Atari Enduro Environment

Overview

nduro consists of maneuvering a race car in the National Enduro, a long-distance endurance race. The object of the race is to pass a certain number of cars each day. Doing so will allow the player to continue racing for the next day. The driver must avoid other racers and pass 200 cars on the first day, and 300 cars with each following day.

As the time in the game passes, visibility changes as well. When it is night in the game the player can only see the oncoming cars’ taillights. As the days progress, cars will become more difficult to avoid as well. Weather and time of day are factors in how to play. During the day the player may drive through an icy patch on the road which would limit control of the vehicle, or a patch of fog may reduce visibility.

Description from Wikipedia

Performances of RL Agents

We list various reinforcement learning algorithms that were tested in this environment. These results are from RL Database. If this page was helpful, please consider giving a star!

Star

Human Starts

Result Algorithm Source
2223.9 PDD DQN Dueling Network Architectures for Deep Reinforcement Learning
2133.4 Distributional DQN Rainbow: Combining Improvements in Deep Reinforcement Learning
2077.4 DuDQN Dueling Network Architectures for Deep Reinforcement Learning
2061.1 Rainbow Rainbow: Combining Improvements in Deep Reinforcement Learning
1884.4 Prioritized DDQN (prop, tuned) Prioritized Experience Replay
1831.0 Prioritized DDQN (rank, tuned) Prioritized Experience Replay
1265.6 Prioritized DQN (rank) Prioritized Experience Replay
1216.6 DDQN (tuned) Deep Reinforcement Learning with Double Q-learning
740.2 Human Massively Parallel Methods for Deep Reinforcement Learning
475.9 DDQN Deep Reinforcement Learning with Double Q-learning
475.6 DQN Massively Parallel Methods for Deep Reinforcement Learning
71.04 Gorila DQN Massively Parallel Methods for Deep Reinforcement Learning
-81.8 Random Massively Parallel Methods for Deep Reinforcement Learning
-82.2 A3C FF 1 day Asynchronous Methods for Deep Reinforcement Learning
-82.5 A3C FF Asynchronous Methods for Deep Reinforcement Learning
-82.5 A3C LSTM Asynchronous Methods for Deep Reinforcement Learning

No-op Starts

Result Algorithm Source
3454 C51 A Distributional Perspective on Reinforcement Learning
2359 IQN Implicit Quantile Networks for Distributional Reinforcement Learning
2357 QR-DQN-0 Distributional Reinforcement Learning with Quantile Regression
2355 QR-DQN-1 Distributional Reinforcement Learning with Quantile Regression
2306.4 PDD DQN Dueling Network Architectures for Deep Reinforcement Learning
2259.3 Distributional DQN Rainbow: Combining Improvements in Deep Reinforcement Learning
2258.2 DuDQN Dueling Network Architectures for Deep Reinforcement Learning
2230.1 Reactor The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning
2224.2 Reactor The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning
2211.3 Reactor ND The Reactor: A fast and sample-efficient Actor-Critic agent for Reinforcement Learning
2125.9 Rainbow Rainbow: Combining Improvements in Deep Reinforcement Learning
2064 DuDQN Noisy Networks for Exploration
2013 NoisyNet DuDQN Noisy Networks for Exploration
1240 NoisyNet DQN Noisy Networks for Exploration
1211.8 DDQN A Distributional Perspective on Reinforcement Learning
971.28 IMPALA (deep, multitask) IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
860.5 Human Dueling Network Architectures for Deep Reinforcement Learning
835 DQN Noisy Networks for Exploration
729.0 DQN A Distributional Perspective on Reinforcement Learning
319.5 DDQN Deep Reinforcement Learning with Double Q-learning
309.6 Human Human-level control through deep reinforcement learning
301.8 DQN Human-level control through deep reinforcement learning
300 NoisyNet A3C Noisy Networks for Exploration
159.4 Contingency Human-level control through deep reinforcement learning
129.1 Linear Human-level control through deep reinforcement learning
114.9 Gorila DQN Massively Parallel Methods for Deep Reinforcement Learning
0.0 ACKTR Scalable trust-region method for deep reinforcement learning using Kronecker-factored approximation
0.0 Random Human-level control through deep reinforcement learning
0 A3C Noisy Networks for Exploration
0.0 IMPALA (shallow) IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
0.0 IMPALA (deep) IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures

Normal Starts

Result Algorithm Source
758.3 PPO Proximal Policy Optimization Algorithm
741 UCC-I Trust Region Policy Optimization
699.8 DQN RL Baselines Zoo b76641e
661 DQN2013 Best Playing Atari with Deep Reinforcement Learning
643.824 PPO RL Baselines Zoo b76641e
534.6 TRPO - single path Trust Region Policy Optimization
479.75 DQN OpenAI Baselines cbd21ef
470 DQN2013 Playing Atari with Deep Reinforcement Learning
430.8 TRPO - vine Trust Region Policy Optimization
368 Human Playing Atari with Deep Reinforcement Learning
350.22 PPO OpenAI Baselines cbd21ef
207.47 PPO (MPI) OpenAI Baselines cbd21ef
159 Contingency Playing Atari with Deep Reinforcement Learning
129 Sarsa Playing Atari with Deep Reinforcement Learning
106 HNeat Best Playing Atari with Deep Reinforcement Learning
91 HNeat Pixel Playing Atari with Deep Reinforcement Learning
24.83 TRPO (MPI) OpenAI Baselines cbd21ef
0 Random Playing Atari with Deep Reinforcement Learning
0.0 A2C Proximal Policy Optimization Algorithm
0.0 ACER Proximal Policy Optimization Algorithm
0.0 ACKTR OpenAI Baselines cbd21ef
0.0 ACER OpenAI Baselines cbd21ef
0.0 A2C OpenAI Baselines cbd21ef
0.0 A2C RL Baselines Zoo b76641e
0.0 ACER RL Baselines Zoo b76641e
0.0 ACKTR RL Baselines Zoo b76641e