fluid.regularizer

SourceEnglish

L1Decay

SourceEnglish

• paddle.fluid.regularizer.L1Decay
• L1DecayRegularizer 的别名

L1DecayRegularizer

SourceEnglish

• class paddle.fluid.regularizer.L1DecayRegularizer(regularization_coeff=0.0)
• 实现 L1 权重衰减正则化。

L1正则将会稀疏化权重矩阵。

• 参数:
• • regularization_coeff (float) – 正则化系数代码示例

import paddle.fluid as fluid
main_prog = fluid.Program()
startup_prog = fluid.Program()
with fluid.program_guard(main_prog, startup_prog):
    data = fluid.layers.data(name='image', shape=[3, 28, 28], dtype='float32')
    label = fluid.layers.data(name='label', shape=[1], dtype='int64')
    hidden = fluid.layers.fc(input=data, size=128, act='relu')
    prediction = fluid.layers.fc(input=hidden, size=10, act='softmax')
    loss = fluid.layers.cross_entropy(input=prediction, label=label)
    avg_loss = fluid.layers.mean(loss)
optimizer = fluid.optimizer.Adagrad(
    learning_rate=1e-4,
    regularization=fluid.regularizer.L1DecayRegularizer(
        regularization_coeff=0.1))
optimizer.minimize(avg_loss)

L2Decay

SourceEnglish

• paddle.fluid.regularizer.L2Decay
• L2DecayRegularizer 的别名

L2DecayRegularizer

SourceEnglish

• class paddle.fluid.regularizer.L2DecayRegularizer(regularization_coeff=0.0)
• 实现L2 权重衰减正则化。

较小的 L2 的有助于防止对训练数据的过度拟合。

• 参数:
• • regularization_coeff (float) – 正则化系数代码示例

import paddle.fluid as fluid
main_prog = fluid.Program()
startup_prog = fluid.Program()
with fluid.program_guard(main_prog, startup_prog):
    data = fluid.layers.data(name='image', shape=[3, 28, 28], dtype='float32')
    label = fluid.layers.data(name='label', shape=[1], dtype='int64')
    hidden = fluid.layers.fc(input=data, size=128, act='relu')
    prediction = fluid.layers.fc(input=hidden, size=10, act='softmax')
    loss = fluid.layers.cross_entropy(input=prediction, label=label)
    avg_loss = fluid.layers.mean(loss)
optimizer = fluid.optimizer.Adagrad(
    learning_rate=1e-4,
    regularization=fluid.regularizer.L2DecayRegularizer(
        regularization_coeff=0.1))
optimizer.minimize(avg_loss)