Reward Calculation for MuJoCo Environments

[arjunneervannan]

I am testing various reinforcement learning algorithms on the MuJoCo environments using Tensorforce, a reinforcement learning library built on Tensorflow. I did not find the reward calculations for these environments on the OpenAI website, and wanted to know if anyone had these calculations. I am using the HalfCheetah-v1 environment, the Walker2d-v1 environment, and the Ant-v1 environment.

[stabgan]

Have you looked inside the source files ?

[zhangchuheng123]

Reward calculation of all of the mentioned environments can be found here,

[jkterry1]

PR #2762 is about to be merged, introducing V4 MuJoCo environments using new bindings and a dramatically newer version of the engine. If this issue still persists with the V4 ones, please create a new issue for it.