locuslab/tofu_ft_phi-1.5

The locuslab/tofu_ft_phi-1.5 is a 1.4 billion parameter Phi-1.5 model, developed by LocusLab, fine-tuned on the TOFU (Task of Fictitious Unlearning) dataset. This model specializes in evaluating and enhancing the ability of large language models to unlearn specific data points. It is designed for research in data privacy, regulatory compliance, and understanding knowledge retention in AI systems, offering a benchmark for selective data unlearning.