Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

fix minor bugs in bijector and distribution utils, fix docs issues

This commit is contained in:
Xun Deng 2020-11-17 15:47:04 -05:00
parent 2424b8bd19
commit b5e05472ce
23 changed files with 130 additions and 107 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
mindspore/nn/probability/bijector/bijector.py
View File

@ -164,6 +164,8 @@ class Bijector(Cell):
self.common_dtype = value_t.dtype
elif value_t.dtype != self.common_dtype:
raise TypeError(f"{name} should have the same dtype as other arguments.")
# check if the parameters are casted into float-type tensors
validator.check_type_name("dtype", value_t.dtype, mstype.float_type, type(self).__name__)
# check if the dtype of the input_parameter agrees with the bijector's dtype
elif value_t.dtype != self.dtype:
raise TypeError(f"{name} should have the same dtype as the bijector's dtype.")

2
mindspore/nn/probability/bijector/power_transform.py
View File

@ -61,7 +61,7 @@ class PowerTransform(Bijector):
"""
def __init__(self,
power=0,
power=0.,
name='PowerTransform'):
param = dict(locals())
param['param_dict'] = {'power': power}

15
mindspore/nn/probability/distribution/_utils/utils.py
View File

@ -159,6 +159,15 @@ def check_prob(p):
raise ValueError('Probabilities should be less than one')
def check_sum_equal_one(probs):
"""
Used in categorical distribution. check if probabilities of each category sum to 1.
"""
if probs is None:
raise ValueError(f'input value cannot be None in check_sum_equal_one')
if isinstance(probs, Parameter):
if not isinstance(probs.data, Tensor):
return
probs = probs.data
prob_sum = np.sum(probs.asnumpy(), axis=-1)
comp = np.equal(np.ones(prob_sum.shape), prob_sum)
if not comp.all():
@ -168,6 +177,12 @@ def check_rank(probs):
"""
Used in categorical distribution. check Rank >=1.
"""
if probs is None:
raise ValueError(f'input value cannot be None in check_rank')
if isinstance(probs, Parameter):
if not isinstance(probs.data, Tensor):
return
probs = probs.data
if probs.asnumpy().ndim == 0:
raise ValueError('probs for Categorical distribution must have rank >= 1.')

3
mindspore/nn/probability/distribution/bernoulli.py
View File

@ -44,7 +44,7 @@ class Bernoulli(Distribution):
>>> # The following creates two independent Bernoulli distributions.
>>> b = msd.Bernoulli([0.5, 0.5], dtype=mstype.int32)
>>>
>>> # A Bernoulli distribution can be initilized without arguments.
>>> # A Bernoulli distribution can be initialized without arguments.
>>> # In this case, `probs` must be passed in through arguments during function calls.
>>> b = msd.Bernoulli(dtype=mstype.int32)
>>>
@ -106,7 +106,6 @@ class Bernoulli(Distribution):
... ans = self.b1.sample((2,3))
... ans = self.b1.sample((2,3), probs_b)
... ans = self.b2.sample((2,3), probs_a)
...
"""
def __init__(self,

1
mindspore/nn/probability/distribution/categorical.py
View File

@ -99,7 +99,6 @@ class Categorical(Distribution):
... ans = self.ca.sample((2,3))
... ans = self.ca.sample((2,3), probs_b)
... ans = self.ca1.sample((2,3), probs_a)
...
"""
def __init__(self,

118
mindspore/nn/probability/distribution/cauchy.py
View File

@ -48,70 +48,70 @@ class Cauchy(Distribution):
>>> # The following creates two independent Cauchy distributions.
>>> cauchy = msd.Cauchy([3.0, 3.0], [4.0, 4.0], dtype=mstype.float32)
>>>
>>> # A Cauchy distribution can be initilize without arguments.
>>> # A Cauchy distribution can be initialized without arguments.
>>> # In this case, 'loc' and `scale` must be passed in through arguments.
>>> cauchy = msd.Cauchy(dtype=mstype.float32)
>>>
>>> # To use a Cauchy distribution in a network.
>>> class net(Cell):
>>> def __init__(self):
>>> super(net, self).__init__():
>>> self.cau1 = msd.Cauchy(0.0, 1.0, dtype=mstype.float32)
>>> self.cau2 = msd.Cauchy(dtype=mstype.float32)
>>>
>>> # The following calls are valid in construct.
>>> def construct(self, value, loc_b, scale_b, loc_a, scale_a):
>>>
>>> # Private interfaces of probability functions corresponding to public interfaces, including
>>> # `prob`, `log_prob`, `cdf`, `log_cdf`, `survival_function`, and `log_survival`, have the same arguments as follows.
>>> # Args:
>>> # value (Tensor): the value to be evaluated.
>>> # loc (Tensor): the location of the distribution. Default: self.loc.
>>> # scale (Tensor): the scale of the distribution. Default: self.scale.
>>>
>>> # Examples of `prob`.
>>> # Similar calls can be made to other probability functions
>>> # by replacing 'prob' by the name of the function
>>> ans = self.cau1.prob(value)
>>> # Evaluate with respect to distribution b.
>>> ans = self.cau1.prob(value, loc_b, scale_b)
>>> # `loc` and `scale` must be passed in during function calls
>>> ans = self.cau2.prob(value, loc_a, scale_a)
>>>
>>> # Functions `mode` and `entropy` have the same arguments.
>>> # Args:
>>> # loc (Tensor): the location of the distribution. Default: self.loc.
>>> # scale (Tensor): the scale of the distribution. Default: self.scale.
>>>
>>> # Example of `mode`.
>>> ans = self.cau1.mode() # return 0.0
>>> ans = self.cau1.mode(loc_b, scale_b) # return loc_b
>>> # `loc` and `scale` must be passed in during function calls.
>>> ans = self.cau2.mode(loc_a, scale_a)
>>>
>>> # Interfaces of 'kl_loss' and 'cross_entropy' are the same:
>>> # Args:
>>> # dist (str): the type of the distributions. Only "Cauchy" is supported.
>>> # loc_b (Tensor): the loc of distribution b.
>>> # scale_b (Tensor): the scale distribution b.
>>> # loc (Tensor): the loc of distribution a. Default: self.loc.
>>> # scale (Tensor): the scale distribution a. Default: self.scale.
>>>
>>> # Examples of `kl_loss`. `cross_entropy` is similar.
>>> ans = self.cau1.kl_loss('Cauchy', loc_b, scale_b)
>>> ans = self.cau1.kl_loss('Cauchy', loc_b, scale_b, loc_a, scale_a)
>>> # Additional `loc` and `scale` must be passed in.
>>> ans = self.cau2.kl_loss('Cauchy', loc_b, scale_b, loc_a, scale_a)
>>>
>>> # Examples of `sample`.
>>> # Args:
>>> # shape (tuple): the shape of the sample. Default: ()
>>> # loc (Tensor): the location of the distribution. Default: self.loc.
>>> # scale (Tensor): the scale of the distribution. Default: self.scale.
>>> ans = self.cau1.sample()
>>> ans = self.cau1.sample((2,3))
>>> ans = self.cau1.sample((2,3), loc_b, s_b)
>>> ans = self.cau2.sample((2,3), loc_a, s_a)
... def __init__(self):
... super(net, self).__init__():
... self.cau1 = msd.Cauchy(0.0, 1.0, dtype=mstype.float32)
... self.cau2 = msd.Cauchy(dtype=mstype.float32)
...
... # The following calls are valid in construct.
... def construct(self, value, loc_b, scale_b, loc_a, scale_a):
...
... # Private interfaces of probability functions corresponding to public interfaces, including
... # `prob`, `log_prob`, `cdf`, `log_cdf`, `survival_function`, and `log_survival`, have the same arguments as follows.
... # Args:
... # value (Tensor): the value to be evaluated.
... # loc (Tensor): the location of the distribution. Default: self.loc.
... # scale (Tensor): the scale of the distribution. Default: self.scale.
...
... # Examples of `prob`.
... # Similar calls can be made to other probability functions
... # by replacing 'prob' by the name of the function
... ans = self.cau1.prob(value)
... # Evaluate with respect to distribution b.
... ans = self.cau1.prob(value, loc_b, scale_b)
... # `loc` and `scale` must be passed in during function calls
... ans = self.cau2.prob(value, loc_a, scale_a)
...
... # Functions `mode` and `entropy` have the same arguments.
... # Args:
... # loc (Tensor): the location of the distribution. Default: self.loc.
... # scale (Tensor): the scale of the distribution. Default: self.scale.
...
... # Example of `mode`.
... ans = self.cau1.mode() # return 0.0
... ans = self.cau1.mode(loc_b, scale_b) # return loc_b
... # `loc` and `scale` must be passed in during function calls.
... ans = self.cau2.mode(loc_a, scale_a)
...
... # Interfaces of 'kl_loss' and 'cross_entropy' are the same:
... # Args:
... # dist (str): the type of the distributions. Only "Cauchy" is supported.
... # loc_b (Tensor): the loc of distribution b.
... # scale_b (Tensor): the scale distribution b.
... # loc (Tensor): the loc of distribution a. Default: self.loc.
... # scale (Tensor): the scale distribution a. Default: self.scale.
...
... # Examples of `kl_loss`. `cross_entropy` is similar.
... ans = self.cau1.kl_loss('Cauchy', loc_b, scale_b)
... ans = self.cau1.kl_loss('Cauchy', loc_b, scale_b, loc_a, scale_a)
... # Additional `loc` and `scale` must be passed in.
... ans = self.cau2.kl_loss('Cauchy', loc_b, scale_b, loc_a, scale_a)
...
... # Examples of `sample`.
... # Args:
... # shape (tuple): the shape of the sample. Default: ()
... # loc (Tensor): the location of the distribution. Default: self.loc.
... # scale (Tensor): the scale of the distribution. Default: self.scale.
... ans = self.cau1.sample()
... ans = self.cau1.sample((2,3))
... ans = self.cau1.sample((2,3), loc_b, s_b)
... ans = self.cau2.sample((2,3), loc_a, s_a)
"""
def __init__(self,

5
mindspore/nn/probability/distribution/exponential.py
View File

@ -46,7 +46,7 @@ class Exponential(Distribution):
>>> # The following creates two independent Exponential distributions.
>>> e = msd.Exponential([0.5, 0.5], dtype=mstype.float32)
>>>
>>> # An Exponential distribution can be initilized without arguments.
>>> # An Exponential distribution can be initialized without arguments.
>>> # In this case, `rate` must be passed in through `args` during function calls.
>>> e = msd.Exponential(dtype=mstype.float32)
>>>
@ -108,7 +108,6 @@ class Exponential(Distribution):
... ans = self.e1.sample((2,3))
... ans = self.e1.sample((2,3), rate_b)
... ans = self.e2.sample((2,3), rate_a)
...
"""
def __init__(self,
@ -187,7 +186,7 @@ class Exponential(Distribution):
def _sd(self, rate=None):
r"""
.. math::
sd(EXP) = \frac{1.0}{\lambda}.
SD(EXP) = \frac{1.0}{\lambda}.
"""
rate = self._check_param_type(rate)
return 1.0 / rate

3
mindspore/nn/probability/distribution/geometric.py
View File

@ -47,7 +47,7 @@ class Geometric(Distribution):
>>> # The following creates two independent Geometric distributions.
>>> n = msd.Geometric([0.5, 0.5], dtype=mstype.int32)
>>>
>>> # A Geometric distribution can be initilized without arguments.
>>> # A Geometric distribution can be initialized without arguments.
>>> # In this case, `probs` must be passed in through arguments during function calls.
>>> n = msd.Geometric(dtype=mstype.int32)
>>>
@ -109,7 +109,6 @@ class Geometric(Distribution):
... ans = self.g1.sample((2,3))
... ans = self.g1.sample((2,3), probs_b)
... ans = self.g2.sample((2,3), probs_a)
...
"""
def __init__(self,

1
mindspore/nn/probability/distribution/gumbel.py
View File

@ -91,7 +91,6 @@ class Gumbel(TransformedDistribution):
...
... ans = self.g1.sample()
... ans = self.g1.sample((2,3))
...
"""
def __init__(self,

3
mindspore/nn/probability/distribution/log_normal.py
View File

@ -47,7 +47,7 @@ class LogNormal(msd.TransformedDistribution):
>>> # The following creates two independent LogNormal distributions.
>>> n = msd.LogNormal([3.0, 3.0], [4.0, 4.0], dtype=mstype.float32)
>>>
>>> # A LogNormal distribution can be initilize without arguments.
>>> # A LogNormal distribution can be initialized without arguments.
>>> # In this case, `loc` and `scale` must be passed in during function calls.
>>> n = msd.LogNormal(dtype=mstype.float32)
>>>
@ -122,7 +122,6 @@ class LogNormal(msd.TransformedDistribution):
... ans = self.n1.sample((2,3))
... ans = self.n1.sample((2,3), loc_b, scale_b)
... ans = self.n2.sample((2,3), loc_a, scale_a)
...
"""
def __init__(self,

3
mindspore/nn/probability/distribution/logistic.py
View File

@ -47,7 +47,7 @@ class Logistic(Distribution):
>>> # The following creates two independent Logistic distributions.
>>> n = msd.Logistic([3.0, 3.0], [4.0, 4.0], dtype=mstype.float32)
>>>
>>> # A Logistic distribution can be initilize without arguments.
>>> # A Logistic distribution can be initialized without arguments.
>>> # In this case, `loc` and `scale` must be passed in through arguments.
>>> n = msd.Logistic(dtype=mstype.float32)
>>>
@ -97,7 +97,6 @@ class Logistic(Distribution):
... ans = self.l1.sample((2,3))
... ans = self.l1.sample((2,3), loc_b, scale_b)
... ans = self.l2.sample((2,3), loc_a, scale_a)
...
"""
def __init__(self,

19
mindspore/nn/probability/distribution/normal.py
View File

@ -47,7 +47,7 @@ class Normal(Distribution):
>>> # The following creates two independent Normal distributions.
>>> n = msd.Normal([3.0, 3.0], [4.0, 4.0], dtype=mstype.float32)
>>>
>>> # A Normal distribution can be initilize without arguments.
>>> # A Normal distribution can be initialized without arguments.
>>> # In this case, `mean` and `sd` must be passed in through arguments.
>>> n = msd.Normal(dtype=mstype.float32)
>>>
@ -55,7 +55,7 @@ class Normal(Distribution):
>>> class net(Cell):
... def __init__(self):
... super(net, self).__init__():
... self.n1 = msd.Nomral(0.0, 1.0, dtype=mstype.float32)
... self.n1 = msd.Normal(0.0, 1.0, dtype=mstype.float32)
... self.n2 = msd.Normal(dtype=mstype.float32)
...
... # The following calls are valid in construct.
@ -65,14 +65,14 @@ class Normal(Distribution):
... # `prob`, `log_prob`, `cdf`, `log_cdf`, `survival_function`, and `log_survival`, have the same arguments as follows.
... # Args:
... # value (Tensor): the value to be evaluated.
... # mean (Tensor): the mean of distribution. Default: self._mean_value.
... # sd (Tensor): the standard deviation of distribution. Default: self._sd_value.
... # mean (Tensor): the mean of the distribution. Default: self._mean_value.
... # sd (Tensor): the standard deviation of the distribution. Default: self._sd_value.
...
... # Examples of `prob`.
... # Similar calls can be made to other probability functions
... # by replacing 'prob' by the name of the function
... ans = self.n1.prob(value)
... # Evaluate with respect to distribution b.
... # Evaluate with respect to the distribution b.
... ans = self.n1.prob(value, mean_b, sd_b)
... # `mean` and `sd` must be passed in during function calls
... ans = self.n2.prob(value, mean_a, sd_a)
@ -80,8 +80,8 @@ class Normal(Distribution):
...
... # Functions `mean`, `sd`, `var`, and `entropy` have the same arguments.
... # Args:
... # mean (Tensor): the mean of distribution. Default: self._mean_value.
... # sd (Tensor): the standard deviation of distribution. Default: self._sd_value.
... # mean (Tensor): the mean of the distribution. Default: self._mean_value.
... # sd (Tensor): the standard deviation of the distribution. Default: self._sd_value.
...
... # Example of `mean`. `sd`, `var`, and `entropy` are similar.
... ans = self.n1.mean() # return 0.0
@ -94,9 +94,9 @@ class Normal(Distribution):
... # Args:
... # dist (str): the type of the distributions. Only "Normal" is supported.
... # mean_b (Tensor): the mean of distribution b.
... # sd_b (Tensor): the standard deviation distribution b.
... # sd_b (Tensor): the standard deviation of distribution b.
... # mean_a (Tensor): the mean of distribution a. Default: self._mean_value.
... # sd_a (Tensor): the standard deviation distribution a. Default: self._sd_value.
... # sd_a (Tensor): the standard deviation of distribution a. Default: self._sd_value.
...
... # Examples of `kl_loss`. `cross_entropy` is similar.
... ans = self.n1.kl_loss('Normal', mean_b, sd_b)
@ -113,7 +113,6 @@ class Normal(Distribution):
... ans = self.n1.sample((2,3))
... ans = self.n1.sample((2,3), mean_b, sd_b)
... ans = self.n2.sample((2,3), mean_a, sd_a)
...
"""
def __init__(self,

1
mindspore/nn/probability/distribution/transformed_distribution.py
View File

@ -67,7 +67,6 @@ class TransformedDistribution(Distribution):
... # Similar calls can be made to other functions
... # by replacing 'sample' by the name of the function.
... ans = self.ln.sample(shape=(2, 3))
...
"""
def __init__(self,

11
mindspore/nn/probability/distribution/uniform.py
View File

@ -46,7 +46,7 @@ class Uniform(Distribution):
>>> # The following creates two independent Uniform distributions.
>>> u = msd.Uniform([0.0, 0.0], [1.0, 2.0], dtype=mstype.float32)
>>>
>>> # A Uniform distribution can be initilized without arguments.
>>> # A Uniform distribution can be initialized without arguments.
>>> # In this case, `high` and `low` must be passed in through arguments during function calls.
>>> u = msd.Uniform(dtype=mstype.float32)
>>>
@ -64,8 +64,8 @@ class Uniform(Distribution):
... # `prob`, `log_prob`, `cdf`, `log_cdf`, `survival_function`, and `log_survival`, have the same arguments.
... # Args:
... # value (Tensor): the value to be evaluated.
... # low (Tensor): the lower bound of distribution. Default: self.low.
... # high (Tensor): the higher bound of distribution. Default: self.high.
... # low (Tensor): the lower bound of the distribution. Default: self.low.
... # high (Tensor): the higher bound of the distribution. Default: self.high.
...
... # Examples of `prob`.
... # Similar calls can be made to other probability functions
@ -79,8 +79,8 @@ class Uniform(Distribution):
...
... # Functions `mean`, `sd`, `var`, and `entropy` have the same arguments.
... # Args:
... # low (Tensor): the lower bound of distribution. Default: self.low.
... # high (Tensor): the higher bound of distribution. Default: self.high.
... # low (Tensor): the lower bound of the distribution. Default: self.low.
... # high (Tensor): the higher bound of the distribution. Default: self.high.
...
... # Examples of `mean`. `sd`, `var`, and `entropy` are similar.
... ans = self.u1.mean() # return 0.5
@ -112,7 +112,6 @@ class Uniform(Distribution):
... ans = self.u1.sample((2,3))
... ans = self.u1.sample((2,3), low_b, high_b)
... ans = self.u2.sample((2,3), low_a, high_a)
...
"""
def __init__(self,

8
tests/st/probability/bijector/test_power_transform.py
View File

@ -35,7 +35,7 @@ class Net(nn.Cell):
return forward
def test_forward():
power = 2
power = 2.
x = np.array([2.0, 3.0, 4.0, 5.0], dtype=np.float32)
tx = Tensor(x, dtype=dtype.float32)
forward = Net(power=power)
@ -57,7 +57,7 @@ class Net1(nn.Cell):
return inverse
def test_inverse():
power = 2
power = 2.
y = np.array([2.0, 3.0, 4.0, 5.0], dtype=np.float32)
ty = Tensor(y, dtype=dtype.float32)
inverse = Net1(power=power)
@ -78,7 +78,7 @@ class Net2(nn.Cell):
return self.bijector.forward_log_jacobian(x_)
def test_forward_jacobian():
power = 2
power = 2.
x = np.array([2.0, 3.0, 4.0, 5.0], dtype=np.float32)
tx = Tensor(x, dtype=dtype.float32)
forward_jacobian = Net2(power=power)
@ -99,7 +99,7 @@ class Net3(nn.Cell):
return self.bijector.inverse_log_jacobian(y_)
def test_inverse_jacobian():
power = 2
power = 2.
y = np.array([2.0, 3.0, 4.0, 5.0], dtype=np.float32)
ty = Tensor(y, dtype=dtype.float32)
inverse_jacobian = Net3(power=power)

22
tests/ut/python/nn/probability/bijector/test_bijector.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
@ -99,6 +99,14 @@ def test_arguments_same_type():
assert isinstance(bijector, msb.Bijector)
bijector = MyBijector(1.0, 2.0)
assert isinstance(bijector, msb.Bijector)
with pytest.raises(TypeError):
MyBijector(1, 2)
with pytest.raises(TypeError):
MyBijector([1, 2], [2, 4])
with pytest.raises(TypeError):
MyBijector(np.array([1, 2]).astype(np.int32), np.array([1, 2]).astype(np.int32))
with pytest.raises(TypeError):
MyBijector(Tensor([1, 2], dtype=dtype.int32), Tensor([1, 2], dtype=dtype.int32))
def test_arguments_with_dtype_specified():
"""
@ -118,12 +126,20 @@ def test_arguments_with_dtype_specified():
MySecondBijector(None, param2_2)
param1_3 = Tensor(1.0, dtype=dtype.float32)
param2_3 = Tensor(2.0, dtype=dtype.float32)
bijector = MyBijector(param1_3, param2_3)
bijector = MySecondBijector(param1_3, param2_3)
assert isinstance(bijector, msb.Bijector)
param1_4 = np.array(2.0).astype(np.float32)
param2_4 = np.array(1.0).astype(np.float32)
bijector = MyBijector(param1_4, param2_4)
bijector = MySecondBijector(param1_4, param2_4)
assert isinstance(bijector, msb.Bijector)
with pytest.raises(TypeError):
MySecondBijector(1, 2)
with pytest.raises(TypeError):
MySecondBijector([1, 2], [2, 4])
with pytest.raises(TypeError):
MySecondBijector(np.array([1, 2]).astype(np.int32), np.array([1, 2]).astype(np.int32))
with pytest.raises(TypeError):
MySecondBijector(Tensor([1, 2], dtype=dtype.int32), Tensor([1, 2], dtype=dtype.int32))
class Net1(nn.Cell):
"""

2
tests/ut/python/nn/probability/bijector/test_exp.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
tests/ut/python/nn/probability/bijector/test_gumbel_cdf.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
tests/ut/python/nn/probability/bijector/test_invert_bijector.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

8
tests/ut/python/nn/probability/bijector/test_power_transform.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
@ -22,7 +22,7 @@ from mindspore import dtype
def test_init():
b = msb.PowerTransform()
assert isinstance(b, msb.Bijector)
b = msb.PowerTransform(1)
b = msb.PowerTransform(1.)
assert isinstance(b, msb.Bijector)
def test_type():
@ -37,7 +37,7 @@ class Net(nn.Cell):
"""
def __init__(self):
super(Net, self).__init__()
self.b1 = msb.PowerTransform(power=0)
self.b1 = msb.PowerTransform(power=0.)
self.b2 = msb.PowerTransform()
def construct(self, x_):
@ -60,7 +60,7 @@ class Jacobian(nn.Cell):
"""
def __init__(self):
super(Jacobian, self).__init__()
self.b1 = msb.PowerTransform(power=0)
self.b1 = msb.PowerTransform(power=0.)
self.b2 = msb.PowerTransform()
def construct(self, x_):

2
tests/ut/python/nn/probability/bijector/test_scalar_affine.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
tests/ut/python/nn/probability/bijector/test_softplus.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.

2
tests/ut/python/nn/probability/distribution/test_distribution.py
View File

@ -1,4 +1,4 @@
# Copyright 2019 Huawei Technologies Co., Ltd
# Copyright 2020 Huawei Technologies Co., Ltd
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.