Lipschitz-constrained Unsupervised Skill Discovery

This repository is the official implementation of

Seohong Park, Jongwook Choi*, Jaekyeom Kim*, Honglak Lee, Gunhee Kim (*: equal contribution). Lipschitz-constrained Unsupervised Skill Discovery. In ICLR, 2022. [paper] [project page] [slides]

The implementation is based on Unsupervised Skill Discovery with Bottleneck Option Learning and garage.

Visit our project page for more results including videos.

Requirements

Python 3.7.8
MuJoCo 1.5

Examples

Install requirements:

pip install -r requirements.txt
pip install -e .
pip install -e garaged

Ant with 2-D continuous skills:

python tests/main.py --run_group EXP --env ant --max_path_length 200 --dim_option 2 --common_lr 0.0001 --seed 0 --normalizer_type ant_preset --use_gpu 1 --traj_batch_size 20 --n_parallel 8 --n_epochs_per_eval 5000 --n_thread 1 --model_master_dim 1024 --record_metric_difference 0 --n_epochs_per_tb 100 --n_epochs_per_save 50000 --n_epochs_per_pt_save 5000 --n_epochs_per_pkl_update 1000 --eval_record_video 1 --n_epochs 200001 --spectral_normalization 1 --n_epochs_per_log 50 --discrete 0 --num_random_trajectories 200 --sac_discount 0.99 --alpha 0.01 --sac_lr_a -1 --lr_te 3e-05 --sac_scale_reward 0 --max_optimization_epochs 1 --trans_minibatch_size 2048 --trans_optimization_epochs 4 --eval_plot_axis -50 50 -50 50

Ant with 16 discrete skills:

python tests/main.py --run_group EXP --env ant --max_path_length 200 --dim_option 16 --common_lr 0.0001 --seed 0 --normalizer_type ant_preset --use_gpu 1 --traj_batch_size 20 --n_parallel 8 --n_epochs_per_eval 5000 --n_thread 1 --model_master_dim 1024 --record_metric_difference 0 --n_epochs_per_tb 100 --n_epochs_per_save 50000 --n_epochs_per_pt_save 5000 --n_epochs_per_pkl_update 1000 --eval_record_video 1 --n_epochs 200001 --spectral_normalization 1 --n_epochs_per_log 50 --discrete 1 --num_random_trajectories 200 --sac_discount 0.99 --alpha 0.003 --sac_lr_a -1 --lr_te 3e-05 --sac_scale_reward 0 --max_optimization_epochs 1 --trans_minibatch_size 2048 --trans_optimization_epochs 4 --eval_plot_axis -50 50 -50 50

Humanoid with 2-D continuous skills:

python tests/main.py --run_group EXP --env humanoid --max_path_length 1000 --dim_option 2 --common_lr 0.0003 --seed 0 --normalizer_type humanoid_preset --use_gpu 1 --traj_batch_size 5 --n_parallel 8 --n_epochs_per_eval 5000 --n_thread 1 --model_master_dim 1024 --record_metric_difference 0 --n_epochs_per_tb 100 --n_epochs_per_save 50000 --n_epochs_per_pt_save 5000 --n_epochs_per_pkl_update 1000 --eval_record_video 1 --n_epochs 200001 --spectral_normalization 1 --n_epochs_per_log 50 --discrete 0 --video_skip_frames 3 --num_random_trajectories 200 --sac_discount 0.99 --alpha 0.03 --sac_lr_a -1 --lr_te 0.0001 --lsd_alive_reward 0.03 --sac_scale_reward 0 --max_optimization_epochs 1 --trans_minibatch_size 2048 --trans_optimization_epochs 4 --sac_replay_buffer 1 --te_max_optimization_epochs 1 --te_trans_optimization_epochs 2

Humanoid with 16 discrete skills:

python tests/main.py --run_group EXP --env humanoid --max_path_length 1000 --dim_option 16 --common_lr 0.0003 --seed 0 --normalizer_type humanoid_preset --use_gpu 1 --traj_batch_size 5 --n_parallel 8 --n_epochs_per_eval 5000 --n_thread 1 --model_master_dim 1024 --record_metric_difference 0 --n_epochs_per_tb 100 --n_epochs_per_save 50000 --n_epochs_per_pt_save 5000 --n_epochs_per_pkl_update 1000 --eval_record_video 1 --n_epochs 200001 --spectral_normalization 1 --n_epochs_per_log 50 --discrete 1 --video_skip_frames 3 --num_random_trajectories 200 --sac_discount 0.99 --alpha 0.03 --sac_lr_a -1 --lr_te 0.0001 --lsd_alive_reward 0.03 --sac_scale_reward 0 --max_optimization_epochs 1 --trans_minibatch_size 2048 --trans_optimization_epochs 4 --sac_replay_buffer 1 --te_max_optimization_epochs 1 --te_trans_optimization_epochs 2

HalfCheetah with 8 discrete skills:

python tests/main.py --run_group EXP --env half_cheetah --max_path_length 200 --dim_option 8 --common_lr 0.0001 --seed 0 --normalizer_type half_cheetah_preset --use_gpu 1 --traj_batch_size 20 --n_parallel 8 --n_epochs_per_eval 5000 --n_thread 1 --model_master_dim 1024 --record_metric_difference 0 --n_epochs_per_tb 100 --n_epochs_per_save 50000 --n_epochs_per_pt_save 5000 --n_epochs_per_pkl_update 1000 --eval_record_video 1 --n_epochs 200001 --spectral_normalization 1 --n_epochs_per_log 50 --discrete 1 --num_random_trajectories 200 --sac_discount 0.99 --alpha 0.01 --sac_lr_a -1 --lr_te 3e-05 --sac_scale_reward 0 --max_optimization_epochs 1 --trans_minibatch_size 2048 --trans_optimization_epochs 4

Lipschitz-constrained Unsupervised Skill Discovery

Related tags

Overview

Lipschitz-constrained Unsupervised Skill Discovery

Requirements

Examples

Owner

Seohong Park

Deep-learning-roadmap - All You Need to Know About Deep Learning - A kick-starter

Godot RL Agents is a fully Open Source packages that allows video game creators

Official repo of the paper "Surface Form Competition: Why the Highest Probability Answer Isn't Always Right"

Social Network Ads Prediction

Neuron Merging: Compensating for Pruned Neurons (NeurIPS 2020)

TorchMultimodal is a PyTorch library for training state-of-the-art multimodal multi-task models at scale.

Production First and Production Ready End-to-End Speech Recognition Toolkit

An open source object detection toolbox based on PyTorch

Franka Emika Panda manipulator kinematics&dynamics simulation

[ICCV 2021] Our work presents a novel neural rendering approach that can efficiently reconstruct geometric and neural radiance fields for view synthesis.

[WACV21] Code for our paper: Samuel, Atzmon and Chechik, "From Generalized zero-shot learning to long-tail with class descriptors"

Simple tutorials using Google's TensorFlow Framework

Boston House Prediction Valuation Tool

LightNet++: Boosted Light-weighted Networks for Real-time Semantic Segmentation

A Python library for Deep Probabilistic Modeling

Multi agent DDPG algorithm written in Python + Pytorch

TensorFlow GNN is a library to build Graph Neural Networks on the TensorFlow platform.

MetaDrive: Composing Diverse Scenarios for Generalizable Reinforcement Learning

Code for "Finding Regions of Heterogeneity in Decision-Making via Expected Conditional Covariance" at NeurIPS 2021

A curated list of awesome Deep Learning tutorials, projects and communities.