Xiao Mu

你好，你的图像生成的很清晰，但我生成的却是以下两个样子的，一个是用savedmodel000000.pt采样的，一个是emasavedmodel_0.9999_000000.pt采用，想知道这是不是训练不够导致的，因为代码中作者在训练中没有给截止条件，这很迷惑  

你解决了吗。之前的框架没有这个问题，但是新版本出现异常了

> emasavedmodel.pt 呢？我发现这比 savedmodel.pt 效果更好。 me too

> @MilkTeaAddicted 感谢您的澄清。您能提到您研究的是什么数据集吗？我对 BRATS2020 进行了 85K 步训练，但仍然得到了黑色样本。 Hello, my dataset is ISIC2016, 60000 epoch, the integration effect of multiple pictures is as follows  The pictures generated each time are...

> 你好，请问可以帮我解决一下，我的这个问题吗？不胜感激 Traceback (most recent call last): File "E:\deep_learning\Segmentation\MedSegDiff-master\scripts\segmentation_sample.py", line 214, in main() File "E:\deep_learning\Segmentation\MedSegDiff-master\scripts\segmentation_sample.py", line 123, in main sample, x_noisy, org, cal, cal_out = sample_fn( File "E:\deep_learning\Segmentation\MedSegDiff-master\guided_diffusion\gaussian_diffusion.py", line 565,...

唔，没做过血管分割相关的工作，不过你的eps的shap是torch.Size([1, 2, 64, 64]) 要不试试沿着第二个维度拆分？调一下输入应该问题不大

If you increase the contrast and brightness, you will find that there are generated contents inside. It seems that you might have used the wrong training model. Using the "emasavedmodel_0.9999_000000.pt"...

How many steps have you trained, and it may be caused by insufficient training time

> ISIC2016 数据集适用于您的代码吗？因为 ISIC2018 太大，我无法成功下载。  Hello, ISIC2016 is applicable; I ran his code and the results are almost identical to what is shown for ISIC2018 dataset in the paper.

这段代码应该是引用的medseg，源代码也没给出停止条件