tms-NET sampling in Python

I have been trying find some pre-implementation of tms-NET sampling in python (as I couldn't implement by myself) for my projects wherein I want to perform sampling on numpy array of shape (num_of_data, embedding_dimensions=100). I couldn't find it.

Can anyone help me with my problem ?

I also try using Latin Hypercube Sampling however the scale of my data in my numpy array is not the with the value sample by pyDOE2 implementation of LHS.

from pyDOE2 import lhs

num_samples = X.shape[0]
num_dimensions = X.shape[1] # 100

lhs_samples = lhs(num_dimensions, samples=5)

The X array has values range from 0 to 1000 while most of the value samples by lhs is around 0.

Additionally, the description of my projects stats that

"If the size of the remaining X is equal to zero, then the selected train sample represents the complete data, and it is considered to be good. Otherwise, the data points are sampled from the remaining X using (t,m,s)-Nets. The samples which are selected using (t,m,s)-Nets are appended to the train sample for improving the representativeness of the data. "

I am not sure if my approach is correct.

If anyone could help, I would be very glad for it.