PyTorch implementation of Twin Delayed DDPG (TD3) tested on the following environments:
-
To test a preTrained network : run
test.py
-
To train a new network : run
train.py
Trained and tested on:
Python 3.6
PyTorch 0.4.1
NumPy 1.15.3
gym 0.10.8
Roboschool 1.0.46
Pillow 5.3.0
BipedalWalker-v2 (800 episodes) | LunarLanderContinuous-v2 (1500 episodes) |
---|---|
RoboschoolWalker2d-v1 (lr=0.002, 1400 episodes) | HalfCheetah-v1 (lr=0.002, 1400 episodes) |
---|---|
*The results are not consistent for BipedalWalker-v2 env