Train_loss与Test_loss的可视化

0.可视化

from matplotlib import pyplot as plt          # 导入需要的包
    
# Train
x_train = range(1, config.num_epochs + 1)
y_train = train_loss_list                     # 每一轮训练集损失平均值                       

# Test
x_test = range(1, config.num_epochs + 1)
y_test = test_loss_list                       # 每一轮测试集损失平均值                 

# 创建一个画布
plt.figure()

# 绘制训练集
plt.plot(x_train, y_train, 'o-', label="Train_loss")
# 绘制测试集
plt.plot(x_test, y_test, 'o-', label="Test_loss")

# 绘制标题及标签
plt.title('Train and Test Loss')
plt.xlabel('Epoch')
plt.ylabel('Loss')

# 设置x轴和y轴刻度值
y_ticks = np.arange(0, 0.2, 1)                # 0-1，0.2为间隔
x_ticks = np.arange(1, 1, config.num_epochs)  # 1-epochs+1, 1为间隔
plt.xticks(x_ticks)
plt.yticks(y_ticks)

# 绘制图例
plt.legend()
# 保存图片
plt.savefig(Train_and_Test_Loss.jpg")
# 显示图片
plt.show()

1.可视化结果样例