Hi I am still new at using docker and also still learning using ubuntu. Currently I am planning to do a machine learning training on a supercomputer shared by many user. I have successfully created an image and mount all the necessary data but when I try to run the training, I got permission denied when the code tries to access a .txt file. Here is the error
user@701c214f9fa9:/app/sfsnet$ python main_mix_training.py
Traceback (most recent call last):
File "main_mix_training.py", line 156, in <module>
main()
File "main_mix_training.py", line 132, in main
with open(args.log_dir+'/details.txt', 'w') as f:
PermissionError: [Errno 13] Permission denied: './results//details.txt'
user@701c214f9fa9:/app/sfsnet$ python main_mix_training.py
Traceback (most recent call last):
File "main_mix_training.py", line 156, in <module>
main()
File "main_mix_training.py", line 132, in main
with open(args.log_dir+'/details.txt', 'w') as f:
PermissionError: [Errno 13] Permission denied: './results//details.txt'
I have checked the mounted directory ownership and it should all be mine (dgx_user6) :
total 308
-rw-r--r-- 1 dgx_user6 dgx_user6 1271 二 14 05:32 checkmodel.py
-rw-r--r-- 1 dgx_user6 dgx_user6 458 二 13 23:51 cropping.py
-rw-r--r-- 1 dgx_user6 dgx_user6 12890 二 14 14:03 data_loading.py
drwxr-xr-x 5 dgx_user6 dgx_user6 4096 二 14 04:26 dataset
-rw-r--r-- 1 dgx_user6 dgx_user6 641 二 13 23:51 generate_dataset_csv.py
-rw-r--r-- 1 dgx_user6 dgx_user6 2687 二 13 23:51 interpolate.py
-rw-r--r-- 1 dgx_user6 dgx_user6 95 二 14 05:32 interpolate.sh
-rw-r--r-- 1 dgx_user6 dgx_user6 0 二 14 05:32 log_file
-rw-r--r-- 1 dgx_user6 dgx_user6 6555 二 13 23:51 main_gen_pseudo-data.py
-rw-r--r-- 1 dgx_user6 dgx_user6 7548 二 13 23:51 main_gen_synthetic_and_full.py
-rw-r--r-- 1 dgx_user6 dgx_user6 7030 二 14 14:12 main_mix_training.py
-rw-r--r-- 1 dgx_user6 dgx_user6 69169 二 14 05:34 'model backup.py'
-rw-r--r-- 1 dgx_user6 dgx_user6 2914 二 14 05:32 model_loading_synchronization.ipynb
-rw-r--r-- 1 dgx_user6 dgx_user6 69339 二 14 05:34 models.py
drwxr-xr-x 2 dgx_user6 dgx_user6 4096 二 13 23:51 pretrained
drwxr-xr-x 2 dgx_user6 dgx_user6 4096 二 14 05:33 __pycache__
-rw-r--r-- 1 dgx_user6 dgx_user6 3490 二 13 23:51 README.md
drwxr-xr-x 9 dgx_user6 dgx_user6 4096 二 14 05:33 results
-rw-r--r-- 1 dgx_user6 dgx_user6 5090 二 14 05:34 shading.py
-rw-r--r-- 1 dgx_user6 dgx_user6 605 二 14 05:34 super_resolution.py
-rw-r--r-- 1 dgx_user6 dgx_user6 39867 二 14 05:32 'train backup.py'
-rw-r--r-- 1 dgx_user6 dgx_user6 32377 二 13 23:51 train.py
-rw-r--r-- 1 dgx_user6 dgx_user6 1685 二 13 23:51 utils.py
When I check the folder ownership via container I got this ID instead
drwxr-xr-x 8 1007 1007 4096 Feb 14 06:12 .
I created the container using this command
nvidia-docker run -e NVIDIA_VISIBLE_DEVICES=3 -t -i --name coba --mount type=bind,src=/home/dgx_user6/SfSNet/,dst=/app/sfsnet sfs_face
I tried to change the ownership via chown command inside the container but got permission is denied, maybe someone can help me resolve this problem