Oct 20, 2023 · We demonstrate that our method can reduce run-time by up to 62.2%, with only a 2% reduction in performance, and in some cases, even improve the ...
Dec 6, 2023 · We propose a novel approach for improving the efficiency of FiD, with a combined ap- proach of Token Filtering and decoder layer reduction, ...
We demonstrate that our method can reduce run-time by up to 62.2%, with only a 2% reduction in performance, and in some cases, even improve the performance ...
This paper optimizes retrieval-augmented language models, specifically the FiD model, to reduce decoding time in open domain question answering (ODQA) tasks. It ...
This work analyzes the contribution and necessity of all the retrieved passages to the performance of reader models, and proposes eliminating some of the ...
This repository contains the implementation of the method presented in the paper "Optimizing Retrieval-augmented Reader Models via Token Elimination".
People also search for
This method prompts agents to sample some potential responses as perturbations, evaluates the impact of these perturbations on the whole RAG system, and ...
View recent discussion. Abstract: Fusion-in-Decoder (FiD) is an effective retrieval-augmented language model applied across a variety of open-domain tasks, ...
In this work, we analyze the contribution and necessity of all the retrieved passages to the performance of reader models, and propose eliminating some of the ...
Fingerprint. Dive into the research topics of 'Optimizing Retrieval-augmented Reader Models via Token Elimination'. Together they form a unique fingerprint.