thebajajra/Ecom-niverse

Ecom-niverse是一个综合电子商务领域的数据集，旨在通过精炼广泛的网络数据来分离出具有零售或购物背景的内容，用于持续预训练LLM和其他Encoder-only模型，使其更好地理解产品描述、价格和其他与商业相关的文本。Ecom-niverse is a comprehensive e-commerce domain dataset designed to isolate content with retail or shopping context from a broad web dataset for continual pre-training of LLMs and other Encoder-only models to better understand product descriptions, prices, and other commerce-related text.