|
--- |
|
library_name: stable-baselines3 |
|
tags: |
|
- BipedalWalkerHardcore-v3 |
|
- deep-reinforcement-learning |
|
- reinforcement-learning |
|
- stable-baselines3 |
|
model-index: |
|
- name: PPO |
|
results: |
|
- metrics: |
|
- type: mean_reward |
|
value: 7.09 +/- 2.73 |
|
name: mean_reward |
|
task: |
|
type: reinforcement-learning |
|
name: reinforcement-learning |
|
dataset: |
|
name: BipedalWalkerHardcore-v3 |
|
type: BipedalWalkerHardcore-v3 |
|
--- |
|
|
|
# parameters <br> |
|
model = A2C(policy = "MlpPolicy", <br> |
|
env = env, <br> |
|
n_steps = 256, <br> |
|
learning_rate = 0.001, <br> |
|
gamma = 0.99, <br> |
|
verbose=1) <br> |