wassname/pytorch-a2c-ppo-acktr
1
0
Fork 0
mirror of https://github.com/wassname/pytorch-a2c-ppo-acktr.git synced 2026-06-27 16:20:05 +08:00
Code Issues Packages Projects Releases Wiki Activity

Add pybullet support

Browse Source
This commit is contained in:
Ilya Kostrikov
2017-10-05 15:54:39 -04:00
parent 041e928c05
commit baf507f6bf
2 changed files with 13 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
README.md
+8
View File
@@ -26,8 +26,16 @@ It's extremely difficult to reproduce results for Reinforcement Learning methods
* Improve performance of KFAC, see kfac.py for more information * Improve performance of KFAC, see kfac.py for more information
* Run evaluation for all games and algorithms * Run evaluation for all games and algorithms
## Requirements
* [PyTorch](http://pytorch.org/)
* [Visdom](https://github.com/facebookresearch/visdom)
* [OpenAI baselines](https://github.com/openai/baselines)
## Usage ## Usage
Start a `Visdom` server with `python -m visdom.server`, it will serve `http://localhost:8097/` by default.
### Atari ### Atari
#### A2C #### A2C
envs.py
+5 -1
View File
@@ -9,7 +9,11 @@ from baselines.common.atari_wrappers import wrap_deepmind
def make_env(env_id, seed, rank, log_dir): def make_env(env_id, seed, rank, log_dir):
def _thunk(): def _thunk():
env = gym.make(env_id) if env_id.find('Bullet') > -1:
import pybullet_envs
env = pybullet_envs.make(env_id)
else:
env = gym.make(env_id)
env.seed(seed + rank) env.seed(seed + rank)
env = bench.Monitor(env, env = bench.Monitor(env,
os.path.join(log_dir, os.path.join(log_dir,