- python 2.7
- tensorflow 0.12
python3 and tensorflow>=1.0 version available at https://github.com/openai/baselines
sh run_train_reacher.sh
Change this line https://github.com/openai/gym/blob/master/gym/envs/mujoco/mujoco_env.py#L118
from
self.viewer = mujoco_py.MjViewer()
to
self.viewer = mujoco_py.MjViewer(visible=False, init_width=160, init_height=210)
Then run
sh run_load_pixels.sh
Pretrained models available at https://www.dropbox.com/sh/6oru65vsafrnjez/AACfc2LQZUoZUIgpT8x0KTWDa?dl=0