DQNCartPole.py

import gym
import collections
import random

import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim

# Hyper parameters
learning_rate = 0.0005
gamma = 0.98
buffer_limit = 50000
batch_size = 32


class ReplayBuffer():
    def __init__(self):
        self.buffer = collections.deque(maxlen=buffer_limit)

    def put(self, transition):
        self.buffer.append(transition)

    def sample(self, n):
        mini_batch = random.sample(self.buffer, n)
        s_lst, a_lst, r_lst, s_prime_lst, done_mask_lst = [], [], [], [], []

        for transition in mini_batch:
            s, a, r, s_prime, done_mask = transition
            s_lst.append(s)
            a_lst.append([a])
            r_lst.append([r])
            s_prime_lst.append(s_prime)
            done_mask_lst.append([done_mask])  # done_mask: 0 if done, 1 o/w

        return torch.tensor(s_lst, dtype=torch.float), torch.tensor(a_lst), torch.tensor(r_lst),\
            torch.tensor(s_prime_lst, dtype=torch.float), torch.tensor(done_mask_lst)

    def size(self):
        return len(self.buffer)


class Qnet(nn.Module):
    def __init__(self):
        super(Qnet, self).__init__()
        self.fc1 = nn.Linear(4, 128)  # state = (cart_pos, cart_vel, pole_degree, pole_acc)
        self.fc2 = nn.Linear(128, 128)
        self.fc3 = nn.Linear(128, 2)  # action 2 (left or right)

    def forward(self, x):
        x = F.relu(self.fc1(x))
        x = F.relu(self.fc2(x))
        x = self.fc3(x)
        return x

    def sample_action(self, obs, eps):
        out = self.forward(obs)  # Q-value
        coin = random.random()
        if coin < eps:
            action = random.randint(0, 1)
        else:
            action = out.argmax().item()
        return action


def train(q, q_target, memory, optimizer):
    for i in range(10):
        s, a, r, s_prime, done_mask = memory.sample(batch_size)

        q_out = q(s)  # forward?
        q_a = q_out.gather(1, a)  # gather(dim, index) # Q(s,a)
        max_q_prime = q_target(s_prime).max(1)[0].unsqueeze(1)
        target = r + gamma * max_q_prime * done_mask
        loss = F.smooth_l1_loss(q_a, target)

        optimizer.zero_grad()
        loss.backward()  # calculate loss gradient
        optimizer.step()


def main():
    env = gym.make('CartPole-v1')
    q = Qnet()
    q_target = Qnet()
    q_target.load_state_dict(q.state_dict())
    memory = ReplayBuffer()

    print_interval = 20
    score = 0.0
    optimizer = optim.Adam(q.parameters(), lr=learning_rate)

    for n_epi in range(10000):
        epsilon = max(0.01, 0.08 - 0.01*(n_epi/200))  # Linear annealing from 0.08 to 0.01 (every 200 epi)
        s = env.reset()
        done = False

        while not done:
            a = q.sample_action(s, epsilon)
            s_prime, r, done, info = env.step(a)
            done_mask = 0.0 if done else 1.0
            memory.put((s, a, r/100.0, s_prime, done_mask))  # reward scaling
            s = s_prime
            score += r
            if done:
                break

        if memory.size() > 2000:
            train(q, q_target, memory, optimizer)

        if n_epi % print_interval == 0 and n_epi != 0:
            q_target.load_state_dict(q.state_dict())
            print("n_episode: {}, score: {:.1f}, n_buffer: {}, eps: {:.1f}%".format(n_epi, score/print_interval,
                                                                                    memory.size(), epsilon * 100))
            score = 0.0
    env.close()

if __name__ == '__main__':
    main()