Coverage for /pythoncovmergedfiles/medio/medio/usr/local/lib/python3.8/site-packages/keras/src/layers/regularization/gaussian_dropout.py: 50%

28 statements  

« prev     ^ index     » next       coverage.py v7.4.0, created at 2024-01-03 07:57 +0000

1# Copyright 2015 The TensorFlow Authors. All Rights Reserved. 

2# 

3# Licensed under the Apache License, Version 2.0 (the "License"); 

4# you may not use this file except in compliance with the License. 

5# You may obtain a copy of the License at 

6# 

7# http://www.apache.org/licenses/LICENSE-2.0 

8# 

9# Unless required by applicable law or agreed to in writing, software 

10# distributed under the License is distributed on an "AS IS" BASIS, 

11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 

12# See the License for the specific language governing permissions and 

13# limitations under the License. 

14# ============================================================================== 

15"""Contains the GaussianDropout layer.""" 

16 

17 

18import numpy as np 

19import tensorflow.compat.v2 as tf 

20 

21from keras.src import backend 

22from keras.src.engine import base_layer 

23from keras.src.utils import tf_utils 

24 

25# isort: off 

26from tensorflow.python.util.tf_export import keras_export 

27 

28 

29@keras_export("keras.layers.GaussianDropout") 

30class GaussianDropout(base_layer.BaseRandomLayer): 

31 """Apply multiplicative 1-centered Gaussian noise. 

32 

33 As it is a regularization layer, it is only active at training time. 

34 

35 Args: 

36 rate: Float, drop probability (as with `Dropout`). 

37 The multiplicative noise will have 

38 standard deviation `sqrt(rate / (1 - rate))`. 

39 seed: Integer, optional random seed to enable deterministic behavior. 

40 

41 Call arguments: 

42 inputs: Input tensor (of any rank). 

43 training: Python boolean indicating whether the layer should behave in 

44 training mode (adding dropout) or in inference mode (doing nothing). 

45 

46 Input shape: 

47 Arbitrary. Use the keyword argument `input_shape` 

48 (tuple of integers, does not include the samples axis) 

49 when using this layer as the first layer in a model. 

50 

51 Output shape: 

52 Same shape as input. 

53 """ 

54 

55 def __init__(self, rate, seed=None, **kwargs): 

56 super().__init__(seed=seed, **kwargs) 

57 self.supports_masking = True 

58 self.rate = rate 

59 self.seed = seed 

60 

61 def call(self, inputs, training=None): 

62 if 0 < self.rate < 1: 

63 

64 def noised(): 

65 stddev = np.sqrt(self.rate / (1.0 - self.rate)) 

66 return inputs * self._random_generator.random_normal( 

67 shape=tf.shape(inputs), 

68 mean=1.0, 

69 stddev=stddev, 

70 dtype=inputs.dtype, 

71 ) 

72 

73 return backend.in_train_phase(noised, inputs, training=training) 

74 return inputs 

75 

76 def get_config(self): 

77 config = {"rate": self.rate, "seed": self.seed} 

78 base_config = super().get_config() 

79 return dict(list(base_config.items()) + list(config.items())) 

80 

81 @tf_utils.shape_type_conversion 

82 def compute_output_shape(self, input_shape): 

83 return input_shape 

84