felipemaiapolo/legalnlp

LegalNLP - Natural Language Processing Methods for the Brazilian Legal Language

About me 🙋:

• 🔎 Statistics PhD student at the University of Michigan, USA.
• 🎓 MSc in Statistics from the University of Sao Paulo, Brazil.
• 😄 Pronouns: he/him

legalnlp's Issues

Incorrect naming convention or vector ngrams upload for FastText (SkipGram, 300 dim) in Google drive

I believe the file fasttext_cbow_size_300_window_15_epochs_20.wv.vectors_ngrams.npy available at this project's Google drive is either incorrectly labeled as a result of "cbow" algorithm (see picture below) or was incorrectly uploaded in place of another file.

This makes the model fail to load:

# gensim==4.2.0
import gensim.models
gensim.models.FastText.load("fasttext_sg_size_300_window_15_epochs_20")

FileNotFoundError: [Errno 2] No such file or directory: 'fasttext_sg_size_300_window_15_epochs_20.wv.vectors_ngrams.npy'

If the file was correctly uploaded, by simply renaming it from "cbow" to "sg" supposedly solves this issue, but I'm not sure if this is the case.

Could you please confirm this information?

Thanks in advance.