Skill UI

Stable Baselines3 supplies PyTorch-based RL algorithms with a scikit-learn-style API for training agents, crafting Gymnasium environments, running vectorized workloads, wiring callbacks, and persisting models—ideal for single-agent experimentation and rapid RL prototyping.

Track running experiments by inspecting remote screen sessions, gathering logs, parsing JSON outputs, and optionally pulling WandB metrics before summarizing baselines, deltas, and costs for Vast.ai or Modal runs.