How to specify a 2-dimension Input layer?

TensorFlow
,
1

I am working on a sine wave GAN model. The generator generates TensorSpec(shape=(None, 2), dtype=tf.float64, name=None) output.
I try this for the discriminator:

        self.model = models.Sequential([
            layers.Input(shape=(,2)),

But it’s an invalid syntax. The first dimension is reserved for the batch size.
When I try layers.Input(shape=(2)),, I get

    raise ValueError(f"Cannot convert '{shape}' to a shape.")
ValueError: Cannot convert '2' to a shape.
3

tf.keras.Input  |  TensorFlow v2.16.1 answers my question. It’s specified as (2,)