← Về thư mục
verl/workers/critic/base.py
- BasePPOCritic · class · L26-L40 — class BasePPOCritic(ABC)
- init · method · L28-L30 — def init(self, config)
- compute_values · method · L33-L35 — def compute_values(self, data: DataProto) -> torch.Tensor
- update_critic · method · L38-L40 — def update_critic(self, data: DataProto)