Consider this code that downloads models and tokenizers to disk and then uses BertTokenizer.from_pretrained to load the tokenizer from disk.

ISSUE: BertTokenizer.from_pretrained() does not seem to be compatible with Python's native pathlib module.

# -*- coding: utf-8 -*-
"""
Created on: 25-04-2020
Author: MacwanJ

ISSUE:

Several parts of the op sec like the main op description, attributes, input and output descriptions become part of the binary that consumes ONNX e.g. onnxruntime causing an increase in its size due to strings that take no part in the execution of the model or its verification.

Setting __ONNX_NO_DOC_STRINGS doesn't really help here since (1) it's not used in the SetDoc(string) overload (s

Split the code from create_sandbox into two separate functions and remove the noqa: C901.

Describe alternatives you've considered
Simplify the function such that no split is required.

Additional context
Code quality:

2020-04-29T13:13:32.5968940Z ./syft/sandbox.py:12:1: C901 'create_sandbox' is too complex (26)
2020-04-29T13:13:32.5969257Z def create_sandbox(gbs,