A conversational (ChatML) dataset for training LLMs to generate residential floor plans following US architectural conventions and IRC (International Residential Code) standards.

Dataset Details

Size: 12,000 examples (11,500 train / 500 test) Format: Conversational ChatML (messages column with system/user/assistant turns) Language: English License: Apache 2.0

	Composition

Source Count Type… See the full description on the dataset page: https://huggingface.co/datasets/Nithins03/us-architectural-floorplan-sft.