A technology firm is developing a single, large language model trained on a vast corpus of text from the internet, with the goal of creating a universal translation and content generation tool. The training data is predominantly in English (60%), with smaller portions of Spanish, French, and Mandarin (10% each), and a very small fraction of Swahili (less than 0.1%). Analyze the specific types of performance issues the model is likely to exhibit when processing Swahili, and explain the fundamental reasons for these anticipated shortcomings based on the principles of model training.