This snippet of code was posted in 2014 and slightly revised in 2016 and 2017. It was more of a quick'n'dirty script than a polished tool. It is made only for Linux and in Python 2, which has since become outdated.
I currently do not use it, and I suggest you avoid it as well. Please do not expect support for using this script.
🔥 If you need an alternative, @glaucocustodio has kindly suggested EasyVPN in this comment.
The rest of the README is left for historical purposed.
The new ModelBuilder class allows user to use direct methods to fit, predict,save, load. A user can create any model as they want to and just inherit the ModelBuilder class and use the predefined methods.
Let's learn more about using an example
class LinearModel(ModelBuilder):
_model_type = 'LinearModel'
version = '0.1'
def _build(self):
# data
| import cloudpickle | |
| from pathlib import Path | |
| import os | |
| import arviz as az | |
| import numpy as np | |
| import pymc as pm | |
| import trace | |
| import logging | |
| import sys |
| import cloudpickle | |
| from pathlib import Path | |
| class base: | |
| def __init__(self): | |
| return | |
| def save(self,file_prefix,filepath,save_format=None): | |
| if save_format == 'h5': | |
| extension = '.hdf5' | |
| else: |