AWS lets you write your own model-training loop on SageMaker without managing GPUs
AWS made training sessions generally available on 22 October for Amazon SageMaker AI serverless model customisation. Developers can write their own post-training loop, with custom loss functions and reward signals, while SageMaker handles GPU allocation, distributed execution, fault recovery and checkpoints. Supported methods include SFT, DPO, RLVR, RLAIF and multi-turn RL, with LoRA or full fine-tuning.
You pay for the tokens SageMaker processes, and billing pauses while your own reward function scores a batch. Supported models at launch are gpt-oss-20b, gemma-4-31B-it, Qwen3.6-27B and Qwen3.8-27B. Data and checkpoints stay in your own S3 buckets, and AWS says it does not use your data or customised weights to train other models.
Custom reinforcement learning used to mean renting and babysitting GPUs. If you can write a check for a good outcome, you can now train on it without running the infrastructure yourself.
Drafted with AI from the original source, then checked and edited by Sridhar Mukkandi. Spotted a mistake? Write to hello@tarkika.com and we'll correct it here.