Diffusion Policy

Installation

cd roboverse_learn/il/utils/diffusion_policy

pip install -e .

cd ../../../../

pip install pandas wandb

Register for a Weights & Biases (wandb) account to obtain an API key.

Workflow

Step 1: Collect and pre-processing data

./roboverse_learn/il/collect_demo.sh

collect_demo.sh collects demos, i.e., metadata, using ~/RoboVerse/scripts/advanced/collect_demo.py and converts the metadata into Zarr format for efficient dataloading. This script can handle both joint position and end effector action and observation spaces.

Outputs: Metadata directory is stored in metadata_dir. Converted dataset is stored in ~/RoboVerse/data_policy

Parameters:

Argument	Description	Example
task_name_set	Name of the task	close_box
sim_set	Name of the selected simulator	isaacsim
max_demo_idx	Maximum index of demos to collect	100
expert_data_num	Number of expert demonstrations to process	100
metadata_dir	Path to the directory containing demonstration metadata saved by collect_demo	~/RoboVerse/roboverse_demo/demo_isaacsim/close_box-/robot-franka
action_space	Type of action space to use (options: ‘joint_pos’ or ‘ee’)	joint_pos
observation_space	Type of observation space to use (options: ‘joint_pos’ or ‘ee’)	joint_pos
delta_ee	(optional) Delta control (0: absolute, 1: delta; default 0)	0
cust_name	User defined name	noDR

Step 2: Training and evaluation

./roboverse_learn/il/dp/dp_run.sh

dp_run.sh uses roboverse_learn/il/dp/main.py and the generated Zarr data, which gets stored in the data_policy/ directory, to train and evaluate the DP model.

Outputs: Training result is stored in ~/RoboVerse/info/outputs/DP. Evaluation result is stored in ~/RoboVerse/tmp

Parameters:

Argument	Description	Example
task_name	Name of the task	close_box
sim_set	Name of the selected simulator	isaacsim
gpu_id	ID of the GPU to use	0
train_enable	Enable training	True
eval_enable	Enable evaluation	True
algo_choose	Choose training or inference algorithm. 0: DDPM, 1: DDIM, 2: Flow matching 3: Score-based	0
eval_path	Path of trained DP model. Need to be set if train_enable = False.	/path/to/your/checkpoint.ckpt
zarr_path	Path to the zarr dataset.	data_policy/${task_name}FrankaL${level}_${extra}_${expert_data_num}.zarr
seed	Random seed for reproducibility	42
num_epochs	Number of training epochs	200
obs_type	Observation type (joint_pos or ee)	joint_pos
action_type	Action type (joint_pos or ee)	joint_pos
delta	Delta control mode (0 for absolute, 1 for delta)	0
device	GPU device to use	"cuda:7"

Important Parameter Overrides:

• horizon, n_obs_steps, and n_action_steps are set directly in dp_runner.sh and override the YAML configurations.

Switching between Joint Position and End Effector Control

• Joint Position Control: Set both obs_space and act_space to joint_pos.

• End Effector Control: Set both obs_space and act_space to ee. You may use delta_ee=1 for delta mode or delta_ee=0 for absolute positioning.

Adjust relevant configuration parameters in:

• roboverse_learn/il/dp/config/.yaml