voc='./voc_emb/voc' / word_embeding samples?

[johndpope]

voc='./voc_emb/voc',
emb_file = './voc_emb/word_embeding',

what data structure is this? where to download?
can you explain how to recreate/ download.

[johndpope]

is it like this

word_vectors = read_glove_vectors(word_vectors_path, token_to_index.keys())
l = list(word_vectors)

voc_emb = {a for (a,b) in l}

len(voc_emb)

[johndpope]

I think I found it.

def load_emb(dataset):
if not hparam.voc:
dataset.generate_vocab()
else:
dataset.load_voc_pickle()

if not hparam.emb_file:
word_emb = dataset.save_emb_pickle()
else:
word_emb = dataset.load_emb_pickle()