Yang Liu, Maosong Sun, Jingfang Xu, Xuancheng Huang
We have not lifted any functions out of this paper's repositories yet, so there is nothing we have run. The repositories linked to it are listed below.
Some links come from the archived Papers with Code dataset (CC BY-SA 4.0): attribution and licence.
Multi-source sequence generation (MSG) is an important kind of sequence generation tasks that takes multiple sources, including automatic post-editing, multi-source translation, multi-document summarization, etc. As MSG tasks suffer from the data scarcity problem and recent pretrained models have been proven to be effective for low-resource downstream tasks, transferring pretrained sequenceto-sequence models to MSG tasks is essential. Although directly finetuning pretrained models on MSG tasks and concatenating multiple sources into a single long sequence is regarded as a simple method to transfer pretrained models to MSG tasks, we conjecture that the direct finetuning method leads to catastrophic forgetting and solely relying on pretrained selfattention layers to capture cross-source information is not sufficient. Therefore, we propose a two-stage finetuning method to alleviate the pretrain-finetune discrepancy and introduce a novel MSG model with a fine encoder to learn better representations in MSG tasks. Experiments show that our approach achieves new state-of-the-art results on the WMT17 APE task and multi-source translation task using the WMT14 test set. When adapted to documentlevel translation, our framework outperforms strong baselines significantly. 1
The same record, over MCP at https://syntology.ai/mcp:
get_harvested_code_for_paper("2105.14809")
get_code_for_paper("2105.14809")
have("2105.14809")
The run record, dated, one paper per request, free:
curl https://syntology.ai/api/ran/2105.14809.json
A badge for a README (the split and the date, never a ratio):
[](https://syntology.ai/paper/2105.14809)
Connect an agent — have() is free.