Remove uniform initialisation

Instead of uniformly initialising all parameters, don't specify any
initialisation and keep the default module-specific intialisations
provided by pytorch.

TODO: implement more sophisticated intialisation methods.

train_model.py

@@ -221,9 +221,6 @@ def initialize_model(opt, src, tgt, train):
     seq2seq = Seq2seq(encoder, decoder)
     seq2seq.to(device)
 
-    for param in seq2seq.parameters():
-        param.data.uniform_(-0.08, 0.08)
-
     return seq2seq, input_vocab, output_vocab