ResearchBenchmarksMeta-Reinforcement Learning on Hopper paramsFollow0.51FLOPs (k)CSP-6.945643.379793.705144.0303Nov 2, 2023Evaluation ResultsMethodMethodLinksFLOPs (k)Inference Time (ms)CSPInference Hardware=Int...Inference Hardware=Intel(R) Xeon(R) Gold 5218R @ 2.10GHz CPU2023.110.514.13HyperInference Hardware=Int...Inference Hardware=Intel(R) Xeon(R) Gold 5218R @ 2.10GHz CPU2023.114.117.2PEARLInference Hardware=Int...Inference Hardware=Intel(R) Xeon(R) Gold 5218R @ 2.10GHz CPU2023.11186.926.6