MCPcopy Create free account
hub / github.com/apache/singa / _encoding_stc

Method _encoding_stc

examples/trans/data.py:127–145  ·  view source on GitHub ↗
(src_tokens, tgt_tokens, src_max_len, tgt_max_len, src_vocab, tgt_vocab)

Source from the content-addressed store, hash-verified

125
126 @staticmethod
127 def _encoding_stc(src_tokens, tgt_tokens, src_max_len, tgt_max_len, src_vocab, tgt_vocab):
128 src_list = []
129 for line in src_tokens:
130 if len(line) > src_max_len:
131 line = line[:src_max_len]
132 lst = src_vocab[line + ['<pad>'] * (src_max_len + 1 - len(line))]
133 src_list.append(lst)
134 tgt_in_list, tgt_out_list = [], []
135 for line in tgt_tokens:
136 if len(line) > tgt_max_len:
137 line = line[:tgt_max_len]
138 in_lst = tgt_vocab[['<bos>'] + line + ['<pad>'] * (tgt_max_len - len(line))]
139 out_lst = tgt_vocab[line + ['<eos>'] + ['<pad>'] * (tgt_max_len - len(line))]
140 tgt_in_list.append(in_lst)
141 tgt_out_list.append(out_lst)
142 src_np = np.asarray(src_list, dtype=np.int32)
143 tgt_in_np = np.asarray(tgt_in_list, dtype=np.int32)
144 tgt_out_np = np.asarray(tgt_out_list, dtype=np.int32)
145 return src_np, tgt_in_np, tgt_out_np
146
147 def get_batch_data(self, batch, mode='train'):
148 assert (mode == 'train' or mode == 'test'), "The mode must be 'train' or 'test'."

Callers 1

__init__Method · 0.80

Calls 1

appendMethod · 0.80

Tested by

no test coverage detected