FiD-Light: Efficient and Effective Retrieval-Augmented Text Generation
This work addresses the complexity and long-input handling issues in retrieval-augmented generation for knowledge-intensive tasks, offering incremental improvements in efficiency and provenance precision.
The authors tackled the efficiency and effectiveness of retrieval-augmented text generation models by introducing FiD-Light, which improved the Pareto frontier between query latency and effectiveness on seven KILT tasks and set new state-of-the-art results on six tasks.
Retrieval-augmented generation models offer many benefits over standalone language models: besides a textual answer to a given query they provide provenance items retrieved from an updateable knowledge base. However, they are also more complex systems and need to handle long inputs. In this work, we introduce FiD-Light to strongly increase the efficiency of the state-of-the-art retrieval-augmented FiD model, while maintaining the same level of effectiveness. Our FiD-Light model constrains the information flow from the encoder (which encodes passages separately) to the decoder (using concatenated encoded representations). Furthermore, we adapt FiD-Light with re-ranking capabilities through textual source pointers, to improve the top-ranked provenance precision. Our experiments on a diverse set of seven knowledge intensive tasks (KILT) show FiD-Light consistently improves the Pareto frontier between query latency and effectiveness. FiD-Light with source pointing sets substantial new state-of-the-art results on six KILT tasks for combined text generation and provenance retrieval evaluation, while maintaining reasonable efficiency.