Multilingual Machine Translation with Hyper-Adapters
This addresses the problem of parameter inefficiency and negative interference in multilingual machine translation for NLP researchers and practitioners, offering an incremental improvement over existing adapter methods.
The paper tackled negative interference in multilingual machine translation by introducing hyper-adapters, which generate adapters from embeddings to improve parameter efficiency and enable positive transfer across languages, achieving up to 12 times fewer parameters for the same performance and consistently outperforming regular adapters with equal parameters and FLOPS.
Multilingual machine translation suffers from negative interference across languages. A common solution is to relax parameter sharing with language-specific modules like adapters. However, adapters of related languages are unable to transfer information, and their total number of parameters becomes prohibitively expensive as the number of languages grows. In this work, we overcome these drawbacks using hyper-adapters -- hyper-networks that generate adapters from language and layer embeddings. While past work had poor results when scaling hyper-networks, we propose a rescaling fix that significantly improves convergence and enables training larger hyper-networks. We find that hyper-adapters are more parameter efficient than regular adapters, reaching the same performance with up to 12 times less parameters. When using the same number of parameters and FLOPS, our approach consistently outperforms regular adapters. Also, hyper-adapters converge faster than alternative approaches and scale better than regular dense networks. Our analysis shows that hyper-adapters learn to encode language relatedness, enabling positive transfer across languages.