In my experiments with pre-trained word2vec models for NLP tasks, I have so far come across two models - one trained on Google News dataset and another which has been trained on Wikipedia text corpus.

Are there other other pre-trained models available for the extraction of word2vec embedding ?

Google search on this issue only comes up with FastText, sentence2vec models etc.

Moreover, can a sentence2vec embedding (obtained from the same dataset) be used as an alternative to word2vec embedding (by naively assuming that the sentence contains only one word.)?

  • $\begingroup$ What do you want to do exactly ? Get an embedding word in order to process a classification ? Get a representation of your sentences ? $\endgroup$ – Kledou Dec 20 '18 at 9:40
  • $\begingroup$ @Kledou exactly. I just want to get a representation for some words from a pre-trained word2vec model for classification tasks $\endgroup$ – Upendra01 Dec 20 '18 at 11:14
  • $\begingroup$ (A) Do you specifically want English embeddings? (B) Why do you particularly want word2vec, as opposed to some other embedding of the word types? Without more information, pre-trained GloVe or FastText embeddings probably suit your needs as well. They’re all approximations to the same function. $\endgroup$ – Arya McCarthy Apr 5 at 0:13

There are several available models, you can check this: https://github.com/RaRe-Technologies/gensim-data


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.