English–Karakalpak Parallel Corpus v8

English–Karakalpak Parallel Corpus v8 is a high-quality bilingual parallel corpus designed for machine translation, natural language processing, and large language model research. The dataset contains sentence-aligned English and Karakalpak text pairs collected from multiple publicly available sources. Each English sentence corresponds to exactly one Karakalpak sentence. The corpus has been manually cleaned and aligned to improve translation quality. It is intended for research on low-resource languages, neural machine translation, and multilingual language models. The latest version of this dataset is maintained on Hugging Face:https://huggingface.co/datasets/bekan/english_karakalpak_parallel_corpus_v8