-
Notifications
You must be signed in to change notification settings - Fork 0
/
agent.py
140 lines (107 loc) · 4.75 KB
/
agent.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
from collections import deque
from os import stat
import numpy as np
import torch
import torch.nn.functional as F
import torch.optim as optim
from model import Actor, Critic
from noise import OUNoise, NormalNoise, ParameterNoise
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
class Agent():
"""Interacts with and learns from the environment."""
def __init__(
self,
state_size,
action_size,
critic_regular,
critic_target,
critic_optimizer,
noise_type="normal",
alpha_actor=1e-4,
gamma=.99,
tau=1e-3,
desired_distance=.3,
scalar=.5,
scalar_decay=.99,
normal_scalar=.4
):
self.hyperparameters = '\n'.join(f"{key:>17}: {value}" for key, value in locals().items() if key not in ['self', 'critic_regular', 'critic_target', 'critic_optimizer'])
self.state_size = state_size
self.action_size = action_size
self.gamma = gamma
self.tau = tau
self.noise_type = noise_type
# parameter noise
self.distances = []
self.desired_distance = desired_distance
self.scalar = scalar
self.scalar_decay = scalar_decay
# normal noise
self.normal_scalar = normal_scalar
# step counter
self.t = 0
# Actor Network (w/ Target Network)
self.actor_regular = Actor(state_size, action_size).to(device)
self.actor_target = Actor(state_size, action_size).to(device)
self.actor_optimizer = optim.Adam(self.actor_regular.parameters(), lr=alpha_actor)
# Critic Network (w/ Target Network)
self.critic_regular = critic_regular
self.critic_target = critic_target
self.critic_optimizer = critic_optimizer
self.actor_noised = Actor(state_size, action_size).to(device)
# hard update to ensure that regular and target start with same values
self.soft_update(self.critic_regular, self.critic_target, 1.)
self.soft_update(self.actor_regular, self.actor_target, 1.)
if noise_type == "ou":
self.noise = OUNoise(self.action_size)
elif noise_type == "param":
self.noise = ParameterNoise(self.actor_noised, desired_distance, scalar, scalar_decay)
else:
self.noise = NormalNoise(self.action_size)
def act(self, state, add_noise=True):
"""Returns actions for given state as per current policy."""
state = torch.from_numpy(state).float().unsqueeze(0).to(device)
self.actor_regular.eval()
self.actor_noised.eval()
with torch.no_grad():
action = self.actor_regular(state).cpu().data.numpy()
if add_noise:
action = self.noise(action, state, self.actor_regular)
self.actor_regular.train()
return np.clip(action, -1, 1)
# needed for ou noise only
def reset(self):
if hasattr(self, 'ou_noise'):
self.ou_noise.reset()
def learn(self, experiences):
states, actions, rewards, next_states, dones = experiences
# ---------------------------- update critic ---------------------------- #
# Get predicted next-state actions and Q values from target models
actions_next = self.actor_target(next_states)
Q_targets_next = self.critic_target(next_states, actions_next)
# Compute Q targets for current states (y_i)
Q_targets = rewards + (self.gamma * Q_targets_next * (1 - dones))
# Compute critic loss
Q_expected = self.critic_regular(states, actions)
critic_loss = F.mse_loss(Q_expected, Q_targets)
# Minimize the loss
self.critic_optimizer.zero_grad()
critic_loss.backward()
self.critic_optimizer.step()
# ---------------------------- update actor ---------------------------- #
# Compute actor loss
actions_pred = self.actor_regular(states)
actor_loss = -self.critic_regular(states, actions_pred).mean()
# Minimize the loss
self.actor_optimizer.zero_grad()
actor_loss.backward()
self.actor_optimizer.step()
# ----------------------- update target networks ----------------------- #
self.soft_update(self.critic_regular, self.critic_target, self.tau)
self.soft_update(self.actor_regular, self.actor_target, self.tau)
@staticmethod
def soft_update(regular_model, target_model, tau):
for target_param, regular_param in zip(target_model.parameters(), regular_model.parameters()):
target_param.data.copy_(tau*regular_param.data + (1.0-tau)*target_param.data)
def __str__(self):
return f"Hyperparameters: \n\n{self.hyperparameters}\n\n{self.actor_regular}"