Fasttext: Pre-trained models - loading GloVe models

[loretoparisi]

I would like to use pre-trained word embedding GloVe models like the Twitter model ( (2B tweets, 27B tokens, 1.2M vocab, uncased, 25d, 50d, 100d, & 200d vectors ) available here
Do I need a conversion between the FastText and the GloVe models?

[gserapio]

I'd like to know this as well!

[cpuhrsch]

Hello @loretoparisi,

You can indeed load these vectors. All you need to add is the number of lines and the dimensionality at the beginning of the GloVe files. You'll see that these then match the .vec format that we use.

Thanks,
Christian

[loretoparisi]

@cpuhrsch thank you. What about Google's word2vec model GoogleNews-vectors-negative300.bin? There are some converters from GloVe, but any easier way?