IMDiffusion issues

UnboundLocalError: local variable 'labels' referenced before assignment and KeyError: 'machine-1-1'

4

run: python compute_score.py --dataset_name SMD issue1: Traceback (most recent call last): File "compute_score.py", line 281, in best_f, best_proper, best_infor, threshold_infor_dict = compute_one_subset_one_strategy( File "compute_score.py", line 215, in compute_one_subset_one_strategy residual_list, labels...

huqiulan

file exist error

15

作者您好，有个小问题，SMD数据集在训练时定义了data_set_number，代码如下： data_set_number = ["3-4",'3-5',"3-10","3-11","1-5","1-8","2-4"] data_set_number += ["1-1","1-2","1-3","1-4","1-5","1-6","1-7","1-8"] data_set_number += ["2-1","2-2","2-3","2-4","2-5","2-6","2-7","2-8","2-9"] data_set_number += ["3-1","3-2","3-3","3-4","3-5","3-6","3-7","3-8","3-9","3-10","3-11"] 请问，第一行代码中由1-5，而第二行代码又加入了一遍1-5，这样是否会导致后续mkdir（name, mode）时显示错误File exists。（因为我运行时出现了这个错误，说train_result/save0/...1-5..文件已经存在）

shaieesss

训练评估尺度不正确

6

size mismatch for diffmodel.input_projection.weight: copying a param with shape torch.Size([64, 1, 1]) from checkpoint, the shape in current model is torch.Size([64, 2, 1]).

Uncle-George-13

训练过程中的掩码策略问题

作者您好，您在论文里提出了栅格掩码，但是在代码里，扩散模型在训练过程中的掩码好像是随机掩码，在验证和测试时使用的是栅格掩码，请问到底使用什么掩码策略啊，代码如下： def forward(self, batch, is_train=1): ( observed_data, observed_mask, observed_tp, gt_mask, for_pattern_mask, _, strategy_type ) = self.process_data(batch) # print("observed data shape is") # print(observed_data.shape) # print("observed mask shape is") # print(observed_mask.shape)...

songxueh