Uploaded December 2025 | Updated September 2026, 2 weeks ago

---

W&B Training offers serverless RL powered by CoreWeave to make reinforcement learning for AI agents both reliable and affordable. In this video, we demonstrate how Serverless RL, the ART fine-tuning framework, and RULER enable you to fine-tune large language models, improve agent performance, and track training progress on CoreWeave’s scalable managed infrastructure.



⏳Timestamps:

0:00 AI applications are hard to productionize
1:04 Post-training LLMs with reinforcement learning
1:46 Challenges with reinforcement learning
2:18 How W&B Training Serverless RL addresses reinforcement learning challenges
4:06 Multi-agent contact center example
5:31 Evaluating agent LLMs using W&B Weave
7:16 W&B Training quickstart resources
8:09 Serverless RL notebook walkthrough
10:39 Reinforcement learning code for our contact center agent
11:33 Viewing post-training results in W&B Models
12:54 Evaluating our post-trained LLM in Weave
14:04 Recap, conclusion, and invitation to try the Weights & Biases AI developer platform