Adaptation of Deep Bidirectional Multilingual Transformers for Russian Language
This work addresses the need for efficient and high-performing language models for Russian, offering an incremental improvement over existing approaches.
The paper tackles the problem of adapting multilingual masked language models to a specific language, showing that transfer learning from a multilingual to a monolingual model results in significant performance growth on tasks like reading comprehension, paraphrase detection, and sentiment analysis, while also reducing training time.
The paper introduces methods of adaptation of multilingual masked language models for a specific language. Pre-trained bidirectional language models show state-of-the-art performance on a wide range of tasks including reading comprehension, natural language inference, and sentiment analysis. At the moment there are two alternative approaches to train such models: monolingual and multilingual. While language specific models show superior performance, multilingual models allow to perform a transfer from one language to another and solve tasks for different languages simultaneously. This work shows that transfer learning from a multilingual model to monolingual model results in significant growth of performance on such tasks as reading comprehension, paraphrase detection, and sentiment analysis. Furthermore, multilingual initialization of monolingual model substantially reduces training time. Pre-trained models for the Russian language are open sourced.