View source on GitHub
|
Computes Relu(x * weight + biases).
tf.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. |