Search Results: sagemaker-training

finetuning

Generates a Jupyter notebook that fine-tunes a base model using SageMaker serverless training jobs. Use when the user says "start training", "fine-tune my model", "I'm ready to train", or when the plan reaches the finetuning step. Supports SFT, DPO, and RLVR trainers, including RLVR Lambda reward function creation.