MetaphorTrans

该数据集包含19,000个训练样本和2,000个测试样本，专注于涉及隐喻或比拟的英文文学句子及其对应的思考和中文翻译。每个样本都包括原始英文句子、作者的思考过程以及相应的中文译文。该数据集的源句子选自英文文学书籍，旨在用于深度推理翻译任务，特别是由于隐喻的复杂性，这类翻译任务需要仔细思考和加工。

This dataset includes 19,000 training samples and 2,000 test samples, focusing on English literary sentences involving metaphors or analogies, as well as their corresponding reasoning processes and Chinese translations. Each sample consists of the original English sentence, the author's reasoning process, and the corresponding Chinese translation. The source sentences of this dataset are selected from English literary books, and it is developed for deep reasoning translation tasks. Particularly, given the complexity of metaphors, such translation tasks require careful thinking and meticulous processing.

搜集汇总

背景与挑战

背景概述

MetaphorTrans是一个专注于深度推理翻译任务的数据集，包含19,000个训练样本和2,000个测试样本，每个样本提供英文文学句子、作者的思考过程及中文译文。该数据集源自英文文学书籍，旨在解决隐喻或比拟的复杂翻译问题，强调翻译过程中的仔细思考和加工。