There is a small typo in the documentation of the [python API](https://github.com/Unity-Technologies/ml-agents/blob/master/docs/Python-API.md). The first argument passed into ```env.reset``` should be ```train_mode``` instead of ```train_model``` as implemented [here](https://github.com/Unity-Technologies/ml-agents/blob/9b1a39982fd03de8f40f85d61f903e6d972fd2cc/ml-agents-envs/mlagents/envs/environment.py#L308).