Files

pzhokhov 36ee5d1707 Import internal changes (#422 )

* import rl-algs from 2e3a166 commit

* extra import of the baselines badge

* exported commit with identity test

* proper rng seeding in the test_identity

* import internal

2018-06-06 11:39:13 -07:00

__init__.py

add __init__.py

2017-08-27 22:36:24 -07:00

cnn_policy.py

Lots of cleanups

2018-01-25 18:54:24 -08:00

mlp_policy.py

fix trpo_mpi bug where logstd wasn’t included

2018-01-25 21:17:40 -08:00

pposgd_simple.py

Import internal changes (#422 )

2018-06-06 11:39:13 -07:00

README.md

Import internal changes (#422 )

2018-06-06 11:39:13 -07:00

run_atari.py

Lots of cleanups

2018-01-25 18:54:24 -08:00

run_humanoid.py

Import internal changes (#422 )

2018-06-06 11:39:13 -07:00

run_mujoco.py

Lots of cleanups

2018-01-25 18:54:24 -08:00

run_robotics.py

Import internal changes (#422 )

2018-06-06 11:39:13 -07:00

README.md

PPOSGD

Original paper: https://arxiv.org/abs/1707.06347
Baselines blog post: https://blog.openai.com/openai-baselines-ppo/
mpirun -np 8 python -m baselines.ppo1.run_atari runs the algorithm for 40M frames = 10M timesteps on an Atari game. See help (-h) for more options.
python -m baselines.ppo1.run_mujoco runs the algorithm for 1M frames on a Mujoco environment.
Train mujoco 3d humanoid (with optimal-ish hyperparameters): mpirun -np 16 python -m baselines.ppo1.run_humanoid --model-path=/path/to/model
Render the 3d humanoid: python -m baselines.ppo1.run_humanoid --play --model-path=/path/to/model