Iker/

Document-Translation-en-es

This dataset contains 10533 news articles from ELiRF/dacsa translated from Spanish to English using GPT-3.5-turbo. The dataset is intended to be used for training a model to translate text from English to Spanish and vicerversa. The dataset is also usefull to evaluate document level machine translation models. We use the following prompt

def get_conversation(text: str, id: int) -> str: messages = { "custom_id": str(id), "method": "POST", "url":… See the full description on the dataset page: https://huggingface.co/datasets/Iker/Document-Translation-en-es.