0. Quick start#

Imitation Learning is a powerful tool for training agents to perform tasks by leveraging expert demonstrations, especially when reward signals are sparse or hard to specify.

Here we will show the ready-to-run commands to quickly install the dependencies of imitation learning and get it up.

1. Installation#

bash roboverse_learn/il/il_setup.sh

This command can install all relevant libraries for diffusion policy and action chunking.

2. Running#

 bash roboverse_learn/il/il_run.sh --task_name_set close_box --algo_choose act --demo_num 100 --sim_set mujoco

This command performs demo collection, model training, and evaluation.

Argument

Description

Example

task_name_set

Name of the task

close_boxstack_cube, pick_cube, pick_butter

sim_set

Name of the selected simulator

isaacsim, mujoco

demo_num

Numbers of demos to collect, train, and eval

100

algo_choose

IL algorithm

act, dp_DDPM, dp_DDIM, dp_FM_UNet, dp_FM_DiT, dp_Score, dp_VITA