21 lines
585 B
Python
21 lines
585 B
Python
"""
|
|
Fast TD3 is a high-performance implementation of Twin Delayed Deep Deterministic Policy Gradient (TD3)
|
|
with distributional critics for reinforcement learning.
|
|
"""
|
|
|
|
# Core model components
|
|
from fast_td3.fast_td3 import Actor, Critic, DistributionalQNetwork
|
|
from fast_td3.fast_td3_utils import EmpiricalNormalization, SimpleReplayBuffer
|
|
from fast_td3.fast_td3_deploy import Policy, load_policy
|
|
|
|
__all__ = [
|
|
# Core model components
|
|
"Actor",
|
|
"Critic",
|
|
"DistributionalQNetwork",
|
|
"EmpiricalNormalization",
|
|
"SimpleReplayBuffer",
|
|
"Policy",
|
|
"load_policy",
|
|
]
|