Fix hidden initialization

[nihilistsumo]

The PyTorch/main.py was throwing the following error:
Traceback (most recent call last):
File "main.py", line 63, in
run_iterations.train_iters()
File "/home/sumanta/Manhattan-LSTM/PyTorch/run_iterations.py", line 58, in train_iters
loss, _ = self.model.train(input_variables, similarity_scores, self.criterion, model_optimizer)
File "/home/sumanta/Manhattan-LSTM/PyTorch/train_network.py", line 32, in train
output_scores = self.manhattan_lstm((sequences_1, sequences_2), hidden).view(-1)
File "/home/sumanta/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/home/sumanta/Manhattan-LSTM/PyTorch/manhattan_lstm.py", line 42, in forward
outputs_1, hidden_1 = self.lstm_1(embedded_1, hidden)
File "/home/sumanta/.local/lib/python3.6/site-packages/torch/nn/modules/module.py", line 532, in call
result = self.forward(*input, **kwargs)
File "/home/sumanta/.local/lib/python3.6/site-packages/torch/nn/modules/rnn.py", line 559, in forward
self.dropout, self.training, self.bidirectional, self.batch_first)
TypeError: lstm() received an invalid combination of arguments - got (Tensor, Tensor, list, bool, int, float, bool, bool, bool), but expected one of:

• (Tensor data, Tensor batch_sizes, tuple of Tensors hx, tuple of Tensors params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional)
  didn't match because some of the arguments have invalid types: (Tensor, Tensor, list, bool, int, float, bool, bool, bool)
• (Tensor input, tuple of Tensors hx, tuple of Tensors params, bool has_biases, int num_layers, float dropout, bool train, bool bidirectional, bool batch_first)
  didn't match because some of the arguments have invalid types: (Tensor, Tensor, list, bool, int, float, bool, bool, bool)

The root cause was erroneous hidden initialization.