AWS Shows SFT + DPO Combo to Improve Small-Model Tool-Calling Accuracy on SageMaker

AWS Machine Learning Blog·medium signal

AWS published a hands-on walkthrough using Supervised Fine-Tuning and Direct Preference Optimization together to raise the tool-calling accuracy of a small language model on SageMaker AI. The recipe is directly actionable for builders trying to get reliable function calling out of cheaper SLMs instead of defaulting to frontier models.