node-word2vec
node-word2vec copied to clipboard
Do I need to train the model or I can use the default values straight away to calculate words' similarity?
Great implementation! I didn't understand from the description if I have to train the model on an existing corpus or if I can also use the default values to calculate similarity. If so, where do the default values come from then? Thanks!
The example vectors.txt included in the fixtures directory of the example directory was created using Shakespeare's Hamlet as a corpus and serves only as an example. There is no default model coming with the library. The link to the project provides several examples of how to obtain a file of pre-trained word vectors that can be used.