Skip to content

Latest commit

 

History

History
 
 

scripts

Try it on Google Colab

This page contains example notebooks to train RL agents on highway-env using several RL libraries.

⚠️ Stable Baselines3 does not currently support gymnasium. As of now, these notebooks are only compatible with previous versions of highway-env, such as highway-env==1.5

Highway with DQN Open In Colab

Train a highway-v0 policy with DQN.

Highway with PPO Open In Colab

Train a highway-v0 policy with PPO.

Highway + DQN using a CNN and image observations Open In Colab

Train a highway-v0 policy with DQN, while using image observations and a CNN model architecture.

Parking with Hindsight Experience Replay Open In Colab

Train a goal-conditioned parking-v0 policy using the HER implementation from stable-baselines and hyperparameters from the RL Zoo.

Trajectory Planning on Highway Open In Colab

Plan a trajectory on highway-v0 using the OPD implementation from rl-agents.

A Model-based Reinforcement Learning tutorial on Parking Open In Colab

A tutorial written for RLSS 2019 and demonstrating the principle of model-based reinforcement learning on the parking-v0 task.

Intersection with DQN and social attention Open In Colab

Train an intersection-v0 crossing policy using the social attention architecture and the DQN implementation from rl-agents.