python-pytorchHow can I use a Recurrent Neural Network (RNN) in Python with PyTorch?

You can use a Recurrent Neural Network (RNN) in Python with PyTorch by following these steps:

1. Import the necessary PyTorch modules:

import torch
import torch.nn as nn

2. Define the RNN model:

class RNN(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(RNN, self).__init__()
        self.hidden_size = hidden_size
        self.i2h = nn.Linear(input_size + hidden_size, hidden_size)
        self.i2o = nn.Linear(input_size + hidden_size, output_size)
        self.softmax = nn.LogSoftmax(dim=1)

    def forward(self, input, hidden):
        combined = torch.cat((input, hidden), 1)
        hidden = self.i2h(combined)
        output = self.i2o(combined)
        output = self.softmax(output)
        return output, hidden

3. Initialize the RNN model:

n_hidden = 128
rnn = RNN(n_letters, n_hidden, n_categories)

4. Define a loss function:

criterion = nn.NLLLoss()

5. Train the model:

for epoch in range(n_epochs):
    # ...
    output, hidden = rnn(category_tensor, input_line_tensor)
    loss = criterion(output, target_line_tensor)
    # ...
    loss.backward()
    # ...

6. Test the model:

# Get the most likely category
_, predicted = torch.max(output, 1)

7. Evaluate the model:

# Calculate the accuracy
correct_count = (predicted == target_line_tensor).sum().item()
accuracy = correct_count / n_test_lines

Helpful links

• PyTorch Tutorials
• PyTorch RNN Tutorial

Edit this code on GitHub