- 博客(6)
- 收藏
- 关注
原创 DDPG(6)_ddpg
1、引用Python库import gymimport tensorflow as tfimport numpy as npfrom ou_noise import OUNoisefrom critic_network import CriticNetwork from actor_network_bn import ActorNetworkfrom replay_buffer imp...
2018-06-19 11:46:56
1579
2
原创 DDPG(5)_filter_env
1、引用python库import numpy as npimport gym2、定义函数def makeFilteredEnv(env): """ crate a new environment class with actions and states normalized to [-1,1] """ acsp = env.action_space obsp = env.obse...
2018-06-19 11:02:31
376
原创 DDPG(4)_ounoise
1、应用python库import numpy as npimport numpy.random as nr2、定义类class OUNoise: """docstring for OUNoise""" def __init__(self,action_dimension,mu=0, theta=0.15, sigma=0.2): self.action_dime...
2018-06-18 22:45:42
3097
原创 DDPG(3)_replay_buffer
1、引用python库from collections import dequeimport random2、定义类class ReplayBuffer(object): def __init__(self, buffer_size): self.buffer_size = buffer_size(以下全为Class ReplayBuffer中的函数)3、初始化函数 ...
2018-06-18 22:25:19
3120
原创 DDPG(2)-critic_network
1、引用python库import tensorflow as tf import numpy as npimport math2、声明参数LAYER1_SIZE = 400LAYER2_SIZE = 300LEARNING_RATE = 1e-3TAU = 0.001L2 = 0.013、定义类class CriticNetwork: """docstring for Criti...
2018-06-18 22:07:19
2036
原创 DDPG(1)-actor_network
1、引用python库import tensorflow as tf from tensorflow.contrib.layers.python.layers import batch_norm as batch_normimport numpy as npimport math2、声明参数# Hyper ParametersLAYER1_SIZE = 400LAYER2_SIZE = ...
2018-06-18 20:55:12
2240
空空如也
空空如也
TA创建的收藏夹 TA关注的收藏夹
TA关注的人