Add OpenReward Intergration

[shamanez]

I wanted to suggest adding support (or at least an example integration) for OpenReward:
👉 https://openreward.ai/

Why this matters

OpenReward is becoming an important component in the RL ecosystem, particularly for:

• Reward modeling
• Evaluation pipelines
• Scalable feedback loops for agent training

While the current ROCK examples are great, having a similar integration or tutorial for OpenReward would:

• Broaden the usability of ROLL across different RL stacks
• Help users experiment with alternative reward systems
• Make it easier to plug ROLL into modern RLHF-style workflows

#401