reglab-rrc/mistral-rrc
TEXT GENERATIONConcurrency Cost:1Model Size:7BQuant:FP8Ctx Length:4kPublished:Nov 23, 2023License:mitArchitecture:Transformer0.0K Open Weights Warm

The reglab-rrc/mistral-rrc is a 7 billion parameter Mistral-based causal language model developed by reglab-rrc, fine-tuned for specialized text classification and extraction. It is designed to identify and extract racial covenants from legal property deeds, supporting legal reform efforts. The model achieves high precision (1.000) and recall (0.994) in detecting these covenants, making it suitable for automating the review of historical legal documents. Its primary strength lies in accurately pinpointing and correcting racially restrictive language within property records.

Loading preview...