Pretrained FastText word vector for English
https://github.com/facebookresearch/fastText
Usage
import fasttext.util
ft = fasttext.load_model('cc.en.300.bin')
ft.get_word_vector('hello')
Inference API (serverless) does not yet support generic models for this pipeline type.