diffusion-point-cloud

diffusion-point-cloud copied to clipboard

GPU Mem Double

1

Hi Shitong, A great job! May I ask why GPU mem double after some iterations? E.g., from 8G to 14G. Thanks,

GloryyrolG

Number of iteration to train?

3

Hi Zhitong, thanks for releasing the code! I am wondering how many training iterations is needed for different model/data to reproduce the results in the paper? Best,

ZENGXH

How long to get fully trained?

1

Hi, I'd like to know what GPUs you used and how long did it take to train on each dataset.

gxwangupc

Met Nan after 150000 Iter

2

Hello, Thanks for your work. May I ask, have you met this problem in your training with train_gen.py [lgan_mmd-CD] nan [lgan_cov-CD] 0.24250001 [lgan_mmd_smp-CD] nan Traceback (most recent call last): File...

JudgeLJX

Hello, thanks for your great work! I am a beginner in generative models, and I want to understand if, in your approach, adding latent code as a condition means that...

wy9933

Question about training

2

Thanks for the fantastic work! However, I trained the Autoencoder with 20k iteration on the Shapenet with all classes, but the performance is not satisfactory. Is there any specific settings...

Explain3D

关于训练自己的数据集

3

我想使用自己数据集训练一个模型，在制作数据集时需要什么特别的操作吗，比如对自己数据集的点数进行限制

Wang-CHZ

关于训练次数

2

想要问一下生成器ITer到多少次会停止啊

Feng881107

感谢您的贡献，这是一项伟大的工作！您好，我是一名新研究员，我想请教一下您论文中的对比实验，如 AtlasNet、PointFlow、ShapeGF 等，您是直接运行它们的源代码，还是根据自己的架构将它们的网络架构运行到encoder文件中？ Thank you for your contribution, this is great work! Hello, I am a new researcher, I would like to ask you about the comparison experiments in your...

DDgogogo

I have a question and would like your reply, why your numstep is 100 instead of 1000, but your generated result is still good

HLJT