This repository contains implementations of ULMFit by Jeremy Howard and Sebastian Ruder applied to NLP tasks for the Spanish Language.
This approach achieves SOTA for Classification of Spanish Tweets as of 2017 in the GeneralTASS Dataset (the test set was not available so a part of the training set was kept separate for testing).
The backbone model weights and int to string mapping can be found here. The TASS dataset can be found here.