XSeg error when trying to apply masks

[JapaTuts]

trafficstars

I've already made the face path in XSeg editor and trained it But now when I try to exectue the file 5.XSeg) data_dst trained mask - apply or 5.XSeg) data_src trained mask - apply the CMD returns this to me

Applying trained XSeg model to aligned/ folder. Traceback (most recent call last): File "D:\.Arquivos\Programs\DFL\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\main.py", line 324, in <module> arguments.func(arguments) File "D:\.Arquivos\Programs\DFL\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\main.py", line 285, in process_xsegapply XSegUtil.apply_xseg (Path(arguments.input_dir), Path(arguments.model_dir)) File "D:\.Arquivos\Programs\DFL\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\mainscripts\XSegUtil.py", line 32, in apply_xseg raise_on_no_model_files=True) File "D:\.Arquivos\Programs\DFL\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\facelib\XSegNet.py", line 73, in __init__ raise Exception(f'{model_file_path} does not exists.') Exception: D:\.Arquivos\Programs\DFL\DeepFaceLab_NVIDIA\workspace\model\XSeg_256.npy does not exists.

I've checked and the file XSeg_256.npy actually exists on the path, so I don't really know where is the bug

[fhrancis]

I'm having the same problem: Applying trained XSeg model to aligned/ folder. Traceback (most recent call last): File "F:\DeepFake\DeepFaceLab\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\main.py", line 324, in <module> arguments.func(arguments) File "F:\DeepFake\DeepFaceLab\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\main.py", line 285, in process_xsegapply XSegUtil.apply_xseg (Path(arguments.input_dir), Path(arguments.model_dir)) File "F:\DeepFake\DeepFaceLab\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\mainscripts\XSegUtil.py", line 32, in apply_xseg raise_on_no_model_files=True) File "F:\DeepFake\DeepFaceLab\DeepFaceLab_NVIDIA\_internal\DeepFaceLab\facelib\XSegNet.py", line 73, in __init__ raise Exception(f'{model_file_path} does not exists.') Exception: F:\DeepFake\DeepFaceLab\DeepFaceLab_NVIDIA\workspace\model\XSeg_256.npy does not exists.

[j-mir-prazak]

also having this error (using deepfacelab_linux)

[minixalpha]

I have the same problem when using CPU to do XSeg apply, if you use CPU, you should change the code of the file "_internal\DeepFaceLab\mainscripts\XSegUtil.py",

from

    xseg = XSegNet(name='XSeg', 
                    load_weights=True,
                    weights_file_root=model_path,
                    data_format=nn.data_format,
                    raise_on_no_model_files=True)

to

    xseg = XSegNet(name='XSeg', 
                    load_weights=True,
                    weights_file_root=model_path,
                    data_format=nn.data_format,
                    place_model_on_cpu=True,
                    run_on_cpu=True,
                    raise_on_no_model_files=True)

this is because XSegUtil.py does not check whether you use CPU or not, and by default the code use GPU to load model, when load model fail, the exception will tell you the model does not exist, even though the model file exists.

[darkrubiks]

Rollback to a previous commit, the last one from May 12, 2021 is giving this error

[Zeesy]

Rollback to a previous commit, the last one from May 12, 2021 is giving this error

Has anyone found a solution to this? This is my traceback:

Traceback (most recent call last):
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\main.py", line 343, in <module>
arguments.func(arguments)
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\main.py", line 304, in process_xsegapply
XSegUtil.apply_xseg (Path(arguments.input_dir), Path(arguments.model_dir))
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\mainscripts\XSegUtil.py", line 57, in apply_xseg
raise_on_no_model_files=True)
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\facelib\XSegNet.py", line 73, in __init__
raise Exception(f'{model_file_path} does not exists.')
Exception: C:\DeepFaceLab_NVIDIA_RTX3000_series\workspace\model\XSeg_256.npy does not exists.

I'm on Windows 10, running on GPU.

[Zeesy]

Rollback to a previous commit, the last one from May 12, 2021 is giving this error

Has anyone found a solution to this? This is my traceback:

Traceback (most recent call last):
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\main.py", line 343, in <module>
arguments.func(arguments)
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\main.py", line 304, in process_xsegapply
XSegUtil.apply_xseg (Path(arguments.input_dir), Path(arguments.model_dir))
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\mainscripts\XSegUtil.py", line 57, in apply_xseg
raise_on_no_model_files=True)
File "C:\DeepFaceLab_NVIDIA_RTX3000_series\_internal\DeepFaceLab\facelib\XSegNet.py", line 73, in __init__
raise Exception(f'{model_file_path} does not exists.')
Exception: C:\DeepFaceLab_NVIDIA_RTX3000_series\workspace\model\XSeg_256.npy does not exists.

I'm on Windows 10, running on GPU.

Solution for me - I needed to move my frames from the worskspace > data_src > frames folder I created to worskspace > data_src and everything worked fine.

[yhsharaf]

My Solution:

I placed the files 'XSeg_256.npy', 'XSeg_data.dat', and 'XSeg_summary.txt' in the 'workspace/model' directory and I was able to run '5_XSeg_data_dst_mask_apply.sh' successfully.

On Ubuntu 20.04

[joolstorrentecalo]

Issue solved / already answered (or it seems like user error), please close it.