Keras Keras3 not working with Sparce Tensors

The code from the manual, "Working with sparse tensors" (link):

x = tf.keras.Input(shape=(4,), sparse=True)
y = tf.keras.layers.Dense(4)(x)
model = tf.keras.Model(x, y)

sparse_data = tf.sparse.SparseTensor(
    indices = [(0,0),(0,1),(0,2),
               (4,3),(5,0),(5,1)],
    values = [1,1,1,1,1,1],
    dense_shape = (6,4)
)

model(sparse_data)

model.predict(sparse_data)

Produces the error,

ValueError: Unrecognized data type: x=SparseTensor(indices=tf.Tensor( ....

This doesnt happen with keras 2.15, only keras 3

Comment From: fchollet

The ability to call fit/predict with sparse tensors is coming. For now, you can use a tf.data.Dataset that yields sparse tensors when calling fit/predict.

Like this:

import tensorflow as tf
import keras

print(keras.version())  # 3.0.5

x = keras.Input(shape=(4,), sparse=True)
y = keras.layers.Dense(4)(x)
model = keras.Model(x, y)

sparse_data = tf.sparse.SparseTensor(
    indices = [(0,0),(0,1),(0,2),
               (4,3),(5,0),(5,1)],
    values = [1,1,1,1,1,1],
    dense_shape = (6,4)
)

model(sparse_data)


sparse_dataset = tf.data.Dataset.from_tensor_slices(sparse_data).batch(1)
model.predict(sparse_dataset)

Comment From: mdhvgoyal

I would like to work on this.