Suitability of Forward-Forward and PEPITA Learning to MLCommons-Tiny benchmarks Home