while the second needs several hours
Text Classifier Algorithms in Machine Learning
Roman Trusov

Are you talking about the Joulin et. al. paper here?

Because, in the paper they mention being able to train 1 billion words in 10 minutes and also being able to beat the char-cnn results by Zhang et. al. using bigram augmentation

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.