L2CS-Net icon indicating copy to clipboard operation
L2CS-Net copied to clipboard

problems occurred when training

Open Fortogether opened this issue 2 years ago • 11 comments

Hello author, first of all, thank you very much for open sourcing your work. My English and code skills are not very good, so here I will briefly talk about the problem when training, I have the following error when training with mpiigaze dataset: get_ignored_params() takes 1 positional argument but 2 were given; when using Adam optimizer get_ ignored_params, get_non_ignored_params, get_fc_params all have the above error, I don't know how to change it, can you give some advice to me, thanks a lot.

Fortogether avatar Apr 18 '22 13:04 Fortogether

Hello author, I have same problem. Waiting for your answer, thanks a lot

MisterLuka avatar May 16 '22 07:05 MisterLuka

Hello author, I have same problem. Waiting for your answer, thanks a lot

MisterLuka avatar May 16 '22 07:05 MisterLuka

作者您好,首先非常感谢您开源您的工作。我的英文和代码能力都不是很好,所以这里简单说一下训练时的问题,我用mpiigaze数据集训练时出现如下错误:get_ignored_pa​​rams() takes 1 positional argument but 2 were given; 在使用Adam优化器时get_ignored_pa​​rams、get_non_ignored_pa​​rams、get_fc_params都出现上述错误,不知道怎么改,能否给点建议,非常感谢。

There is an unspecified value in his train.py, and I used 180。

AimersS avatar Jul 21 '22 09:07 AimersS

Hello author, first of all, thank you very much for open sourcing your work. My English and code skills are not very good, so here I will briefly talk about the problem when training, I have the following error when training with mpiigaze dataset: get_ignored_params() takes 1 positional argument but 2 were given; when using Adam optimizer get_ ignored_params, get_non_ignored_params, get_fc_params all have the above error, I don't know how to change it, can you give some advice to me, thanks a lot.

您好,看到您在训练时出现了问题,想必数据的前处理已经搞定了,想请问您的MPIIFaceGaze数据前处理是怎么做的呢?我在这个地方卡了很久了,期望得到您的回复

Xiaosong020 avatar Nov 03 '22 11:11 Xiaosong020

BCAA GazeHub@Phi-ai Lab.provides data preprocessing code, which you can find online,u need to download both mpiigaze and mpiifacegaze

AimersS avatar Nov 03 '22 11:11 AimersS

BCAA GazeHub@Phi-ai Lab.provides data preprocessing code, which you can find online,u need to download both mpiigaze and mpiifacegaze

我按照给定数据处理步骤,运行data_processing_mpii.py时,出现了如下问题,我下载的Origianal MPIIFaceGaze数据集没有annotation.txt,

File "/home/nixiaosong/projects/GazeDetection/L2CS-Net/data_preprocess/MPIIFaceGaze/data_processing_mpii.py", line 154, in ImageProcessing_MPII() File "/home/nixiaosong/projects/GazeDetection/L2CS-Net/data_preprocess/MPIIFaceGaze/data_processing_mpii.py", line 31, in ImageProcessing_MPII ImageProcessing_Person(im_root, sample_list, im_outpath, label_outpath, person) File "/home/nixiaosong/projects/GazeDetection/L2CS-Net/data_preprocess/MPIIFaceGaze/data_processing_mpii.py", line 50, in ImageProcessing_Person with open(annotation) as infile: FileNotFoundError: [Errno 2] No such file or directory: '/ssd1/datasets/GazeDetection/Original/MPIIFaceGaze/Origin/Data/Original/p00/day38/annotation.txt'

Xiaosong020 avatar Nov 04 '22 07:11 Xiaosong020

BCAA GazeHub@Phi-ai Lab.provides data preprocessing code, which you can find online,u need to download both mpiigaze and mpiifacegaze 已经解决了,谢谢

Xiaosong020 avatar Nov 04 '22 08:11 Xiaosong020

I suggest you carefully read what Dr. Chen wrote on the page. MPII contains two datasets, which u need to download and preprocess at the same time

AimersS avatar Nov 04 '22 08:11 AimersS

Hello author, I have same problem. Waiting for your answer, thanks a lot

Hello , I have same problem. Could you give me some advice, thanks a lot

flinya avatar Mar 09 '23 01:03 flinya

Hello author, first of all, thank you very much for open sourcing your work. My English and code skills are not very good, so here I will briefly talk about the problem when training, I have the following error when training with mpiigaze dataset: get_ignored_params() takes 1 positional argument but 2 were given; when using Adam optimizer get_ ignored_params, get_non_ignored_params, get_fc_params all have the above error, I don't know how to change it, can you give some advice to me, thanks a lot.

Hello, I have same problem, did you solve the problem? Could you give me some suggestions, thanks a lot.

Foggyfish avatar Jan 03 '24 08:01 Foggyfish

作者您好,首先非常感谢您开源您的工作。我的英文和代码能力都不是很好,所以这里简单说一下训练时的问题,我用mpiigaze数据集训练时出现如下错误:get_ignored_pa​​rams() takes 1 positional argument but 2 were given; 在使用Adam优化器时get_ignored_pa​​rams、get_non_ignored_pa​​rams、get_fc_params都出现上述错误,不知道怎么改,能否给点建议,非常感谢。

There is an unspecified value in his train.py, and I used 180。

How did you specifically solve this problem? Can you provide a detailed explanation? Thank you very much

Foggyfish avatar Mar 11 '24 07:03 Foggyfish