NLTK POS classifier using custom provided data

Question

I was given code in the following format:

1   i PRP
2   'd  MD
3   like    VB
4   to  TO
5   go  VB
6   to  IN
7   a   DT
8   fancy   JJ
9   restaurant  NN
10  .   .

I am having trouble getting it into a format that the nltk.Unigram model will accept. Here is what I've tried so far.

from nltk import word_tokenize
import numpy as np


f = open('POS-training.txt', 'r')
content = f.read()
text = word_tokenize(content)

# Removing all digits from the text
cleaned_txt = [text for text in text if not text.isdigit()]
 
# Using an iterator and the zip() method to zip all the words with their respective tag 
it = iter(cleaned_txt)
tuple_txt = [*zip(it, it)]      

print(tuple_txt[:50])

Output:

[('i', 'PRP'), ("'d", 'MD'), ('like', 'VB'), ('to', 'TO'), ('go', 'VB'), ('to', 'IN'), ('a', 'DT'), ('fancy', 'JJ'), ('restaurant', 'NN'), ('.', '.'), ('i', 'PRP'), ("'d", 'MD'), ('like', 'VB'), ('french', 'JJ'), ('food', 'NN'), ('.', '.'), ('next', 'JJ'), ('thursday', 'NN'), ('.', '.'), ('next', 'JJ'), ('thursday', 'NN'), ('.', '.')]

It provides a list of tuples as expected but when trying the following:

uni = nltk.UnigramTagger(train=tuple_txt)

I get the following error:

ValueError: not enough values to unpack (expected 2, got 1)

I've looked through documentation and can't seem to find anything. Anyone know how?