Installs as a private draft. Your edits and self-improvement runs do not change the published bundle.
Track and switch between reference model, current policy, and reward model versions during training
Track and switch between reference model, current policy, and reward model versions during training
RL has multiple models in play simultaneously
RL Training Infrastructure > Serving & Inference for RL
--- name: "Model Versioning For Rl" version: "1.0.0" description: "Track and switch between reference model, current policy, and reward model versions during training" tags: - rl-skills-universe - "rl-training-infrastructure" - "serving-inference-for-rl" trigger: "model-versioning-for-rl" --- # Model Versioning For Rl Track and switch between reference model, current policy, and reward model versions during training ## Why It Matters RL has multiple models in play simultaneously ## Domain **RL Training Infrastructure** > Serving & Inference for RL