洌泉_就这样吧

【PyTorch】生成对抗网络/GAN(generative adversarial network)

1 模型介绍

GAN(generative adversarial network)自2014年被提出以来就引起广泛关注，下面是来自百度百科的词条内容：

生成式对抗网络（GAN, Generative Adversarial Networks ）是一种深度学习模型，是近年来复杂分布上无监督学习最具前景的方法之一。模型通过框架中（至少）两个模块：生成模型（Generative Model）和判别模型（Discriminative Model）的互相博弈学习产生相当好的输出。原始 GAN 理论中，并不要求 G 和 D 都是神经网络，只需要是能拟合相应生成和判别的函数即可。但实用中一般均使用深度神经网络作为 G 和 D 。一个优秀的GAN应用需要有良好的训练方法，否则可能由于神经网络模型的自由性而导致输出不理想。

Ian J. Goodfellow等人发表的论文链接：https://arxiv.org/abs/1406.2661

2 具体代码

# generative adversarial network(GAN)
import os 
import torch
import torchvision 
import torch.nn as nn 
from torchvision import transforms
from torchvision.utils import save_image

# Device configuration 
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')

# Hyper-parameters
latent_size = 64
hidden_size = 256
image_size = 784
num_epochs = 200
batch_size = 100
sample_dir = 'samples'

# Create a directory if not exists
if not os.path.exists(sample_dir):
    os.makedirs(sample_dir)

# Image processing, 3 for RGB channels, 1 for greyscale channels
# transform = transforms.Compose([
#   transforms.ToTensor(),
#   transforms.Normalize(mean=(0.5, 0.5, 0.5),   
#   std=(0.5, 0.5, 0.5))])

transform = transforms.Compose([
    transforms.ToTensor(),
    transforms.Normalize(mean=[0.5], std=[0.5])])

# MNIST dataset
mnist = torchvision.datasets.MNIST(
    root='data/',
    train=True,
    transform=transform,
    download=True)
# Data loader
data_loader = torch.utils.data.DataLoader(
    dataset=mnist,
    batch_size=batch_size, 
    shuffle=True)

# Discriminator
D = nn.Sequential(
    nn.Linear(image_size, hidden_size),
    nn.LeakyReLU(0.2),
    nn.Linear(hidden_size, hidden_size),
    nn.LeakyReLU(0.2),
    nn.Linear(hidden_size, 1),
    nn.Sigmoid())

# Generator 
G = nn.Sequential(
    nn.Linear(latent_size, hidden_size),
    nn.ReLU(),
    nn.Linear(hidden_size, hidden_size),
    nn.ReLU(),
    nn.Linear(hidden_size, image_size),
    nn.Tanh())

# Device setting
D = D.to(device)
G = G.to(device)

# Binary cross entropy loss and optimizer
criterion = nn.BCELoss()
d_optimizer = torch.optim.Adam(D.parameters(), lr=0.0002)
g_optimizer = torch.optim.Adam(G.parameters(), lr=0.0002)

def denorm(x):
    out = (x + 1) / 2
    return out.clamp(0, 1)

def reset_grad():
    d_optimizer.zero_grad()
    g_optimizer.zero_grad()

# Start training
total_step = len(data_loader)
for epoch in range(num_epochs):
    for i, (images, _) in enumerate(data_loader):
        images = images.reshape(batch_size, -1).to(device)        

        # Create the labels which are later used as input for the BCE loss
        real_labels = torch.ones(batch_size, 1).to(device)
        fake_labels = torch.zeros(batch_size, 1).to(device)

        # Train the discriminator           
        # Compute BCE_Loss using real images where BCE_Loss(x, y): - y * log(D(x)) - (1-y) * log(1 - D(x))
        # Second term of the loss is always zero since real_labels == 1
        outputs = D(images)
        d_loss_real = criterion(outputs, real_labels)
        real_score = outputs

        # Compute BCELoss using fake images
        # First term of the loss is always zero since fake_labels == 0
        z = torch.randn(batch_size, latent_size).to(device)
        fake_images = G(z)
        outputs = D(fake_images)
        d_loss_fake = criterion(outputs, fake_labels)
        fake_score = outputs

        # Backprop and optimize
        d_loss = d_loss_real + d_loss_fake
        reset_grad()
        d_loss.backward()
        d_optimizer.step()

        #  Train the generator 
        # Compute loss with fake images
        z = torch.randn(batch_size, latent_size).to(device)
        fake_images = G(z)
        outputs = D(fake_images)

        # We train G to maximize log(D(G(z)) instead of minimizing log(1-D(G(z)))
        # For the reason, see the last paragraph of section 3. https://arxiv.org/pdf/1406.2661.pdf
        g_loss = criterion(outputs, real_labels)
        
        # Backprop and optimize
        reset_grad()
        g_loss.backward()
        g_optimizer.step()

        if (i+1) % 200 == 0:
            print('Epoch [{}/{}], Step [{}/{}], d_loss: {:.4f}, g_loss: {:.4f}, D(x): {:.2f}, D(G(z)): {:.2f}' 
                  .format(
                      epoch, 
                      num_epochs, 
                      i+1, 
                      total_step, 
                      d_loss.item(), 
                      g_loss.item(), 
                      real_score.mean().item(), fake_score.mean().item()))
        
    # Save real images
    if (epoch+1) == 1:
        images = images.reshape(images.size(0), 1, 28, 28)
        save_image(denorm(images), os.path.join(sample_dir, 'real_images.png'))
    
    # Save sampled images
    fake_images = fake_images.reshape(fake_images.size(0), 1, 28, 28)
    save_image(denorm(fake_images), os.path.join(sample_dir, 'fake_images-{}.png'.format(epoch+1)))

# Save the model checkpoints 
torch.save(G.state_dict(), 'G.ckpt')
torch.save(D.state_dict(), 'D.ckpt')

3 程序输出

训练过程就是不断增强生成模型G的能力，达到“以假乱真”的效果。从训练输出可以看出d_loss增大、D(x)降低，说明判别模型D越来越分辨不出real image和fake image；g_loss下降、D(G(z))增大，说明生成模型G生成的fake image越来越接近real image。
模型每步都保存了生成的fake image，这里展示其中几个epoch（0，50，100，150，199），以及展示real image。
– epoch = 0

– epoch = 50

– epoch = 100

– epoch = 150

– epoch = 199

– real image

Epoch [0/200], Step [200/600], d_loss: 0.0470, g_loss: 4.4273, D(x): 0.99, D(G(z)): 0.04
Epoch [0/200], Step [400/600], d_loss: 0.1453, g_loss: 6.1361, D(x): 0.96, D(G(z)): 0.07
Epoch [0/200], Step [600/600], d_loss: 0.0254, g_loss: 5.2651, D(x): 0.99, D(G(z)): 0.02
Epoch [1/200], Step [200/600], d_loss: 0.0161, g_loss: 6.4054, D(x): 0.99, D(G(z)): 0.01
Epoch [1/200], Step [400/600], d_loss: 0.2298, g_loss: 4.2960, D(x): 0.94, D(G(z)): 0.08
Epoch [1/200], Step [600/600], d_loss: 0.4214, g_loss: 4.9261, D(x): 0.93, D(G(z)): 0.23
Epoch [2/200], Step [200/600], d_loss: 0.0552, g_loss: 4.5034, D(x): 0.98, D(G(z)): 0.03
Epoch [2/200], Step [400/600], d_loss: 0.0684, g_loss: 5.3159, D(x): 0.97, D(G(z)): 0.04
Epoch [2/200], Step [600/600], d_loss: 0.3504, g_loss: 2.7364, D(x): 0.85, D(G(z)): 0.09
Epoch [3/200], Step [200/600], d_loss: 0.8830, g_loss: 3.0232, D(x): 0.69, D(G(z)): 0.14
Epoch [3/200], Step [400/600], d_loss: 0.2208, g_loss: 4.2910, D(x): 0.93, D(G(z)): 0.06
Epoch [3/200], Step [600/600], d_loss: 0.4205, g_loss: 2.9028, D(x): 0.85, D(G(z)): 0.11
Epoch [4/200], Step [200/600], d_loss: 0.2909, g_loss: 3.5080, D(x): 0.96, D(G(z)): 0.20
Epoch [4/200], Step [400/600], d_loss: 0.0816, g_loss: 5.0248, D(x): 0.97, D(G(z)): 0.03
Epoch [4/200], Step [600/600], d_loss: 0.2410, g_loss: 4.5674, D(x): 0.92, D(G(z)): 0.07
Epoch [5/200], Step [200/600], d_loss: 0.4174, g_loss: 3.5316, D(x): 0.89, D(G(z)): 0.12
Epoch [5/200], Step [400/600], d_loss: 0.3851, g_loss: 4.3338, D(x): 0.87, D(G(z)): 0.07
Epoch [5/200], Step [600/600], d_loss: 0.4061, g_loss: 4.1645, D(x): 0.89, D(G(z)): 0.09
Epoch [6/200], Step [200/600], d_loss: 0.1964, g_loss: 5.1655, D(x): 0.91, D(G(z)): 0.03
Epoch [6/200], Step [400/600], d_loss: 0.2616, g_loss: 3.4105, D(x): 0.92, D(G(z)): 0.08
Epoch [6/200], Step [600/600], d_loss: 0.2332, g_loss: 3.8159, D(x): 0.94, D(G(z)): 0.10
Epoch [7/200], Step [200/600], d_loss: 0.1943, g_loss: 3.9958, D(x): 0.92, D(G(z)): 0.06
Epoch [7/200], Step [400/600], d_loss: 0.0882, g_loss: 5.2199, D(x): 0.96, D(G(z)): 0.04
Epoch [7/200], Step [600/600], d_loss: 0.0824, g_loss: 4.9151, D(x): 0.96, D(G(z)): 0.02
Epoch [8/200], Step [200/600], d_loss: 0.3849, g_loss: 4.6934, D(x): 0.86, D(G(z)): 0.02
Epoch [8/200], Step [400/600], d_loss: 0.2212, g_loss: 3.4482, D(x): 0.96, D(G(z)): 0.11
Epoch [8/200], Step [600/600], d_loss: 0.1743, g_loss: 5.7623, D(x): 0.96, D(G(z)): 0.07
Epoch [9/200], Step [200/600], d_loss: 0.2993, g_loss: 4.2877, D(x): 0.95, D(G(z)): 0.10
Epoch [9/200], Step [400/600], d_loss: 0.1218, g_loss: 5.1353, D(x): 0.95, D(G(z)): 0.02
Epoch [9/200], Step [600/600], d_loss: 0.4142, g_loss: 3.8948, D(x): 0.88, D(G(z)): 0.07
Epoch [10/200], Step [200/600], d_loss: 0.1586, g_loss: 4.3726, D(x): 0.98, D(G(z)): 0.09
Epoch [10/200], Step [400/600], d_loss: 0.1054, g_loss: 5.0776, D(x): 0.96, D(G(z)): 0.02
Epoch [10/200], Step [600/600], d_loss: 0.2254, g_loss: 5.7384, D(x): 0.91, D(G(z)): 0.01
Epoch [11/200], Step [200/600], d_loss: 0.0880, g_loss: 4.9024, D(x): 0.96, D(G(z)): 0.04
Epoch [11/200], Step [400/600], d_loss: 0.2936, g_loss: 2.4484, D(x): 0.93, D(G(z)): 0.08
Epoch [11/200], Step [600/600], d_loss: 0.1335, g_loss: 4.1263, D(x): 0.96, D(G(z)): 0.05
Epoch [12/200], Step [200/600], d_loss: 0.1299, g_loss: 5.0566, D(x): 0.95, D(G(z)): 0.02
Epoch [12/200], Step [400/600], d_loss: 0.0839, g_loss: 4.9301, D(x): 0.96, D(G(z)): 0.02
Epoch [12/200], Step [600/600], d_loss: 0.1568, g_loss: 4.9410, D(x): 0.96, D(G(z)): 0.06
Epoch [13/200], Step [200/600], d_loss: 0.0838, g_loss: 4.4768, D(x): 0.98, D(G(z)): 0.02
Epoch [13/200], Step [400/600], d_loss: 0.2606, g_loss: 5.2801, D(x): 0.97, D(G(z)): 0.15
Epoch [13/200], Step [600/600], d_loss: 0.1078, g_loss: 6.9736, D(x): 0.95, D(G(z)): 0.02
Epoch [14/200], Step [200/600], d_loss: 0.2917, g_loss: 7.4783, D(x): 0.89, D(G(z)): 0.00
Epoch [14/200], Step [400/600], d_loss: 0.1246, g_loss: 5.6440, D(x): 0.99, D(G(z)): 0.06
Epoch [14/200], Step [600/600], d_loss: 0.1669, g_loss: 4.6825, D(x): 0.97, D(G(z)): 0.08
Epoch [15/200], Step [200/600], d_loss: 0.0684, g_loss: 5.4430, D(x): 0.96, D(G(z)): 0.01
Epoch [15/200], Step [400/600], d_loss: 0.1261, g_loss: 4.1301, D(x): 0.97, D(G(z)): 0.05
Epoch [15/200], Step [600/600], d_loss: 0.3928, g_loss: 5.0515, D(x): 0.89, D(G(z)): 0.04
Epoch [16/200], Step [200/600], d_loss: 0.0822, g_loss: 5.1707, D(x): 0.97, D(G(z)): 0.03
Epoch [16/200], Step [400/600], d_loss: 0.1735, g_loss: 5.7974, D(x): 0.92, D(G(z)): 0.02
Epoch [16/200], Step [600/600], d_loss: 0.1906, g_loss: 4.7995, D(x): 0.98, D(G(z)): 0.12
Epoch [17/200], Step [200/600], d_loss: 0.3929, g_loss: 2.9602, D(x): 0.90, D(G(z)): 0.12
Epoch [17/200], Step [400/600], d_loss: 0.3379, g_loss: 4.5416, D(x): 0.92, D(G(z)): 0.13
Epoch [17/200], Step [600/600], d_loss: 0.3320, g_loss: 4.0064, D(x): 0.90, D(G(z)): 0.05
Epoch [18/200], Step [200/600], d_loss: 0.1296, g_loss: 4.5214, D(x): 0.99, D(G(z)): 0.09
Epoch [18/200], Step [400/600], d_loss: 0.1229, g_loss: 4.5100, D(x): 0.93, D(G(z)): 0.02
Epoch [18/200], Step [600/600], d_loss: 0.1938, g_loss: 4.3949, D(x): 0.92, D(G(z)): 0.04
Epoch [19/200], Step [200/600], d_loss: 0.5453, g_loss: 4.8068, D(x): 0.98, D(G(z)): 0.26
Epoch [19/200], Step [400/600], d_loss: 0.2208, g_loss: 4.7370, D(x): 0.95, D(G(z)): 0.07
Epoch [19/200], Step [600/600], d_loss: 0.0719, g_loss: 5.3687, D(x): 0.98, D(G(z)): 0.05
Epoch [20/200], Step [200/600], d_loss: 0.2448, g_loss: 5.9713, D(x): 0.92, D(G(z)): 0.04
Epoch [20/200], Step [400/600], d_loss: 0.1789, g_loss: 6.3533, D(x): 0.96, D(G(z)): 0.08
Epoch [20/200], Step [600/600], d_loss: 0.1881, g_loss: 5.5990, D(x): 0.95, D(G(z)): 0.07
Epoch [21/200], Step [200/600], d_loss: 0.1302, g_loss: 6.8167, D(x): 0.96, D(G(z)): 0.06
Epoch [21/200], Step [400/600], d_loss: 0.1292, g_loss: 6.2911, D(x): 0.96, D(G(z)): 0.05
Epoch [21/200], Step [600/600], d_loss: 0.3090, g_loss: 4.1408, D(x): 0.90, D(G(z)): 0.09
Epoch [22/200], Step [200/600], d_loss: 0.2777, g_loss: 5.8237, D(x): 0.95, D(G(z)): 0.13
Epoch [22/200], Step [400/600], d_loss: 0.1448, g_loss: 4.4736, D(x): 0.95, D(G(z)): 0.04
Epoch [22/200], Step [600/600], d_loss: 0.5411, g_loss: 5.5632, D(x): 0.82, D(G(z)): 0.03
Epoch [23/200], Step [200/600], d_loss: 0.2966, g_loss: 5.2273, D(x): 0.92, D(G(z)): 0.10
Epoch [23/200], Step [400/600], d_loss: 0.2568, g_loss: 3.5231, D(x): 0.90, D(G(z)): 0.05
Epoch [23/200], Step [600/600], d_loss: 0.3062, g_loss: 4.3809, D(x): 0.91, D(G(z)): 0.12
Epoch [24/200], Step [200/600], d_loss: 0.2343, g_loss: 4.5202, D(x): 0.89, D(G(z)): 0.03
Epoch [24/200], Step [400/600], d_loss: 0.1924, g_loss: 3.9855, D(x): 0.95, D(G(z)): 0.09
Epoch [24/200], Step [600/600], d_loss: 0.4676, g_loss: 5.2467, D(x): 0.83, D(G(z)): 0.03
Epoch [25/200], Step [200/600], d_loss: 0.2428, g_loss: 3.3750, D(x): 0.93, D(G(z)): 0.09
Epoch [25/200], Step [400/600], d_loss: 0.4260, g_loss: 4.1176, D(x): 0.86, D(G(z)): 0.11
Epoch [25/200], Step [600/600], d_loss: 0.3781, g_loss: 3.8905, D(x): 0.87, D(G(z)): 0.10
Epoch [26/200], Step [200/600], d_loss: 0.3581, g_loss: 4.3727, D(x): 0.90, D(G(z)): 0.08
Epoch [26/200], Step [400/600], d_loss: 0.3745, g_loss: 4.4046, D(x): 0.90, D(G(z)): 0.10
Epoch [26/200], Step [600/600], d_loss: 0.2697, g_loss: 3.4132, D(x): 0.93, D(G(z)): 0.11
Epoch [27/200], Step [200/600], d_loss: 0.3638, g_loss: 4.0925, D(x): 0.90, D(G(z)): 0.10
Epoch [27/200], Step [400/600], d_loss: 0.2662, g_loss: 3.7483, D(x): 0.92, D(G(z)): 0.09
Epoch [27/200], Step [600/600], d_loss: 0.3261, g_loss: 2.8968, D(x): 0.97, D(G(z)): 0.19
Epoch [28/200], Step [200/600], d_loss: 0.4217, g_loss: 2.8956, D(x): 0.90, D(G(z)): 0.12
Epoch [28/200], Step [400/600], d_loss: 0.6127, g_loss: 3.2918, D(x): 0.94, D(G(z)): 0.30
Epoch [28/200], Step [600/600], d_loss: 0.2736, g_loss: 3.9999, D(x): 0.94, D(G(z)): 0.13
Epoch [29/200], Step [200/600], d_loss: 0.2467, g_loss: 4.0165, D(x): 0.96, D(G(z)): 0.12
Epoch [29/200], Step [400/600], d_loss: 0.4403, g_loss: 3.2785, D(x): 0.91, D(G(z)): 0.16
Epoch [29/200], Step [600/600], d_loss: 0.3136, g_loss: 3.7659, D(x): 0.91, D(G(z)): 0.11
Epoch [30/200], Step [200/600], d_loss: 0.3471, g_loss: 2.9367, D(x): 0.90, D(G(z)): 0.13
Epoch [30/200], Step [400/600], d_loss: 0.3247, g_loss: 3.4481, D(x): 0.92, D(G(z)): 0.14
Epoch [30/200], Step [600/600], d_loss: 0.5462, g_loss: 2.6313, D(x): 0.87, D(G(z)): 0.20
Epoch [31/200], Step [200/600], d_loss: 0.5558, g_loss: 3.1563, D(x): 0.83, D(G(z)): 0.11
Epoch [31/200], Step [400/600], d_loss: 0.4707, g_loss: 2.9219, D(x): 0.85, D(G(z)): 0.12
Epoch [31/200], Step [600/600], d_loss: 0.5093, g_loss: 4.5862, D(x): 0.80, D(G(z)): 0.04
Epoch [32/200], Step [200/600], d_loss: 0.4759, g_loss: 3.3408, D(x): 0.87, D(G(z)): 0.15
Epoch [32/200], Step [400/600], d_loss: 0.3769, g_loss: 2.8612, D(x): 0.85, D(G(z)): 0.08
Epoch [32/200], Step [600/600], d_loss: 0.5780, g_loss: 1.8472, D(x): 0.83, D(G(z)): 0.18
Epoch [33/200], Step [200/600], d_loss: 0.5509, g_loss: 2.8013, D(x): 0.87, D(G(z)): 0.20
Epoch [33/200], Step [400/600], d_loss: 0.3934, g_loss: 3.4992, D(x): 0.84, D(G(z)): 0.06
Epoch [33/200], Step [600/600], d_loss: 0.3825, g_loss: 2.5781, D(x): 0.91, D(G(z)): 0.18
Epoch [34/200], Step [200/600], d_loss: 0.6687, g_loss: 3.3864, D(x): 0.78, D(G(z)): 0.15
Epoch [34/200], Step [400/600], d_loss: 0.5931, g_loss: 1.7835, D(x): 0.86, D(G(z)): 0.23
Epoch [34/200], Step [600/600], d_loss: 0.4132, g_loss: 2.6908, D(x): 0.86, D(G(z)): 0.13
Epoch [35/200], Step [200/600], d_loss: 0.4721, g_loss: 3.1934, D(x): 0.84, D(G(z)): 0.10
Epoch [35/200], Step [400/600], d_loss: 0.6144, g_loss: 3.2895, D(x): 0.88, D(G(z)): 0.24
Epoch [35/200], Step [600/600], d_loss: 0.2995, g_loss: 3.3611, D(x): 0.91, D(G(z)): 0.13
Epoch [36/200], Step [200/600], d_loss: 0.6170, g_loss: 2.6114, D(x): 0.80, D(G(z)): 0.15
Epoch [36/200], Step [400/600], d_loss: 0.6916, g_loss: 2.7102, D(x): 0.79, D(G(z)): 0.18
Epoch [36/200], Step [600/600], d_loss: 0.6275, g_loss: 3.8490, D(x): 0.78, D(G(z)): 0.11
Epoch [37/200], Step [200/600], d_loss: 0.3609, g_loss: 3.4735, D(x): 0.91, D(G(z)): 0.15
Epoch [37/200], Step [400/600], d_loss: 0.5123, g_loss: 3.4943, D(x): 0.78, D(G(z)): 0.08
Epoch [37/200], Step [600/600], d_loss: 0.7515, g_loss: 2.9656, D(x): 0.80, D(G(z)): 0.19
Epoch [38/200], Step [200/600], d_loss: 0.4069, g_loss: 3.1408, D(x): 0.93, D(G(z)): 0.21
Epoch [38/200], Step [400/600], d_loss: 0.6983, g_loss: 3.0050, D(x): 0.80, D(G(z)): 0.19
Epoch [38/200], Step [600/600], d_loss: 0.6511, g_loss: 3.3372, D(x): 0.82, D(G(z)): 0.16
Epoch [39/200], Step [200/600], d_loss: 0.5633, g_loss: 3.1432, D(x): 0.86, D(G(z)): 0.21
Epoch [39/200], Step [400/600], d_loss: 0.5349, g_loss: 4.1383, D(x): 0.84, D(G(z)): 0.17
Epoch [39/200], Step [600/600], d_loss: 0.5642, g_loss: 2.8495, D(x): 0.79, D(G(z)): 0.14
Epoch [40/200], Step [200/600], d_loss: 0.4702, g_loss: 4.4767, D(x): 0.88, D(G(z)): 0.17
Epoch [40/200], Step [400/600], d_loss: 0.5330, g_loss: 2.8938, D(x): 0.85, D(G(z)): 0.18
Epoch [40/200], Step [600/600], d_loss: 0.4769, g_loss: 3.6471, D(x): 0.84, D(G(z)): 0.14
Epoch [41/200], Step [200/600], d_loss: 0.5435, g_loss: 2.7885, D(x): 0.85, D(G(z)): 0.16
Epoch [41/200], Step [400/600], d_loss: 0.6488, g_loss: 3.1947, D(x): 0.81, D(G(z)): 0.21
Epoch [41/200], Step [600/600], d_loss: 0.4791, g_loss: 3.1396, D(x): 0.87, D(G(z)): 0.18
Epoch [42/200], Step [200/600], d_loss: 0.5799, g_loss: 2.7526, D(x): 0.83, D(G(z)): 0.23
Epoch [42/200], Step [400/600], d_loss: 0.5212, g_loss: 2.5783, D(x): 0.81, D(G(z)): 0.14
Epoch [42/200], Step [600/600], d_loss: 0.8305, g_loss: 2.5786, D(x): 0.79, D(G(z)): 0.26
Epoch [43/200], Step [200/600], d_loss: 0.4356, g_loss: 2.7466, D(x): 0.90, D(G(z)): 0.20
Epoch [43/200], Step [400/600], d_loss: 0.3823, g_loss: 4.8288, D(x): 0.84, D(G(z)): 0.09
Epoch [43/200], Step [600/600], d_loss: 0.5135, g_loss: 2.4546, D(x): 0.87, D(G(z)): 0.20
Epoch [44/200], Step [200/600], d_loss: 0.5855, g_loss: 2.9187, D(x): 0.82, D(G(z)): 0.21
Epoch [44/200], Step [400/600], d_loss: 0.4671, g_loss: 2.7824, D(x): 0.84, D(G(z)): 0.17
Epoch [44/200], Step [600/600], d_loss: 0.5226, g_loss: 2.8698, D(x): 0.94, D(G(z)): 0.28
Epoch [45/200], Step [200/600], d_loss: 0.4413, g_loss: 3.1040, D(x): 0.84, D(G(z)): 0.13
Epoch [45/200], Step [400/600], d_loss: 0.4432, g_loss: 2.1684, D(x): 0.84, D(G(z)): 0.13
Epoch [45/200], Step [600/600], d_loss: 0.7264, g_loss: 1.9863, D(x): 0.77, D(G(z)): 0.17
Epoch [46/200], Step [200/600], d_loss: 0.5404, g_loss: 2.6853, D(x): 0.80, D(G(z)): 0.12
Epoch [46/200], Step [400/600], d_loss: 0.6838, g_loss: 2.1826, D(x): 0.84, D(G(z)): 0.25
Epoch [46/200], Step [600/600], d_loss: 0.5273, g_loss: 3.0864, D(x): 0.84, D(G(z)): 0.16
Epoch [47/200], Step [200/600], d_loss: 0.4595, g_loss: 2.7024, D(x): 0.83, D(G(z)): 0.14
Epoch [47/200], Step [400/600], d_loss: 0.5056, g_loss: 2.7431, D(x): 0.83, D(G(z)): 0.19
Epoch [47/200], Step [600/600], d_loss: 0.7370, g_loss: 2.7867, D(x): 0.77, D(G(z)): 0.17
Epoch [48/200], Step [200/600], d_loss: 0.4393, g_loss: 2.7028, D(x): 0.88, D(G(z)): 0.18
Epoch [48/200], Step [400/600], d_loss: 0.5673, g_loss: 2.7980, D(x): 0.92, D(G(z)): 0.29
Epoch [48/200], Step [600/600], d_loss: 0.5334, g_loss: 2.1539, D(x): 0.85, D(G(z)): 0.22
Epoch [49/200], Step [200/600], d_loss: 0.5661, g_loss: 3.0253, D(x): 0.78, D(G(z)): 0.10
Epoch [49/200], Step [400/600], d_loss: 0.4719, g_loss: 2.5743, D(x): 0.86, D(G(z)): 0.17
Epoch [49/200], Step [600/600], d_loss: 0.6065, g_loss: 2.5545, D(x): 0.82, D(G(z)): 0.22
Epoch [50/200], Step [200/600], d_loss: 0.4225, g_loss: 2.4486, D(x): 0.84, D(G(z)): 0.15
Epoch [50/200], Step [400/600], d_loss: 0.6071, g_loss: 2.4751, D(x): 0.76, D(G(z)): 0.09
Epoch [50/200], Step [600/600], d_loss: 0.6256, g_loss: 2.4212, D(x): 0.80, D(G(z)): 0.20
Epoch [51/200], Step [200/600], d_loss: 0.5850, g_loss: 3.3030, D(x): 0.77, D(G(z)): 0.13
Epoch [51/200], Step [400/600], d_loss: 0.4128, g_loss: 2.2547, D(x): 0.85, D(G(z)): 0.14
Epoch [51/200], Step [600/600], d_loss: 0.5473, g_loss: 2.4636, D(x): 0.85, D(G(z)): 0.23
Epoch [52/200], Step [200/600], d_loss: 0.6625, g_loss: 1.8152, D(x): 0.80, D(G(z)): 0.24
Epoch [52/200], Step [400/600], d_loss: 0.7375, g_loss: 3.1472, D(x): 0.88, D(G(z)): 0.34
Epoch [52/200], Step [600/600], d_loss: 0.6016, g_loss: 2.0477, D(x): 0.76, D(G(z)): 0.15
Epoch [53/200], Step [200/600], d_loss: 0.8320, g_loss: 2.4250, D(x): 0.71, D(G(z)): 0.20
Epoch [53/200], Step [400/600], d_loss: 0.6171, g_loss: 2.1886, D(x): 0.77, D(G(z)): 0.19
Epoch [53/200], Step [600/600], d_loss: 0.7682, g_loss: 1.6478, D(x): 0.74, D(G(z)): 0.21
Epoch [54/200], Step [200/600], d_loss: 0.5863, g_loss: 2.3504, D(x): 0.85, D(G(z)): 0.24
Epoch [54/200], Step [400/600], d_loss: 0.8063, g_loss: 2.3055, D(x): 0.72, D(G(z)): 0.21
Epoch [54/200], Step [600/600], d_loss: 0.7454, g_loss: 2.2996, D(x): 0.74, D(G(z)): 0.16
Epoch [55/200], Step [200/600], d_loss: 0.5397, g_loss: 2.3880, D(x): 0.83, D(G(z)): 0.19
Epoch [55/200], Step [400/600], d_loss: 0.7615, g_loss: 2.0757, D(x): 0.80, D(G(z)): 0.28
Epoch [55/200], Step [600/600], d_loss: 0.4954, g_loss: 2.7377, D(x): 0.85, D(G(z)): 0.21
Epoch [56/200], Step [200/600], d_loss: 0.5220, g_loss: 2.6000, D(x): 0.87, D(G(z)): 0.24
Epoch [56/200], Step [400/600], d_loss: 0.7056, g_loss: 1.6327, D(x): 0.81, D(G(z)): 0.28
Epoch [56/200], Step [600/600], d_loss: 0.7858, g_loss: 2.8862, D(x): 0.82, D(G(z)): 0.28
Epoch [57/200], Step [200/600], d_loss: 0.4511, g_loss: 2.6597, D(x): 0.88, D(G(z)): 0.20
Epoch [57/200], Step [400/600], d_loss: 0.5671, g_loss: 2.5462, D(x): 0.79, D(G(z)): 0.17
Epoch [57/200], Step [600/600], d_loss: 0.5693, g_loss: 2.0140, D(x): 0.80, D(G(z)): 0.18
Epoch [58/200], Step [200/600], d_loss: 0.4825, g_loss: 2.1138, D(x): 0.87, D(G(z)): 0.22
Epoch [58/200], Step [400/600], d_loss: 0.4156, g_loss: 3.4756, D(x): 0.91, D(G(z)): 0.19
Epoch [58/200], Step [600/600], d_loss: 0.5624, g_loss: 2.8075, D(x): 0.79, D(G(z)): 0.15
Epoch [59/200], Step [200/600], d_loss: 0.6324, g_loss: 2.9195, D(x): 0.85, D(G(z)): 0.21
Epoch [59/200], Step [400/600], d_loss: 0.6008, g_loss: 2.3756, D(x): 0.83, D(G(z)): 0.23
Epoch [59/200], Step [600/600], d_loss: 0.4797, g_loss: 2.4052, D(x): 0.84, D(G(z)): 0.16
Epoch [60/200], Step [200/600], d_loss: 0.7230, g_loss: 2.6456, D(x): 0.73, D(G(z)): 0.15
Epoch [60/200], Step [400/600], d_loss: 0.7388, g_loss: 1.8178, D(x): 0.79, D(G(z)): 0.26
Epoch [60/200], Step [600/600], d_loss: 0.4551, g_loss: 2.7809, D(x): 0.87, D(G(z)): 0.18
Epoch [61/200], Step [200/600], d_loss: 0.4204, g_loss: 3.1016, D(x): 0.89, D(G(z)): 0.17
Epoch [61/200], Step [400/600], d_loss: 0.5638, g_loss: 2.5622, D(x): 0.83, D(G(z)): 0.19
Epoch [61/200], Step [600/600], d_loss: 0.6783, g_loss: 3.2631, D(x): 0.76, D(G(z)): 0.16
Epoch [62/200], Step [200/600], d_loss: 0.4386, g_loss: 2.8836, D(x): 0.82, D(G(z)): 0.12
Epoch [62/200], Step [400/600], d_loss: 0.6454, g_loss: 2.2365, D(x): 0.81, D(G(z)): 0.21
Epoch [62/200], Step [600/600], d_loss: 0.6741, g_loss: 2.0690, D(x): 0.85, D(G(z)): 0.29
Epoch [63/200], Step [200/600], d_loss: 0.5414, g_loss: 2.1266, D(x): 0.77, D(G(z)): 0.13
Epoch [63/200], Step [400/600], d_loss: 0.7652, g_loss: 2.1745, D(x): 0.72, D(G(z)): 0.21
Epoch [63/200], Step [600/600], d_loss: 0.7641, g_loss: 2.1227, D(x): 0.82, D(G(z)): 0.30
Epoch [64/200], Step [200/600], d_loss: 0.6524, g_loss: 2.6566, D(x): 0.77, D(G(z)): 0.17
Epoch [64/200], Step [400/600], d_loss: 0.8171, g_loss: 2.5850, D(x): 0.69, D(G(z)): 0.18
Epoch [64/200], Step [600/600], d_loss: 0.6750, g_loss: 2.3300, D(x): 0.74, D(G(z)): 0.17
Epoch [65/200], Step [200/600], d_loss: 0.7582, g_loss: 1.9702, D(x): 0.76, D(G(z)): 0.24
Epoch [65/200], Step [400/600], d_loss: 0.9523, g_loss: 2.4954, D(x): 0.70, D(G(z)): 0.24
Epoch [65/200], Step [600/600], d_loss: 0.6429, g_loss: 1.7366, D(x): 0.77, D(G(z)): 0.20
Epoch [66/200], Step [200/600], d_loss: 0.7050, g_loss: 2.5913, D(x): 0.70, D(G(z)): 0.15
Epoch [66/200], Step [400/600], d_loss: 0.5947, g_loss: 2.0411, D(x): 0.85, D(G(z)): 0.25
Epoch [66/200], Step [600/600], d_loss: 0.4883, g_loss: 2.6249, D(x): 0.84, D(G(z)): 0.17
Epoch [67/200], Step [200/600], d_loss: 0.8298, g_loss: 1.3956, D(x): 0.77, D(G(z)): 0.30
Epoch [67/200], Step [400/600], d_loss: 0.7305, g_loss: 2.7201, D(x): 0.75, D(G(z)): 0.18
Epoch [67/200], Step [600/600], d_loss: 0.5734, g_loss: 2.0877, D(x): 0.84, D(G(z)): 0.26
Epoch [68/200], Step [200/600], d_loss: 0.5154, g_loss: 2.9730, D(x): 0.78, D(G(z)): 0.14
Epoch [68/200], Step [400/600], d_loss: 0.8208, g_loss: 2.4197, D(x): 0.68, D(G(z)): 0.15
Epoch [68/200], Step [600/600], d_loss: 0.7707, g_loss: 2.4385, D(x): 0.77, D(G(z)): 0.23
Epoch [69/200], Step [200/600], d_loss: 0.7308, g_loss: 2.4949, D(x): 0.77, D(G(z)): 0.25
Epoch [69/200], Step [400/600], d_loss: 0.6022, g_loss: 1.7119, D(x): 0.80, D(G(z)): 0.19
Epoch [69/200], Step [600/600], d_loss: 0.7737, g_loss: 2.0528, D(x): 0.76, D(G(z)): 0.23
Epoch [70/200], Step [200/600], d_loss: 0.6713, g_loss: 2.1575, D(x): 0.74, D(G(z)): 0.18
Epoch [70/200], Step [400/600], d_loss: 0.4558, g_loss: 2.6740, D(x): 0.81, D(G(z)): 0.14
Epoch [70/200], Step [600/600], d_loss: 0.6664, g_loss: 2.3781, D(x): 0.82, D(G(z)): 0.26
Epoch [71/200], Step [200/600], d_loss: 0.4208, g_loss: 2.7306, D(x): 0.84, D(G(z)): 0.13
Epoch [71/200], Step [400/600], d_loss: 0.5696, g_loss: 2.3525, D(x): 0.86, D(G(z)): 0.28
Epoch [71/200], Step [600/600], d_loss: 0.6642, g_loss: 2.1666, D(x): 0.89, D(G(z)): 0.34
Epoch [72/200], Step [200/600], d_loss: 0.7516, g_loss: 2.1152, D(x): 0.85, D(G(z)): 0.34
Epoch [72/200], Step [400/600], d_loss: 0.8580, g_loss: 2.2404, D(x): 0.76, D(G(z)): 0.28
Epoch [72/200], Step [600/600], d_loss: 0.8393, g_loss: 1.7819, D(x): 0.85, D(G(z)): 0.37
Epoch [73/200], Step [200/600], d_loss: 0.5924, g_loss: 2.6861, D(x): 0.76, D(G(z)): 0.17
Epoch [73/200], Step [400/600], d_loss: 0.7924, g_loss: 2.5014, D(x): 0.80, D(G(z)): 0.30
Epoch [73/200], Step [600/600], d_loss: 0.8842, g_loss: 2.0168, D(x): 0.75, D(G(z)): 0.31
Epoch [74/200], Step [200/600], d_loss: 0.6369, g_loss: 2.6035, D(x): 0.77, D(G(z)): 0.16
Epoch [74/200], Step [400/600], d_loss: 1.0137, g_loss: 1.9856, D(x): 0.74, D(G(z)): 0.31
Epoch [74/200], Step [600/600], d_loss: 0.6401, g_loss: 2.8239, D(x): 0.83, D(G(z)): 0.24
Epoch [75/200], Step [200/600], d_loss: 0.7222, g_loss: 2.6499, D(x): 0.76, D(G(z)): 0.21
Epoch [75/200], Step [400/600], d_loss: 0.5535, g_loss: 2.6794, D(x): 0.76, D(G(z)): 0.16
Epoch [75/200], Step [600/600], d_loss: 0.7484, g_loss: 1.8749, D(x): 0.76, D(G(z)): 0.24
Epoch [76/200], Step [200/600], d_loss: 0.6172, g_loss: 2.1332, D(x): 0.84, D(G(z)): 0.26
Epoch [76/200], Step [400/600], d_loss: 0.9073, g_loss: 1.7258, D(x): 0.74, D(G(z)): 0.32
Epoch [76/200], Step [600/600], d_loss: 0.6465, g_loss: 3.3903, D(x): 0.74, D(G(z)): 0.15
Epoch [77/200], Step [200/600], d_loss: 0.7583, g_loss: 2.1551, D(x): 0.78, D(G(z)): 0.25
Epoch [77/200], Step [400/600], d_loss: 0.6474, g_loss: 1.8975, D(x): 0.80, D(G(z)): 0.21
Epoch [77/200], Step [600/600], d_loss: 0.8241, g_loss: 1.8567, D(x): 0.68, D(G(z)): 0.20
Epoch [78/200], Step [200/600], d_loss: 0.6767, g_loss: 2.6657, D(x): 0.71, D(G(z)): 0.16
Epoch [78/200], Step [400/600], d_loss: 0.5965, g_loss: 2.0876, D(x): 0.82, D(G(z)): 0.23
Epoch [78/200], Step [600/600], d_loss: 0.8346, g_loss: 1.7815, D(x): 0.80, D(G(z)): 0.34
Epoch [79/200], Step [200/600], d_loss: 0.9473, g_loss: 1.3489, D(x): 0.77, D(G(z)): 0.35
Epoch [79/200], Step [400/600], d_loss: 0.6000, g_loss: 1.9143, D(x): 0.81, D(G(z)): 0.24
Epoch [79/200], Step [600/600], d_loss: 0.6208, g_loss: 2.1870, D(x): 0.75, D(G(z)): 0.18
Epoch [80/200], Step [200/600], d_loss: 0.7634, g_loss: 2.1038, D(x): 0.73, D(G(z)): 0.19
Epoch [80/200], Step [400/600], d_loss: 0.6876, g_loss: 2.3661, D(x): 0.78, D(G(z)): 0.25
Epoch [80/200], Step [600/600], d_loss: 0.7142, g_loss: 2.4693, D(x): 0.78, D(G(z)): 0.25
Epoch [81/200], Step [200/600], d_loss: 0.6182, g_loss: 1.9420, D(x): 0.83, D(G(z)): 0.24
Epoch [81/200], Step [400/600], d_loss: 0.6391, g_loss: 2.0722, D(x): 0.76, D(G(z)): 0.19
Epoch [81/200], Step [600/600], d_loss: 0.7186, g_loss: 2.5009, D(x): 0.70, D(G(z)): 0.15
Epoch [82/200], Step [200/600], d_loss: 0.7566, g_loss: 2.4882, D(x): 0.77, D(G(z)): 0.25
Epoch [82/200], Step [400/600], d_loss: 0.8851, g_loss: 2.2362, D(x): 0.77, D(G(z)): 0.33
Epoch [82/200], Step [600/600], d_loss: 0.7781, g_loss: 1.6748, D(x): 0.79, D(G(z)): 0.30
Epoch [83/200], Step [200/600], d_loss: 0.7610, g_loss: 2.0682, D(x): 0.80, D(G(z)): 0.28
Epoch [83/200], Step [400/600], d_loss: 0.5442, g_loss: 1.9091, D(x): 0.84, D(G(z)): 0.24
Epoch [83/200], Step [600/600], d_loss: 0.9009, g_loss: 1.7996, D(x): 0.77, D(G(z)): 0.30
Epoch [84/200], Step [200/600], d_loss: 0.5958, g_loss: 2.4650, D(x): 0.81, D(G(z)): 0.22
Epoch [84/200], Step [400/600], d_loss: 0.6394, g_loss: 1.7314, D(x): 0.78, D(G(z)): 0.24
Epoch [84/200], Step [600/600], d_loss: 0.7294, g_loss: 2.2409, D(x): 0.80, D(G(z)): 0.29
Epoch [85/200], Step [200/600], d_loss: 1.0039, g_loss: 2.1785, D(x): 0.76, D(G(z)): 0.36
Epoch [85/200], Step [400/600], d_loss: 0.7370, g_loss: 1.9103, D(x): 0.83, D(G(z)): 0.33
Epoch [85/200], Step [600/600], d_loss: 0.6967, g_loss: 2.4991, D(x): 0.75, D(G(z)): 0.22
Epoch [86/200], Step [200/600], d_loss: 0.6571, g_loss: 1.6934, D(x): 0.85, D(G(z)): 0.31
Epoch [86/200], Step [400/600], d_loss: 0.9010, g_loss: 1.7845, D(x): 0.67, D(G(z)): 0.21
Epoch [86/200], Step [600/600], d_loss: 0.7478, g_loss: 1.8386, D(x): 0.78, D(G(z)): 0.26
Epoch [87/200], Step [200/600], d_loss: 0.7735, g_loss: 1.5609, D(x): 0.80, D(G(z)): 0.30
Epoch [87/200], Step [400/600], d_loss: 0.8374, g_loss: 1.3501, D(x): 0.73, D(G(z)): 0.26
Epoch [87/200], Step [600/600], d_loss: 0.8446, g_loss: 1.9913, D(x): 0.75, D(G(z)): 0.28
Epoch [88/200], Step [200/600], d_loss: 0.7278, g_loss: 2.1532, D(x): 0.71, D(G(z)): 0.16
Epoch [88/200], Step [400/600], d_loss: 0.7137, g_loss: 2.1165, D(x): 0.69, D(G(z)): 0.18
Epoch [88/200], Step [600/600], d_loss: 0.8887, g_loss: 1.2378, D(x): 0.70, D(G(z)): 0.28
Epoch [89/200], Step [200/600], d_loss: 0.8917, g_loss: 1.9705, D(x): 0.69, D(G(z)): 0.26
Epoch [89/200], Step [400/600], d_loss: 0.7657, g_loss: 2.4673, D(x): 0.77, D(G(z)): 0.26
Epoch [89/200], Step [600/600], d_loss: 0.6419, g_loss: 1.9880, D(x): 0.82, D(G(z)): 0.27
Epoch [90/200], Step [200/600], d_loss: 0.7286, g_loss: 1.9025, D(x): 0.78, D(G(z)): 0.23
Epoch [90/200], Step [400/600], d_loss: 0.5856, g_loss: 1.6288, D(x): 0.84, D(G(z)): 0.25
Epoch [90/200], Step [600/600], d_loss: 0.6936, g_loss: 2.2498, D(x): 0.81, D(G(z)): 0.26
Epoch [91/200], Step [200/600], d_loss: 0.7015, g_loss: 2.2918, D(x): 0.70, D(G(z)): 0.18
Epoch [91/200], Step [400/600], d_loss: 0.6961, g_loss: 1.4990, D(x): 0.76, D(G(z)): 0.24
Epoch [91/200], Step [600/600], d_loss: 1.0444, g_loss: 1.7366, D(x): 0.72, D(G(z)): 0.33
Epoch [92/200], Step [200/600], d_loss: 0.8074, g_loss: 2.1368, D(x): 0.67, D(G(z)): 0.20
Epoch [92/200], Step [400/600], d_loss: 1.0336, g_loss: 1.5150, D(x): 0.67, D(G(z)): 0.29
Epoch [92/200], Step [600/600], d_loss: 0.7923, g_loss: 1.6682, D(x): 0.71, D(G(z)): 0.22
Epoch [93/200], Step [200/600], d_loss: 0.9705, g_loss: 1.5378, D(x): 0.85, D(G(z)): 0.44
Epoch [93/200], Step [400/600], d_loss: 0.9318, g_loss: 1.4094, D(x): 0.74, D(G(z)): 0.34
Epoch [93/200], Step [600/600], d_loss: 0.8385, g_loss: 2.1395, D(x): 0.69, D(G(z)): 0.18
Epoch [94/200], Step [200/600], d_loss: 0.8333, g_loss: 1.7765, D(x): 0.67, D(G(z)): 0.19
Epoch [94/200], Step [400/600], d_loss: 0.7310, g_loss: 2.2643, D(x): 0.73, D(G(z)): 0.21
Epoch [94/200], Step [600/600], d_loss: 0.8712, g_loss: 2.3476, D(x): 0.68, D(G(z)): 0.22
Epoch [95/200], Step [200/600], d_loss: 0.6457, g_loss: 1.8868, D(x): 0.78, D(G(z)): 0.24
Epoch [95/200], Step [400/600], d_loss: 0.8926, g_loss: 1.7284, D(x): 0.72, D(G(z)): 0.30
Epoch [95/200], Step [600/600], d_loss: 0.7586, g_loss: 1.8511, D(x): 0.75, D(G(z)): 0.27
Epoch [96/200], Step [200/600], d_loss: 0.6954, g_loss: 2.1118, D(x): 0.76, D(G(z)): 0.20
Epoch [96/200], Step [400/600], d_loss: 0.9447, g_loss: 1.9479, D(x): 0.69, D(G(z)): 0.26
Epoch [96/200], Step [600/600], d_loss: 0.8760, g_loss: 1.6268, D(x): 0.69, D(G(z)): 0.26
Epoch [97/200], Step [200/600], d_loss: 0.8981, g_loss: 1.8473, D(x): 0.70, D(G(z)): 0.26
Epoch [97/200], Step [400/600], d_loss: 0.8313, g_loss: 2.1134, D(x): 0.79, D(G(z)): 0.34
Epoch [97/200], Step [600/600], d_loss: 0.9799, g_loss: 1.7327, D(x): 0.77, D(G(z)): 0.38
Epoch [98/200], Step [200/600], d_loss: 0.7928, g_loss: 2.4119, D(x): 0.73, D(G(z)): 0.23
Epoch [98/200], Step [400/600], d_loss: 0.7950, g_loss: 1.7916, D(x): 0.77, D(G(z)): 0.27
Epoch [98/200], Step [600/600], d_loss: 0.9023, g_loss: 1.8960, D(x): 0.77, D(G(z)): 0.34
Epoch [99/200], Step [200/600], d_loss: 0.7460, g_loss: 1.4753, D(x): 0.77, D(G(z)): 0.25
Epoch [99/200], Step [400/600], d_loss: 1.0268, g_loss: 2.0465, D(x): 0.67, D(G(z)): 0.28
Epoch [99/200], Step [600/600], d_loss: 1.0442, g_loss: 1.4376, D(x): 0.71, D(G(z)): 0.31
Epoch [100/200], Step [200/600], d_loss: 0.9897, g_loss: 1.7333, D(x): 0.66, D(G(z)): 0.29
Epoch [100/200], Step [400/600], d_loss: 0.7949, g_loss: 2.5213, D(x): 0.73, D(G(z)): 0.26
Epoch [100/200], Step [600/600], d_loss: 0.8020, g_loss: 1.8205, D(x): 0.80, D(G(z)): 0.34
Epoch [101/200], Step [200/600], d_loss: 0.8511, g_loss: 1.7354, D(x): 0.68, D(G(z)): 0.23
Epoch [101/200], Step [400/600], d_loss: 0.6992, g_loss: 1.7517, D(x): 0.77, D(G(z)): 0.25
Epoch [101/200], Step [600/600], d_loss: 0.7008, g_loss: 1.9080, D(x): 0.75, D(G(z)): 0.24
Epoch [102/200], Step [200/600], d_loss: 0.9531, g_loss: 1.8946, D(x): 0.67, D(G(z)): 0.28
Epoch [102/200], Step [400/600], d_loss: 0.7998, g_loss: 2.2541, D(x): 0.77, D(G(z)): 0.30
Epoch [102/200], Step [600/600], d_loss: 0.7117, g_loss: 1.7053, D(x): 0.78, D(G(z)): 0.24
Epoch [103/200], Step [200/600], d_loss: 0.8539, g_loss: 1.6787, D(x): 0.73, D(G(z)): 0.28
Epoch [103/200], Step [400/600], d_loss: 0.9104, g_loss: 2.1986, D(x): 0.81, D(G(z)): 0.39
Epoch [103/200], Step [600/600], d_loss: 0.9180, g_loss: 2.4920, D(x): 0.68, D(G(z)): 0.25
Epoch [104/200], Step [200/600], d_loss: 0.9855, g_loss: 1.8069, D(x): 0.59, D(G(z)): 0.17
Epoch [104/200], Step [400/600], d_loss: 0.8452, g_loss: 2.0168, D(x): 0.70, D(G(z)): 0.24
Epoch [104/200], Step [600/600], d_loss: 0.9090, g_loss: 1.9037, D(x): 0.72, D(G(z)): 0.30
Epoch [105/200], Step [200/600], d_loss: 0.9647, g_loss: 1.5274, D(x): 0.64, D(G(z)): 0.26
Epoch [105/200], Step [400/600], d_loss: 0.7125, g_loss: 1.8630, D(x): 0.75, D(G(z)): 0.22
Epoch [105/200], Step [600/600], d_loss: 0.7479, g_loss: 1.8036, D(x): 0.79, D(G(z)): 0.31
Epoch [106/200], Step [200/600], d_loss: 0.9350, g_loss: 1.9197, D(x): 0.73, D(G(z)): 0.33
Epoch [106/200], Step [400/600], d_loss: 0.8513, g_loss: 1.5831, D(x): 0.79, D(G(z)): 0.32
Epoch [106/200], Step [600/600], d_loss: 0.8434, g_loss: 2.5236, D(x): 0.68, D(G(z)): 0.19
Epoch [107/200], Step [200/600], d_loss: 0.9225, g_loss: 1.6804, D(x): 0.78, D(G(z)): 0.39
Epoch [107/200], Step [400/600], d_loss: 0.8651, g_loss: 1.8469, D(x): 0.74, D(G(z)): 0.29
Epoch [107/200], Step [600/600], d_loss: 1.0709, g_loss: 1.8910, D(x): 0.61, D(G(z)): 0.26
Epoch [108/200], Step [200/600], d_loss: 0.8335, g_loss: 1.8592, D(x): 0.75, D(G(z)): 0.31
Epoch [108/200], Step [400/600], d_loss: 0.8644, g_loss: 1.8383, D(x): 0.65, D(G(z)): 0.20
Epoch [108/200], Step [600/600], d_loss: 0.7784, g_loss: 2.1106, D(x): 0.81, D(G(z)): 0.30
Epoch [109/200], Step [200/600], d_loss: 0.7713, g_loss: 1.8474, D(x): 0.78, D(G(z)): 0.30
Epoch [109/200], Step [400/600], d_loss: 0.6287, g_loss: 2.1118, D(x): 0.77, D(G(z)): 0.20
Epoch [109/200], Step [600/600], d_loss: 0.8158, g_loss: 1.9784, D(x): 0.68, D(G(z)): 0.21
Epoch [110/200], Step [200/600], d_loss: 0.8989, g_loss: 1.6065, D(x): 0.76, D(G(z)): 0.35
Epoch [110/200], Step [400/600], d_loss: 0.8740, g_loss: 1.4742, D(x): 0.71, D(G(z)): 0.30
Epoch [110/200], Step [600/600], d_loss: 0.6629, g_loss: 2.0034, D(x): 0.84, D(G(z)): 0.30
Epoch [111/200], Step [200/600], d_loss: 0.8124, g_loss: 1.7153, D(x): 0.80, D(G(z)): 0.34
Epoch [111/200], Step [400/600], d_loss: 0.8479, g_loss: 2.0981, D(x): 0.68, D(G(z)): 0.20
Epoch [111/200], Step [600/600], d_loss: 0.7136, g_loss: 1.8688, D(x): 0.72, D(G(z)): 0.21
Epoch [112/200], Step [200/600], d_loss: 0.7248, g_loss: 1.7079, D(x): 0.71, D(G(z)): 0.20
Epoch [112/200], Step [400/600], d_loss: 0.9196, g_loss: 1.7992, D(x): 0.64, D(G(z)): 0.20
Epoch [112/200], Step [600/600], d_loss: 0.6023, g_loss: 1.9134, D(x): 0.77, D(G(z)): 0.19
Epoch [113/200], Step [200/600], d_loss: 0.9980, g_loss: 1.7837, D(x): 0.70, D(G(z)): 0.31
Epoch [113/200], Step [400/600], d_loss: 0.9119, g_loss: 1.4296, D(x): 0.74, D(G(z)): 0.33
Epoch [113/200], Step [600/600], d_loss: 0.8514, g_loss: 1.8519, D(x): 0.72, D(G(z)): 0.24
Epoch [114/200], Step [200/600], d_loss: 0.9286, g_loss: 1.6051, D(x): 0.74, D(G(z)): 0.35
Epoch [114/200], Step [400/600], d_loss: 0.8345, g_loss: 1.9047, D(x): 0.72, D(G(z)): 0.28
Epoch [114/200], Step [600/600], d_loss: 0.7357, g_loss: 1.6866, D(x): 0.76, D(G(z)): 0.26
Epoch [115/200], Step [200/600], d_loss: 1.1178, g_loss: 1.8709, D(x): 0.73, D(G(z)): 0.39
Epoch [115/200], Step [400/600], d_loss: 0.9369, g_loss: 1.4690, D(x): 0.70, D(G(z)): 0.31
Epoch [115/200], Step [600/600], d_loss: 0.6943, g_loss: 1.7425, D(x): 0.77, D(G(z)): 0.24
Epoch [116/200], Step [200/600], d_loss: 0.7689, g_loss: 1.6584, D(x): 0.78, D(G(z)): 0.30
Epoch [116/200], Step [400/600], d_loss: 0.8996, g_loss: 1.3530, D(x): 0.80, D(G(z)): 0.38
Epoch [116/200], Step [600/600], d_loss: 0.9487, g_loss: 1.3110, D(x): 0.79, D(G(z)): 0.41
Epoch [117/200], Step [200/600], d_loss: 0.7204, g_loss: 2.5085, D(x): 0.81, D(G(z)): 0.27
Epoch [117/200], Step [400/600], d_loss: 0.9537, g_loss: 1.8315, D(x): 0.67, D(G(z)): 0.26
Epoch [117/200], Step [600/600], d_loss: 0.6763, g_loss: 1.7682, D(x): 0.80, D(G(z)): 0.27
Epoch [118/200], Step [200/600], d_loss: 0.9276, g_loss: 1.7115, D(x): 0.66, D(G(z)): 0.22
Epoch [118/200], Step [400/600], d_loss: 0.9136, g_loss: 1.4024, D(x): 0.73, D(G(z)): 0.31
Epoch [118/200], Step [600/600], d_loss: 0.8714, g_loss: 1.6280, D(x): 0.80, D(G(z)): 0.37
Epoch [119/200], Step [200/600], d_loss: 0.8546, g_loss: 1.5792, D(x): 0.73, D(G(z)): 0.31
Epoch [119/200], Step [400/600], d_loss: 0.7035, g_loss: 1.8312, D(x): 0.71, D(G(z)): 0.19
Epoch [119/200], Step [600/600], d_loss: 0.8878, g_loss: 1.7036, D(x): 0.69, D(G(z)): 0.25
Epoch [120/200], Step [200/600], d_loss: 0.9592, g_loss: 1.3375, D(x): 0.61, D(G(z)): 0.21
Epoch [120/200], Step [400/600], d_loss: 0.9011, g_loss: 2.0829, D(x): 0.66, D(G(z)): 0.22
Epoch [120/200], Step [600/600], d_loss: 0.8830, g_loss: 1.5800, D(x): 0.73, D(G(z)): 0.31
Epoch [121/200], Step [200/600], d_loss: 1.0095, g_loss: 1.6000, D(x): 0.64, D(G(z)): 0.31
Epoch [121/200], Step [400/600], d_loss: 0.7881, g_loss: 1.6243, D(x): 0.71, D(G(z)): 0.21
Epoch [121/200], Step [600/600], d_loss: 0.8295, g_loss: 1.6557, D(x): 0.71, D(G(z)): 0.24
Epoch [122/200], Step [200/600], d_loss: 0.9404, g_loss: 1.3748, D(x): 0.60, D(G(z)): 0.18
Epoch [122/200], Step [400/600], d_loss: 0.8459, g_loss: 2.3138, D(x): 0.66, D(G(z)): 0.19
Epoch [122/200], Step [600/600], d_loss: 0.9420, g_loss: 1.5667, D(x): 0.67, D(G(z)): 0.25
Epoch [123/200], Step [200/600], d_loss: 1.1139, g_loss: 1.4878, D(x): 0.56, D(G(z)): 0.20
Epoch [123/200], Step [400/600], d_loss: 0.7282, g_loss: 1.6294, D(x): 0.73, D(G(z)): 0.24
Epoch [123/200], Step [600/600], d_loss: 0.8722, g_loss: 1.2667, D(x): 0.68, D(G(z)): 0.26
Epoch [124/200], Step [200/600], d_loss: 0.7568, g_loss: 1.9428, D(x): 0.67, D(G(z)): 0.21
Epoch [124/200], Step [400/600], d_loss: 0.8482, g_loss: 2.0735, D(x): 0.72, D(G(z)): 0.30
Epoch [124/200], Step [600/600], d_loss: 0.9920, g_loss: 1.4466, D(x): 0.72, D(G(z)): 0.33
Epoch [125/200], Step [200/600], d_loss: 1.0061, g_loss: 2.2152, D(x): 0.64, D(G(z)): 0.27
Epoch [125/200], Step [400/600], d_loss: 1.0078, g_loss: 1.6272, D(x): 0.74, D(G(z)): 0.37
Epoch [125/200], Step [600/600], d_loss: 0.9771, g_loss: 2.1177, D(x): 0.71, D(G(z)): 0.29
Epoch [126/200], Step [200/600], d_loss: 1.0673, g_loss: 1.3473, D(x): 0.62, D(G(z)): 0.31
Epoch [126/200], Step [400/600], d_loss: 1.0108, g_loss: 1.8402, D(x): 0.67, D(G(z)): 0.30
Epoch [126/200], Step [600/600], d_loss: 0.9529, g_loss: 1.7117, D(x): 0.77, D(G(z)): 0.38
Epoch [127/200], Step [200/600], d_loss: 0.9988, g_loss: 1.5464, D(x): 0.71, D(G(z)): 0.35
Epoch [127/200], Step [400/600], d_loss: 0.7894, g_loss: 1.7470, D(x): 0.78, D(G(z)): 0.33
Epoch [127/200], Step [600/600], d_loss: 0.6126, g_loss: 2.2699, D(x): 0.79, D(G(z)): 0.23
Epoch [128/200], Step [200/600], d_loss: 0.8969, g_loss: 1.7038, D(x): 0.76, D(G(z)): 0.36
Epoch [128/200], Step [400/600], d_loss: 1.0108, g_loss: 1.3678, D(x): 0.67, D(G(z)): 0.33
Epoch [128/200], Step [600/600], d_loss: 0.8283, g_loss: 1.8264, D(x): 0.74, D(G(z)): 0.30
Epoch [129/200], Step [200/600], d_loss: 0.7680, g_loss: 1.9022, D(x): 0.69, D(G(z)): 0.18
Epoch [129/200], Step [400/600], d_loss: 0.9077, g_loss: 1.5271, D(x): 0.76, D(G(z)): 0.36
Epoch [129/200], Step [600/600], d_loss: 0.7863, g_loss: 1.5050, D(x): 0.75, D(G(z)): 0.30
Epoch [130/200], Step [200/600], d_loss: 0.7647, g_loss: 1.6246, D(x): 0.76, D(G(z)): 0.29
Epoch [130/200], Step [400/600], d_loss: 1.0322, g_loss: 1.4377, D(x): 0.82, D(G(z)): 0.44
Epoch [130/200], Step [600/600], d_loss: 0.9320, g_loss: 2.2156, D(x): 0.67, D(G(z)): 0.24
Epoch [131/200], Step [200/600], d_loss: 1.0010, g_loss: 1.8664, D(x): 0.65, D(G(z)): 0.29
Epoch [131/200], Step [400/600], d_loss: 0.8542, g_loss: 1.8232, D(x): 0.70, D(G(z)): 0.22
Epoch [131/200], Step [600/600], d_loss: 0.7486, g_loss: 1.7270, D(x): 0.78, D(G(z)): 0.28
Epoch [132/200], Step [200/600], d_loss: 0.9719, g_loss: 1.2860, D(x): 0.68, D(G(z)): 0.29
Epoch [132/200], Step [400/600], d_loss: 0.9132, g_loss: 1.5629, D(x): 0.65, D(G(z)): 0.24
Epoch [132/200], Step [600/600], d_loss: 0.9548, g_loss: 1.8049, D(x): 0.68, D(G(z)): 0.29
Epoch [133/200], Step [200/600], d_loss: 0.6619, g_loss: 1.7538, D(x): 0.76, D(G(z)): 0.23
Epoch [133/200], Step [400/600], d_loss: 1.0047, g_loss: 1.5824, D(x): 0.74, D(G(z)): 0.37
Epoch [133/200], Step [600/600], d_loss: 0.8980, g_loss: 1.5372, D(x): 0.70, D(G(z)): 0.29
Epoch [134/200], Step [200/600], d_loss: 0.9166, g_loss: 1.2988, D(x): 0.80, D(G(z)): 0.37
Epoch [134/200], Step [400/600], d_loss: 0.9848, g_loss: 1.6209, D(x): 0.72, D(G(z)): 0.35
Epoch [134/200], Step [600/600], d_loss: 0.8396, g_loss: 1.5322, D(x): 0.69, D(G(z)): 0.28
Epoch [135/200], Step [200/600], d_loss: 0.8331, g_loss: 1.5928, D(x): 0.68, D(G(z)): 0.24
Epoch [135/200], Step [400/600], d_loss: 0.9271, g_loss: 1.7751, D(x): 0.68, D(G(z)): 0.28
Epoch [135/200], Step [600/600], d_loss: 1.2802, g_loss: 1.3046, D(x): 0.63, D(G(z)): 0.37
Epoch [136/200], Step [200/600], d_loss: 1.0288, g_loss: 1.6774, D(x): 0.68, D(G(z)): 0.32
Epoch [136/200], Step [400/600], d_loss: 0.9027, g_loss: 1.6579, D(x): 0.77, D(G(z)): 0.37
Epoch [136/200], Step [600/600], d_loss: 1.0080, g_loss: 1.3883, D(x): 0.67, D(G(z)): 0.30
Epoch [137/200], Step [200/600], d_loss: 0.9043, g_loss: 1.5809, D(x): 0.69, D(G(z)): 0.30
Epoch [137/200], Step [400/600], d_loss: 0.8684, g_loss: 1.7851, D(x): 0.76, D(G(z)): 0.31
Epoch [137/200], Step [600/600], d_loss: 0.9694, g_loss: 1.5669, D(x): 0.71, D(G(z)): 0.33
Epoch [138/200], Step [200/600], d_loss: 0.7400, g_loss: 1.6188, D(x): 0.72, D(G(z)): 0.24
Epoch [138/200], Step [400/600], d_loss: 0.9084, g_loss: 1.6095, D(x): 0.67, D(G(z)): 0.27
Epoch [138/200], Step [600/600], d_loss: 0.8807, g_loss: 1.4123, D(x): 0.72, D(G(z)): 0.30
Epoch [139/200], Step [200/600], d_loss: 0.6726, g_loss: 1.3809, D(x): 0.78, D(G(z)): 0.25
Epoch [139/200], Step [400/600], d_loss: 0.9256, g_loss: 1.1885, D(x): 0.73, D(G(z)): 0.33
Epoch [139/200], Step [600/600], d_loss: 0.9302, g_loss: 1.2736, D(x): 0.69, D(G(z)): 0.28
Epoch [140/200], Step [200/600], d_loss: 0.9735, g_loss: 1.3551, D(x): 0.69, D(G(z)): 0.31
Epoch [140/200], Step [400/600], d_loss: 1.0635, g_loss: 1.8120, D(x): 0.65, D(G(z)): 0.27
Epoch [140/200], Step [600/600], d_loss: 0.9437, g_loss: 1.2334, D(x): 0.69, D(G(z)): 0.33
Epoch [141/200], Step [200/600], d_loss: 0.9152, g_loss: 1.8172, D(x): 0.64, D(G(z)): 0.26
Epoch [141/200], Step [400/600], d_loss: 0.8368, g_loss: 1.5303, D(x): 0.71, D(G(z)): 0.26
Epoch [141/200], Step [600/600], d_loss: 0.8082, g_loss: 1.7512, D(x): 0.73, D(G(z)): 0.30
Epoch [142/200], Step [200/600], d_loss: 1.0856, g_loss: 1.4494, D(x): 0.64, D(G(z)): 0.31
Epoch [142/200], Step [400/600], d_loss: 1.0798, g_loss: 1.6213, D(x): 0.61, D(G(z)): 0.29
Epoch [142/200], Step [600/600], d_loss: 1.0647, g_loss: 1.6840, D(x): 0.72, D(G(z)): 0.37
Epoch [143/200], Step [200/600], d_loss: 1.0429, g_loss: 1.7174, D(x): 0.73, D(G(z)): 0.38
Epoch [143/200], Step [400/600], d_loss: 0.7951, g_loss: 1.7997, D(x): 0.68, D(G(z)): 0.23
Epoch [143/200], Step [600/600], d_loss: 1.0584, g_loss: 1.6945, D(x): 0.63, D(G(z)): 0.31
Epoch [144/200], Step [200/600], d_loss: 0.8060, g_loss: 2.1261, D(x): 0.70, D(G(z)): 0.24
Epoch [144/200], Step [400/600], d_loss: 0.8945, g_loss: 1.5272, D(x): 0.70, D(G(z)): 0.28
Epoch [144/200], Step [600/600], d_loss: 0.7505, g_loss: 1.8432, D(x): 0.75, D(G(z)): 0.28
Epoch [145/200], Step [200/600], d_loss: 1.1672, g_loss: 1.6065, D(x): 0.66, D(G(z)): 0.35
Epoch [145/200], Step [400/600], d_loss: 0.8520, g_loss: 1.4128, D(x): 0.82, D(G(z)): 0.37
Epoch [145/200], Step [600/600], d_loss: 0.8785, g_loss: 1.4892, D(x): 0.75, D(G(z)): 0.33
Epoch [146/200], Step [200/600], d_loss: 0.8565, g_loss: 1.3905, D(x): 0.75, D(G(z)): 0.31
Epoch [146/200], Step [400/600], d_loss: 1.0247, g_loss: 1.2543, D(x): 0.70, D(G(z)): 0.38
Epoch [146/200], Step [600/600], d_loss: 0.9116, g_loss: 1.5564, D(x): 0.61, D(G(z)): 0.20
Epoch [147/200], Step [200/600], d_loss: 0.9788, g_loss: 1.6320, D(x): 0.67, D(G(z)): 0.32
Epoch [147/200], Step [400/600], d_loss: 1.0233, g_loss: 1.2688, D(x): 0.64, D(G(z)): 0.31
Epoch [147/200], Step [600/600], d_loss: 0.8218, g_loss: 1.3464, D(x): 0.73, D(G(z)): 0.30
Epoch [148/200], Step [200/600], d_loss: 0.8826, g_loss: 1.8604, D(x): 0.69, D(G(z)): 0.29
Epoch [148/200], Step [400/600], d_loss: 0.8574, g_loss: 1.2975, D(x): 0.72, D(G(z)): 0.31
Epoch [148/200], Step [600/600], d_loss: 0.8590, g_loss: 1.2500, D(x): 0.71, D(G(z)): 0.29
Epoch [149/200], Step [200/600], d_loss: 0.9476, g_loss: 1.3186, D(x): 0.70, D(G(z)): 0.32
Epoch [149/200], Step [400/600], d_loss: 1.1153, g_loss: 1.4313, D(x): 0.66, D(G(z)): 0.36
Epoch [149/200], Step [600/600], d_loss: 0.9207, g_loss: 1.5918, D(x): 0.64, D(G(z)): 0.24
Epoch [150/200], Step [200/600], d_loss: 1.0620, g_loss: 1.9464, D(x): 0.67, D(G(z)): 0.36
Epoch [150/200], Step [400/600], d_loss: 0.8902, g_loss: 2.1322, D(x): 0.62, D(G(z)): 0.18
Epoch [150/200], Step [600/600], d_loss: 0.8853, g_loss: 1.5706, D(x): 0.71, D(G(z)): 0.30
Epoch [151/200], Step [200/600], d_loss: 1.1444, g_loss: 1.7974, D(x): 0.60, D(G(z)): 0.30
Epoch [151/200], Step [400/600], d_loss: 0.9993, g_loss: 1.4988, D(x): 0.63, D(G(z)): 0.26
Epoch [151/200], Step [600/600], d_loss: 1.0538, g_loss: 1.8101, D(x): 0.63, D(G(z)): 0.26
Epoch [152/200], Step [200/600], d_loss: 0.8820, g_loss: 1.8347, D(x): 0.76, D(G(z)): 0.33
Epoch [152/200], Step [400/600], d_loss: 0.9782, g_loss: 1.5429, D(x): 0.63, D(G(z)): 0.26
Epoch [152/200], Step [600/600], d_loss: 1.2152, g_loss: 1.7726, D(x): 0.60, D(G(z)): 0.33
Epoch [153/200], Step [200/600], d_loss: 0.9680, g_loss: 1.2081, D(x): 0.68, D(G(z)): 0.33
Epoch [153/200], Step [400/600], d_loss: 0.9256, g_loss: 1.6799, D(x): 0.69, D(G(z)): 0.30
Epoch [153/200], Step [600/600], d_loss: 0.9704, g_loss: 1.2176, D(x): 0.76, D(G(z)): 0.38
Epoch [154/200], Step [200/600], d_loss: 1.0162, g_loss: 1.3491, D(x): 0.72, D(G(z)): 0.36
Epoch [154/200], Step [400/600], d_loss: 0.9492, g_loss: 2.0773, D(x): 0.62, D(G(z)): 0.21
Epoch [154/200], Step [600/600], d_loss: 0.8634, g_loss: 1.4710, D(x): 0.71, D(G(z)): 0.29
Epoch [155/200], Step [200/600], d_loss: 0.8810, g_loss: 1.6961, D(x): 0.69, D(G(z)): 0.25
Epoch [155/200], Step [400/600], d_loss: 1.0829, g_loss: 1.4673, D(x): 0.58, D(G(z)): 0.25
Epoch [155/200], Step [600/600], d_loss: 1.0839, g_loss: 1.4223, D(x): 0.73, D(G(z)): 0.40
Epoch [156/200], Step [200/600], d_loss: 1.1851, g_loss: 1.6687, D(x): 0.63, D(G(z)): 0.37
Epoch [156/200], Step [400/600], d_loss: 0.9716, g_loss: 1.6617, D(x): 0.64, D(G(z)): 0.28
Epoch [156/200], Step [600/600], d_loss: 0.8191, g_loss: 1.7810, D(x): 0.69, D(G(z)): 0.24
Epoch [157/200], Step [200/600], d_loss: 1.0335, g_loss: 1.3579, D(x): 0.61, D(G(z)): 0.24
Epoch [157/200], Step [400/600], d_loss: 1.0019, g_loss: 1.3144, D(x): 0.71, D(G(z)): 0.35
Epoch [157/200], Step [600/600], d_loss: 0.8015, g_loss: 1.4819, D(x): 0.75, D(G(z)): 0.31
Epoch [158/200], Step [200/600], d_loss: 0.9141, g_loss: 1.5450, D(x): 0.73, D(G(z)): 0.32
Epoch [158/200], Step [400/600], d_loss: 1.0250, g_loss: 1.6567, D(x): 0.70, D(G(z)): 0.34
Epoch [158/200], Step [600/600], d_loss: 1.1062, g_loss: 1.2812, D(x): 0.72, D(G(z)): 0.41
Epoch [159/200], Step [200/600], d_loss: 0.9579, g_loss: 1.7507, D(x): 0.65, D(G(z)): 0.29
Epoch [159/200], Step [400/600], d_loss: 1.0090, g_loss: 1.6796, D(x): 0.66, D(G(z)): 0.30
Epoch [159/200], Step [600/600], d_loss: 0.9404, g_loss: 1.4110, D(x): 0.62, D(G(z)): 0.24
Epoch [160/200], Step [200/600], d_loss: 0.8459, g_loss: 1.5923, D(x): 0.79, D(G(z)): 0.35
Epoch [160/200], Step [400/600], d_loss: 1.0983, g_loss: 1.5172, D(x): 0.61, D(G(z)): 0.30
Epoch [160/200], Step [600/600], d_loss: 0.9830, g_loss: 1.4812, D(x): 0.69, D(G(z)): 0.35
Epoch [161/200], Step [200/600], d_loss: 0.8541, g_loss: 1.5836, D(x): 0.66, D(G(z)): 0.25
Epoch [161/200], Step [400/600], d_loss: 1.0266, g_loss: 1.3839, D(x): 0.65, D(G(z)): 0.32
Epoch [161/200], Step [600/600], d_loss: 0.9394, g_loss: 1.6997, D(x): 0.70, D(G(z)): 0.33
Epoch [162/200], Step [200/600], d_loss: 0.8215, g_loss: 1.5584, D(x): 0.71, D(G(z)): 0.29
Epoch [162/200], Step [400/600], d_loss: 0.9102, g_loss: 1.5626, D(x): 0.71, D(G(z)): 0.33
Epoch [162/200], Step [600/600], d_loss: 0.9159, g_loss: 1.8538, D(x): 0.71, D(G(z)): 0.31
Epoch [163/200], Step [200/600], d_loss: 1.0573, g_loss: 1.7526, D(x): 0.78, D(G(z)): 0.45
Epoch [163/200], Step [400/600], d_loss: 0.7909, g_loss: 1.6003, D(x): 0.68, D(G(z)): 0.23
Epoch [163/200], Step [600/600], d_loss: 0.9629, g_loss: 1.4818, D(x): 0.73, D(G(z)): 0.35
Epoch [164/200], Step [200/600], d_loss: 1.0584, g_loss: 1.8487, D(x): 0.62, D(G(z)): 0.29
Epoch [164/200], Step [400/600], d_loss: 0.9867, g_loss: 1.7461, D(x): 0.71, D(G(z)): 0.36
Epoch [164/200], Step [600/600], d_loss: 1.0074, g_loss: 1.5841, D(x): 0.68, D(G(z)): 0.31
Epoch [165/200], Step [200/600], d_loss: 0.9651, g_loss: 1.5523, D(x): 0.72, D(G(z)): 0.35
Epoch [165/200], Step [400/600], d_loss: 0.8979, g_loss: 1.5189, D(x): 0.66, D(G(z)): 0.23
Epoch [165/200], Step [600/600], d_loss: 0.8604, g_loss: 1.4859, D(x): 0.79, D(G(z)): 0.35
Epoch [166/200], Step [200/600], d_loss: 0.9645, g_loss: 1.7473, D(x): 0.61, D(G(z)): 0.23
Epoch [166/200], Step [400/600], d_loss: 0.6825, g_loss: 1.3785, D(x): 0.84, D(G(z)): 0.33
Epoch [166/200], Step [600/600], d_loss: 0.9094, g_loss: 1.3634, D(x): 0.73, D(G(z)): 0.34
Epoch [167/200], Step [200/600], d_loss: 0.8262, g_loss: 1.6001, D(x): 0.72, D(G(z)): 0.30
Epoch [167/200], Step [400/600], d_loss: 0.9909, g_loss: 1.9477, D(x): 0.55, D(G(z)): 0.18
Epoch [167/200], Step [600/600], d_loss: 1.0432, g_loss: 1.3547, D(x): 0.60, D(G(z)): 0.26
Epoch [168/200], Step [200/600], d_loss: 0.9359, g_loss: 2.0457, D(x): 0.70, D(G(z)): 0.32
Epoch [168/200], Step [400/600], d_loss: 1.0142, g_loss: 1.6429, D(x): 0.66, D(G(z)): 0.32
Epoch [168/200], Step [600/600], d_loss: 0.9667, g_loss: 1.3652, D(x): 0.75, D(G(z)): 0.39
Epoch [169/200], Step [200/600], d_loss: 0.8659, g_loss: 1.8797, D(x): 0.66, D(G(z)): 0.23
Epoch [169/200], Step [400/600], d_loss: 0.9469, g_loss: 1.5848, D(x): 0.77, D(G(z)): 0.39
Epoch [169/200], Step [600/600], d_loss: 0.9253, g_loss: 1.6434, D(x): 0.67, D(G(z)): 0.27
Epoch [170/200], Step [200/600], d_loss: 0.9876, g_loss: 1.5204, D(x): 0.73, D(G(z)): 0.40
Epoch [170/200], Step [400/600], d_loss: 1.0286, g_loss: 1.3180, D(x): 0.64, D(G(z)): 0.31
Epoch [170/200], Step [600/600], d_loss: 1.0189, g_loss: 1.3023, D(x): 0.68, D(G(z)): 0.33
Epoch [171/200], Step [200/600], d_loss: 0.9299, g_loss: 1.4886, D(x): 0.77, D(G(z)): 0.36
Epoch [171/200], Step [400/600], d_loss: 0.8501, g_loss: 1.1939, D(x): 0.72, D(G(z)): 0.31
Epoch [171/200], Step [600/600], d_loss: 0.9355, g_loss: 1.5083, D(x): 0.61, D(G(z)): 0.25
Epoch [172/200], Step [200/600], d_loss: 0.9333, g_loss: 1.3659, D(x): 0.76, D(G(z)): 0.37
Epoch [172/200], Step [400/600], d_loss: 0.8872, g_loss: 1.5996, D(x): 0.69, D(G(z)): 0.27
Epoch [172/200], Step [600/600], d_loss: 0.9234, g_loss: 1.5076, D(x): 0.64, D(G(z)): 0.23
Epoch [173/200], Step [200/600], d_loss: 0.8251, g_loss: 1.9171, D(x): 0.72, D(G(z)): 0.26
Epoch [173/200], Step [400/600], d_loss: 1.1377, g_loss: 1.5771, D(x): 0.57, D(G(z)): 0.28
Epoch [173/200], Step [600/600], d_loss: 0.9027, g_loss: 1.7397, D(x): 0.65, D(G(z)): 0.27
Epoch [174/200], Step [200/600], d_loss: 0.8281, g_loss: 1.6004, D(x): 0.68, D(G(z)): 0.26
Epoch [174/200], Step [400/600], d_loss: 0.8857, g_loss: 1.7043, D(x): 0.72, D(G(z)): 0.33
Epoch [174/200], Step [600/600], d_loss: 0.9295, g_loss: 1.8799, D(x): 0.68, D(G(z)): 0.28
Epoch [175/200], Step [200/600], d_loss: 0.9017, g_loss: 1.7627, D(x): 0.70, D(G(z)): 0.30
Epoch [175/200], Step [400/600], d_loss: 0.8745, g_loss: 1.3993, D(x): 0.68, D(G(z)): 0.28
Epoch [175/200], Step [600/600], d_loss: 1.0171, g_loss: 1.4064, D(x): 0.74, D(G(z)): 0.40
Epoch [176/200], Step [200/600], d_loss: 0.8911, g_loss: 1.4087, D(x): 0.78, D(G(z)): 0.38
Epoch [176/200], Step [400/600], d_loss: 1.0525, g_loss: 1.3220, D(x): 0.67, D(G(z)): 0.34
Epoch [176/200], Step [600/600], d_loss: 0.7854, g_loss: 2.0379, D(x): 0.68, D(G(z)): 0.22
Epoch [177/200], Step [200/600], d_loss: 0.9156, g_loss: 1.9224, D(x): 0.73, D(G(z)): 0.34
Epoch [177/200], Step [400/600], d_loss: 0.7522, g_loss: 1.7134, D(x): 0.68, D(G(z)): 0.20
Epoch [177/200], Step [600/600], d_loss: 0.8522, g_loss: 1.3662, D(x): 0.70, D(G(z)): 0.27
Epoch [178/200], Step [200/600], d_loss: 0.9038, g_loss: 1.3921, D(x): 0.70, D(G(z)): 0.27
Epoch [178/200], Step [400/600], d_loss: 0.9785, g_loss: 1.6295, D(x): 0.68, D(G(z)): 0.33
Epoch [178/200], Step [600/600], d_loss: 0.9072, g_loss: 1.6979, D(x): 0.69, D(G(z)): 0.31
Epoch [179/200], Step [200/600], d_loss: 1.0421, g_loss: 2.0187, D(x): 0.65, D(G(z)): 0.31
Epoch [179/200], Step [400/600], d_loss: 0.8482, g_loss: 1.6220, D(x): 0.74, D(G(z)): 0.31
Epoch [179/200], Step [600/600], d_loss: 0.9346, g_loss: 1.7184, D(x): 0.66, D(G(z)): 0.28
Epoch [180/200], Step [200/600], d_loss: 0.9424, g_loss: 1.6355, D(x): 0.65, D(G(z)): 0.26
Epoch [180/200], Step [400/600], d_loss: 0.9139, g_loss: 1.5548, D(x): 0.70, D(G(z)): 0.31
Epoch [180/200], Step [600/600], d_loss: 0.7622, g_loss: 1.4710, D(x): 0.72, D(G(z)): 0.27
Epoch [181/200], Step [200/600], d_loss: 1.1145, g_loss: 1.4237, D(x): 0.66, D(G(z)): 0.36
Epoch [181/200], Step [400/600], d_loss: 1.0776, g_loss: 1.4094, D(x): 0.75, D(G(z)): 0.40
Epoch [181/200], Step [600/600], d_loss: 0.7935, g_loss: 1.5159, D(x): 0.72, D(G(z)): 0.26
Epoch [182/200], Step [200/600], d_loss: 0.7814, g_loss: 1.3784, D(x): 0.75, D(G(z)): 0.28
Epoch [182/200], Step [400/600], d_loss: 1.0844, g_loss: 1.1643, D(x): 0.66, D(G(z)): 0.37
Epoch [182/200], Step [600/600], d_loss: 0.8495, g_loss: 1.8156, D(x): 0.69, D(G(z)): 0.28
Epoch [183/200], Step [200/600], d_loss: 1.0465, g_loss: 1.5846, D(x): 0.66, D(G(z)): 0.34
Epoch [183/200], Step [400/600], d_loss: 1.0113, g_loss: 1.5419, D(x): 0.66, D(G(z)): 0.31
Epoch [183/200], Step [600/600], d_loss: 0.9660, g_loss: 1.3313, D(x): 0.75, D(G(z)): 0.38
Epoch [184/200], Step [200/600], d_loss: 0.9482, g_loss: 1.5999, D(x): 0.68, D(G(z)): 0.29
Epoch [184/200], Step [400/600], d_loss: 0.9748, g_loss: 1.3257, D(x): 0.66, D(G(z)): 0.29
Epoch [184/200], Step [600/600], d_loss: 1.0018, g_loss: 1.5651, D(x): 0.70, D(G(z)): 0.35
Epoch [185/200], Step [200/600], d_loss: 0.9115, g_loss: 1.6923, D(x): 0.71, D(G(z)): 0.32
Epoch [185/200], Step [400/600], d_loss: 0.7879, g_loss: 1.7067, D(x): 0.70, D(G(z)): 0.26
Epoch [185/200], Step [600/600], d_loss: 0.9752, g_loss: 1.6181, D(x): 0.65, D(G(z)): 0.27
Epoch [186/200], Step [200/600], d_loss: 0.9705, g_loss: 1.5116, D(x): 0.65, D(G(z)): 0.29
Epoch [186/200], Step [400/600], d_loss: 0.8693, g_loss: 1.5713, D(x): 0.75, D(G(z)): 0.35
Epoch [186/200], Step [600/600], d_loss: 1.0797, g_loss: 1.4829, D(x): 0.68, D(G(z)): 0.38
Epoch [187/200], Step [200/600], d_loss: 0.8853, g_loss: 1.6581, D(x): 0.70, D(G(z)): 0.30
Epoch [187/200], Step [400/600], d_loss: 0.8808, g_loss: 1.7821, D(x): 0.72, D(G(z)): 0.33
Epoch [187/200], Step [600/600], d_loss: 0.9180, g_loss: 1.7416, D(x): 0.74, D(G(z)): 0.35
Epoch [188/200], Step [200/600], d_loss: 0.8069, g_loss: 1.4954, D(x): 0.73, D(G(z)): 0.29
Epoch [188/200], Step [400/600], d_loss: 0.9400, g_loss: 1.8438, D(x): 0.69, D(G(z)): 0.29
Epoch [188/200], Step [600/600], d_loss: 0.9805, g_loss: 1.5522, D(x): 0.72, D(G(z)): 0.35
Epoch [189/200], Step [200/600], d_loss: 0.8633, g_loss: 1.7176, D(x): 0.69, D(G(z)): 0.30
Epoch [189/200], Step [400/600], d_loss: 0.8756, g_loss: 1.5109, D(x): 0.71, D(G(z)): 0.30
Epoch [189/200], Step [600/600], d_loss: 1.1062, g_loss: 1.7254, D(x): 0.64, D(G(z)): 0.32
Epoch [190/200], Step [200/600], d_loss: 0.9787, g_loss: 1.3468, D(x): 0.70, D(G(z)): 0.35
Epoch [190/200], Step [400/600], d_loss: 0.7695, g_loss: 1.5876, D(x): 0.70, D(G(z)): 0.22
Epoch [190/200], Step [600/600], d_loss: 0.9003, g_loss: 1.5295, D(x): 0.72, D(G(z)): 0.35
Epoch [191/200], Step [200/600], d_loss: 1.0498, g_loss: 1.5651, D(x): 0.57, D(G(z)): 0.24
Epoch [191/200], Step [400/600], d_loss: 0.9835, g_loss: 1.5292, D(x): 0.62, D(G(z)): 0.26
Epoch [191/200], Step [600/600], d_loss: 1.0058, g_loss: 1.8599, D(x): 0.61, D(G(z)): 0.26
Epoch [192/200], Step [200/600], d_loss: 1.0540, g_loss: 1.4902, D(x): 0.62, D(G(z)): 0.32
Epoch [192/200], Step [400/600], d_loss: 0.9754, g_loss: 1.3995, D(x): 0.70, D(G(z)): 0.37
Epoch [192/200], Step [600/600], d_loss: 0.9655, g_loss: 1.3887, D(x): 0.68, D(G(z)): 0.31
Epoch [193/200], Step [200/600], d_loss: 0.9888, g_loss: 1.6682, D(x): 0.63, D(G(z)): 0.26
Epoch [193/200], Step [400/600], d_loss: 0.8445, g_loss: 1.5723, D(x): 0.73, D(G(z)): 0.33
Epoch [193/200], Step [600/600], d_loss: 0.9094, g_loss: 1.8187, D(x): 0.69, D(G(z)): 0.30
Epoch [194/200], Step [200/600], d_loss: 0.9689, g_loss: 1.7502, D(x): 0.61, D(G(z)): 0.23
Epoch [194/200], Step [400/600], d_loss: 0.8786, g_loss: 1.3475, D(x): 0.75, D(G(z)): 0.35
Epoch [194/200], Step [600/600], d_loss: 0.8869, g_loss: 2.3168, D(x): 0.67, D(G(z)): 0.26
Epoch [195/200], Step [200/600], d_loss: 0.9529, g_loss: 1.9417, D(x): 0.71, D(G(z)): 0.33
Epoch [195/200], Step [400/600], d_loss: 1.0048, g_loss: 1.2613, D(x): 0.75, D(G(z)): 0.40
Epoch [195/200], Step [600/600], d_loss: 0.9013, g_loss: 1.4521, D(x): 0.66, D(G(z)): 0.30
Epoch [196/200], Step [200/600], d_loss: 0.8513, g_loss: 1.5400, D(x): 0.74, D(G(z)): 0.29
Epoch [196/200], Step [400/600], d_loss: 0.9727, g_loss: 1.3992, D(x): 0.64, D(G(z)): 0.30
Epoch [196/200], Step [600/600], d_loss: 0.9113, g_loss: 1.4926, D(x): 0.77, D(G(z)): 0.40
Epoch [197/200], Step [200/600], d_loss: 0.9602, g_loss: 1.4885, D(x): 0.68, D(G(z)): 0.32
Epoch [197/200], Step [400/600], d_loss: 1.0403, g_loss: 1.8077, D(x): 0.65, D(G(z)): 0.32
Epoch [197/200], Step [600/600], d_loss: 0.8746, g_loss: 1.8007, D(x): 0.74, D(G(z)): 0.33
Epoch [198/200], Step [200/600], d_loss: 0.9678, g_loss: 1.2999, D(x): 0.62, D(G(z)): 0.26
Epoch [198/200], Step [400/600], d_loss: 0.9882, g_loss: 1.3104, D(x): 0.71, D(G(z)): 0.37
Epoch [198/200], Step [600/600], d_loss: 0.9346, g_loss: 1.3032, D(x): 0.74, D(G(z)): 0.39
Epoch [199/200], Step [200/600], d_loss: 0.9317, g_loss: 1.5710, D(x): 0.69, D(G(z)): 0.31
Epoch [199/200], Step [400/600], d_loss: 1.0818, g_loss: 1.3135, D(x): 0.56, D(G(z)): 0.22
Epoch [199/200], Step [600/600], d_loss: 0.8854, g_loss: 1.0944, D(x): 0.69, D(G(z)): 0.31

你可能感兴趣的:(PyTorch)

[实践应用] 深度学习之优化器 YuanDaima2048 深度学习工具使用 pytorch 深度学习人工智能机器学习 python 优化器
文章总览：YuanDaiMa2048博客文章总览深度学习之优化器1.随机梯度下降（SGD）2.动量优化（Momentum）3.自适应梯度（Adagrad）4.自适应矩估计（Adam）5.RMSprop总结其他介绍在深度学习中，优化器用于更新模型的参数，以最小化损失函数。常见的优化函数有很多种，下面是几种主流的优化器及其特点、原理和PyTorch实现：1.随机梯度下降（SGD）原理:随机梯度下降通过
个人学习笔记7-6：动手学深度学习pytorch版-李沐浪子L 深度学习深度学习笔记计算机视觉 python 人工智能神经网络 pytorch
#人工智能##深度学习##语义分割##计算机视觉##神经网络#计算机视觉13.11全卷积网络全卷积网络（fullyconvolutionalnetwork，FCN）采用卷积神经网络实现了从图像像素到像素类别的变换。引入l转置卷积（transposedconvolution）实现的，输出的类别预测与输入图像在像素级别上具有一一对应关系：通道维的输出即该位置对应像素的类别预测。13.11.1构造模型下
【安装环境】配置MMTracking环境 xuanyu22 安装环境机器学习神经网络深度学习 python
版本v0.14.0安装torchnumpy的版本不能太高，否则后面安装时会发生冲突。先安装numpy，因为pytorch的安装会自动配置高版本numpy。condainstallnumpy=1.21.5mmtracking支持的torch版本有限，需要找到合适的condainstallpytorch==1.11.0torchvision==0.12.0cudatoolkit=10.2-cpytor
Python(PyTorch)和MATLAB及Rust和C++结构相似度指数测量导图亚图跨际 Python 交叉知识算法量化检查图像压缩质量低分辨率多光谱峰值信噪比端到端优化图像压缩手术机器人三维实景实时可微分渲染重建三维可视化
要点量化检查图像压缩质量低分辨率多光谱和高分辨率图像实现超分辨率分析图像质量图像索引/多尺度结构相似度指数和光谱角映射器及视觉信息保真度多种指标峰值信噪比和结构相似度指数测量结构相似性图像分类PNG和JPEG图像相似性近似算法图像压缩，视频压缩、端到端优化图像压缩、神经图像压缩、GPU变速图像压缩手术机器人深度估计算法重建三维可视化推理图像超分辨率算法模型三维实景实时可微分渲染算法MATLAB结构
【深度学习】训练过程中一个OOM的问题，太难查了 weixin_40293999 深度学习深度学习人工智能
现象：各位大佬又遇到过ubuntu的这个问题么？现象是在训练过程中，ssh上不去了，能ping通，没死机，但是ubunutu的pc侧的显示器，鼠标啥都不好用了。只能重启。问题原因：OOM了95G，尼玛！！！！pytorch爆内存了，然后journald假死了，在journald被watchdog干掉之后，系统就崩溃了。这种规模的爆内存一般，即使被oomkill了，也要卡半天的，确实会这样，能不能配
Pyorch中 nn.Conv1d 与 nn.Linear 的区别迪三 #NN_Layer 神经网络
即一维卷积层和全联接层的区别nn.Conv1d和nn.Linear都是PyTorch中的层，它们用于不同的目的，主要区别在于它们处理输入数据的方式和执行的操作类型。nn.Conv1d通过应用滑动过滤器来捕捉序列数据中的局部模式，适用于处理具有时间或序列结构的数据。nn.Linear通过将每个输入与每个输出相连接，捕捉全局关系，适用于将输入数据作为整体处理的任务。1.维度与输入nn.Conv1d（一
图片中的上采样，下采样和通道融合(up-sample, down-sample, channel confusion) 迪三 #图像处理_PyTorch 计算机视觉深度学习人工智能
前言以conv2d为例（即图片），Pytorch中输入的数据格式为tensor，格式为:[N,C,W,H,W]第一维N.代表图片个数，类似一个batch里面有N张图片第二维C.代表通道数，在模型中输入如果为彩色，常用RGB三色图，那么就是3维，即C=3。如果是黑白的，即灰度图，那么只有一个通道，即C=1第三维H.代表图片的高度，H的数量是图片像素的列数第四维W.代表图片的宽度，W的数量是图片像素的
深度学习：怎么看pth文件的参数奥利给少年深度学习人工智能
.pth文件是PyTorch模型的权重文件，它通常包含了训练好的模型的参数。要查看或使用这个文件，你可以按照以下步骤操作：1.确保你有模型的定义你需要有创建这个.pth文件时所用的模型的代码。这意味着你需要有模型的类定义和架构。2.加载模型权重使用PyTorch的load_state_dict方法来加载权重。这里是如何操作的：importtorchimporttorch.nnasnn#定义模型结构
每天五分钟玩转深度学习PyTorch：模型参数优化器torch.optim 幻风_huanfeng 深度学习框架pytorch 深度学习 pytorch 人工智能神经网络机器学习优化算法
本文重点在机器学习或者深度学习中，我们需要通过修改参数使得损失函数最小化(或最大化)，优化算法就是一种调整模型参数更新的策略。在pytorch中定义了优化器optim，我们可以使用它调用封装好的优化算法，然后传递给它神经网络模型参数，就可以对模型进行优化。本文是学习第6步(优化器)，参考链接pytorch的学习路线随机梯度下降算法在深度学习和机器学习中，梯度下降算法是最常用的参数更新方法，它的公式
【深度学习】【OnnxRuntime】【Python】模型转化、环境搭建以及模型部署的详细教程牙牙要健康深度学习 onnx onnxruntime 深度学习 python 人工智能
【深度学习】【OnnxRuntime】【Python】模型转化、环境搭建以及模型部署的详细教程提示:博主取舍了很多大佬的博文并亲测有效,分享笔记邀大家共同学习讨论文章目录【深度学习】【OnnxRuntime】【Python】模型转化、环境搭建以及模型部署的详细教程前言模型转换--pytorch转onnxWindows平台搭建依赖环境onnxruntime调用onnx模型ONNXRuntime推理核
天下苦英伟达久矣！PyTorch官方免CUDA加速推理，Triton时代要来？诗者才子酒中仙物联网 /互联网 /人工智能 /其他 pytorch 人工智能 python
在做大语言模型（LLM）的训练、微调和推理时，使用英伟达的GPU和CUDA是常见的做法。在更大的机器学习编程与计算范畴，同样严重依赖CUDA，使用它加速的机器学习模型可以实现更大的性能提升。虽然CUDA在加速计算领域占据主导地位，并成为英伟达重要的护城河之一。但其他一些工作的出现正在向CUDA发起挑战，比如OpenAI推出的Triton，它在可用性、内存开销、AI编译器堆栈构建等方面具有一定的优势
pytorch安装(windows) m0_62244898 windows 人工智能
（1）下载pycharmPyCharm:thePythonIDEforProfessionalDevelopersbyJetBrains(2)下载anacondaAnaconda|TheWorld'sMostPopularDataSciencePlatform(3)创建一个新环境：torchcondacreate-ntorch-y(4)进入新环境condaactivatetorch(5)加入清华源
深度学习入门篇：PyTorch实现手写数字识别 AI_Guru人工智能深度学习 pytorch 人工智能
深度学习作为机器学习的一个分支，近年来在图像识别、自然语言处理等领域取得了显著的成就。在众多的深度学习框架中，PyTorch以其动态计算图、易用性强和灵活度高等特点，受到了广泛的喜爱。本篇文章将带领大家使用PyTorch框架，实现一个手写数字识别的基础模型。手写数字识别简介手写数字识别是计算机视觉领域的一个经典问题，目的是让计算机能够识别并理解手写数字图像。这个问题通常作为深度学习入门的练习，因为
【ShuQiHere】小白也能懂的 TensorFlow 和 PyTorch GPU 配置教程 ShuQiHere tensorflow pytorch 人工智能
【ShuQiHere】在深度学习中，GPU的使用对于加速模型训练至关重要。然而，对于许多刚刚入门的小白来说，如何在TensorFlow和PyTorch中指定使用GPU进行训练可能会感到困惑。在本文中，我将详细介绍如何在这两个主流的深度学习框架中指定使用GPU进行训练，并确保每一个步骤都简单易懂，跟着我的步骤来，你也能轻松上手！1.安装所需库首先，确保你已经安装了TensorFlow或PyTorch
解决ModuleNotFoundError: No module named ‘torch的方法梅菊林各种问题解决方案开发语言
ModuleNotFoundError:Nomodulenamed‘torch’错误是Python在尝试导入名为torch的模块时找不到该模块而抛出的异常。torch是PyTorch深度学习框架的核心库，如果你的Python环境中没有安装这个库，尝试导入时就会遇到这个错误。文章目录报错问题报错原因解决方法报错问题当你尝试在Python脚本或交互式环境中执行以下命令时：importtorch如果Py
Python中item()和items()的用处 ~|Bernard| 深度学习疑点总结 python pytorch 深度学习
item()区别一:在pytorch训练时，一般用到.item()。比如loss.item()。我们可以做个简单测试代码看看它的区别:importtorchx=torch.randn(2,2)print(x)print(x[1,1])print(x[1,1].item())运行结果:tensor([[-2.0743,0.1675],[0.7016,-0.6779]])tensor(-0.6779)
GPU版pytorch安装普通攻击往后拉 python tips 神经网络基础模型关键点
由于经常重装系统，导致电脑的环境需要经常重新配置，其中尤其是cudatorch比较难以安装，因此记录一下安装GPU版本torch的过程。1）安装CUDAtoolkit这个可以看做是N卡所有cuda计算的基础，一般都会随驱动的更新自动安装，但是不全，仍然需要安装toolkit，并不需要先看已有版本是哪个，反正下载完后会自动覆盖原有的cuda。下载网站两个：国内网站：只能下载最新的toolkit，但是
轻松升级：Ollama + OpenWebUI 安装与配置【AIStarter】 ai_xiaogui AI作画 AI软件人工智能 AI写作 AIStarter
Ollama是一个开源项目，用于构建和训练大规模语言模型，而OpenWebUI则提供了一个方便的前端界面来管理和监控这些模型。本文将指导你如何更新这两个工具，并顺利完成配置。准备工作确保你的系统已安装Git和Python环境。安装必要的依赖库，如TensorFlow或PyTorch等。更新步骤克隆项目：使用Git命令行工具克隆最新的Ollama和OpenWebUI仓库到本地。更新代码：确保你正在使
conda环境管理 Johnson0722 python python conda 环境管理
Anaconda使用软件包管理系统Conda进行包管理，为用户对不同版本、不同功能的工具包的环境进行配置和管理提供便利。来看一看使用conda来进行环境管理的基本命令创建环境创建一个名为test的python环境，指定python版本是3.7.3，并在test环境中安装pytorchcondacreate--nametestpython=3.7.3pytorch查看系统中的所有环境用户安装的不同环
R-Drop pytorch实现 warpin 深度学习深度学习 pytorch
Pytorch实现了R-Drop，可以用于训练分类模型。#-*-coding:utf-8-*-"""Description:AnimplementationofR-Drop(https://arxiv.org/pdf/2106.14448.pdf).Authors:lihpCreateDate:2021/8/24"""fromtorchimportnnfromtorch.nnimportfunct
Transformer模型：WordEmbedding实现 Galaxy.404 Transformer transformer 深度学习人工智能 embedding
前言最近在学Transformer，学了理论的部分之后就开始学代码的实现，这里是跟着b站的up主的视频记的笔记，视频链接：19、Transformer模型Encoder原理精讲及其PyTorch逐行实现_哔哩哔哩_bilibili正文首先导入所需要的包：importtorchimportnumpyasnpimporttorch.nnasnnimporttorch.nn.functionalasF关
如何使用Pytorch-Metric-Learning？鱼儿也有烦恼 PyTorch pytorch
文章目录如何使用Pytorch-Metric-Learning？1.Pytorch-Metric-Learning库9个模块的功能1.1Sampler模块1.2Miner模块1.3Loss模块1.4Reducer模块1.5Distance模块1.6Regularizer模块1.7Trainer模块1.8Tester模块1.9Utils模块2.如何使用PyTorchMetricLearning库中的
每天五分钟玩转深度学习框架PyTorch：获取神经网络模型的参数幻风_huanfeng 深度学习框架pytorch 深度学习 pytorch 神经网络人工智能模型参数 python
本文重点当我们定义好神经网络之后，这个网络是由多个网络层构成的，每层都有参数，我们如何才能获取到这些参数呢？我们将再下面介绍几个方法来获取神经网络的模型参数，此文我们是为了学习第6步（优化器）。获取所有参数Parametersfromtorchimportnnnet=nn.Sequential(nn.Linear(4,2),nn.Linear(2,2))print(list(net.paramet
一维数组 list 呢，怎么转换成 (批次句子长度特征值 )三维向量 python pytorch lstm 编程人工智能 zhangfeng1133 python pytorch 人工智能数据挖掘
一、介绍对于一维数组，如果你想将其转换成适合深度学习模型（如LSTM）输入的格式，你需要考虑将其扩展为三维张量。这通常涉及到批次大小（batchsize）、序列长度（sequencelength）和特征数量（numberoffeatures）的维度。以下是如何将一维数组转换为这种格式的步骤：###1.确定维度-**批次大小（BatchSize）**：这是你一次处理的样本数量。-**序列长度（Seq
每天五分钟玩转深度学习框架PyTorch：将nn的神经网络层连接起来幻风_huanfeng 深度学习框架pytorch 深度学习 pytorch 神经网络人工智能机器学习 python
本文重点前面我们学习pytorch中已经封装好的神经网络层，有全连接层，激活层，卷积层等等，我们可以直接使用。如代码所示我们直接使用了两个nn.Linear（），这两个linear之间并没有组合在一起，所以forward的之后，分别调用了，在实际使用中我们常常将几个神经层组合在一起，这样不仅操作方便，而且代码清晰。这里介绍一下Sequential()和ModuleList()，它们可以将多个神经网
项目实训十四 qq_51946537 项目实训 python
将pytorch模型封装成接口由于前面对于模型的构建、训练、评估都以完成，接下来要做的就是将按照项目要求，将模型封装成接口，供后端直接调用。我需要做的是后端直接调用系统命令pythonprase.py-img图片便可以直接得到解析结果。由于前面的测试模型的正确率都是批量处理过的图片，而现在前端只会传过来要解析的图片或者图片路径，而且图片也是未经处理过的，显然直接输入不会得到好的结果，并且性能也会比
pytorch矩阵乘法 weixin_45694975 pytorch 深度学习神经网络
一、torch.bmminput1shape:(batch_size,seq1_len,emb_dim)input2shape:(batch_size,emb_dim,seq2_len)outputshape:(batch_size,seq1_len,seq2_len)注意：torch.bmm只适合三维tensor做矩阵运算特别地，torch.bmm支持tenso广播运算input1shape:(
pytorch矩阵乘法总结 chenxi yan PyTorch 学习 pytorch 矩阵深度学习
1.element-wise（*）按元素相乘，支持广播，等价于torch.mul()a=torch.tensor([[1,2],[3,4]])b=torch.tensor([[2,3],[4,5]])c=a*b#等价于torch.mul(a,b)#tensor([[2,6],#[12,20]])a*torch.tensor([1,2])#广播,等价于torch.mul(a,torch.tensor
推荐开源项目：PyTorch-Metric-Learning 潘惟妍
推荐开源项目：PyTorch-Metric-Learningpytorch-metric-learningTheeasiestwaytousedeepmetriclearninginyourapplication.Modular,flexible,andextensible.WritteninPyTorch.项目地址:https://gitcode.com/gh_mirrors/py/pytorc
pytroch2.4 提示到不到fbgemm.dll bziyue python pytorch
#python/pytorch/问题记录```>>>importtorchTraceback(mostrecentcalllast):File"",line1,inFile"C:\Users\95416\AppData\Local\Programs\Python\Python312\Lib\site-packages\torch\__init__.py",line148,inraiseerrOSE
springmvc 下 freemarker页面枚举的遍历输出杨白白 enum freemarker
spring mvc freemarker 中遍历枚举 1枚举类型有一个本地方法叫values（），这个方法可以直接返回枚举数组。所以可以利用这个遍历。 enum public enum BooleanEnum { TRUE(Boolean.TRUE, "是"), FALSE(Boolean.FALSE, "否");
实习简要总结 byalias 工作
来白虹不知不觉中已经一个多月了，因为项目还在需求分析及项目架构阶段，自己在这段时间都是在学习相关技术知识，现在对这段时间的工作及学习情况做一个总结：（1）工作技能方面大体分为两个阶段，Java Web 基础阶段和Java EE阶段 1）Java Web阶段在这个阶段，自己主要着重学习了 JSP, Servlet, JDBC, MySQL，这些知识的核心点都过了一遍，也
Quartz——DateIntervalTrigger触发器 eksliang quartz
转载请出自出处：http://eksliang.iteye.com/blog/2208559 一.概述 simpleTrigger 内部实现机制是通过计算间隔时间来计算下次的执行时间，这就导致他有不适合调度的定时任务。例如我们想每天的 1：00AM 执行任务，如果使用 SimpleTrigger，间隔时间就是一天。注意这里就会有一个问题，即当有 misfired 的任务并且恢复执行时，该执行时间
Unix快捷键 18289753290 unix Unix；快捷键;
复制，删除，粘贴： dd:删除光标所在的行 &nbs
获取Android设备屏幕的相关参数酷的飞上天空 android
包含屏幕的分辨率以及屏幕宽度的最大dp 高度最大dp TextView text = (TextView)findViewById(R.id.text); DisplayMetrics dm = new DisplayMetrics(); text.append("getResources().ge
要做物联网？先保护好你的数据蓝儿唯美数据
根据Beecham Research的说法，那些在行业中希望利用物联网的关键领域需要提供更好的安全性。在Beecham的物联网安全威胁图谱上，展示了那些可能产生内外部攻击并且需要通过快速发展的物联网行业加以解决的关键领域。 Beecham Research的技术主管Jon Howes说：“之所以我们目前还没有看到与物联网相关的严重安全事件，是因为目前还没有在大型客户和企业应用中进行部署，也就
Java取模（求余）运算随便小屋 java
整数之间的取模求余运算很好求，但几乎没有遇到过对负数进行取模求余，直接看下面代码： /** * * @author Logic * */ public class Test { public static void main(String[] args) { // TODO A
SQL注入介绍 aijuans sql注入
二、SQL注入范例这里我们根据用户登录页面 <form action="" > 用户名：<input type="text" name="username"><br/> 密码：<input type="password" name="passwor
优雅代码风格 aoyouzi 代码
总结了几点关于优雅代码风格的描述：代码简单：不隐藏设计者的意图，抽象干净利落，控制语句直截了当。接口清晰：类型接口表现力直白，字面表达含义，API 相互呼应以增强可测试性。依赖项少：依赖关系越少越好，依赖少证明内聚程度高，低耦合利于自动测试，便于重构。没有重复：重复代码意味着某些概念或想法没有在代码中良好的体现，及时重构消除重复。战术分层：代码分层清晰，隔离明确，
布尔数组百合不是茶 java 布尔数组
androi中提到了布尔数组; 布尔数组默认的是false, 并且只会打印false或者是true 布尔数组的例子; 根据字符数组创建布尔数组 char[] c = {'p','u','b','l','i','c'}; //根据字符数组的长度创建布尔数组的个数 boolean[] b = new bool
web.xml之welcome-file-list、error-page bijian1013 java web.xml servlet error-page
welcome-file-list 1.定义： <welcome-file-list> <welcome-file>login.jsp</welcome> </welcome-file-list> 2.作用：用来指定WEB应用首页名称。 error-page1.定义： <error-page&g
richfaces 4 fileUpload组件删除上传的文件 sunjing clear Richfaces 4 fileupload
页面代码 <h:form id="fileForm"> <rich:
技术文章备忘 bit1129 技术文章
Zookeeper http://wenku.baidu.com/view/bab171ffaef8941ea76e05b8.html http://wenku.baidu.com/link?url=8thAIwFTnPh2KL2b0p1V7XSgmF9ZEFgw4V_MkIpA9j8BX2rDQMPgK5l3wcs9oBTxeekOnm5P3BK8c6K2DWynq9nfUCkRlTt9uV
org.hibernate.hql.ast.QuerySyntaxException: unexpected token: on near line 1解决方案白糖_ Hibernate
文章摘自：http://blog.csdn.net/yangwawa19870921/article/details/7553181 在编写HQL时，可能会出现这种代码： select a.name,b.age from TableA a left join TableB b on a.id=b.id 如果这是HQL，那么这段代码就是错误的，因为HQL不支持
sqlserver按照字段内容进行排序 bozch 按照内容排序
在做项目的时候，遇到了这样的一个需求：从数据库中取出的数据集，首先要将某个数据或者多个数据按照地段内容放到前面显示，例如:从学生表中取出姓李的放到数据集的前面； select * fro
编程珠玑-第一章-位图排序 bylijinnan java 编程珠玑
import java.io.BufferedWriter; import java.io.File; import java.io.FileWriter; import java.io.IOException; import java.io.Writer; import java.util.Random; public class BitMapSearch {
Java关于==和equals chenbowen00 java
关于==和equals概念其实很简单，一个是比较内存地址是否相同，一个比较的是值内容是否相同。虽然理解上不难，但是有时存在一些理解误区，如下情况： 1、 String a = "aaa"; a=="aaa"; ==> true 2、 new String("aaa")==new String("aaa
[IT与资本]软件行业需对外界投资热情保持警惕 comsci it
我还是那个看法,软件行业需要增强内生动力,尽量依靠自有资金和营业收入来进行经营,避免在资本市场上经受各种不同类型的风险,为企业自主研发核心技术和产品提供稳定,温和的外部环境... 如果我们在自己尚未掌握核心技术之前,企图依靠上市来筹集资金,然后使劲往某个领域砸钱,然
oracle 数据块结构 daizj oracle 块数据块块结构行目录
oracle 数据块是数据库存储的最小单位，一般为操作系统块的N倍。其结构为：块头－－〉空行－－〉数据，其实际为纵行结构。块的标准大小由初始化参数DB_BLOCK_SIZE指定。具有标准大小的块称为标准块（Standard Block）。块的大小和标准块的大小不同的块叫非标准块（Nonstandard Block）。同一数据库中，Oracle9i及以上版本支持同一数据库中同时使用标
github上一些觉得对自己工作有用的项目收集 dengkane github
github上一些觉得对自己工作有用的项目收集技能类 markdown语法中文说明回到顶部全文检索 elasticsearch bigdesk elasticsearch管理插件回到顶部 nosql mapdb 支持亿级别map, list, 支持事务. 可考虑做为缓存使用 C
初二上学期难记单词二 dcj3sjt126com english word
dangerous 危险的 panda 熊猫 lion 狮子 elephant 象 monkey 猴子 tiger 老虎 deer 鹿 snake 蛇 rabbit 兔子 duck 鸭 horse 马 forest 森林 fall 跌倒；落下 climb 爬；攀登 finish 完成；结束 cinema 电影院；电影 seafood 海鲜；海产食品 bank 银行
8、mysql外键(FOREIGN KEY)的简单使用 dcj3sjt126com mysql
一、基本概念 1、MySQL中“键”和“索引”的定义相同，所以外键和主键一样也是索引的一种。不同的是MySQL会自动为所有表的主键进行索引，但是外键字段必须由用户进行明确的索引。用于外键关系的字段必须在所有的参照表中进行明确地索引，InnoDB不能自动地创建索引。 2、外键可以是一对一的，一个表的记录只能与另一个表的一条记录连接，或者是一对多的，一个表的记录与另一个表的多条记录连接。 3、如
java循环标签 Foreach shuizhaosi888 标签 java循环 foreach
1. 简单的for循环 public static void main(String[] args) { for (int i = 1, y = i + 10; i < 5 && y < 12; i++, y = i * 2) { System.err.println("i=" + i + " y="
Spring Security（05）——异常信息本地化 234390216 exception Spring Security 异常信息本地化
异常信息本地化 Spring Security支持将展现给终端用户看的异常信息本地化，这些信息包括认证失败、访问被拒绝等。而对于展现给开发者看的异常信息和日志信息（如配置错误）则是不能够进行本地化的，它们是以英文硬编码在Spring Security的代码中的。在Spring-Security-core-x
DUBBO架构服务端告警Failed to send message Response javamingtingzhao 架构 DUBBO
废话不多说，警告日志如下，不知道有哪位遇到过，此异常在服务端抛出(服务器启动第一次运行会有这个警告)，后续运行没问题，找了好久真心不知道哪里错了。 WARN 2015-07-18 22:31:15,272 com.alibaba.dubbo.remoting.transport.dispatcher.ChannelEventRunnable.run(84)
JS中Date对象中几个用法 leeqq JavaScript Date 最后一天
近来工作中遇到这样的两个需求 1. 给个Date对象，找出该时间所在月的第一天和最后一天 2. 给个Date对象，找出该时间所在周的第一天和最后一天需求1中的找月第一天很简单，我记得api中有setDate方法可以使用使用setDate方法前，先看看getDate var date = new Date(); console.log(date); // Sat J
MFC中使用ado技术操作数据库你不认识的休道人 sql mfc
1.在stdafx.h中导入ado动态链接库 #import"C:\Program Files\Common Files\System\ado\msado15.dll" no_namespace rename("EOF","end")2.在CTestApp文件的InitInstance()函数中domodal之前写::CoIniti
Android Studio加速 rensanning android studio
Android Studio慢、吃内存！启动时后会立即通过Gradle来sync & build工程。（1）设置Android Studio a) 禁用插件 File -> Settings... Plugins 去掉一些没有用的插件。比如：Git Integration、GitHub、Google Cloud Testing、Google Cloud
各数据库的批量Update操作 tomcat_oracle java oracle sql mysql sqlite
MyBatis的update元素的用法与insert元素基本相同，因此本篇不打算重复了。本篇仅记录批量update操作的 sql语句，懂得SQL语句，那么MyBatis部分的操作就简单了。　　注意：下列批量更新语句都是作为一个事务整体执行，要不全部成功，要不全部回滚。 MSSQL的SQL语句　WITH R AS（　　SELECT 'John' as name, 18 as
html禁止清除input文本输入缓存 xp9802 input
多数浏览器默认会缓存input的值，只有使用ctl+F5强制刷新的才可以清除缓存记录。如果不想让浏览器缓存input的值，有2种方法：方法一：在不想使用缓存的input中添加 autocomplete="off"; eg: <input type="text" autocomplete="off" name