尝试将列表作为参数传递给 hyperas 时出错

Question

我正在使用 keras 和 hyperas library 进行超参数优化。

我正在尝试传递一个列表而不是硬编码列表，但它会引发错误。在 hyperas 中，您将模板值包装在“{{ some value }}”中，您可以将代码卡为列表。示例：

model.add(Dense({{choice([258,512,1024])}}))

错误：

File "hy.py", line 89, in
trials=Trials())
File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 67, in minimize
verbose=verbose)
File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 118, in base_minimizer
space=get_space(),
File "/home/peachy/Documents/tensorflow/temp_model.py", line 119, in get_space
NameError: name 'dropout' is not defined

使用一个基本示例，这是模型类。我正在从加载列表的 yml 文件加载，但这不起作用，所以我只是试图在模型类中创建一个列表并将其传递进去。它不会工作。

from __future__ import print_function

from hyperopt import Trials, STATUS_OK, tpe
from keras.datasets import mnist
from keras.layers.core import Dense, Dropout, Activation
from keras.models import Sequential
from keras.utils import np_utils

from hyperas import optim
from hyperas.distributions import choice, uniform, conditional
import common_config as cfg

def data():
    """
    Data providing function:
    This function is separated from model() so that hyperopt
    won't reload data for each evaluation run.
    """
    (x_train, y_train), (x_test, y_test) = mnist.load_data()
    x_train = x_train.reshape(60000, 784)
    x_test = x_test.reshape(10000, 784)
    x_train = x_train.astype('float32')
    x_test = x_test.astype('float32')
    x_train /= 255
    x_test /= 255
    nb_classes = 10
    y_train = np_utils.to_categorical(y_train, nb_classes)
    y_test = np_utils.to_categorical(y_test, nb_classes)
    return x_train, y_train, x_test, y_test


def model(x_train, y_train, x_test, y_test):
    """
    Model providing function:
    Create Keras model with double curly brackets dropped-in as needed.
    Return value has to be a valid python dictionary with two customary keys:
        - loss: Specify a numeric evaluation metric to be minimized
        - status: Just use STATUS_OK and see hyperopt documentation if not feasible
    The last one is optional, though recommended, namely:
        - model: specify the model just created so that we can later use it again.
    """
    ModelConfig = cfg.ModelConfig
    dropout = [0,1]
    print (dropout)
    print (type(dropout))




    model = Sequential()
    model.add(Dense(512, input_shape=(784,)))
    model.add(Activation(ModelConfig.activation))
    model.add(Dropout({{uniform(dropout)}}))
    model.add(Dense({{choice([258,512,1024])}}))
    model.add(Activation({{choice(['relu', 'sigmoid'])}}))
    model.add(Dropout({{uniform(0,1)}}))

    # If we choose 'four', add an additional fourth layer
    if conditional({{choice(['three', 'four'])}}) == 'four':
        model.add(Dense(100))

        # We can also choose between complete sets of layers

        model.add({{choice([Dropout(0.5), Activation('linear')])}})
        model.add(Activation('relu'))

    model.add(Dense(10))
    model.add(Activation('softmax'))

    model.compile(loss='categorical_crossentropy', metrics=['accuracy'],
                  optimizer={{choice(['rmsprop', 'adam', 'sgd'])}})

    model.fit(x_train, y_train,
              batch_size={{choice([64, 128])}},
              epochs=1,
              verbose=2,
              validation_data=(x_test, y_test))
    score, acc = model.evaluate(x_test, y_test, verbose=0)
    print('Test accuracy:', acc)
    return {'loss': -acc, 'status': STATUS_OK, 'model': model}


if __name__ == '__main__':
    best_run, best_model = optim.minimize(model=model,
                                          data=data,
                                          algo=tpe.suggest,
                                          max_evals=10,
                                          trials=Trials())
    X_train, Y_train, X_test, Y_test = data()
    print("Evalutation of best performing model:")
    print(best_model.evaluate(X_test, Y_test))
    print("Best performing model chosen hyper-parameters:")
    print(best_run)

值是如何从模板发送到 optim.py 文件的？ 我已经检查过了，正则表达式应该能够处理列表。

我不希望任何人知道，但你将如何调试它，因为它很可能是一个库问题

编辑

解决方案 1 和 2 只会引发类型错误，这在源代码中是一个非常奇怪的检查，以查看它是否是某个字符串类型

Hyperopt 代码 pyll_util.py

def validate_label(f):
    @wraps(f)
    def wrapper(label, *args, **kwargs):
        is_real_string = isinstance(label, basestring)
        is_literal_string = (isinstance(label, Literal) and
                             isinstance(label.obj, basestring))
        if not is_real_string and not is_literal_string:
            raise TypeError('require string label')
        return f(label, *args, **kwargs)
    return wrapper

错误

Traceback (most recent call last):
  File "hy.py", line 107, in <module>
    trials=Trials())
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 67, in minimize
    verbose=verbose)
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 136, in base_minimizer
    return_argmin=True),
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/fmin.py", line 307, in fmin
    return_argmin=return_argmin,
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/base.py", line 635, in fmin
    return_argmin=return_argmin)
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/fmin.py", line 320, in fmin
    rval.exhaust()
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/fmin.py", line 199, in exhaust
    self.run(self.max_evals - n_done, block_until_done=self.async)
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/fmin.py", line 173, in run
    self.serial_evaluate()
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/fmin.py", line 92, in serial_evaluate
    result = self.domain.evaluate(spec, ctrl)
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/base.py", line 840, in evaluate
    rval = self.fn(pyll_rval)
  File "/home/peachy/Documents/tensorflow/temp_model.py", line 101, in keras_fmin_fnct
  File "<string>", line 1, in <module>
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperopt/pyll_utils.py", line 21, in wrapper
    raise TypeError('require string label')
TypeError: require string label

通过合并列表的解决方案 3 会导致以下错误，该错误表明 %s 已与 dropout_1 一起基于生成的 temp_model 文件中。

Unexpected error: <class 'SyntaxError'>
Traceback (most recent call last):
  File "hy.py", line 107, in <module>
    trials=Trials())
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 67, in minimize
    verbose=verbose)
  File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 104, in base_minimizer
    from temp_model import keras_fmin_fnct, get_space
  File "/home/peachy/Documents/tensorflow/temp_model.py", line 133
    'Dropout_1': hp.uniform('Dropout_1', %s),
                                         ^

Answer 1

问题出在这一行：

model.add(Dropout({{uniform(dropout)}}))

变量dropout 是在model() 函数的范围内定义的，但在双括号{{ }} 内，您有一个完全不同的范围，由hyperas 库处理。因此，当 hyperas 尝试获取应用此超优化器的空间时，您遇到的错误是：

File "/home/peachy/Documents/tensorflow/tf/lib/python3.6/site-packages/hyperas-0.4-py3.6.egg/hyperas/optim.py", line 118, in base_minimizer
space=get_space(),
File "/home/peachy/Documents/tensorflow/temp_model.py", line 119, in get_space
NameError: name 'dropout' is not defined

变量 dropout 没有定义在 hyperas 的范围内（{{ }} 内的范围）。它只定义在model()函数的范围内。

您会注意到，在hyperas README.md 中提供的所有示例中，传递给超优化器的值都是硬编码的字符串或数字。

解决方案 1

如果您需要改变这些参数，您可以使用eval 将参数作为参数。例如，要修复上面有问题的行并在 dropout 上使用变量下限/上限：

lo = 0.2
hi = 0.8
cmd = "model.add(Dropout({{uniform(%.1f,%.1f)}}))"%( lo, hi )
eval(cmd)

解决方案 2

或者，使用.format()（在最后的字符串中每个括号都有两个括号，总共4个）：

lo = 0.2
hi = 0.8
cmd = "model.add(Dropout({{{{uniform({:.1f},{:.1f})}}}}))".format(lo,hi)
eval(cmd)

解决方案 3

如果列表中有lo 和hi，请使用.join() 将它们组合成一个字符串并将它们传递给eval() 函数：

mylist = [0.2, 0.8]
argstring = ",".join( str(x) for x in mylist )
cmd = "model.add(Dropout({{uniform(%s)}}))"%(argstring)
eval(cmd)

编辑 1：添加了两个附加示例 - 一个替代字符串格式化函数和一个列表示例。

编辑 2： 将字符串格式化步骤与 eval() 步骤分开。