racineai/

VDR_Geotechnie

	VDR - Organized, Grouped, Cleaned




	
	Geotechnie Vision DSE

Intended for image/text to vector (DSE)

Made with https://github.com/RacineAIOS/VDR_pdf-to-parquet This dataset was created by scraping PDF documents from online sources and generating relevant synthetic queries. We used Google's Gemini 2.0 Flash Lite model in our custom pipeline to produce the queries, allowing us to create a diverse set of questions based on the document content.… See the full description on the dataset page: https://huggingface.co/datasets/racineai/VDR_Geotechnie.