CTU-Hacktivist-RU_EN-1: A Dataset with Russian English Translation

This dataset was created as part of the research on using LLMs to translate hacktivist messages from Russian to English. We split the dataset in two: CTU-Hacktivist-Translations-FineTuning-RU_EN-v0.2.jsonl: The dataset used to fine-tune the LLM model in JSONL format. The prompt used for fine-tuning the LLM The original text in Russian from the hacktivist group NoName(057)16 or vocabulary annotations created by a language expert. The English translation text for each Russian message or vocabulary CTU-Hacktivist-Translations-Evaluation-RU_EN-v0.5.jsonl: The dataset used to evaluate the fine-tuned LLM Model in JSONL format. The original text in Russian The ground-truth translation in English The translation by the LLM model without fine-tuning The translation by the LLM model with fine-tuning The translation by Google Translate The translation by DeepL The ground-truth translation was generated with specific instructions so that the translations will preserve the original punctuation, including the use of guillemets «».