During the training phase of the cargo-demand forecasting model, the PM's data science lead asks whether the PM needs to personally review each hyperparameter setting before training runs. What is the PM's appropriate role during this phase?

A PM navigates the training phase effectively by managing the project-level concerns — resourcing (compute, data access), scheduling, and milestone/Go-No-Go checkpoints — while leaving detailed technical choices like specific hyperparameter values to the data science team, consistent with the CPMAI framing of a PM overseeing the work without personally implementing it. Personally approving every hyperparameter is wrong and outside the PM's role as defined throughout this exam — that is data-science implementation depth the practitioner constraints explicitly place out of scope for a PM. Staying entirely uninvolved is also wrong — the PM still owns resourcing, timeline, and milestone tracking during this phase, which is real project management work, not something to fully delegate away. Requiring the PM to reproduce the model's math before allowing training to proceed is an unreasonable and impractical bar that would stall every AI project a non-data-scientist PM oversees, contradicting the CPMAI premise that a PM can manage AI initiatives without being an ML engineer.