This paper presents the contribution of the LGI2P (Labo-ratoire de Génie Informatique et d'Ingénierie de Production) team from IMT Mines Alès to the Hate Speech and Offensive Content Identification in Indo-European Languages (HASOC) 2019 shared task. This challenge aims at automatically identifying hate speech content in social media through three sub-tasks, each available in three different languages (En-glish, German and Hindi). We are interested in sub-tasks A and B, requiring to (A) classify tweets as offensive or as non offensive, and (B) to further classify offensive tweets from sub-task A as hate speech, offensive speech or profane. We trained a fastText model for each proposed language and obtained promising results on the Hindi dataset for both sub-tasks A and B.