sdaioje
sdaioje
the same question, does anyone know how to fix it, please detailed answer
> torch 1.4的无法下载了
> > 作者您好： 我按照您发布的代码和数据集进行了PSM、SMD和SWaT的复现实验，现在得到SMD的实验结果与您论文中发布的结果相当，但是PSM和SWaT数据集的复现实验结果分别比论文中低了5.7%和10%。请问您在进行这两个数据集的实验中是否调整了部分实验参数？ > > 请问您后来有复现出来嘛，我在swat和psm数据集上复现也有较大的偏差 其实这种情况很普遍，一种解释就是电脑的差距，但从大多数论文的复现来看，结果不会有太大偏差，另一种解释就是为了发论文压低数值了，这也可以理解，也挺常见的
问题可能也是处在config文件里面的命名
> Do you have these problems when you run it and how do you solve them, kindly help me! run: python compute_score.py --dataset_name SMD issue1: Traceback (most recent call last):...
in the past few weeks i want to study diffusion model and i also find it is not many available code so i change the direction but i welcome any...
shebro1 | | msezwx | | ***@***.*** | ---- Replied Message ---- | From | ***@***.***> | | Date | 09/18/2024 10:36 | | To | 17000cyh/IMDiffusion ***@***.***> | |...
> I am also researching diffusion models and would like to connect with you both to exchange ideas. Can we connect on WeChat? shebro1
我其实最近没有整这个，因为这个方向做的确实一塌糊涂，我一般是图学习，时序异常检测，现在在接触vae生成，微信shebro1 | | msezwx | | ***@***.*** | ---- Replied Message ---- | From | ***@***.***> | | Date | 03/21/2025 11:32 | | To | 17000cyh/IMDiffusion ***@***.***> | |...