test(mge): move `NUMBER` config to pytest.init

GitOrigin-RevId: 1d82209c409ccf884ed9f7917a51ea4da62ec00b

python_module/megengine/functional/math.py

@@ -190,7 +190,6 @@ def sqrt(inp: Tensor) -> Tensor:
     Outputs:
 
     .. testoutput::
-        :options: +NUMBER
 
         [[0.      1.     1.4142]
          [1.7321  2.     2.2361 ]]

python_module/megengine/functional/nn.py

@@ -636,7 +636,6 @@ def interpolate(
     Outputs:
 
     .. testoutput::
-        :options: +NUMBER
 
         [[[[1.   1.25 1.75 2.  ]
            [1.5  1.75 2.25 2.5 ]

python_module/megengine/functional/sort.py

@@ -39,7 +39,6 @@ def argsort(inp: Tensor, descending: bool = False) -> Tuple[Tensor, Tensor]:
     Outputs:
 
     .. testoutput::
-        :options: +NUMBER
 
         [1. 2.] [0 1]
 
@@ -93,7 +92,6 @@ def top_k(
     Outputs:
 
     .. testoutput::
-        :options: +NUMBER
 
         [1. 2. 3. 4. 5.] [7 0 6 1 5]
 

python_module/megengine/functional/utils.py

@@ -50,7 +50,6 @@ def accuracy(logits: Tensor, target: Tensor, topk: Union[int, Iterable[int]] = 1
     Outputs:
 
     .. testoutput::
-        :options: +NUMBER
 
         [0.] [0.375]
     """

python_module/megengine/module/activation.py

@@ -20,7 +20,7 @@ class Softmax(Module):
     .. math::
             \text{Softmax}(x_{i}) = \frac{exp(x_i)}{\sum_j exp(x_j)}
 
-    It is applied to an n-dimensional input Tensor and rescaling them so that the elements of the 
+    It is applied to an n-dimensional input Tensor and rescaling them so that the elements of the
     n-dimensional output Tensor lie in the range of `[0, 1]` and sum to 1.
 
     :param axis: An axis along which softmax will be applied. By default,
@@ -137,8 +137,8 @@ class PReLU(Module):
         ax, & \text{ otherwise }
         \end{cases}
 
-    Here :math:`a` is a learnable parameter. When called without arguments, `PReLU()` uses 
-    a single paramter :math:`a` across all input channel. If called with `PReLU(num_of_channels)`, 
+    Here :math:`a` is a learnable parameter. When called without arguments, `PReLU()` uses
+    a single paramter :math:`a` across all input channel. If called with `PReLU(num_of_channels)`,
     a seperate :math:`a` is used for each input channle.
 
     :param num_parameters: number of :math:`a` to learn, there is only two
@@ -218,7 +218,6 @@ class LeakyReLU(Module):
     Outputs:
 
     .. testoutput::
-        :options: +NUMBER
 
         [-0.08   -0.12  6.   10.  ]
 

python_module/megengine/module/embedding.py

@@ -21,7 +21,7 @@ class Embedding(Module):
     A simple lookup table that stores embeddings of a fixed dictionary and size.
 
     This module is often used to store word embeddings and retrieve them using indices.
-    The input to the module is a list of indices, and the output is the corresponding word embeddings. 
+    The input to the module is a list of indices, and the output is the corresponding word embeddings.
     The indices should less than num_embeddings.
 
     :param num_embeddings: size of embedding dictionary.
@@ -138,7 +138,6 @@ class Embedding(Module):
         Outputs:
 
         .. testoutput::
-            :options: +NUMBER
 
             [[[1.2 2.3 3.4 4.5 5.6]
               [0.1 1.1 2.1 3.1 4.1]