tf.compat.v1.nn.xw_plus_b

View source on GitHub

Computes matmul(x, weights) + biases.

tf.compat.v1.nn.xw_plus_b(
    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 "xw_plus_b" is used.

Returns:

A 2-D Tensor computing matmul(x, weights) + biases. Dimensions typically: batch, out_units.