils.constants import ACTION, OBS_STR
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say
@@ -74,23 +74,40 @@ def main():
     listener, events = init_keyboard_listener()
     init_rerun(session_name="lekiwi_record")
 
-    try:
-        if not robot.is_connected or not leader_arm.is_connected or not keyboard.is_connected:
-            raise ValueError("Robot or teleop is not connected!")
+    if not robot.is_connected or not leader_arm.is_connected or not keyboard.is_connected:
+        raise ValueError("Robot or teleop is not connected!")
 
-        print("Starting record loop...")
-        recorded_episodes = 0
-        while recorded_episodes < NUM_EPISODES and not events["stop_recording"]:
-            log_say(f"Recording episode {recorded_episodes}")
+    print("Starting record loop...")
+    recorded_episodes = 0
+    while recorded_episodes < NUM_EPISODES and not events["stop_recording"]:
+        log_say(f"Recording episode {recorded_episodes}")
 
-            # Main record loop
+        # Main record loop
+        record_loop(
+            robot=robot,
+            events=events,
+            fps=FPS,
+            dataset=dataset,
+            teleop=[leader_arm, keyboard],
+            control_time_s=EPISODE_TIME_SEC,
+            single_task=TASK_DESCRIPTION,
+            display_data=True,
+            teleop_action_processor=teleop_action_processor,
+            robot_action_processor=robot_action_processor,
+            robot_observation_processor=robot_observation_processor,
+        )
+
+        # Reset the environment if not stopping or re-recording
+        if not events["stop_recording"] and (
+            (recorded_episodes < NUM_EPISODES - 1) or events["rerecord_episode"]
+        ):
+            log_say("Reset the environment")
             record_loop(
                 robot=robot,
                 events=events,
                 fps=FPS,
-                dataset=dataset,
                 teleop=[leader_arm, keyboard],
-                control_time_s=EPISODE_TIME_SEC,
+                control_time_s=RESET_TIME_SEC,
                 single_task=TASK_DESCRIPTION,
                 display_data=True,
                 teleop_action_processor=teleop_action_processor,
@@ -98,44 +115,26 @@ def main():
                 robot_observation_processor=robot_observation_processor,
             )
 
-            # Reset the environment if not stopping or re-recording
-            if not events["stop_recording"] and (
-                (recorded_episodes < NUM_EPISODES - 1) or events["rerecord_episode"]
-            ):
-                log_say("Reset the environment")
-                record_loop(
-                    robot=robot,
-                    events=events,
-                    fps=FPS,
-                    teleop=[leader_arm, keyboard],
-                    control_time_s=RESET_TIME_SEC,
-                    single_task=TASK_DESCRIPTION,
-                    display_data=True,
-                    teleop_action_processor=teleop_action_processor,
-                    robot_action_processor=robot_action_processor,
-                    robot_observation_processor=robot_observation_processor,
-                )
+        if events["rerecord_episode"]:
+            log_say("Re-record episode")
+            events["rerecord_episode"] = False
+            events["exit_early"] = False
+            dataset.clear_episode_buffer()
+            continue
 
-            if events["rerecord_episode"]:
-                log_say("Re-record episode")
-                events["rerecord_episode"] = False
-                events["exit_early"] = False
-                dataset.clear_episode_buffer()
-                continue
+        # Save episode
+        dataset.save_episode()
+        recorded_episodes += 1
 
-            # Save episode
-            dataset.save_episode()
-            recorded_episodes += 1
-    finally:
-        # Clean up
-        log_say("Stop recording")
-        robot.disconnect()
-        leader_arm.disconnect()
-        keyboard.disconnect()
-        listener.stop()
+    # Clean up
+    log_say("Stop recording")
+    robot.disconnect()
+    leader_arm.disconnect()
+    keyboard.disconnect()
+    listener.stop()
 
-        dataset.finalize()
-        dataset.push_to_hub()
+    dataset.finalize()
+    dataset.push_to_hub()
 
 
 if __name__ == "__main__":

examples/lekiwi/replay.py

@@ -35,32 +35,32 @@ def main():
 
     # Fetch the dataset to replay
     dataset = LeRobotDataset("<hf_username>/<dataset_repo_id>", episodes=[EPISODE_IDX])
-    actions = dataset.select_columns(ACTION)
+    # Filter dataset to only include frames from the specified episode since episodes are chunked in dataset V3.0
+    episode_frames = dataset.hf_dataset.filter(lambda x: x["episode_index"] == EPISODE_IDX)
+    actions = episode_frames.select_columns(ACTION)
 
     # Connect to the robot
     robot.connect()
 
-    try:
-        if not robot.is_connected:
-            raise ValueError("Robot is not connected!")
+    if not robot.is_connected:
+        raise ValueError("Robot is not connected!")
 
-        print("Starting replay loop...")
-        log_say(f"Replaying episode {EPISODE_IDX}")
-        for idx in range(dataset.num_frames):
-            t0 = time.perf_counter()
+    print("Starting replay loop...")
+    log_say(f"Replaying episode {EPISODE_IDX}")
+    for idx in range(len(episode_frames)):
+        t0 = time.perf_counter()
 
-            # Get recorded action from dataset
-            action = {
-                name: float(actions[idx][ACTION][i])
-                for i, name in enumerate(dataset.features[ACTION]["names"])
-            }
+        # Get recorded action from dataset
+        action = {
+            name: float(actions[idx][ACTION][i]) for i, name in enumerate(dataset.features[ACTION]["names"])
+        }
 
-            # Send action to robot
-            _ = robot.send_action(action)
+        # Send action to robot
+        _ = robot.send_action(action)
 
-            precise_sleep(max(1.0 / dataset.fps - (time.perf_counter() - t0), 0.0))
-    finally:
-        robot.disconnect()
+        precise_sleep(max(1.0 / dataset.fps - (time.perf_counter() - t0), 0.0))
+
+    robot.disconnect()
 
 
 if __name__ == "__main__":

examples/lekiwi/teleoperate.py

@@ -18,7 +18,7 @@ import time
 
 from lerobot.robots.lekiwi import LeKiwiClient, LeKiwiClientConfig
 from lerobot.teleoperators.keyboard.teleop_keyboard import KeyboardTeleop, KeyboardTeleopConfig
-from lerobot.teleoperators.so_leader import SO100Leader, SO100LeaderConfig
+from lerobot.teleoperators.so100_leader import SO100Leader, SO100LeaderConfig
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.visualization_utils import init_rerun, log_rerun_data
 

examples/phone_to_so100/evaluate.py

@@ -16,13 +16,15 @@
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
 from lerobot.configs.types import FeatureType, PolicyFeature
-from lerobot.datasets.feature_utils import combine_feature_dicts
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.pipeline_features import aggregate_pipeline_dataset_features, create_initial_features
+from lerobot.datasets.utils import combine_feature_dicts
 from lerobot.model.kinematics import RobotKinematics
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors
 from lerobot.processor import (
+    RobotAction,
+    RobotObservation,
     RobotProcessorPipeline,
     make_default_teleop_action_processor,
 )
@@ -32,13 +34,13 @@ from lerobot.processor.converters import (
     transition_to_observation,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     ForwardKinematicsJointsToEE,
     InverseKinematicsEEToJoints,
 )
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.scripts.lerobot_record import record_loop
-from lerobot.types import RobotAction, RobotObservation
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say
 from lerobot.utils.visualization_utils import init_rerun
@@ -141,24 +143,38 @@ def main():
     listener, events = init_keyboard_listener()
     init_rerun(session_name="phone_so100_evaluate")
 
-    try:
-        if not robot.is_connected:
-            raise ValueError("Robot is not connected!")
+    if not robot.is_connected:
+        raise ValueError("Robot is not connected!")
 
-        print("Starting evaluate loop...")
-        episode_idx = 0
-        for episode_idx in range(NUM_EPISODES):
-            log_say(f"Running inference, recording eval episode {episode_idx + 1} of {NUM_EPISODES}")
+    print("Starting evaluate loop...")
+    episode_idx = 0
+    for episode_idx in range(NUM_EPISODES):
+        log_say(f"Running inference, recording eval episode {episode_idx + 1} of {NUM_EPISODES}")
 
-            # Main record loop
+        # Main record loop
+        record_loop(
+            robot=robot,
+            events=events,
+            fps=FPS,
+            policy=policy,
+            preprocessor=preprocessor,  # Pass the pre and post policy processors
+            postprocessor=postprocessor,
+            dataset=dataset,
+            control_time_s=EPISODE_TIME_SEC,
+            single_task=TASK_DESCRIPTION,
+            display_data=True,
+            teleop_action_processor=make_default_teleop_action_processor(),
+            robot_action_processor=robot_ee_to_joints_processor,
+            robot_observation_processor=robot_joints_to_ee_pose_processor,
+        )
+
+        # Reset the environment if not stopping or re-recording
+        if not events["stop_recording"] and ((episode_idx < NUM_EPISODES - 1) or events["rerecord_episode"]):
+            log_say("Reset the environment")
             record_loop(
                 robot=robot,
                 events=events,
                 fps=FPS,
-                policy=policy,
-                preprocessor=preprocessor,  # Pass the pre and post policy processors
-                postprocessor=postprocessor,
-                dataset=dataset,
                 control_time_s=EPISODE_TIME_SEC,
                 single_task=TASK_DESCRIPTION,
                 display_data=True,
@@ -167,41 +183,24 @@ def main():
                 robot_observation_processor=robot_joints_to_ee_pose_processor,
             )
 
-            # Reset the environment if not stopping or re-recording
-            if not events["stop_recording"] and (
-                (episode_idx < NUM_EPISODES - 1) or events["rerecord_episode"]
-            ):
-                log_say("Reset the environment")
-                record_loop(
-                    robot=robot,
-                    events=events,
-                    fps=FPS,
-                    control_time_s=EPISODE_TIME_SEC,
-                    single_task=TASK_DESCRIPTION,
-                    display_data=True,
-                    teleop_action_processor=make_default_teleop_action_processor(),
-                    robot_action_processor=robot_ee_to_joints_processor,
-                    robot_observation_processor=robot_joints_to_ee_pose_processor,
-                )
+        if events["rerecord_episode"]:
+            log_say("Re-record episode")
+            events["rerecord_episode"] = False
+            events["exit_early"] = False
+            dataset.clear_episode_buffer()
+            continue
 
-            if events["rerecord_episode"]:
-                log_say("Re-record episode")
-                events["rerecord_episode"] = False
-                events["exit_early"] = False
-                dataset.clear_episode_buffer()
-                continue
+        # Save episode
+        dataset.save_episode()
+        episode_idx += 1
 
-            # Save episode
-            dataset.save_episode()
-            episode_idx += 1
-    finally:
-        # Clean up
-        log_say("Stop recording")
-        robot.disconnect()
-        listener.stop()
+    # Clean up
+    log_say("Stop recording")
+    robot.disconnect()
+    listener.stop()
 
-        dataset.finalize()
-        dataset.push_to_hub()
+    dataset.finalize()
+    dataset.push_to_hub()
 
 
 if __name__ == "__main__":

examples/phone_to_so100/record.py

@@ -15,30 +15,30 @@
 # limitations under the License.
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
-from lerobot.datasets.feature_utils import combine_feature_dicts
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.pipeline_features import aggregate_pipeline_dataset_features, create_initial_features
+from lerobot.datasets.utils import combine_feature_dicts
 from lerobot.model.kinematics import RobotKinematics
-from lerobot.processor import RobotProcessorPipeline
+from lerobot.processor import RobotAction, RobotObservation, RobotProcessorPipeline
 from lerobot.processor.converters import (
     observation_to_transition,
     robot_action_observation_to_transition,
     transition_to_observation,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     EEBoundsAndSafety,
     EEReferenceAndDelta,
     ForwardKinematicsJointsToEE,
     GripperVelocityToJoint,
     InverseKinematicsEEToJoints,
 )
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.scripts.lerobot_record import record_loop
 from lerobot.teleoperators.phone.config_phone import PhoneConfig, PhoneOS
 from lerobot.teleoperators.phone.phone_processor import MapPhoneActionToRobotAction
 from lerobot.teleoperators.phone.teleop_phone import Phone
-from lerobot.types import RobotAction, RobotObservation
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say
 from lerobot.utils.visualization_utils import init_rerun
@@ -150,23 +150,38 @@ def main():
     listener, events = init_keyboard_listener()
     init_rerun(session_name="phone_so100_record")
 
-    try:
-        if not robot.is_connected or not phone.is_connected:
-            raise ValueError("Robot or teleop is not connected!")
+    if not robot.is_connected or not phone.is_connected:
+        raise ValueError("Robot or teleop is not connected!")
 
-        print("Starting record loop. Move your phone to teleoperate the robot...")
-        episode_idx = 0
-        while episode_idx < NUM_EPISODES and not events["stop_recording"]:
-            log_say(f"Recording episode {episode_idx + 1} of {NUM_EPISODES}")
+    print("Starting record loop. Move your phone to teleoperate the robot...")
+    episode_idx = 0
+    while episode_idx < NUM_EPISODES and not events["stop_recording"]:
+        log_say(f"Recording episode {episode_idx + 1} of {NUM_EPISODES}")
 
-            # Main record loop
+        # Main record loop
+        record_loop(
+            robot=robot,
+            events=events,
+            fps=FPS,
+            teleop=phone,
+            dataset=dataset,
+            control_time_s=EPISODE_TIME_SEC,
+            single_task=TASK_DESCRIPTION,
+            display_data=True,
+            teleop_action_processor=phone_to_robot_ee_pose_processor,
+            robot_action_processor=robot_ee_to_joints_processor,
+            robot_observation_processor=robot_joints_to_ee_pose,
+        )
+
+        # Reset the environment if not stopping or re-recording
+        if not events["stop_recording"] and (episode_idx < NUM_EPISODES - 1 or events["rerecord_episode"]):
+            log_say("Reset the environment")
             record_loop(
                 robot=robot,
                 events=events,
                 fps=FPS,
                 teleop=phone,
-                dataset=dataset,
-                control_time_s=EPISODE_TIME_SEC,
+                control_time_s=RESET_TIME_SEC,
                 single_task=TASK_DESCRIPTION,
                 display_data=True,
                 teleop_action_processor=phone_to_robot_ee_pose_processor,
@@ -174,43 +189,25 @@ def main():
                 robot_observation_processor=robot_joints_to_ee_pose,
             )
 
-            # Reset the environment if not stopping or re-recording
-            if not events["stop_recording"] and (
-                episode_idx < NUM_EPISODES - 1 or events["rerecord_episode"]
-            ):
-                log_say("Reset the environment")
-                record_loop(
-                    robot=robot,
-                    events=events,
-                    fps=FPS,
-                    teleop=phone,
-                    control_time_s=RESET_TIME_SEC,
-                    single_task=TASK_DESCRIPTION,
-                    display_data=True,
-                    teleop_action_processor=phone_to_robot_ee_pose_processor,
-                    robot_action_processor=robot_ee_to_joints_processor,
-                    robot_observation_processor=robot_joints_to_ee_pose,
-                )
+        if events["rerecord_episode"]:
+            log_say("Re-recording episode")
+            events["rerecord_episode"] = False
+            events["exit_early"] = False
+            dataset.clear_episode_buffer()
+            continue
 
-            if events["rerecord_episode"]:
-                log_say("Re-recording episode")
-                events["rerecord_episode"] = False
-                events["exit_early"] = False
-                dataset.clear_episode_buffer()
-                continue
+        # Save episode
+        dataset.save_episode()
+        episode_idx += 1
 
-            # Save episode
-            dataset.save_episode()
-            episode_idx += 1
-    finally:
-        # Clean up
-        log_say("Stop recording")
-        robot.disconnect()
-        phone.disconnect()
-        listener.stop()
+    # Clean up
+    log_say("Stop recording")
+    robot.disconnect()
+    phone.disconnect()
+    listener.stop()
 
-        dataset.finalize()
-        dataset.push_to_hub()
+    dataset.finalize()
+    dataset.push_to_hub()
 
 
 if __name__ == "__main__":

examples/phone_to_so100/replay.py

@@ -18,16 +18,16 @@ import time
 
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.model.kinematics import RobotKinematics
-from lerobot.processor import RobotProcessorPipeline
+from lerobot.processor import RobotAction, RobotObservation, RobotProcessorPipeline
 from lerobot.processor.converters import (
     robot_action_observation_to_transition,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     InverseKinematicsEEToJoints,
 )
-from lerobot.types import RobotAction, RobotObservation
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.utils.constants import ACTION
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.utils import log_say
@@ -67,39 +67,39 @@ def main():
 
     # Fetch the dataset to replay
     dataset = LeRobotDataset(HF_REPO_ID, episodes=[EPISODE_IDX])
-    actions = dataset.select_columns(ACTION)
+    # Filter dataset to only include frames from the specified episode since episodes are chunked in dataset V3.0
+    episode_frames = dataset.hf_dataset.filter(lambda x: x["episode_index"] == EPISODE_IDX)
+    actions = episode_frames.select_columns(ACTION)
 
     # Connect to the robot
     robot.connect()
 
-    try:
-        if not robot.is_connected:
-            raise ValueError("Robot is not connected!")
+    if not robot.is_connected:
+        raise ValueError("Robot is not connected!")
 
-        print("Starting replay loop...")
-        log_say(f"Replaying episode {EPISODE_IDX}")
-        for idx in range(dataset.num_frames):
-            t0 = time.perf_counter()
+    print("Starting replay loop...")
+    log_say(f"Replaying episode {EPISODE_IDX}")
+    for idx in range(len(episode_frames)):
+        t0 = time.perf_counter()
 
-            # Get recorded action from dataset
-            ee_action = {
-                name: float(actions[idx][ACTION][i])
-                for i, name in enumerate(dataset.features[ACTION]["names"])
-            }
+        # Get recorded action from dataset
+        ee_action = {
+            name: float(actions[idx][ACTION][i]) for i, name in enumerate(dataset.features[ACTION]["names"])
+        }
 
-            # Get robot observation
-            robot_obs = robot.get_observation()
+        # Get robot observation
+        robot_obs = robot.get_observation()
 
-            # Dataset EE -> robot joints
-            joint_action = robot_ee_to_joints_processor((ee_action, robot_obs))
+        # Dataset EE -> robot joints
+        joint_action = robot_ee_to_joints_processor((ee_action, robot_obs))
 
-            # Send action to robot
-            _ = robot.send_action(joint_action)
+        # Send action to robot
+        _ = robot.send_action(joint_action)
 
-            precise_sleep(max(1.0 / dataset.fps - (time.perf_counter() - t0), 0.0))
-    finally:
-        # Clean up
-        robot.disconnect()
+        precise_sleep(1.0 / dataset.fps - (time.perf_counter() - t0))
+
+    # Clean up
+    robot.disconnect()
 
 
 if __name__ == "__main__":

examples/phone_to_so100/teleoperate.py

@@ -16,22 +16,22 @@
 import time
 
 from lerobot.model.kinematics import RobotKinematics
-from lerobot.processor import RobotProcessorPipeline
+from lerobot.processor import RobotAction, RobotObservation, RobotProcessorPipeline
 from lerobot.processor.converters import (
     robot_action_observation_to_transition,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     EEBoundsAndSafety,
     EEReferenceAndDelta,
     GripperVelocityToJoint,
     InverseKinematicsEEToJoints,
 )
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.teleoperators.phone.config_phone import PhoneConfig, PhoneOS
 from lerobot.teleoperators.phone.phone_processor import MapPhoneActionToRobotAction
 from lerobot.teleoperators.phone.teleop_phone import Phone
-from lerobot.types import RobotAction, RobotObservation
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.visualization_utils import init_rerun, log_rerun_data
 

examples/port_datasets/port_droid.py

@@ -22,8 +22,7 @@ from pathlib import Path
 import numpy as np
 import tensorflow_datasets as tfds
 
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.lerobot_dataset import LeRobotDataset, LeRobotDatasetMetadata
 from lerobot.utils.utils import get_elapsed_time_in_days_hours_minutes_seconds
 
 DROID_SHARDS = 2048

examples/port_datasets/slurm_upload.py

@@ -26,7 +26,7 @@ from huggingface_hub import HfApi
 from huggingface_hub.constants import REPOCARD_NAME
 from port_droid import DROID_SHARDS
 
-from lerobot.datasets.dataset_metadata import CODEBASE_VERSION, LeRobotDatasetMetadata
+from lerobot.datasets.lerobot_dataset import CODEBASE_VERSION, LeRobotDatasetMetadata
 from lerobot.datasets.utils import create_lerobot_dataset_card
 from lerobot.utils.utils import init_logging
 
@@ -155,7 +155,7 @@ class UploadDataset(PipelineStep):
         from datasets.utils.tqdm import disable_progress_bars
         from huggingface_hub import CommitOperationAdd, preupload_lfs_files
 
-        from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
+        from lerobot.datasets.lerobot_dataset import LeRobotDatasetMetadata
         from lerobot.utils.utils import init_logging
 
         init_logging()

examples/rtc/eval_dataset.py

@@ -27,8 +27,8 @@ measuring consistency and ground truth alignment.
 Usage:
     # Basic usage with smolvla policy
     uv run python examples/rtc/eval_dataset.py \
-        --policy.path=<USER>/smolvla_check_rtc_last3 \
-        --dataset.repo_id=<USER>/check_rtc \
+        --policy.path=helper2424/smolvla_check_rtc_last3 \
+        --dataset.repo_id=helper2424/check_rtc \
         --rtc.execution_horizon=8 \
         --device=mps \
         --rtc.max_guidance_weight=10.0 \
@@ -58,16 +58,16 @@ Usage:
         --device=cuda
 
     uv run python examples/rtc/eval_dataset.py \
-        --policy.path=<USER>/reuben_pi0 \
-        --dataset.repo_id=<USER>/so101_cube_in_cup \
+        --policy.path=lipsop/reuben_pi0 \
+        --dataset.repo_id=ReubenLim/so101_cube_in_cup \
         --rtc.execution_horizon=8 \
         --device=cuda
 
     # With torch.compile for faster inference (PyTorch 2.0+)
     # Note: CUDA graphs disabled by default due to in-place ops in denoising loop
     uv run python examples/rtc/eval_dataset.py \
-        --policy.path=<USER>/smolvla_check_rtc_last3 \
-        --dataset.repo_id=<USER>/check_rtc \
+        --policy.path=helper2424/smolvla_check_rtc_last3 \
+        --dataset.repo_id=helper2424/check_rtc \
         --rtc.execution_horizon=8 \
         --device=mps \
         --use_torch_compile=true \
@@ -75,8 +75,8 @@ Usage:
 
     # With torch.compile on CUDA (CUDA graphs disabled by default)
     uv run python examples/rtc/eval_dataset.py \
-        --policy.path=<USER>/smolvla_check_rtc_last3 \
-        --dataset.repo_id=<USER>/check_rtc \
+        --policy.path=helper2424/smolvla_check_rtc_last3 \
+        --dataset.repo_id=helper2424/check_rtc \
         --rtc.execution_horizon=8 \
         --device=cuda \
         --use_torch_compile=true \
@@ -84,8 +84,8 @@ Usage:
 
     # Enable CUDA graphs (advanced - may cause tensor aliasing errors)
     uv run python examples/rtc/eval_dataset.py \
-        --policy.path=<USER>/smolvla_check_rtc_last3 \
-        --dataset.repo_id=<USER>/check_rtc \
+        --policy.path=helper2424/smolvla_check_rtc_last3 \
+        --dataset.repo_id=helper2424/check_rtc \
         --use_torch_compile=true \
         --torch_compile_backend=inductor \
         --torch_compile_mode=max-autotune \
@@ -113,9 +113,8 @@ from lerobot.configs import parser
 from lerobot.configs.default import DatasetConfig
 from lerobot.configs.policies import PreTrainedConfig
 from lerobot.configs.types import RTCAttentionSchedule
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
 from lerobot.datasets.factory import resolve_delta_timestamps
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.lerobot_dataset import LeRobotDataset, LeRobotDatasetMetadata
 from lerobot.policies.factory import get_policy_class, make_pre_post_processors
 from lerobot.policies.rtc.configuration_rtc import RTCConfig
 from lerobot.policies.rtc.debug_visualizer import RTCDebugVisualizer

examples/rtc/eval_with_real_robot.py

@@ -28,7 +28,7 @@ For simulation environments, see eval_with_simulation.py
 Usage:
     # Run RTC with Real robot with RTC
     uv run examples/rtc/eval_with_real_robot.py \
-        --policy.path=<USER>/smolvla_check_rtc_last3 \
+        --policy.path=helper2424/smolvla_check_rtc_last3 \
         --policy.device=mps \
         --rtc.enabled=true \
         --rtc.execution_horizon=20 \
@@ -41,7 +41,7 @@ Usage:
 
     # Run RTC with Real robot without RTC
     uv run examples/rtc/eval_with_real_robot.py \
-        --policy.path=<USER>/smolvla_check_rtc_last3 \
+        --policy.path=helper2424/smolvla_check_rtc_last3 \
         --policy.device=mps \
         --rtc.enabled=false \
         --robot.type=so100_follower \
@@ -53,7 +53,7 @@ Usage:
 
     # Run RTC with Real robot with pi0.5 policy
     uv run examples/rtc/eval_with_real_robot.py \
-        --policy.path=<USER>/pi05_check_rtc \
+        --policy.path=helper2424/pi05_check_rtc \
         --policy.device=mps \
         --rtc.enabled=true \
         --rtc.execution_horizon=20 \
@@ -63,31 +63,6 @@ Usage:
         --robot.cameras="{ gripper: {type: opencv, index_or_path: 0, width: 640, height: 480, fps: 30}, front: {type: opencv, index_or_path: 1, width: 640, height: 480, fps: 30}}" \
         --task="Move green small object into the purple platform" \
         --duration=120
-
-    # Run RTC with bi_openarm_follower (dual-arm OpenArms) and pi0.5 policy
-    python examples/rtc/eval_with_real_robot.py \
-        --policy.path=lerobot-data-collection/folding_final \
-        --robot.type=bi_openarm_follower \
-        --robot.cameras='{left_wrist: {type: opencv, index_or_path: "/dev/video4", width: 1280, height: 720, fps: 30}, base: {type: opencv, index_or_path: "/dev/video2", width: 640, height: 480, fps: 30}, right_wrist: {type: opencv, index_or_path: "/dev/video0", width: 1280, height: 720, fps: 30}}' \
-        --robot.left_arm_config.port=can0 \
-        --robot.left_arm_config.side=left \
-        --robot.left_arm_config.can_interface=socketcan \
-        --robot.left_arm_config.disable_torque_on_disconnect=true \
-        --robot.left_arm_config.max_relative_target=8.0 \
-        --robot.right_arm_config.port=can1 \
-        --robot.right_arm_config.side=right \
-        --robot.right_arm_config.can_interface=socketcan \
-        --robot.right_arm_config.disable_torque_on_disconnect=true \
-        --robot.right_arm_config.max_relative_target=8.0 \
-        --task="Fold the T-shirt properly" \
-        --fps=30 \
-        --duration=2000 \
-        --interpolation_multiplier=3 \
-        --rtc.enabled=true \
-        --rtc.execution_horizon=20 \
-        --rtc.max_guidance_weight=5.0 \
-        --rtc.prefix_attention_schedule=LINEAR \
-        --device=cuda
 """
 
 import logging
@@ -103,36 +78,28 @@ from torch import Tensor
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig  # noqa: F401
 from lerobot.cameras.realsense.configuration_realsense import RealSenseCameraConfig  # noqa: F401
-from lerobot.cameras.zmq.configuration_zmq import ZMQCameraConfig  # noqa: F401
 from lerobot.configs import parser
 from lerobot.configs.policies import PreTrainedConfig
 from lerobot.configs.types import RTCAttentionSchedule
-from lerobot.datasets.feature_utils import build_dataset_frame, hw_to_dataset_features
+from lerobot.datasets.utils import build_dataset_frame, hw_to_dataset_features
 from lerobot.policies.factory import get_policy_class, make_pre_post_processors
-from lerobot.policies.rtc import ActionInterpolator, ActionQueue, LatencyTracker, RTCConfig
-from lerobot.processor import (
-    NormalizerProcessorStep,
-    RelativeActionsProcessorStep,
-    TransitionKey,
-    create_transition,
-)
+from lerobot.policies.rtc.action_queue import ActionQueue
+from lerobot.policies.rtc.configuration_rtc import RTCConfig
+from lerobot.policies.rtc.latency_tracker import LatencyTracker
 from lerobot.processor.factory import (
     make_default_robot_action_processor,
     make_default_robot_observation_processor,
 )
-from lerobot.processor.relative_action_processor import to_relative_actions
 from lerobot.rl.process import ProcessSignalHandler
 from lerobot.robots import (  # noqa: F401
     Robot,
     RobotConfig,
-    bi_openarm_follower,
-    bi_so_follower,
     koch_follower,
-    so_follower,
-    unitree_g1,
+    so100_follower,
+    so101_follower,
 )
 from lerobot.robots.utils import make_robot_from_config
-from lerobot.utils.constants import OBS_IMAGES, OBS_STATE
+from lerobot.utils.constants import OBS_IMAGES
 from lerobot.utils.hub import HubMixin
 from lerobot.utils.utils import init_logging
 
@@ -184,7 +151,6 @@ class RTCDemoConfig(HubMixin):
     # Demo parameters
     duration: float = 30.0  # Duration to run the demo (seconds)
     fps: float = 10.0  # Action execution frequency (Hz)
-    interpolation_multiplier: int = 1  # Control rate multiplier (1=off, 2=2x, 3=3x)
 
     # Compute device
     device: str | None = None  # Device to run on (cuda, cpu, auto)
@@ -244,35 +210,6 @@ def is_image_key(k: str) -> bool:
     return k.startswith(OBS_IMAGES)
 
 
-def _reanchor_relative_rtc_prefix(
-    prev_actions_absolute: Tensor,
-    current_state: Tensor,
-    relative_step: RelativeActionsProcessorStep,
-    normalizer_step: NormalizerProcessorStep | None,
-    policy_device: torch.device | str,
-) -> Tensor:
-    """Convert absolute leftovers into model-space for relative-action RTC policies.
-
-    When a policy uses relative actions, the RTC prefix (leftover actions from
-    the previous chunk) is stored in absolute space. Before feeding it back to
-    the policy we need to re-express it relative to the *current* robot state
-    and then re-normalize.
-    """
-    state = current_state.detach().cpu()
-    if state.dim() == 1:
-        state = state.unsqueeze(0)
-
-    action_cpu = prev_actions_absolute.detach().cpu()
-    mask = relative_step._build_mask(action_cpu.shape[-1])
-    relative_actions = to_relative_actions(action_cpu, state, mask)
-
-    transition = create_transition(action=relative_actions)
-    if normalizer_step is not None:
-        transition = normalizer_step(transition)
-
-    return transition[TransitionKey.ACTION].to(policy_device)
-
-
 def get_actions(
     policy,
     robot: RobotWrapper,
@@ -298,15 +235,7 @@ def get_actions(
         fps = cfg.fps
         time_per_chunk = 1.0 / fps
 
-        # Only keep .pos joints + camera streams if the policy was trained on positions,
-        # not the full pos/vel/torque state the robot exposes.
-        observation_features_hw = {
-            key: value
-            for key, value in robot.observation_features().items()
-            if key.endswith(".pos") or isinstance(value, tuple)
-        }
-
-        dataset_features = hw_to_dataset_features(observation_features_hw, "observation")
+        dataset_features = hw_to_dataset_features(robot.observation_features(), "observation")
         policy_device = policy.config.device
 
         # Load preprocessor and postprocessor from pretrained files
@@ -324,25 +253,6 @@ def get_actions(
 
         logger.info("[GET_ACTIONS] Preprocessor/postprocessor loaded successfully with embedded stats")
 
-        relative_step = next(
-            (s for s in preprocessor.steps if isinstance(s, RelativeActionsProcessorStep) and s.enabled),
-            None,
-        )
-        normalizer_step = next(
-            (s for s in preprocessor.steps if isinstance(s, NormalizerProcessorStep)),
-            None,
-        )
-        if relative_step is not None:
-            if relative_step.action_names is None:
-                cfg_names = getattr(cfg.policy, "action_feature_names", None)
-                if cfg_names:
-                    relative_step.action_names = list(cfg_names)
-                else:
-                    relative_step.action_names = [
-                        k for k in robot.robot.action_features if k.endswith(".pos")
-                    ]
-            logger.info("[GET_ACTIONS] Relative actions enabled: will re-anchor RTC prefix")
-
         get_actions_threshold = cfg.action_queue_size_to_get_new_actions
 
         if not cfg.rtc.enabled:
@@ -385,28 +295,6 @@ def get_actions(
 
                 preproceseded_obs = preprocessor(obs_with_policy_features)
 
-                # Re-anchor leftover actions for relative-action policies.
-                # We need the *postprocessed* (absolute) leftover, not the original
-                # (normalized/relative) one that get_left_over() returns.
-                if (
-                    prev_actions is not None
-                    and relative_step is not None
-                    and OBS_STATE in obs_with_policy_features
-                ):
-                    with action_queue.lock:
-                        if action_queue.queue is not None:
-                            prev_actions_abs = action_queue.queue[action_queue.last_index :].clone()
-                        else:
-                            prev_actions_abs = None
-                    if prev_actions_abs is not None and prev_actions_abs.numel() > 0:
-                        prev_actions = _reanchor_relative_rtc_prefix(
-                            prev_actions_absolute=prev_actions_abs,
-                            current_state=obs_with_policy_features[OBS_STATE],
-                            relative_step=relative_step,
-                            normalizer_step=normalizer_step,
-                            policy_device=policy_device,
-                        )
-
                 # Generate actions WITH RTC
                 actions = policy.predict_action_chunk(
                     preproceseded_obs,
@@ -462,26 +350,21 @@ def actor_control(
     try:
         logger.info("[ACTOR] Starting actor thread")
 
-        action_keys = [k for k in robot.action_features() if k.endswith(".pos")]
-
         action_count = 0
-        interpolator = ActionInterpolator(multiplier=cfg.interpolation_multiplier)
-        action_interval = interpolator.get_control_interval(cfg.fps)
+        action_interval = 1.0 / cfg.fps
 
         while not shutdown_event.is_set():
             start_time = time.perf_counter()
 
-            if interpolator.needs_new_action():
-                new_action = action_queue.get()
-                if new_action is not None:
-                    interpolator.add(new_action.cpu())
+            # Try to get an action from the queue with timeout
+            action = action_queue.get()
 
-            action = interpolator.get()
             if action is not None:
                 action = action.cpu()
-                action_dict = {key: action[i].item() for i, key in enumerate(action_keys)}
+                action_dict = {key: action[i].item() for i, key in enumerate(robot.action_features())}
                 action_processed = robot_action_processor((action_dict, None))
                 robot.send_action(action_processed)
+
                 action_count += 1
 
             dt_s = time.perf_counter() - start_time

examples/so100_to_so100_EE/evaluate.py

@@ -16,13 +16,15 @@
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
 from lerobot.configs.types import FeatureType, PolicyFeature
-from lerobot.datasets.feature_utils import combine_feature_dicts
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.pipeline_features import aggregate_pipeline_dataset_features, create_initial_features
+from lerobot.datasets.utils import combine_feature_dicts
 from lerobot.model.kinematics import RobotKinematics
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors
 from lerobot.processor import (
+    RobotAction,
+    RobotObservation,
     RobotProcessorPipeline,
     make_default_teleop_action_processor,
 )
@@ -32,13 +34,13 @@ from lerobot.processor.converters import (
     transition_to_observation,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     ForwardKinematicsJointsToEE,
     InverseKinematicsEEToJoints,
 )
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.scripts.lerobot_record import record_loop
-from lerobot.types import RobotAction, RobotObservation
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say
 from lerobot.utils.visualization_utils import init_rerun
@@ -141,24 +143,38 @@ def main():
     listener, events = init_keyboard_listener()
     init_rerun(session_name="so100_so100_evaluate")
 
-    try:
-        if not robot.is_connected:
-            raise ValueError("Robot is not connected!")
+    if not robot.is_connected:
+        raise ValueError("Robot is not connected!")
 
-        print("Starting evaluate loop...")
-        episode_idx = 0
-        for episode_idx in range(NUM_EPISODES):
-            log_say(f"Running inference, recording eval episode {episode_idx + 1} of {NUM_EPISODES}")
+    print("Starting evaluate loop...")
+    episode_idx = 0
+    for episode_idx in range(NUM_EPISODES):
+        log_say(f"Running inference, recording eval episode {episode_idx + 1} of {NUM_EPISODES}")
 
-            # Main record loop
+        # Main record loop
+        record_loop(
+            robot=robot,
+            events=events,
+            fps=FPS,
+            policy=policy,
+            preprocessor=preprocessor,  # Pass the pre and post policy processors
+            postprocessor=postprocessor,
+            dataset=dataset,
+            control_time_s=EPISODE_TIME_SEC,
+            single_task=TASK_DESCRIPTION,
+            display_data=True,
+            teleop_action_processor=make_default_teleop_action_processor(),
+            robot_action_processor=robot_ee_to_joints_processor,
+            robot_observation_processor=robot_joints_to_ee_pose_processor,
+        )
+
+        # Reset the environment if not stopping or re-recording
+        if not events["stop_recording"] and ((episode_idx < NUM_EPISODES - 1) or events["rerecord_episode"]):
+            log_say("Reset the environment")
             record_loop(
                 robot=robot,
                 events=events,
                 fps=FPS,
-                policy=policy,
-                preprocessor=preprocessor,  # Pass the pre and post policy processors
-                postprocessor=postprocessor,
-                dataset=dataset,
                 control_time_s=EPISODE_TIME_SEC,
                 single_task=TASK_DESCRIPTION,
                 display_data=True,
@@ -167,41 +183,24 @@ def main():
                 robot_observation_processor=robot_joints_to_ee_pose_processor,
             )
 
-            # Reset the environment if not stopping or re-recording
-            if not events["stop_recording"] and (
-                (episode_idx < NUM_EPISODES - 1) or events["rerecord_episode"]
-            ):
-                log_say("Reset the environment")
-                record_loop(
-                    robot=robot,
-                    events=events,
-                    fps=FPS,
-                    control_time_s=EPISODE_TIME_SEC,
-                    single_task=TASK_DESCRIPTION,
-                    display_data=True,
-                    teleop_action_processor=make_default_teleop_action_processor(),
-                    robot_action_processor=robot_ee_to_joints_processor,
-                    robot_observation_processor=robot_joints_to_ee_pose_processor,
-                )
+        if events["rerecord_episode"]:
+            log_say("Re-record episode")
+            events["rerecord_episode"] = False
+            events["exit_early"] = False
+            dataset.clear_episode_buffer()
+            continue
 
-            if events["rerecord_episode"]:
-                log_say("Re-record episode")
-                events["rerecord_episode"] = False
-                events["exit_early"] = False
-                dataset.clear_episode_buffer()
-                continue
+        # Save episode
+        dataset.save_episode()
+        episode_idx += 1
 
-            # Save episode
-            dataset.save_episode()
-            episode_idx += 1
-    finally:
-        # Clean up
-        log_say("Stop recording")
-        robot.disconnect()
-        listener.stop()
+    # Clean up
+    log_say("Stop recording")
+    robot.disconnect()
+    listener.stop()
 
-        dataset.finalize()
-        dataset.push_to_hub()
+    dataset.finalize()
+    dataset.push_to_hub()
 
 
 if __name__ == "__main__":

examples/so100_to_so100_EE/record.py

@@ -16,26 +16,27 @@
 
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
-from lerobot.datasets.feature_utils import combine_feature_dicts
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.datasets.pipeline_features import aggregate_pipeline_dataset_features, create_initial_features
+from lerobot.datasets.utils import combine_feature_dicts
 from lerobot.model.kinematics import RobotKinematics
-from lerobot.processor import RobotProcessorPipeline
+from lerobot.processor import RobotAction, RobotObservation, RobotProcessorPipeline
 from lerobot.processor.converters import (
     observation_to_transition,
     robot_action_observation_to_transition,
     transition_to_observation,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     EEBoundsAndSafety,
     ForwardKinematicsJointsToEE,
     InverseKinematicsEEToJoints,
 )
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.scripts.lerobot_record import record_loop
-from lerobot.teleoperators.so_leader import SO100Leader, SO100LeaderConfig
-from lerobot.types import RobotAction, RobotObservation
+from lerobot.teleoperators.so100_leader.config_so100_leader import SO100LeaderConfig
+from lerobot.teleoperators.so100_leader.so100_leader import SO100Leader
 from lerobot.utils.control_utils import init_keyboard_listener
 from lerobot.utils.utils import log_say
 from lerobot.utils.visualization_utils import init_rerun
@@ -147,23 +148,38 @@ def main():
     listener, events = init_keyboard_listener()
     init_rerun(session_name="recording_phone")
 
-    try:
-        if not leader.is_connected or not follower.is_connected:
-            raise ValueError("Robot or teleop is not connected!")
+    if not leader.is_connected or not follower.is_connected:
+        raise ValueError("Robot or teleop is not connected!")
 
-        print("Starting record loop...")
-        episode_idx = 0
-        while episode_idx < NUM_EPISODES and not events["stop_recording"]:
-            log_say(f"Recording episode {episode_idx + 1} of {NUM_EPISODES}")
+    print("Starting record loop...")
+    episode_idx = 0
+    while episode_idx < NUM_EPISODES and not events["stop_recording"]:
+        log_say(f"Recording episode {episode_idx + 1} of {NUM_EPISODES}")
 
-            # Main record loop
+        # Main record loop
+        record_loop(
+            robot=follower,
+            events=events,
+            fps=FPS,
+            teleop=leader,
+            dataset=dataset,
+            control_time_s=EPISODE_TIME_SEC,
+            single_task=TASK_DESCRIPTION,
+            display_data=True,
+            teleop_action_processor=leader_joints_to_ee,
+            robot_action_processor=ee_to_follower_joints,
+            robot_observation_processor=follower_joints_to_ee,
+        )
+
+        # Reset the environment if not stopping or re-recording
+        if not events["stop_recording"] and (episode_idx < NUM_EPISODES - 1 or events["rerecord_episode"]):
+            log_say("Reset the environment")
             record_loop(
                 robot=follower,
                 events=events,
                 fps=FPS,
                 teleop=leader,
-                dataset=dataset,
-                control_time_s=EPISODE_TIME_SEC,
+                control_time_s=RESET_TIME_SEC,
                 single_task=TASK_DESCRIPTION,
                 display_data=True,
                 teleop_action_processor=leader_joints_to_ee,
@@ -171,44 +187,25 @@ def main():
                 robot_observation_processor=follower_joints_to_ee,
             )
 
-            # Reset the environment if not stopping or re-recording
-            if not events["stop_recording"] and (
-                episode_idx < NUM_EPISODES - 1 or events["rerecord_episode"]
-            ):
-                log_say("Reset the environment")
-                record_loop(
-                    robot=follower,
-                    events=events,
-                    fps=FPS,
-                    teleop=leader,
-                    control_time_s=RESET_TIME_SEC,
-                    single_task=TASK_DESCRIPTION,
-                    display_data=True,
-                    teleop_action_processor=leader_joints_to_ee,
-                    robot_action_processor=ee_to_follower_joints,
-                    robot_observation_processor=follower_joints_to_ee,
-                )
+        if events["rerecord_episode"]:
+            log_say("Re-recording episode")
+            events["rerecord_episode"] = False
+            events["exit_early"] = False
+            dataset.clear_episode_buffer()
+            continue
 
-            if events["rerecord_episode"]:
-                log_say("Re-recording episode")
-                events["rerecord_episode"] = False
-                events["exit_early"] = False
-                dataset.clear_episode_buffer()
-                continue
+        # Save episode
+        dataset.save_episode()
+        episode_idx += 1
 
-            # Save episode
-            dataset.save_episode()
-            episode_idx += 1
+    # Clean up
+    log_say("Stop recording")
+    leader.disconnect()
+    follower.disconnect()
+    listener.stop()
 
-    finally:
-        # Clean up
-        log_say("Stop recording")
-        leader.disconnect()
-        follower.disconnect()
-        listener.stop()
-
-        dataset.finalize()
-        dataset.push_to_hub()
+    dataset.finalize()
+    dataset.push_to_hub()
 
 
 if __name__ == "__main__":

examples/so100_to_so100_EE/replay.py

@@ -19,16 +19,16 @@ import time
 
 from lerobot.datasets.lerobot_dataset import LeRobotDataset
 from lerobot.model.kinematics import RobotKinematics
-from lerobot.processor import RobotProcessorPipeline
+from lerobot.processor import RobotAction, RobotObservation, RobotProcessorPipeline
 from lerobot.processor.converters import (
     robot_action_observation_to_transition,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     InverseKinematicsEEToJoints,
 )
-from lerobot.types import RobotAction, RobotObservation
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 from lerobot.utils.constants import ACTION
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.utils import log_say
@@ -68,40 +68,39 @@ def main():
 
     # Fetch the dataset to replay
     dataset = LeRobotDataset(HF_REPO_ID, episodes=[EPISODE_IDX])
-    actions = dataset.select_columns(ACTION)
+    # Filter dataset to only include frames from the specified episode since episodes are chunked in dataset V3.0
+    episode_frames = dataset.hf_dataset.filter(lambda x: x["episode_index"] == EPISODE_IDX)
+    actions = episode_frames.select_columns(ACTION)
 
     # Connect to the robot
     robot.connect()
 
-    try:
-        if not robot.is_connected:
-            raise ValueError("Robot is not connected!")
+    if not robot.is_connected:
+        raise ValueError("Robot is not connected!")
 
-        print("Starting replay loop...")
-        log_say(f"Replaying episode {EPISODE_IDX}")
-        for idx in range(dataset.num_frames):
-            t0 = time.perf_counter()
+    print("Starting replay loop...")
+    log_say(f"Replaying episode {EPISODE_IDX}")
+    for idx in range(len(episode_frames)):
+        t0 = time.perf_counter()
 
-            # Get recorded action from dataset
-            ee_action = {
-                name: float(actions[idx][ACTION][i])
-                for i, name in enumerate(dataset.features[ACTION]["names"])
-            }
+        # Get recorded action from dataset
+        ee_action = {
+            name: float(actions[idx][ACTION][i]) for i, name in enumerate(dataset.features[ACTION]["names"])
+        }
 
-            # Get robot observation
-            robot_obs = robot.get_observation()
+        # Get robot observation
+        robot_obs = robot.get_observation()
 
-            # Dataset EE -> robot joints
-            joint_action = robot_ee_to_joints_processor((ee_action, robot_obs))
+        # Dataset EE -> robot joints
+        joint_action = robot_ee_to_joints_processor((ee_action, robot_obs))
 
-            # Send action to robot
-            _ = robot.send_action(joint_action)
+        # Send action to robot
+        _ = robot.send_action(joint_action)
 
-            precise_sleep(max(1.0 / dataset.fps - (time.perf_counter() - t0), 0.0))
+        precise_sleep(1.0 / dataset.fps - (time.perf_counter() - t0))
 
-    finally:
-        # Clean up
-        robot.disconnect()
+    # Clean up
+    robot.disconnect()
 
 
 if __name__ == "__main__":

examples/so100_to_so100_EE/teleoperate.py

@@ -17,20 +17,21 @@
 import time
 
 from lerobot.model.kinematics import RobotKinematics
-from lerobot.processor import RobotProcessorPipeline
+from lerobot.processor import RobotAction, RobotObservation, RobotProcessorPipeline
 from lerobot.processor.converters import (
     robot_action_observation_to_transition,
     robot_action_to_transition,
     transition_to_robot_action,
 )
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
-from lerobot.robots.so_follower.robot_kinematic_processor import (
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.robot_kinematic_processor import (
     EEBoundsAndSafety,
     ForwardKinematicsJointsToEE,
     InverseKinematicsEEToJoints,
 )
-from lerobot.teleoperators.so_leader import SO100Leader, SO100LeaderConfig
-from lerobot.types import RobotAction, RobotObservation
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
+from lerobot.teleoperators.so100_leader.config_so100_leader import SO100LeaderConfig
+from lerobot.teleoperators.so100_leader.so100_leader import SO100Leader
 from lerobot.utils.robot_utils import precise_sleep
 from lerobot.utils.visualization_utils import init_rerun, log_rerun_data
 

examples/training/train_policy.py

@@ -19,9 +19,8 @@ from pathlib import Path
 import torch
 
 from lerobot.configs.types import FeatureType
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
-from lerobot.datasets.feature_utils import dataset_to_policy_features
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.lerobot_dataset import LeRobotDataset, LeRobotDatasetMetadata
+from lerobot.datasets.utils import dataset_to_policy_features
 from lerobot.policies.diffusion.configuration_diffusion import DiffusionConfig
 from lerobot.policies.diffusion.modeling_diffusion import DiffusionPolicy
 from lerobot.policies.factory import make_pre_post_processors

examples/training/train_with_streaming.py

@@ -20,9 +20,9 @@ from pathlib import Path
 import torch
 
 from lerobot.configs.types import FeatureType
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
-from lerobot.datasets.feature_utils import dataset_to_policy_features
+from lerobot.datasets.lerobot_dataset import LeRobotDatasetMetadata
 from lerobot.datasets.streaming_dataset import StreamingLeRobotDataset
+from lerobot.datasets.utils import dataset_to_policy_features
 from lerobot.policies.act.configuration_act import ACTConfig
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors

examples/tutorial/act/act_training_example.py

@@ -5,9 +5,8 @@ from pathlib import Path
 import torch
 
 from lerobot.configs.types import FeatureType
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
-from lerobot.datasets.feature_utils import dataset_to_policy_features
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.lerobot_dataset import LeRobotDataset, LeRobotDatasetMetadata
+from lerobot.datasets.utils import dataset_to_policy_features
 from lerobot.policies.act.configuration_act import ACTConfig
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors

examples/tutorial/act/act_using_example.py

@@ -1,11 +1,12 @@
 import torch
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
+from lerobot.datasets.lerobot_dataset import LeRobotDatasetMetadata
 from lerobot.policies.act.modeling_act import ACTPolicy
 from lerobot.policies.factory import make_pre_post_processors
 from lerobot.policies.utils import build_inference_frame, make_robot_action
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 
 MAX_EPISODES = 5
 MAX_STEPS_PER_EPISODE = 20

examples/tutorial/async-inf/robot_client.py

@@ -4,7 +4,7 @@ from lerobot.async_inference.configs import RobotClientConfig
 from lerobot.async_inference.helpers import visualize_action_queue_size
 from lerobot.async_inference.robot_client import RobotClient
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
-from lerobot.robots.so_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower import SO100FollowerConfig
 
 
 def main():
@@ -30,7 +30,6 @@ def main():
         robot=robot_cfg,
         server_address=server_address,
         policy_device="mps",
-        client_device="cpu",
         policy_type="act",
         pretrained_name_or_path="<user>/robot_learning_tutorial_act",
         chunk_size_threshold=0.5,  # g

examples/tutorial/diffusion/diffusion_training_example.py

@@ -5,9 +5,8 @@ from pathlib import Path
 import torch
 
 from lerobot.configs.types import FeatureType
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
-from lerobot.datasets.feature_utils import dataset_to_policy_features
-from lerobot.datasets.lerobot_dataset import LeRobotDataset
+from lerobot.datasets.lerobot_dataset import LeRobotDataset, LeRobotDatasetMetadata
+from lerobot.datasets.utils import dataset_to_policy_features
 from lerobot.policies.diffusion.configuration_diffusion import DiffusionConfig
 from lerobot.policies.diffusion.modeling_diffusion import DiffusionPolicy
 from lerobot.policies.factory import make_pre_post_processors

examples/tutorial/diffusion/diffusion_using_example.py

@@ -1,11 +1,12 @@
 import torch
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
-from lerobot.datasets.dataset_metadata import LeRobotDatasetMetadata
+from lerobot.datasets.lerobot_dataset import LeRobotDatasetMetadata
 from lerobot.policies.diffusion.modeling_diffusion import DiffusionPolicy
 from lerobot.policies.factory import make_pre_post_processors
 from lerobot.policies.utils import build_inference_frame, make_robot_action
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 
 MAX_EPISODES = 5
 MAX_STEPS_PER_EPISODE = 20

examples/tutorial/pi0/using_pi0_example.py

@@ -1,11 +1,12 @@
 import torch
 
 from lerobot.cameras.opencv.configuration_opencv import OpenCVCameraConfig
-from lerobot.datasets.feature_utils import hw_to_dataset_features
+from lerobot.datasets.utils import hw_to_dataset_features
 from lerobot.policies.factory import make_pre_post_processors
 from lerobot.policies.pi0.modeling_pi0 import PI0Policy
 from lerobot.policies.utils import build_inference_frame, make_robot_action
-from lerobot.robots.so_follower import SO100Follower, SO100FollowerConfig
+from lerobot.robots.so100_follower.config_so100_follower import SO100FollowerConfig
+from lerobot.robots.so100_follower.so100_follower import SO100Follower
 
 MAX_EPISODES = 5
 MAX_STEPS_PER_EPISODE = 20