> For clean Markdown content of this page, append .md to this URL. For the complete documentation index, see https://docs.nvidia.com/dynamo/llms.txt. For full content including API reference and SDK examples, see https://docs.nvidia.com/dynamo/llms-full.txt.

# Reinforcement Learning

## Docs

- [Reinforcement Learning](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/overview.md): Use Dynamo as the rollout-serving layer for RL workloads
- [KV-Aware Load Balancing for RL Rollouts](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/kv-aware-load-balancing.md): Balance cache reuse, live worker load, and queue pressure
- [Distribute and Update Rollout Weights](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/rollout-weight-updates.md): Move policy weights into the fleet, then coordinate refresh and recovery
- [Profile and Simulate RL Rollouts](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/rl-profiling-and-simulation.md): Inspect a live rollout in Perfetto, then replay its request-plane workload
- [verl Integration](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/verl-integration.md): Run the public verl-recipe Dynamo rollout backend
- [NeMo RL Integration](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/ne-mo-rl-integration.md): Run NeMo RL's managed Dynamo vLLM backend on Slurm
- [Slime Integration](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/slime-integration.md): Run Slime external rollouts through Dynamo on Kubernetes
- [External Trainer Integration](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/external-trainer-integration.md): Connect a trainer to Dynamo generation, worker control, and policy updates
- [RL Integration Reference](https://docs.nvidia.com/dynamo/v1.5.1/reinforcement-learning/rl-integration-reference.md): Generation, worker discovery, administration, and compatibility contracts for RL frameworks