test_weight_decay_extend.py 7.0 KB
Newer Older
C
chengduo 已提交
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
#   Copyright (c) 2019 PaddlePaddle 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 unittest
from functools import partial
import numpy as np
import paddle
import paddle.fluid as fluid
import contextlib

P
pangyoki 已提交
22 23
paddle.enable_static()

24 25
SEED = 2020

C
chengduo 已提交
26

27 28 29 30 31
def fake_imdb_reader(word_dict_size,
                     sample_num,
                     lower_seq_len=100,
                     upper_seq_len=200,
                     class_dim=2):
32

33
    def __reader__():
34
        for _ in range(sample_num):
35 36 37 38 39 40 41 42 43
            length = np.random.random_integers(low=lower_seq_len,
                                               high=upper_seq_len,
                                               size=[1])[0]
            ids = np.random.random_integers(low=0,
                                            high=word_dict_size - 1,
                                            size=[length]).astype('int64')
            label = np.random.random_integers(low=0,
                                              high=class_dim - 1,
                                              size=[1]).astype('int64')[0]
44 45 46 47 48
            yield ids, label

    return __reader__


C
chengduo 已提交
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
def get_places():
    places = [fluid.CPUPlace()]
    if fluid.core.is_compiled_with_cuda():
        places.append(fluid.CUDAPlace(0))
    return places


@contextlib.contextmanager
def prog_scope_guard(main_prog, startup_prog):
    scope = fluid.core.Scope()
    with fluid.unique_name.guard():
        with fluid.scope_guard(scope):
            with fluid.program_guard(main_prog, startup_prog):
                yield


def bow_net(data,
            label,
            dict_dim,
            is_sparse=False,
            emb_dim=128,
            hid_dim=128,
            hid_dim2=96,
            class_dim=2):
    """
    BOW net
    This model is from https://github.com/PaddlePaddle/models:
    fluid/PaddleNLP/text_classification/nets.py
    """
78 79 80
    emb = fluid.layers.embedding(input=data,
                                 is_sparse=is_sparse,
                                 size=[dict_dim, emb_dim])
C
chengduo 已提交
81 82 83 84 85 86
    bow = fluid.layers.sequence_pool(input=emb, pool_type='sum')
    bow_tanh = fluid.layers.tanh(bow)
    fc_1 = fluid.layers.fc(input=bow_tanh, size=hid_dim, act="tanh")
    fc_2 = fluid.layers.fc(input=fc_1, size=hid_dim2, act="tanh")
    prediction = fluid.layers.fc(input=[fc_2], size=class_dim, act="softmax")
    cost = fluid.layers.cross_entropy(input=prediction, label=label)
87
    avg_cost = paddle.mean(x=cost)
C
chengduo 已提交
88 89 90 91 92

    return avg_cost


class TestWeightDecay(unittest.TestCase):
93

C
chengduo 已提交
94
    def setUp(self):
95 96 97 98 99
        # set seed
        np.random.seed(SEED)
        paddle.seed(SEED)
        paddle.framework.random._manual_program_seed(SEED)
        # configs
100 101 102 103 104
        self.word_dict_len = 5147
        batch_size = 2
        reader = fake_imdb_reader(self.word_dict_len, batch_size * 100)
        reader = paddle.batch(reader, batch_size=batch_size)()
        self.train_data = [next(reader) for _ in range(3)]
C
chengduo 已提交
105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128
        self.learning_rate = .5

    def run_program(self, place, feed_list):
        exe = fluid.Executor(place)
        feeder = fluid.DataFeeder(feed_list=feed_list, place=place)
        exe.run(fluid.default_startup_program())

        main_prog = fluid.default_main_program()
        param_list = [var.name for var in main_prog.block(0).all_parameters()]

        param_sum = []
        for data in self.train_data:
            out = exe.run(main_prog,
                          feed=feeder.feed(data),
                          fetch_list=param_list)
            p_sum = 0
            for v in out:
                p_sum += np.sum(np.abs(v))
            param_sum.append(p_sum)
        return param_sum

    def check_weight_decay(self, place, model):
        main_prog = fluid.framework.Program()
        startup_prog = fluid.framework.Program()
L
Leo Chen 已提交
129

C
chengduo 已提交
130
        with prog_scope_guard(main_prog=main_prog, startup_prog=startup_prog):
131 132 133 134
            data = fluid.layers.data(name="words",
                                     shape=[1],
                                     dtype="int64",
                                     lod_level=1)
C
chengduo 已提交
135
            label = fluid.layers.data(name="label", shape=[1], dtype="int64")
136
            avg_cost = model(data, label, self.word_dict_len)
C
chengduo 已提交
137 138 139
            AdamW = fluid.contrib.extend_with_decoupled_weight_decay(
                fluid.optimizer.Adam)

140 141
            optimizer = AdamW(learning_rate=self.learning_rate,
                              weight_decay=self.learning_rate)
C
chengduo 已提交
142 143 144 145 146 147 148 149 150

            optimizer.minimize(avg_cost)
            param_sum = self.run_program(place, [data, label])

        return param_sum

    def check_weight_decay2(self, place, model):
        main_prog = fluid.framework.Program()
        startup_prog = fluid.framework.Program()
L
Leo Chen 已提交
151

C
chengduo 已提交
152
        with prog_scope_guard(main_prog=main_prog, startup_prog=startup_prog):
153 154 155 156
            data = fluid.layers.data(name="words",
                                     shape=[1],
                                     dtype="int64",
                                     lod_level=1)
C
chengduo 已提交
157 158
            label = fluid.layers.data(name="label", shape=[1], dtype="int64")

159
            avg_cost = model(data, label, self.word_dict_len)
C
chengduo 已提交
160

161 162 163 164
            optimizer = fluid.optimizer.Adam(learning_rate=self.learning_rate)

            params_grads = optimizer.backward(avg_cost)

C
chengduo 已提交
165 166 167 168
            param_list = [(var, var * self.learning_rate)
                          for var in main_prog.block(0).all_parameters()]

            for params in param_list:
169 170
                updated_p = fluid.layers.elementwise_sub(x=params[0],
                                                         y=params[1])
C
chengduo 已提交
171 172
                fluid.layers.assign(input=updated_p, output=params[0])

173 174
            optimizer.apply_optimize(avg_cost, startup_prog, params_grads)

C
chengduo 已提交
175 176 177 178 179 180 181 182 183 184
            param_sum = self.run_program(place, [data, label])
        return param_sum

    def test_weight_decay(self):
        for place in get_places():
            model = partial(bow_net, is_sparse=False)
            param_sum1 = self.check_weight_decay(place, model)
            param_sum2 = self.check_weight_decay2(place, model)

            for i in range(len(param_sum1)):
185 186 187 188 189 190
                np.testing.assert_allclose(
                    param_sum1[i],
                    param_sum2[i],
                    rtol=1e-05,
                    err_msg='Current place: {}, i: {}, sum1: {}, sum2: {}'.
                    format(
191 192 193 194
                        place, i, param_sum1[i]
                        [~np.isclose(param_sum1[i], param_sum2[i])],
                        param_sum2[i]
                        [~np.isclose(param_sum1[i], param_sum2[i])]))
C
chengduo 已提交
195 196 197 198


if __name__ == '__main__':
    unittest.main()