Full fine-tuned LLaMA-3.2-3B-Instruct on medical instruction datasets plus 15% MQR data from MedCalc-Bench-v1.0 to improve medical quantitative reasoning.
Model Details
Base model: meta-llama/Llama-3.2-3B-Instruct
Developed by: Caleb King
Type: Text generation / Instruction fine-tuned language model
License: llama3.2
Language: English
Training
Dataset: Med-MathInstruct (MedInstruct combined from BioInstruct and AlpaCare-MedInstruct-52k, plus MQR data from MedCalc-Bench-v1.0)
