广西民族大学高级人工智能课程—头歌实践教学实践平台-RNN 基础

本文链接：https://blog.csdn.net/gxmzuai/article/details/135015984

本文介绍了如何使用 PyTorch 构建和训练基本的 RNN 模型，涵盖 RNN 的发展历史、原理以及如何在 PyTorch 中实现。通过预测正弦曲线的下一个值来展示 RNN 的应用。

摘要由CSDN通过智能技术生成

代码文件

import random
import torch
import numpy as np
import matplotlib.pyplot as plt
import warnings


warnings.filterwarnings("ignore")

# 数据生成
random.seed(42)
np.random.seed(42)
torch.manual_seed(42)

x = np.linspace(-10, 10, 1000)
y = np.sin(x) + np.random.uniform(-0.5, 0.5, 1000)

# 将numpy数组转换为torch张量
x = torch.tensor(x).float().view(-1, 1)
y = torch.tensor(y).float().view(-1, 1)

# 定义RNN模型
class RNN(torch.nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(RNN, self).__init__()
        self.hidden_size = hidden_size
        self.rnn = torch.nn.RNN(input_size, hidden_size, batch_first=True)
        self.linear = torch.nn.Linear(hidden_size, output_size)

    def forward(self, x, hidden):
        out, hidden = self.rnn(x, hidden)
        out = self.linear(out[:, -1, :])
        return out, hidden


# 初始化模型、优化器和损失函数
################### Begin #############################
model = RNN(1, 32, 1)
optimizer = torch.optim.Adam(model.parameters(), lr=0.01)
criterion = torch.nn.MSELoss()


###################  End  #############################

# 开始训练
seq_len = 10
for epoch in range(21):
    for i in range(x.size(0) - seq_len):
        inputs = x[i:i+seq_len]
        targets = y[i+1:i+seq_len+1]
        hidden = torch.zeros(1, 1, 32)
        output, hidden = model(inputs.unsqueeze(0), hidden)
        loss = criterion(output, targets)
        optimizer.zero_grad()
        loss.backward()
        optimizer.step()
    if (epoch+1) % 5 == 0:
        print('Epoch: {}, Loss: {:.4f}'.format(epoch+1, loss.item()))


# 评估
predictions = []
hidden = torch.zeros(1, 1, 32) 
for i in range(x.size(0) - 1):
    input1 = x[i:i+1]
    output, hidden = model(input1.unsqueeze(0), hidden)
    predictions.append(output.detach().numpy())

predictions = torch.tensor(predictions).float().view(-1, 1)
print("predictions:", predictions)

# 我们可以通过绘制预测值和真实值的曲线来直观地看到预测效果
plt.plot(x.numpy(), y.numpy(), label='true'