def forward(self, input_seq):
lstm_out, self.hidden_cell = self.lstm(
input_seq.view(len(input_seq), 1, -1), self.hidden_cell)
predictions = self.linear(lstm_out.view(len(input_seq), -1))
return predictions[-1]
epochs = 50
model = LSTM(1, 100, 1)
loss = nn.MSELoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.001)
for i in range(epochs + 1):
for seq, label in train_inout_seq:
optimizer.zero_grad()
model.hidden_cell = (torch.zeros(1, 1, model.hidden_layer),
torch.zeros(1, 1, model.hidden_layer))
predictions = model(seq)
single_loss = loss(predictions, label)
single_loss.backward()
optimizer.step()
if i % 10 == 0:
print("Epoch: {0} Loss: {1:10.10f}".format(i, single_loss))
to_pred = 9128
test_inputs = target_norm[-train_window:].tolist()
for i in range(to_pred):
seq = torch.FloatTensor(test_inputs[-train_window:])
with torch.no_grad():
model.hidden = (torch.zeros(1, 1, model.hidden_layer),
