AWS Adds a Hands-On Multi-Turn RL Guide for SageMaker AI
AWS Machine Learning Blog published an end-to-end guide to multi-turn reinforcement learning (MTRL) on Amazon SageMaker AI, covering agent deployment, dataset preparation, training, MLflow monitoring and evaluation at serving settings. The workflow trains a policy using a reward for the outcome of an entire multi-step episode, without requiring the correct action trajectory, and supports agents hosted on Bedrock AgentCore or customer infrastructure. In AWS’s Wordle example, the company reports that gpt-oss-20b improved from 61% to 84% solve rate at default reasoning effort and from 15% to 69% at its trained low-effort setting, costing about $339 in training tokens.
The guide provides a concrete training workflow for business-specific, multi-step AI agents where only the final outcome can be scored, while also documenting reward-design and training failures.
Most fine-tuning teaches a model to produce a better answer to a single prompt. Agentic tasks are different. An agent that files an expense report or triages a support ticket takes a sequence of actions, and each step depends on what came before. There’s rarely one correct sequence to imitate, and the only reliable feedback is whether the whole sequence succeeded. Supervised fine-tuning has no labels for that, and single-turn reinforcement learning cannot credit a decision made three steps before the outcome.
Multi-Turn Reinforcement Learning (MTRL) on Amazon SageMaker AI trains models on exactly that signal. SageMaker AI sends prompts to an agent you host on Amazon Bedrock AgentCore, or on your own infrastructure, such as Amazon Elastic Kubernetes Service (Amazon EKS), Amazon Elastic Compute Cloud (Amazon EC2), or AWS Fargate, through an AWS Lambda forwarder. The example here uses AgentCore. The agent calls the policy model being trained, acts in your environment over as many turns as the task needs, and reports one reward when the episode ends. You don’t specify the right trajectory, only how to score the result.
#Amazon SageMaker AI#Amazon Bedrock AgentCore#gpt-oss-20b#AWS MLflow#Amazon Elastic Kubernetes Service#Amazon Elastic Compute Cloud#AWS Fargate#Strands