You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
77 lines
2.6 KiB
77 lines
2.6 KiB
#edit-mode: -*- python -*-
|
|
# Copyright (c) 2016 Baidu, Inc. 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.
|
|
|
|
from paddle.trainer_config_helpers import *
|
|
|
|
######################## data source ################################
|
|
define_py_data_sources2(train_list='gserver/tests/Sequence/dummy.list',
|
|
test_list=None,
|
|
module='rnn_data_provider',
|
|
obj='process_subseq2')
|
|
|
|
|
|
settings(batch_size=2, learning_rate=0.01)
|
|
######################## network configure ################################
|
|
dict_dim = 10
|
|
word_dim = 8
|
|
hidden_dim = 8
|
|
label_dim = 3
|
|
|
|
data = data_layer(name="word", size=dict_dim)
|
|
|
|
emb = embedding_layer(input=data, size=word_dim)
|
|
|
|
# This hierachical RNN is designed to be equivalent to the simple RNN in
|
|
# sequence_rnn.conf
|
|
|
|
def outer_step(wid, x):
|
|
outer_mem = memory(name="outer_rnn_state", size=hidden_dim)
|
|
def inner_step(y, wid):
|
|
z = embedding_layer(input=wid, size=word_dim)
|
|
inner_mem = memory(name="inner_rnn_state",
|
|
size=hidden_dim,
|
|
boot_layer=outer_mem)
|
|
out = fc_layer(input=[y, z, inner_mem],
|
|
size=hidden_dim,
|
|
act=TanhActivation(),
|
|
bias_attr=True,
|
|
name="inner_rnn_state")
|
|
return out
|
|
|
|
inner_rnn_output = recurrent_group(
|
|
step=inner_step,
|
|
name="inner",
|
|
input=[x, wid])
|
|
last = last_seq(input=inner_rnn_output, name="outer_rnn_state")
|
|
|
|
# "return last" should also work. But currently RecurrentGradientMachine
|
|
# does not handle it, and will report error: In hierachical RNN, all out
|
|
# links should be from sequences now.
|
|
return inner_rnn_output
|
|
|
|
out = recurrent_group(
|
|
name="outer",
|
|
step=outer_step,
|
|
input=[SubsequenceInput(data), SubsequenceInput(emb)])
|
|
|
|
rep = last_seq(input=out)
|
|
prob = fc_layer(size=label_dim,
|
|
input=rep,
|
|
act=SoftmaxActivation(),
|
|
bias_attr=True)
|
|
|
|
outputs(classification_cost(input=prob,
|
|
label=data_layer(name="label", size=label_dim)))
|