HF RL Explorer

I want the Python scoring API in navsim.evaluate.pdm score to expose pdm score(metric cache: MetricCache…

I want the Python scoring API in navsim.evaluate.pdm score to expose pdm score(metric cache: MetricCache…: a task in MiMo-V2.6-RL-harbor-code: MiMo-V2.6-RL Code (Harbor) (Harbor dataset). The single-trajectory call should take the metric cache's initial ego state and reference PDM trajectory plus…

The task

The single-trajectory call should take the metric cache's initial ego state and reference PDM trajectory plus the agent trajectory in ego-relative poses, evaluate both the reference and prediction through the supplied simulator and scorer, and return the prediction row as a `PDMResults` object. For example, if the…

Part of FineEnvs/MiMo-V2.6-RL-harbor-code.