【深度学习实战（8）】如何绘制loss曲线图

news2024/12/24 2:31:10

一、步骤

我们先定义一个dict，每一个key对应的value都是一个list。

loss_history = dict((k, []) for k in ["epoch", "train_loss", "val_loss"])

每一轮或者每一次迭代的损失都通过list记录下来。

loss_history["epoch"].append(1)
loss_history["train_loss"].append(0.1)
loss_history["val_loss"].append(0.1)

loss_history["epoch"].append(2)
loss_history["train_loss"].append(0.05)
loss_history["val_loss"].append(0.05)

使用matplotlib.pyplot工具来绘制曲线图

plt.figure()
plt.title('loss during training')  #标题
plt.plot(loss_history["epoch"], loss_history["train_loss"], label="train_loss")
plt.plot(loss_history["epoch"], loss_history["val_loss"], label="valid_loss")
plt.legend()
plt.grid()
plt.show()
plt.savefig('loss figure')

二、完整代码

import matplotlib.pyplot as plt


# dict + list
loss_history = dict((k, []) for k in ["epoch", "train_loss", "val_loss"])


loss_history["epoch"].append(1)
loss_history["train_loss"].append(0.1)
loss_history["val_loss"].append(0.1)

loss_history["epoch"].append(2)
loss_history["train_loss"].append(0.05)
loss_history["val_loss"].append(0.05)

plt.figure()
plt.title('loss during training')  #标题
plt.plot(loss_history["epoch"], loss_history["train_loss"], label="train_loss")
plt.plot(loss_history["epoch"], loss_history["val_loss"], label="valid_loss")
plt.legend()
plt.grid()
plt.show()
plt.savefig('loss figure')