Hey guys!
I’m currently building a bidirectional LSTM for text classification. I’m pretty new to coding with pytorch and I’m wondering which layers I need for the bi-LSTM to work correctly. It’s currently running without errors, but I wonder if the bidirectional part is also working, or if it’s still a normal LSTM:
class LSTM(nn.Module):
def __init__(self, embedding_dim, hidden_dim, vocab_size, label_size, batch_size):
super(LSTM, self).__init__()
self.hidden_dim = hidden_dim
self.batch_size = batch_size
self.word_embeddings = nn.Embedding(vocab_size, embedding_dim)
self.lstm = nn.LSTM(embedding_dim, hidden_dim, bidirectional=True)
self.hidden2label = nn.Linear(hidden_dim * 2, label_size) # Ä: hidden_dim * 2 for bidirectional
self.hidden = self.init_hidden()
def init_hidden(self):
h0 = Variable(torch.zeros(1 * 2, self.batch_size, self.hidden_dim))
c0 = Variable(torch.zeros(1 * 2, self.batch_size, self.hidden_dim))
return h0, c0
def forward(self, sentence):
embeds = self.word_embeddings(sentence)
x = embeds.view(len(sentence), self.batch_size, -1)
lstm_out, self.hidden = self.lstm(x, self.hidden)
y = self.hidden2label(lstm_out[-1])
return y
Does bidirectional=True do anything by itself? Or do I need to add another backward_lstm for doing the same thing, but the other way around?
And if so, how can I feed the backward layer the input in reverse?
Sorry I’m pretty new to all this stuff and I’m sometimes really unsure about what is doing on in pytorch (._.)
Thank you!