• Followed the Flowster forum BERT classification tutorial in Colab. Learned how to use Colab to train the models.
• Learned how to use the models to generate post embeddings.
• Learned how to test recommender/classifier models, and deal with data imbalances in classifers.
Tools: Simple Transformers Library, Selenium Webdriver, Numpy, Pandas ,Matplotlib, Scikit, Wordcloud, NLTK, requests, genism, seaborn, Google Colab
• Touched base with a few team members through groups to gauge my progress and also learn how far our team is in consolidating everything into our final project.
• I tried to find the best language model that played nice with the Drowned dataset. This was largely done manually.