yuhang li

RuntimeError: Expected all tensors to be on the same device, but found at least two devices, cuda:0 and cuda:1! 下面是代码 model = DeepFM(linear_feature_columns=linear_feature_columns, dnn_feature_columns=dnn_feature_columns, task='binary', device=device, gpus=[0, 1], dnn_dropout=0.5)

` self._M = 2 * (self.G[:, self.partial_vars] - self.G[:, self.other_vars] @ (self._A_other_inv @ self._A_partial))` Can you tell me what this line of code does, I don't understand what it means