Summary of Cross-lingual Transfer Of Multilingual Models on Low Resource African Languages, by Harish Thangaraj et al.
Cross-lingual transfer of multilingual models on low resource African Languages
by Harish Thangaraj, Ananya Chenat, Jaskaran Singh Walia, Vukosi Marivate
First submitted to arxiv on: 17 Sep 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper investigates the performance of large multilingual and monolingual natural language processing (NLP) models on low-resource languages. The authors benchmark transformer-based architectures like Multilingual BERT (mBERT), AfriBERT, and BantuBERTa against neural-based models such as BiGRU, CNN, and char-CNN. They train the models on Kinyarwanda and test them on Kirundi, with fine-tuning applied to assess performance improvement and catastrophic forgetting. The results show that AfriBERT achieved the highest cross-lingual accuracy of 88.3% after fine-tuning, while BiGRU emerged as the best-performing neural model with 83.3% accuracy. This study highlights the strong cross-lingual transfer capabilities of multilingual models in resource-limited settings. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This research compares big language models to see how well they work on languages that don’t have a lot of data available. The models were trained on one language and tested on another. The best-performing model was AfriBERT, which got 88% of the answers correct after adjusting its training. Another model called BiGRU also did well, getting 83% of the answers right. This study shows that these big language models can be very good at understanding languages even if they’re not as common or don’t have a lot of data available. |
Keywords
» Artificial intelligence » Bert » Cnn » Fine tuning » Natural language processing » Nlp » Transformer
