Unbabel/M-Feedback-Collection

M-Feedback-Collection是一个多语言反馈收集数据集，包含英语、法语、葡萄牙语、中文、希腊语和印地语等6种语言。此外，它还包含80k个涉及8种语言对的机器翻译评估数据，包括英语-德语、英语-捷克语、英语-西班牙语、英语-乌克兰语、英语-俄语、英语-中文、英语-日语和英语-印地语。The M-Feedback-Collection is a multilingual feedback collection dataset that includes English, French, Portuguese, Chinese, Greek, and Hindi. In addition, it contains 80k machine translation evaluation instances across 8 language pairs: English-German, English-Czech, English-Spanish, English-Ukrainian, English-Russian, English-Chinese, English-Japanese, and English-Hindi.