A good name is all you need

presentation : https://drive.google.com/file/d/1-9wwoZZPWz_EVhi28o-vo9GJOOkK69wL/view

example

• way 1

    input:鍾 target:佳 
    input:佳 target:紋 
    input:紋 target:。 
  

• way 2

    input:鍾 target:鍾佳
    input:鍾佳 target:鍾佳紋
    input:鍾佳紋 target:鍾佳紋。
  

model

class RNN(nn.Module):
    def __init__(self, input_size, hidden_size, output_size):
        super(RNN, self).__init__()
        self.input_size = input_size
        self.hidden_size = hidden_size
        self.output_size = output_size
        self.gru = nn.GRU(self.input_size, self.hidden_size,dropout = 0.5)
        self.h2o = nn.Linear(self.hidden_size, self.output_size)
        self.softmax = nn.LogSoftmax() 
    def forward(self, input):
        out,_ = self.gru(input)
        out = self.h2o(out).squeeze(0).squeeze(0)
        output = self.softmax(out)
        return output
    def initHidden(self):
        return torch.zeros(1, self.hidden_size)

char embedding

import pickle
f=  open ('./char_embedding.pkl', 'rb')
dict_char = pickle.load(f)
f.close()

data

crawler data from https://findbiz.nat.gov.tw/fts/query/QueryBar/queryInit.do;jsessionid=F906D81F3071A0B848AC0633D47165D7