Question:
Name data is not defined. How do I define it? Ghostwriter made this code and I put it verbatim into my Repl. I then did pip install torch. Starting on line 64.
Repl link:
https://replit.com/@CoderElijah/ai#main.py
Code
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
# define the neural network architecture
class Net(nn.Module):
def __init__(self, vocab_size, embedding_dim, hidden_dim):
super(Net, self).__init__()
self.embedding = nn.Embedding(vocab_size, embedding_dim)
self.lstm = nn.LSTM(embedding_dim, hidden_dim)
self.fc = nn.Linear(hidden_dim, vocab_size)
def forward(self, x, hidden):
x = self.embedding(x)
x, hidden = self.lstm(x.unsqueeze(1), hidden)
x = self.fc(x.squeeze(1))
return x, hidden
# define the training function
def train(model, iterator, optimizer, criterion, device):
model.train()
hidden = (torch.zeros(1, 1, model.hidden_dim).to(device),
torch.zeros(1, 1, model.hidden_dim).to(device))
for batch in iterator:
optimizer.zero_grad()
text = batch.text.to(device)
target = batch.target.to(device)
output, hidden = model(text, hidden)
loss = criterion(output.view(-1, model.vocab_size), target.view(-1))
loss.backward()
optimizer.step()
# define the testing function
def test(model, iterator, criterion, device):
model.eval()
hidden = (torch.zeros(1, 1, model.hidden_dim).to(device),
torch.zeros(1, 1, model.hidden_dim).to(device))
with torch.no_grad():
loss = 0
for batch in iterator:
text = batch.text.to(device)
target = batch.target.to(device)
output, hidden = model(text, hidden)
loss += criterion(output.view(-1, model.vocab_size), target.view(-1)).item()
loss /= len(iterator)
return loss
# define the main function
def main():
# set up the hyperparameters
train_file = "train.txt"
test_file = "test.txt"
batch_size = 16
embedding_dim = 128
hidden_dim = 256
num_epochs = 10
lr = 0.001
# set up the device
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# set up the field and the dataset
TEXT = data.Field(tokenize=lambda x: x.split(), lower=True)
train_data, test_data = data.TabularDataset.splits(
path='', train=train_file, test=test_file, format='csv',
fields=[('text', TEXT), ('target', TEXT)])
TEXT.build_vocab(train_data)
# set up the iterators
train_iterator, test_iterator = data.BucketIterator.splits(
(train_data, test_data), batch_size=batch_size, device=device)
# set up the model, criterion and optimizer
model = Net(len(TEXT.vocab), embedding_dim, hidden_dim).to(device)
criterion = nn.CrossEntropyLoss()
optimizer = optim.Adam(model.parameters(), lr=lr)
# train the model
for epoch in range(num_epochs):
train(model, train_iterator, optimizer, criterion, device)
train_loss = test(model, train_iterator, criterion, device)
test_loss = test(model, test_iterator, criterion, device)
print(f'Epoch: {epoch+1:02} | Train Loss: {train_loss:.3f} | Test Loss: {test_loss:.3f}')
if __name__ == '__main__':
main()
