Rio Yokota

A new method shows variational learning can excel in training complex models.

A robust Japanese corpus created from Common Crawl data improves LLM performance.

Enhancing Japanese language models using English knowledge boosts performance significantly.

Explore local learning methods transforming neural network training.

Researchers develop methods to teach computers to process invoices while protecting privacy.

A guide to speeding up large language model training with parallelism and memory management.

A look at how IVON improves AI model performance efficiently.

SuperGCN improves training efficiency for large graphs using advanced CPU techniques.

Learn how to reduce communication overhead in deep learning models to improve training speed.

Exploring the importance of developing large language models in local languages.