File size: 12,460 Bytes
0b8359d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
# Copyright 2017 The TensorFlow Authors All Rights Reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# ==============================================================================

import tensorflow as tf
from util.nn import fc_layer as fc


def _get_lstm_cell(num_layers, lstm_dim):
  cell_list = [
      tf.contrib.rnn.BasicLSTMCell(lstm_dim, state_is_tuple=True)
      for _ in range(num_layers)
  ]
  cell = tf.contrib.rnn.MultiRNNCell(cell_list, state_is_tuple=True)
  return cell


class AttentionSeq2Seq:

  def __init__(self,
               config,
               text_seq_batch,
               seq_length_batch,
               num_vocab_txt,
               num_vocab_nmn,
               EOS_token,
               decoder_sampling,
               embedding_mat,
               use_gt_layout=None,
               gt_layout_batch=None,
               scope='encoder_decoder',
               reuse=None):
    self.T_decoder = config.T_decoder
    self.encoder_num_vocab = num_vocab_txt
    self.encoder_embed_dim = config.embed_dim_txt
    self.decoder_num_vocab = num_vocab_nmn
    self.decoder_embed_dim = config.embed_dim_nmn
    self.lstm_dim = config.lstm_dim
    self.num_layers = config.num_layers
    self.EOS_token = EOS_token
    self.decoder_sampling = decoder_sampling
    self.embedding_mat = embedding_mat

    with tf.variable_scope(scope, reuse=reuse):
      self._build_encoder(text_seq_batch, seq_length_batch)
      self._build_decoder(use_gt_layout, gt_layout_batch)

  def _build_encoder(self,
                     text_seq_batch,
                     seq_length_batch,
                     scope='encoder',
                     reuse=None):
    lstm_dim = self.lstm_dim
    num_layers = self.num_layers

    with tf.variable_scope(scope, reuse=reuse):
      T = tf.shape(text_seq_batch)[0]
      N = tf.shape(text_seq_batch)[1]
      self.T_encoder = T
      self.N = N

      # text_seq has shape [T, N] and embedded_seq has shape [T, N, D]
      embedded_seq = tf.nn.embedding_lookup(self.embedding_mat, text_seq_batch)
      self.embedded_input_seq = embedded_seq

      # The RNN
      cell = _get_lstm_cell(num_layers, lstm_dim)

      # encoder_outputs has shape [T, N, lstm_dim]
      encoder_outputs, encoder_states = tf.nn.dynamic_rnn(
          cell,
          embedded_seq,
          seq_length_batch,
          dtype=tf.float32,
          time_major=True,
          scope='lstm')
      self.encoder_outputs = encoder_outputs
      self.encoder_states = encoder_states

      # transform the encoder outputs for further attention alignments
      # encoder_outputs_flat has shape [T, N, lstm_dim]
      encoder_h_transformed = fc(
          'encoder_h_transform',
          tf.reshape(encoder_outputs, [-1, lstm_dim]),
          output_dim=lstm_dim)
      encoder_h_transformed = tf.reshape(encoder_h_transformed,
                                         [T, N, lstm_dim])
      self.encoder_h_transformed = encoder_h_transformed

      # seq_not_finished is a shape [T, N, 1] tensor,
      # where seq_not_finished[t, n]
      # is 1 iff sequence n is not finished at time t, and 0 otherwise
      seq_not_finished = tf.less(
          tf.range(T)[:, tf.newaxis, tf.newaxis],
          seq_length_batch[:, tf.newaxis])
      seq_not_finished = tf.cast(seq_not_finished, tf.float32)
      self.seq_not_finished = seq_not_finished

  def _build_decoder(self,
                     use_gt_layout,
                     gt_layout_batch,
                     scope='decoder',
                     reuse=None):
    # The main difference from before is that the decoders now takes another
    # input (the attention) when computing the next step
    # T_max is the maximum length of decoded sequence (including <eos>)
    #
    # This function is for decoding only. It performs greedy search or sampling.
    # the first input is <go> (its embedding vector) and the subsequent inputs
    # are the outputs from previous time step
    # num_vocab does not include <go>
    #
    # use_gt_layout is None or a bool tensor, and gt_layout_batch is a tensor
    # with shape [T_max, N].
    # If use_gt_layout is not None, then when use_gt_layout is true, predict
    # exactly the tokens in gt_layout_batch, regardless of actual probability.
    # Otherwise, if sampling is True, sample from the token probability
    # If sampling is False, do greedy decoding (beam size 1)
    N = self.N
    encoder_states = self.encoder_states
    T_max = self.T_decoder
    lstm_dim = self.lstm_dim
    num_layers = self.num_layers
    EOS_token = self.EOS_token
    sampling = self.decoder_sampling

    with tf.variable_scope(scope, reuse=reuse):
      embedding_mat = tf.get_variable(
          'embedding_mat', [self.decoder_num_vocab, self.decoder_embed_dim])
      # we use a separate embedding for <go>, as it is only used in the
      # beginning of the sequence
      go_embedding = tf.get_variable('go_embedding',
                                     [1, self.decoder_embed_dim])

      with tf.variable_scope('att_prediction'):
        v = tf.get_variable('v', [lstm_dim])
        W_a = tf.get_variable(
            'weights', [lstm_dim, lstm_dim],
            initializer=tf.contrib.layers.xavier_initializer())
        b_a = tf.get_variable(
            'biases', lstm_dim, initializer=tf.constant_initializer(0.))

      # The parameters to predict the next token
      with tf.variable_scope('token_prediction'):
        W_y = tf.get_variable(
            'weights', [lstm_dim * 2, self.decoder_num_vocab],
            initializer=tf.contrib.layers.xavier_initializer())
        b_y = tf.get_variable(
            'biases',
            self.decoder_num_vocab,
            initializer=tf.constant_initializer(0.))

      # Attentional decoding
      # Loop function is called at time t BEFORE the cell execution at time t,
      # and its next_input is used as the input at time t (not t+1)
      # c.f. https://www.tensorflow.org/api_docs/python/tf/nn/raw_rnn
      mask_range = tf.reshape(
          tf.range(self.decoder_num_vocab, dtype=tf.int32), [1, -1])
      all_eos_pred = EOS_token * tf.ones([N], tf.int32)
      all_one_prob = tf.ones([N], tf.float32)
      all_zero_entropy = tf.zeros([N], tf.float32)
      if use_gt_layout is not None:
        gt_layout_mult = tf.cast(use_gt_layout, tf.int32)
        pred_layout_mult = 1 - gt_layout_mult

      def loop_fn(time, cell_output, cell_state, loop_state):
        if cell_output is None:  # time == 0
          next_cell_state = encoder_states
          next_input = tf.tile(go_embedding, [N, 1])
        else:  # time > 0
          next_cell_state = cell_state

          # compute the attention map over the input sequence
          # a_raw has shape [T, N, 1]
          att_raw = tf.reduce_sum(
              tf.tanh(
                  tf.nn.xw_plus_b(cell_output, W_a, b_a) +
                  self.encoder_h_transformed) * v,
              axis=2,
              keep_dims=True)
          # softmax along the first dimension (T) over not finished examples
          # att has shape [T, N, 1]
          att = tf.nn.softmax(att_raw, dim=0) * self.seq_not_finished
          att = att / tf.reduce_sum(att, axis=0, keep_dims=True)
          # d has shape [N, lstm_dim]
          d2 = tf.reduce_sum(att * self.encoder_outputs, axis=0)

          # token_scores has shape [N, num_vocab]
          token_scores = tf.nn.xw_plus_b(
              tf.concat([cell_output, d2], axis=1), W_y, b_y)
          # predict the next token (behavior depending on parameters)
          if sampling:
            # predicted_token has shape [N]
            logits = token_scores
            predicted_token = tf.cast(
                tf.reshape(tf.multinomial(token_scores, 1), [-1]), tf.int32)
          else:
            # predicted_token has shape [N]
            predicted_token = tf.cast(tf.argmax(token_scores, 1), tf.int32)
          if use_gt_layout is not None:
            predicted_token = (gt_layout_batch[time - 1] * gt_layout_mult +
                               predicted_token * pred_layout_mult)

          # token_prob has shape [N], the probability of the predicted token
          # although token_prob is not needed for predicting the next token
          # it is needed in output (for policy gradient training)
          # [N, num_vocab]
          # mask has shape [N, num_vocab]
          mask = tf.equal(mask_range, tf.reshape(predicted_token, [-1, 1]))
          all_token_probs = tf.nn.softmax(token_scores)
          token_prob = tf.reduce_sum(
              all_token_probs * tf.cast(mask, tf.float32), axis=1)
          neg_entropy = tf.reduce_sum(
              all_token_probs * tf.log(all_token_probs), axis=1)

          # is_eos_predicted is a [N] bool tensor, indicating whether
          # <eos> has already been predicted previously in each sequence
          is_eos_predicted = loop_state[2]
          predicted_token_old = predicted_token
          # if <eos> has already been predicted, now predict <eos> with
          # prob 1
          predicted_token = tf.where(is_eos_predicted, all_eos_pred,
                                     predicted_token)
          token_prob = tf.where(is_eos_predicted, all_one_prob, token_prob)
          neg_entropy = tf.where(is_eos_predicted, all_zero_entropy,
                                 neg_entropy)
          is_eos_predicted = tf.logical_or(is_eos_predicted,
                                           tf.equal(predicted_token_old,
                                                    EOS_token))

          # the prediction is from the cell output of the last step
          # timestep (t-1), feed it as input into timestep t
          next_input = tf.nn.embedding_lookup(embedding_mat, predicted_token)

        elements_finished = tf.greater_equal(time, T_max)

        # loop_state is a 5-tuple, representing
        #   1) the predicted_tokens
        #   2) the prob of predicted_tokens
        #   3) whether <eos> has already been predicted
        #   4) the negative entropy of policy (accumulated across timesteps)
        #   5) the attention
        if loop_state is None:  # time == 0
          # Write the predicted token into the output
          predicted_token_array = tf.TensorArray(
              dtype=tf.int32, size=T_max, infer_shape=False)
          token_prob_array = tf.TensorArray(
              dtype=tf.float32, size=T_max, infer_shape=False)
          att_array = tf.TensorArray(
              dtype=tf.float32, size=T_max, infer_shape=False)
          next_loop_state = (predicted_token_array, token_prob_array, tf.zeros(
              [N], dtype=tf.bool), tf.zeros([N], dtype=tf.float32), att_array)
        else:  # time > 0
          t_write = time - 1
          next_loop_state = (
              loop_state[0].write(t_write, predicted_token),
              loop_state[1].write(t_write, token_prob),
              is_eos_predicted,
              loop_state[3] + neg_entropy,
              loop_state[4].write(t_write, att))
        return (elements_finished, next_input, next_cell_state, cell_output,
                next_loop_state)

      # The RNN
      cell = _get_lstm_cell(num_layers, lstm_dim)
      _, _, decodes_ta = tf.nn.raw_rnn(cell, loop_fn, scope='lstm')
      predicted_tokens = decodes_ta[0].stack()
      token_probs = decodes_ta[1].stack()
      neg_entropy = decodes_ta[3]
      # atts has shape [T_decoder, T_encoder, N, 1]
      atts = decodes_ta[4].stack()
      self.atts = atts
      # word_vec has shape [T_decoder, N, D]
      word_vecs = tf.reduce_sum(atts * self.embedded_input_seq, axis=1)

      predicted_tokens.set_shape([None, None])
      token_probs.set_shape([None, None])
      neg_entropy.set_shape([None])
      word_vecs.set_shape([None, None, self.encoder_embed_dim])

      self.predicted_tokens = predicted_tokens
      self.token_probs = token_probs
      self.neg_entropy = neg_entropy
      self.word_vecs = word_vecs