This guide provides a standardized operating procedure (SOP) for running OpenVLA robotic manipulation evaluations (LIBERO Benchmark) on AMD GPU high-performance computing (HPC) clusters using Apptainer/Singularity and SLURM.
Additional installation steps for old repo.
PPO, AC
Reinforcement Learning Note. [Model-based | Model-free Value-based | Model-free Policy-based | AC | Continuous Action Space]
Walkthrough more SOTA compact models.