Computes Relu(x * weight + biases).
tf.compat.v1.nn.relu_layer( x, weights, biases, name=None )
Args | |
---|---|
x | a 2D tensor. Dimensions typically: batch, in_units |
weights | a 2D tensor. Dimensions typically: in_units, out_units |
biases | a 1D tensor. Dimensions: out_units |
name | A name for the operation (optional). If not specified "nn_relu_layer" is used. |
Returns | |
---|---|
A 2-D Tensor computing relu(matmul(x, weights) + biases). Dimensions typically: batch, out_units. |
© 2020 The TensorFlow Authors. All rights reserved.
Licensed under the Creative Commons Attribution License 3.0.
Code samples licensed under the Apache 2.0 License.
https://www.tensorflow.org/versions/r2.3/api_docs/python/tf/compat/v1/nn/relu_layer