Propose a list of other word vector algorithms to train on that corpus.

lang-uk / word-vectors-uk

Scripts, stats and tasks for ubertext vectors trained using different models

1 stars 0 forks source link

Propose a list of other word vector algorithms to train on that corpus. #3

Open dchaplinsky opened 3 years ago

dchaplinsky commented 3 years ago

For ubertext 1.0 we've used following algorithms:

LexVec (it seems that now it works with subwords, yay!)
Word2Vec
GloVe

Params that was previously used for training is described in article, linked above. Should we add something to the list (ELMo?). Should we remove something? I'm open for your ideas.

Mayya98 commented 3 years ago

The list of other word vector algorithms to train on that corpus:

Embedding Layer https://machinelearningmastery.com/what-are-word-embeddings/,
ELMO,
BERT https://www.kaggle.com/residentmario/notes-on-word-embedding-algorithms.

dchaplinsky commented 3 years ago

Great, thanks.

What about word2vec/lexvec/GloVe?

On Sun, Jul 11, 2021 at 5:40 PM Mayya98 @.***> wrote:

The list of other word vector algorithms to train on that corpus:

Embedding Layer,

ELMO,

BERT https://www.kaggle.com/residentmario/notes-on-word-embedding-algorithms .

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/lang-uk/word-vectors-uk/issues/3#issuecomment-877811024, or unsubscribe https://github.com/notifications/unsubscribe-auth/AABAA4UOKWD6K62OAQ43RRDTXGUNPANCNFSM5AC75PTQ .

Mayya98 commented 3 years ago

Great, thanks. What about word2vec/lexvec/GloVe? …

They are mentioned above. I only added other word vector algorithms to the existing ones.