The following is an evaluation of UnigramTagger, one the simpler taggers offered by the NLTK.

Ancient Greek

To replicate this, clone https://github.com/cltk/greek_treebank_perseus and copy the .pos into your current directory, naming it greek_training_set.pos.

In [1]: 
from nltk.corpus.reader import TaggedCorpusReader
from nltk.tag import UnigramTagger
from nltk.tokenize import wordpunct_tokenize
In [2]: 
reader = TaggedCorpusReader('.', 'greek_training_set.pos')
train_sents = reader.tagged_sents()
In [3]: 
tagger = UnigramTagger(train_sents)
tagger.evaluate(train_sents)
Out[3]:
0.9196123340065213

Latin

To replicate this, clone https://github.com/cltk/latin_treebank_perseus and copy the .pos into your current directory, naming it latin_training_set.pos.

In [6]: 
reader = TaggedCorpusReader('.', 'latin_training_set.pos')
train_sents = reader.tagged_sents()
In [8]: 
unigram_tagger = UnigramTagger(train_sents)
unigram_tagger.evaluate(train_sents)
Out[8]:
0.8873793350017877