Function Differences with tf.nn.relu

tf.nn.relu

tf.nn.relu(features, name=None) -> Tensor

For more information, see tf.nn.relu.

mindspore.nn.ReLU

class mindspore.nn.ReLU()(x) -> Tensor

For more information, see mindspore.nn.ReLU.

Differences

TensorFlow: PReLU activation function.

MindSpore: MindSpore API implements the same function as TensorFlow, but the parameter setting is different, and the operator needs to be instantiated first.

Categories	Subcategories	TensorFlow	MindSpore	Differences
Parameters	parameter 1	features	x	Same function, different parameter names
	parameter 2	name	-	Not involved

Code Example

The two APIs implement the same function, but the TensorFlow operator is functional and can accept input directly. The operator in MindSpore needs to be instantiated first.

# TensorFlow
import tensorflow as tf

x = tf.constant([[-1.0, 2.2], [3.3, -4.0]], dtype=tf.float16)
out = tf.nn.relu(x).numpy()
print(out)
# [[0.  2.2]
#  [3.3 0. ]]

# MindSpore
import mindspore
import mindspore.nn as nn
from mindspore import Tensor
import numpy as np

x = Tensor(np.array([[-1.0, 2.2], [3.3, -4.0]]), mindspore.float16)
relu = nn.ReLU()
output = relu(x)
print(output)
# [[0.  2.2]
#  [3.3 0. ]]