def setup(self): # Creating the values for two tensors val1 = np.array([[1, 5], [10, -3]]) val2 = np.array([[2, 6, 9], [-1, 2, 4]]) # Declaring the two input tensors tensor1 = self.declare_input('tensor1', val=val1) tensor2 = self.declare_input('tensor2', val=val2) # Creating the output for matrix multiplication self.register_output('ElementwiseMinWrongSize', ot.min(tensor1, tensor2))
def setup(self): m = 2 n = 3 # Shape of the three tensors is (2,3) shape = (m, n) # Creating the values for two tensors val1 = np.array([[1, 5, -8], [10, -3, -5]]) val2 = np.array([[2, 6, 9], [-1, 2, 4]]) # Declaring the two input tensors tensor1 = self.declare_input('tensor1', val=val1) tensor2 = self.declare_input('tensor2', val=val2) # Creating the output for matrix multiplication self.register_output('ElementwiseMin', ot.min(tensor1, tensor2))
def setup(self): m = 2 n = 3 o = 4 p = 5 q = 6 # Shape of a tensor tensor_shape = (m, n, o, p, q) num_of_elements = np.prod(tensor_shape) # Creating the values of the tensor val = np.arange(num_of_elements).reshape(tensor_shape) # Declaring the tensor as an input ten = self.declare_input('tensor', val=val) # Computing the minimum across the entire tensor, returns single value self.register_output('ScalarMin', ot.min(ten))
def setup(self): m = 2 n = 3 o = 4 p = 5 q = 6 # Shape of a tensor tensor_shape = (m, n, o, p, q) num_of_elements = np.prod(tensor_shape) # Creating the values of the tensor val = np.arange(num_of_elements).reshape(tensor_shape) # Declaring the tensor as an input ten = self.declare_input('tensor', val=val) # Computing the axiswise minimum on the tensor axis = 1 self.register_output('AxiswiseMin', ot.min(ten, axis=axis))