added readme, dataset recording guide and ACT policy training for so101 arm

examples/lerobot-dataset-record/Readme.md

@@ -0,0 +1,137 @@
+# Dataset Recording with SO101 Robot Arm
+
+Guide for recording imitation learning datasets with the SO101 arm using Dora and train policies with LeRobot.
+
+## Overview
+
+To begin recording, edit the dataflow.yml file. You must update camera settings, robot ports, and dataset configuration based on your setup.
+
+### Required Changes
+
+Update the camera device id, you can also add or remove cameras based on your setup:
+
+```yaml
+# Laptop camera
+CAPTURE_PATH: "0"  # Laptop camera is defaults to 0
+# External camera
+CAPTURE_PATH: "1"  # Change this to match your external camera device
+```
+### SO101 Arms
+
+Identify and set the correct USB ports for both the leader and follower SO101 arms
+```yaml
+PORT: "/dev/ttyACM0" # change this
+```
+### Dataset Recorder Configuration
+
+>Edit the following fields:
+
+```yaml
+REPO_ID: "your_username/so101_dataset"  # HuggingFace dataset path
+SINGLE_TASK: "Pick up the red cube and place it in the blue box"  # Your task description
+CAMERA_NAMES: "laptop, front"  # Name your camera sources depending on your setup
+CAMERA_LAPTOP_RESOLUTION: "480,640,3"
+CAMERA_FRONT_RESOLUTION: "480,640,3"
+```
+
+You can adjust the following parameters as needed:
+
+```yaml
+TOTAL_EPISODES: "2"         # Number of episodes 
+#you may want to try with 2-3 episodes to test, then atleast 50 episodes for training is recommended
+EPISODE_DURATION_S: "60"     # Duration of each episode (in seconds) - depends on complexity of task
+RESET_DURATION_S: "15"       # Time to reset the environment between episodes
+FPS: "30"                    # Should match camera fps
+PUSH_TO_HUB: "false"         # Set to "true" to auto-upload dataset to HuggingFace
+ROOT_PATH: "full path where you want to save the dataset" 
+# if not defined then it will be stored at ~/.cache/huggingface/lerobot/repo_id
+```
+
+Once everything is updated in `dataflow.yml`, you are ready to record your dataset.
+
+## Start Recording
+Build and Run
+
+```bash
+dora build dataflow.yml
+dora run dataflow.yml
+```
+
+## Recording Process
+
+In the rerun window you can see the the info regarding Start of episodes, start of reset phase and saving of episodes.
+
+#### During Recording
+
+1. **Episode Active**:
+   - Use the **leader robot** to demonstrate/perform the task
+   - Move smoothly and naturally
+   - Complete the full task within the time limit
+
+2. **Reset Phase**:
+   - move the leader arm to initial position
+   - Reset objects to starting positions
+   - Prepare workspace for next demonstration
+
+3. **Repeat** until all episodes are complete
+
+**Recording Tips**
+
+- **Practice First**: Move the leader robot around to get comfortable before recording
+- **Smooth Movements**: Avoid jerky or sudden movements
+- **Complete Tasks**: Try to finish the task within each episode
+- **Consistent Setup**: Keep object positions and lighting consistent
+- **Monitor Rerun**: Watch the visualization to ensure cameras and robot data are flowing
+
+## After Recording
+
+Your dataset will be saved locally. Check the recording was successful:
+>It will be stored in ~/.cache/huggingface/lerobot/repo_id
+
+
+# Training and Testing Policies using the recorded dataset
+
+After successfully recording your dataset, we will be training imitation learning policies and deploying them on your SO101 robot arm.
+
+## Training Your Policy
+
+#### Install LeRobot Training Dependencies
+
+Easiest way to train your policy is to use lerobots training scripts
+```bash
+# Install training requirements and lerobot repo
+git clone https://github.com/huggingface/lerobot.git
+pip install lerobot[training]
+pip install tensorboard wandb  # For monitoring (Optional)
+```
+
+### Choose Your Policy
+
+**ACT (Recommended for SO101)**
+- Good for manipulation tasks
+- Handles multi-modal data well
+- Faster training (for me it took 7hrs 🥺 to train on 50 episodes for pick and place, 3050 laptop gpu)
+
+**Diffusion Policy**
+- Better for complex behaviors
+- More robust to distribution shift
+- Longer training
+
+
+### Start Training
+
+```bash
+cd lerobot
+
+python lerobot/scripts/train.py \
+  --dataset.repo_id=${HF_USER}/your_repo_id \ # provide full path of your dataset
+  --policy.type=act \
+  --output_dir=outputs/train/act_so101_test \
+  --job_name=act_so101_test \
+  --policy.device=cuda \
+  --wandb.enable=true \
+  --policy.repo_id=${HF_USER}/my_policy
+```
+
+You can monitor your training progress on wandb
+> For more details regarding training check [lerobot](https://huggingface.co/docs/lerobot/en/il_robots#train-a-policy) guide on Imitation learning for SO101 and testing your policy

examples/lerobot-dataset-record/dataflow.yml

@@ -20,7 +20,7 @@ nodes:
     outputs:
       - image
     env:
-      CAPTURE_PATH: "2"
+      CAPTURE_PATH: "1"
       ENCODING: "rgb8"
       IMAGE_WIDTH: "640"
       IMAGE_HEIGHT: "480"
@@ -48,7 +48,7 @@ nodes:
       PORT: "/dev/ttyACM0"
       IDS: "1,2,3,4,5,6"
 
-  - id: lerobot_recorder
+  - id: dataset_recorder
     build: pip install -e ../../node-hub/dora-dataset-record
     path: dora-dataset-record
     inputs:
@@ -59,12 +59,12 @@ nodes:
     outputs:
       - text
     env:
-      REPO_ID: "HF_username/name_of_dataset"
+      REPO_ID: "HF_username/Name_your_dataset" 
       SINGLE_TASK: "Pick up the cube and place it in the box"
-      ROBOT_TYPE: "your robot type" # e.g., "so101_follower", "ur5e" etc
+      ROBOT_TYPE: "so101_follower" # e.g., "koch", "franka", "ur5e" etc
 
       FPS: "30"
-      TOTAL_EPISODES: "5"
+      TOTAL_EPISODES: "50"
       EPISODE_DURATION_S: "60"
       RESET_DURATION_S: "15"
       
@@ -85,4 +85,4 @@ nodes:
     inputs:
       image_laptop: laptop_cam/image
       image_front: front_cam/image
-      text: lerobot_recorder/text
+      text: dataset_recorder/text

node-hub/dora-dataset-record/README.md

@@ -1,40 +1,107 @@
 # dora-dataset-record
 
-## Getting started
+Node for recording robot datasets in LeRobot format. You can captures synchronized camera feeds and robot poses to create high-quality datasets for imitation learning and robot training.
 
-- Install it with uv:
+- **Robot pose recording** - Capture both state and action data
+- **Multi-camera support** - Record from multiple cameras simultaneously
+- **LeRobot dataset format (v2.1)** - Direct integration with HuggingFace LeRobot datasets
+- **Episode management** - Automatic episode segmentation with reset phases
+
+## Quick Start
+
+### 1. Installation
 
 ```bash
-uv venv -p 3.11 --seed
+# Source your venv 
+cd dora/node-hub/dora-dataset-record
 uv pip install -e .
 ```
 
-## Contribution Guide
+### 2. Usage Guide
 
-- Format with [ruff](https://docs.astral.sh/ruff/):
+Create a dataflow file, see `examples/lerobot-dataset-record/dataflow.yml`:
 
-```bash
-uv pip install ruff
-uv run ruff check . --fix
+```yaml
+nodes:
+  # Dataset recorder
+  - id: dataset_recorder
+    build: pip install -e ../../node-hub/dora-dataset-record
+    path: dora-dataset-record
+    inputs:
+      laptop: laptop_cam/image
+      front: front_cam/image
+      robot_state: robot_follower/pose
+      robot_action: leader_interface/pose
+    outputs:
+      - text
+    env:
+      # Required settings
+      REPO_ID: "your_username/your_dataset_name"
+      SINGLE_TASK: "Pick up the cube and place it in the box"
+      ROBOT_TYPE: "your_robot_type"
+      
+      # Recording settings
+      FPS: "30"
+      TOTAL_EPISODES: "50" 
+      EPISODE_DURATION_S: "60"
+      RESET_DURATION_S: "15"
+      
+      # Camera configuration
+      CAMERA_NAMES: "laptop,front"
+      CAMERA_LAPTOP_RESOLUTION: "480,640,3"
+      CAMERA_FRONT_RESOLUTION: "480,640,3"
+      
+      # Robot configuration
+      ROBOT_JOINTS: "joint1,joint2,joint3,joint4,joint5,gripper"
+      
+      # Optional settings
+      USE_VIDEOS: "true"
+      PUSH_TO_HUB: "false" 
+      PRIVATE: "false"
+      TAGS: "robotics,manipulation,imitation_learning"
+
+  # Visualization with rerun
+  - id: plot
+    build: pip install dora-rerun
+    path: dora-rerun
+    inputs:
+      text: dataset_recorder/text
 ```
 
-- Lint with ruff:
+### 3. Start Recording the dataset
 
 ```bash
-uv run ruff check .
+dora build dataflow.yml
+dora run dataflow.yml
 ```
 
-- Test with [pytest](https://github.com/pytest-dev/pytest)
+The node will send instructions on dora-rerun, about episode starting, reset time, Saving episodes etc.
 
-```bash
-uv pip install pytest
-uv run pytest . # Test
-```
+## Configuration
+
+### Required Environment Variables
+
+| Variable | Description | Example |
+|----------|-------------|---------|
+| `REPO_ID` | HuggingFace dataset repo | `"username/dataset_name"` |
+| `SINGLE_TASK` | Task description | `"Pick and place objects"` |
+| `CAMERA_NAMES` | Comma-separated camera names | `"laptop,front,top"` |
+| `CAMERA_*_RESOLUTION` | Resolution for each camera | `"480,640,3"` |
+| `ROBOT_JOINTS` | Comma-separated joint names | `"joint1,joint2,gripper"` |
 
-## YAML Specification
+### Optional Settings
 
-## Examples
+| Variable | Default | Description |
+|----------|---------|-------------|
+| `FPS` | `30` | Recording frame rate (match camera fps) |
+| `TOTAL_EPISODES` | `10` | Number of episodes to record |
+| `EPISODE_DURATION_S` | `60` | Episode length in seconds |
+| `RESET_DURATION_S` | `15` | Break between episodes to reset the environment |
+| `USE_VIDEOS` | `true` | Encode as MP4 videos, else saves images |
+| `PUSH_TO_HUB` | `false` | Upload to HuggingFace Hub |
+| `PRIVATE` | `false` | Make dataset private |
+| `ROOT_PATH` | `~/.cache/huggingface/lerobot/your_repo_id` | Local storage path where you want to save the dataset |
 
 ## License
 
-dora-dataset-record's code are released under the MIT License
+This project is released under the MIT License.

node-hub/dora-dataset-record/dora_dataset_record/main.py

@@ -311,6 +311,7 @@ class DoraLeRobotRecorder:
         """Output message."""
         # Put message in queue to send
         self.message_queue.put(message)
+        print(message)
 
     def get_pending_messages(self):
         """Get all pending messages from the queue."""

node-hub/dora-dataset-record/pyproject.toml

@@ -27,5 +27,4 @@ ignore = [
 ]
 
 [tool.uv.sources]
-lerobot = {git = "https://github.com/huggingface/lerobot.git" }
-
+lerobot = {git = "https://github.com/huggingface/lerobot.git", rev = "main"}