Existing long-text generation methods primarily concentrate on producing lengthy texts from short inputs, neglecting the long-input and long-output tasks. Such tasks have numerous practical applications while lacking available benchmarks. Moreover, as the input grows in length, existing methods inevitably encounter the "lost-in-the-middle" phenomenon. In this paper, we first introduce a Long Input and Output Benchmark (LongInOutBench), including a synthetic dataset and a comprehensive evaluation framework, addressing the challenge of the missing benchmark. We then develop the Retrieval-Augmented Long-Text Writer (RAL-Writer), which retrieves and restates important yet overlooked content, mitigating the "lost-in-the-middle" issue by constructing explicit prompts. We finally employ the proposed LongInOutBench to evaluate our RAL-Writer against comparable baselines, and the results demonstrate the effectiveness of our approach. Our code has been released at https://github.com/OnlyAR/RAL-Writer.

Large language models (LLMs) (Zhao et al., 2023) We minimize the dependencies of these modules present remarkable reasoning capabilities and empower on specific models and datasets, allowing the framework a wide range of applications, such as question to flexibly scale to hundreds of models and answering (Jiang et al., 2023b), machine translation datasets. Concretely, we first establish a model registry (Wang et al., 2023c; Jiao et al., 2023a), and where the Model Loader can precisely attach information extraction (Jiao et al., 2023b). Subsequently, adapters to the pre-trained models by identifying a substantial number of LLMs are developed exact layers. Then we develop a data description and accessible through open-source communities.