Simple way to reproduce Keras default initialisation in a typical pyTorch NLP model

pytorch_weight_initialization.py

def init_weights(self):
    """
    Here we reproduce Keras default initialization weights to initialize Embeddings/LSTM weights
    """
    ih = (param.data for name, param in self.named_parameters() if 'weight_ih' in name)
    hh = (param.data for name, param in self.named_parameters() if 'weight_hh' in name)
    b = (param.data for name, param in self.named_parameters() if 'bias' in name)
    nn.init.uniform(self.embed.weight.data, a=-0.5, b=0.5)
    for t in ih:
        nn.init.xavier_uniform(t)
    for t in hh:
        nn.init.orthogonal(t)
    for t in b:
        nn.init.constant(t, 0)