wheel_legged_genesis v0.0.7

Complete joint domain randomization, better control command tracking on complex terrain, and relatively stable training, perfect victory over Mujoco
It is recommended to test a model with around 4,000 rounds, 10,000 because 10,000 rounds were trained
完整的关节域随机化，更好的复杂地形上控制命令跟踪，比较稳定的训练，完美战胜mujoco
建议测试4000轮左右的模型，10000是因为训练了10000轮

wheel_legged_genesis v0.0.5

Terrain training, custom terrain,add keyboard and better curriculum
添加地形训练，自定义地形，键盘控制，更稳定的课程学习。自定义地形见 assets/terrain，基础地形提供连续上下坡和崎岖路面。

wheel_legged_genesis v0.0.3

After the RD is opened, the Conmands course is more stable, and the iteration speed is faster(WARNING: Do not turn on the terrain)
开启域随机化后，相较于0.0.2更稳定，收敛更快，效果更好（注意不要开启地形）

wheel_legged_genesis v0.0.2

add domain rand

wheel_legged_genesis v0.0.1

Currently, only flat, rudimentary course learning is supported, and domain randomization is not joined

Full Changelog: https://github.com/Albusgive/wheel_legged_genesis/commits/v0.0.1