Ctrl+K

dropout#

ivy.dropout(x, prob, /, *, scale=True, dtype=None, training=True, seed=None, noise_shape=None, out=None)[source]#

Randomly setting a fraction of input tensor to zeroes with probability.

prob at each update during training time to prevent possible overfitting. The inputs not set to 0 are scaled up 1 / (1 - prob) by default, so that overall sum is unchanged at training time and inference time.

Parameters:

  • x (Union[Array, NativeArray]) – The input array x to perform dropout on.

  • prob (float) – The probability of zeroing out each array element, float between 0 and 1.

  • scale (bool, default: True) – Whether to scale the output by 1/(1-prob). Default is True.

  • dtype (Optional[Union[Dtype, NativeDtype]], default: None) – output array data type. If dtype is None, the output array data type must be inferred from x. Default is None.

  • training (bool, default: True) – Turn on dropout if training, turn off otherwise. Default is True.

  • seed (Optional[int], default: None) – Set a default seed for random number generating (for reproducibility). Default is None.

  • noise_shape (Optional[Sequence[int]], default: None) – a sequence representing the shape of the binary dropout mask that will be multiplied with the input. A shape dimension set to None means that a different mask value will be applied to each element of the input across that dimension. A dimension set to 1 means the same mask value will be applied to all elements of the input across that dimension.

  • out (Optional[Array], default: None) – optional output array, for writing the result to. It must have a shape that the inputs broadcast to.

Return type:

Array

Returns:

  • ret – Result array after dropout is performed.

  • Both the description and the type hints above assumes an array input for simplicity,

  • but this function is nestable, and therefore also accepts ivy.Container

  • instances in place of any of the arguments.

Examples

With ivy.Array input:

>>> x = ivy.array([[1., 2., 3.],
...                [4., 5., 6.],
...                [7., 8., 9.],
...                [10., 11., 12.]])
>>> y = ivy.dropout(x,0.3)
>>> print(y)
ivy.array([[ 1.42857146,  2.85714293,  4.28571415],
       [ 0.        ,  7.14285755,  8.5714283 ],
       [10.        , 11.4285717 ,  0.        ],
       [14.2857151 ,  0.        , 17.1428566 ]])

>>> x = ivy.array([[1.5, 2.6],
...                [4.9, 6.6],
...                [7.2, 8.7]])
>>> y = ivy.dropout(x,0.5)
>>> print(y)
ivy.array([[ 0.        ,  5.19999981],
           [ 0.        ,  0.        ],
           [ 0.        , 17.39999962]])

>>> x = ivy.array([[1., 2., 3.],
...                [4., 5., 6.],
...                [7., 8., 9.],
...                [10., 11., 12.]])
>>> y = ivy.dropout(x,0.3,scale=False)
>>> print(y)
ivy.array([[ 1.,  2., 3.],
           [ 4.,  5., 0.],
           [ 7.,  0., 9.],
           [10., 11., 0.]])

>>> x = ivy.array([[1.5, 2.6],
...                [4.9, 6.6],
...                [7.2, 8.7]])
>>> y = ivy.dropout(x,0.5,scale=False)
>>> print(y)
ivy.array([[0., 2.6],
           [0., 0. ],
           [0., 8.7]])

With ivy.Container input:

>>> x = ivy.Container(a=ivy.array([[1., 2., 3.], [4., 5., 6.]]),
...                   b=ivy.array([7., 8., 9.]))
>>> y = ivy.dropout(x,0.3)
>>> print(y)
{
a: ivy.array([[0., 0., 4.28571415],
              [5.71428585, 7.14285755, 0.]]),
b: ivy.array([0., 11.4285717, 12.8571434])
}

>>> x = ivy.Container(a=ivy.array([[1.1, 2.2, 3.3], [11., 22., 33.]]),
...                   b=ivy.array([[1.245, 0.278, 4.105], [7., 13., 17.]]))
>>> y = ivy.dropout(x,0.5)
>>> print(y)
{
    a: ivy.array([[0., 4.4000001, 6.5999999],
                  [22., 44., 0.]]),
    b: ivy.array([[2.49000001, 0.55599999, 8.21000004],
                  [14., 0., 0.]])
}

>>> x = ivy.Container(a=ivy.array([[1., 2., 3.], [4., 5., 6.]]),
...                   b=ivy.array([7., 8., 9.]))
>>> y = ivy.dropout(x,0.3)
>>> print(y)
{
    a: ivy.array([[0., 0., 3.],
                  [4., 5., 0.]]),
    b: ivy.array([0., 8., 9.])
}

>>> x = ivy.Container(a=ivy.array([[1.1, 2.2, 3.3], [11., 22., 33.]]),
...                   b=ivy.array([[1.245, 0.278, 4.105], [7., 13., 17.]]))
>>> y = ivy.dropout(x,0.5)
>>> print(y)
{
    a: ivy.array([[0., 2.2, 3.3],
                  [11., 22., 0.]]),
    b: ivy.array([[1.245, 0.278, 4.105],
                  [7., 0., 0.]])
}
Array.dropout(self, prob, /, *, scale=True, dtype=None, training=True, seed=None, noise_shape=None, out=None)[source]#

ivy.Array instance method variant of ivy.dropout. This method simply wraps the function, and so the docstring for ivy.dropout also applies to this method with minimal changes.

Parameters:

  • self (Array) – The input array x to perform dropout on.

  • prob (float) – The probability of zeroing out each array element, float between 0 and 1.

  • scale (bool, default: True) – Whether to scale the output by 1/(1-prob), default is True.

  • dtype (Optional[Union[Dtype, NativeDtype]], default: None) – output array data type. If dtype is None, the output array data type must be inferred from x. Default: None.

  • training (bool, default: True) – Turn on dropout if training, turn off otherwise. Default is True.

  • seed (Optional[int], default: None) – Set a default seed for random number generating (for reproducibility).Default is None.

  • noise_shape (Optional[Sequence[int]], default: None) – a sequence representing the shape of the binary dropout mask that will be multiplied with the input.

  • out (Optional[Array], default: None) – optional output array, for writing the result to. It must have a shape that the inputs broadcast to.

Return type:

Array

Returns:

ret – Result array of the output after dropout is performed.

Examples

With ivy.Array instances:

>>> x = ivy.array([[1., 2., 3.],
...                [4., 5., 6.],
...                [7., 8., 9.],
...                [10., 11., 12.]])
>>> y = x.dropout(0.3)
>>> print(y)
ivy.array([[ 1.42857146,  2.85714293,  4.28571415],
           [ 5.71428585,  7.14285755,  8.5714283 ],
           [ 0.        , 11.4285717 , 12.8571434 ],
           [14.2857151 ,  0.        ,  0.        ]])

>>> x = ivy.array([[1., 2., 3.],
...                [4., 5., 6.],
...                [7., 8., 9.],
...                [10., 11., 12.]])
>>> y = x.dropout(0.3, scale=False)
>>> print(y)
ivy.array([[ 1.,  2., 3.],
           [ 4.,  5., 0.],
           [ 7.,  0., 9.],
           [10., 11., 0.]])
Container.dropout(self, prob, /, *, scale=True, dtype=None, training=True, seed=None, noise_shape=None, key_chains=None, to_apply=True, prune_unapplied=False, map_sequences=False, out=None)[source]#

ivy.Container instance method variant of ivy.dropout. This method simply wraps the function, and so the docstring for ivy.dropout also applies to this method with minimal changes.

Parameters:

  • self (Container) – The input container to perform dropout on.

  • prob (Union[float, Container]) – The probability of zeroing out each array element, float between 0 and 1.

  • scale (Union[bool, Container], default: True) – Whether to scale the output by 1/(1-prob), default is True.

  • dtype (Optional[Union[Dtype, Container]], default: None) – output array data type. If dtype is None, the output array data type must be inferred from x. Default: None.

  • training (Union[bool, Container], default: True) – Turn on dropout if training, turn off otherwise. Default is True.

  • seed (Optional[Union[int, Container]], default: None) – Set a default seed for random number generating (for reproducibility). Default is None.

  • noise_shape (Optional[Union[Sequence[int], Container]], default: None) – a sequence representing the shape of the binary dropout mask that will be multiplied with the input.

  • key_chains (Optional[Union[List[str], Dict[str, str], Container]], default: None) – The key-chains to apply or not apply the method to. Default is None.

  • to_apply (Union[bool, Container], default: True) – If True, the method will be applied to key_chains, otherwise key_chains will be skipped. Default is True.

  • prune_unapplied (Union[bool, Container], default: False) – Whether to prune key_chains for which the function was not applied. Default is False.

  • map_sequences (Union[bool, Container], default: False) – Whether to also map method to sequences (lists, tuples). Default is False.

  • out (Optional[Container], default: None) – optional output array, for writing the result to. It must have a shape that the inputs broadcast to.

Return type:

Container

Returns:

ret – Result array of the output after dropout is performed.

Examples

>>> x = ivy.Container(a=ivy.array([[1., 2., 3.], [4., 5., 6.]]),
...                   b=ivy.array([7., 8., 9.]))
>>> y = x.dropout(0.3)
>>> print(y)
{
    a: ivy.array([[0., 0., 4.28571415],
                  [5.71428585, 7.14285755, 0.]]),
    b: ivy.array([0., 11.4285717, 12.8571434])
}
On this page