privateboss/Car_Race_AI_V5

这是一个包含30,000,000个训练时间步的智能体赛车数据集，使用Tensorflow 2.xx和PPO算法在Ubuntu Linux环境下训练。数据集基于CarRacing-v3 Gymnasium环境，这是一个OpenAI Gym库的维护分支。智能体学会了使用FIFO角落策略进行赛车，这是一种高风险和高回报的驾驶方式。This is an agent car racing dataset with 30,000,000 training time-steps, trained using Tensorflow 2.xx and PPO algorithm in Ubuntu Linux environment. The dataset is based on the CarRacing-v3 Gymnasium environment, which is a maintained fork of OpenAIs Gym library. The agent has learned to use the FIFO cornering strategy in racing, which is a high risk and high reward driving style.