datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
RAG_Evaluation_DatasetOdia_English_Sentences_Aligned_81k
🌐 Vaqas AI: Odia-English Sentence-Aligned Precision Corpus
Created and Maintained by Vaqas AI Creator & Owner: Vaqas Ahmed
🚀 Why This Dataset?
Standard parallel corpora often consist of large paragraphs that exceed the 4k or 8k context windows typical of LLM training, leading to truncated data and poor model performance.
The Vaqas AI Sentence-Aligned Corpus was engineered to solve this. We transformed the original vaqasai/Odia_English_News dataset by decomposing dense… See the full description on the dataset page: https://huggingface.co/datasets/vaqasai/Odia_English_Sentences_Aligned_81k.wat24_text_to_text_translation