Deep Learning Essentials

FastText

FastText (https://fasttext.cc/) is a library for efficient learning of word representations and sentence classification. The main advantage of FastText embeddings over Word2Vec is to take into account the internal structure of words while learning word representations, which could be very useful for morphologically rich languages, and also for words that occur rarely.

The main difference between Word2Vec and FastText is that for Word2Vec, the atomic entity is each word, which is the smallest unit to train on. On the contrary, in FastText, the smallest unit is character-level n-grams, and each word is treated as being composed of character n-grams. For example, the word vector of happy with an n-gram of minimum size three and maximum ...