Implement keras-core ops in as methods of a class for easier method chaining.
from keras_core import ops
from keras_tensor_pp import KerasTensorPP
(KerasTensorPP(ops.arange(10))
.clip(4, 6)
.cumsum()
.log()
)
# <KerasTensorPP<tf.Tensor: shape=(10,), dtype=float64, numpy=
# array([1.38629436, 2.07944154, 2.48490665, 2.77258872, 2.99573227,
# 3.21887582, 3.4339872 , 3.61091791, 3.76120012, 3.8918203 ])>>