Creates a tensor filled with a scalar value.
tf.fill( dims, value, name=None )
This operation creates a tensor of shape
dims and fills it with
# Output tensor has shape [2, 3]. fill([2, 3], 9) ==> [[9, 9, 9] [9, 9, 9]]
tf.fillonly supports scalar contents, whereas
tf.constantsupports Tensor values.
tf.fillcreates an Op in the computation graph that constructs the actual Tensor value at runtime. This is in contrast to
tf.constantwhich embeds the entire Tensor into the graph with a
tf.fillevaluates at graph runtime, it supports dynamic shapes based on other runtime Tensors, unlike
Tensor. Must be one of the following types:
int64. 1-D. Represents the shape of the output tensor.
Tensor. 0-D (scalar). Value to fill the returned tensor.
name: A name for the operation (optional).
Tensor. Has the same type as
Equivalent to np.full