This is the repository containing LoRA checkpoints tuned on math datasets with 1.5B foundation model, served as training data for DnD.
