Finetuning Technique

This guides you through picking a fine-tuning technique (SFT, DPO, RLVR, or RLAIF) based on your use case, then validates whether your chosen model actually supports it on SageMaker. It assumes you've already selected a base model and have a use case spec ready. The workflow is straightforward: consults a reference guide to recommend a technique, runs a Python script to check compatibility against the model's available recipes, and either confirms you're good to go or sends you back to pick a different model. Useful safety check before you invest time in a fine-tuning approach that won't work with your model.