LOCOST: State-Space Models for Long Document Abstractive Summarization
Florian Le Bronnec, Song Duong, Mathieu Ravaut, Alexandre Allauzen, Nancy F. Chen, Vincent Guigue, Alberto Lumbreras, Laure Soulier, Patrick Gallinari
Code Available — Be the first to reproduce this paper.
ReproduceCode
- github.com/flbbb/locost-summarizationOfficialIn paperpytorch★ 29
Abstract
State-space models are a low-complexity alternative to transformers for encoding long sequences and capturing long-term dependencies. We propose LOCOST: an encoder-decoder architecture based on state-space models for conditional text generation with long context inputs. With a computational complexity of O(L L), this architecture can handle significantly longer sequences than state-of-the-art models that are based on sparse attention patterns. We evaluate our model on a series of long document abstractive summarization tasks. The model reaches a performance level that is 93-96% comparable to the top-performing sparse transformers of the same size while saving up to 50% memory during training and up to 87% during inference. Additionally, LOCOST effectively handles input texts exceeding 600K tokens at inference time, setting new state-of-the-art results on full-book summarization and opening new perspectives for long input processing.