Training BERT Text Classifier on Tensor Processing Unit (TPU)
Training hugging face most famous model on TPU for social media Tunisian Arabizi sentiment analysis.
Introduction
The Arabic speakers usually express themself in local dialect on social media, so Tunisians use Tunisian Arabizi which consists of Arabic written in form of Latin alphabets. The sentiment analysis relies on cultural knowledge and word sense with contextual information. We will be using both Arabizi dialect and sentimental analysis to solve the problem in this project.
The competition is hosted on Zindi which is organized by AI4D. In this competition, we are provided with train and test datasets. The train data set consists of Arabizi text with the sentiment of that text. We need to build a model and improve our metric to score the top position in the competition leader board.
The dataset was collected from different social media platforms and the objective of this project is to predict labels of the test dataset. The label will determine whether our sentence has positive, negative,