Note the leading zero indicating that the total number of columns is more than 9. But since we are not tracking this, the moment we use the seed_input argument, the transformation only infers from the given data and not from the data used during training.
The same is true in this case. The transformation should note that the hhsize variable has values that may exceed 10, so it should truncate the leading 0 as shown in the second image.
Note the leading zero indicating that the total number of columns is more than 9. But since we are not tracking this, the moment we use the
seed_input
argument, the transformation only infers from the given data and not from the data used during training.The same is true in this case. The transformation should note that the
hhsize
variable has values that may exceed 10, so it should truncate the leading 0 as shown in the second image.