莫烦 pytorch GAN

我的视频学习笔记

视频地址：https://www.bilibili.com/video/av15997678?p=29
源地址：https://github.com/MorvanZhou/PyTorch-Tutorial/blob/master/tutorial-contents/406_GAN.py

讲解：https://morvanzhou.github.io/tutorials/machine-learning/torch/4-06-GAN/

import torch
import matplotlib.pyplot as plt
import numpy as np
import torch.nn as nn

# Hyper Parameters
BATCH_SIZE = 64
LR_G = 0.0001
LR_D = 0.0001
N_IDEAS = 5
ART_COMPONENTS = 15
PAINT_POINTS = np.vstack([np.linspace(-1, 1, ART_COMPONENTS) for _ in range(BATCH_SIZE)])  # -1，1的坐标上有15个点


def artist_works():
    a = np.random.uniform(1, 2, size=BATCH_SIZE)[:, np.newaxis]  # 加一个维度
    paintings = a * np.power(PAINT_POINTS, 2) + (a - 1)
    paintings = torch.from_numpy(paintings).float()
    return paintings


# 用随机灵感创造出一幅画
G = nn.Sequential(
    nn.Linear(N_IDEAS, 128),
    nn.ReLU(),
    nn.Linear(128, ART_COMPONENTS)
)

D = nn.Sequential(
    nn.Linear(ART_COMPONENTS, 128),
    nn.ReLU(),
    nn.Linear(128, 1),
    nn.Sigmoid(),
)

opt_D = torch.optim.Adam(D.parameters(), lr=LR_D)
opt_G = torch.optim.Adam(G.parameters(), lr=LR_G)

plt.ion()   # something about continuous plotting

for step in range(10000):
    artist_paintings = artist_works()
    G_ideas = torch.randn(BATCH_SIZE, N_IDEAS)  # 随机生成BATCH_SIZE组，每组N_IDEAS个
    G_paintings = G(G_ideas)

    prob_artist0 = D(artist_paintings)
    prob_artist1 = D(G_paintings)

    D_loss = -  torch.mean(torch.log(prob_artist0) + torch.log(1 - prob_artist1))  # 增加著名画家的概率
    # 减少认为是著名画家的新手画家画的概率
    G_loss = torch.mean(torch.log(1 - prob_artist1))

    opt_D.zero_grad()
    D_loss.backward(retain_graph=True)  # 保留参数给后面
    opt_D.step()

    opt_G.zero_grad()
    G_loss.backward()
    opt_G.step()

    if step % 50 == 0:  # plotting
        plt.cla()
        plt.plot(PAINT_POINTS[0], G_paintings.data.numpy()[0], c='#4AD631', lw=3, label='Generated painting', )
        plt.plot(PAINT_POINTS[0], 2 * np.power(PAINT_POINTS[0], 2) + 1, c='#74BCFF', lw=3, label='upper bound')
        plt.plot(PAINT_POINTS[0], 1 * np.power(PAINT_POINTS[0], 2) + 0, c='#FF9359', lw=3, label='lower bound')
        plt.text(-.5, 2.3, 'D accuracy=%.2f (0.5 for D to converge)' % prob_artist0.data.numpy().mean(),
                 fontdict={'size': 13})
        plt.text(-.5, 2, 'D score= %.2f (-1.38 for G to converge)' % -D_loss.data.numpy(), fontdict={'size': 13})

        plt.ylim((0, 3))
        plt.legend(loc='upper right', fontsize=10)
        plt.draw()
        plt.pause(0.01)

plt.ioff()
plt.show()