wjj5881005

I think the oob score computed in the fit function is wrong. The authors get the oob sample indices by "mask = ~samples", and then apply X[mask, :] to get...

class TrainStep: def __init__(self, model, opt, device=None): self.model = model self.opt = opt self.device = device def __call__(self, engine, batch): x = batch[0] if self.device: x = x.to(self.device) pred_output =...