Closed bryanlyon closed 6 years ago
Hmm import backend as K
just loads the configured backend, and the backend is chosen by your own config. Are you sure your keras backend config is correct?
Ahhh, you're right. In that case, I'd suggest advising people to set their config to tensorflow instead.
Ah ! Nice if you solved your issue. If you have a good way to warn people, fell free to propose, otherwise this issue should help those who have the problem (and take the time to search...)
I'd suggest adding the notice into INSTALL. Since keras defaults (at least on Ubuntu) to Theano, which wont work due to the older version of keras we're using, we should add the notice to the INSTALL.md file. I can issue a PR if necessary.
Feel free to propose the best solution you are thinking about. If updating keras version can solve this, maybe we can look into this too
Traceback (most recent call last): File "faceswap.py", line 10, in
from scripts.extract import ExtractTrainingData
File "/superm/data/Projects/faceswap/scripts/extract.py", line 7, in
from lib.cli import DirectoryProcessor
File "/superm/data/Projects/faceswap/lib/cli.py", line 7, in
from lib.faces_detect import detect_faces, DetectedFace
File "/superm/data/Projects/faceswap/lib/faces_detect.py", line 1, in
from lib import FaceLandmarksExtractor
File "/superm/data/Projects/faceswap/lib/FaceLandmarksExtractor/init.py", line 1, in
from .FaceLandmarksExtractor import extract
File "/superm/data/Projects/faceswap/lib/FaceLandmarksExtractor/FaceLandmarksExtractor.py", line 6, in
import keras
File "/local/lib/python3.5/site-packages/keras/init.py", line 3, in
from . import utils
File "/local/lib/python3.5/site-packages/keras/utils/init.py", line 6, in
from . import conv_utils
File "/local/lib/python3.5/site-packages/keras/utils/conv_utils.py", line 3, in
from .. import backend as K
File "/local/lib/python3.5/site-packages/keras/backend/init.py", line 80, in
from .theano_backend import *
File "/home/nikkelitous/.local/lib/python3.5/site-packages/keras/backend/theano_backend.py", line 3, in
import theano
File "/usr/local/lib/python3.5/dist-packages/theano/init.py", line 88, in
from theano.configdefaults import config
File "/usr/local/lib/python3.5/dist-packages/theano/configdefaults.py", line 137, in
in_c_key=False)
File "/usr/local/lib/python3.5/dist-packages/theano/configparser.py", line 287, in AddConfigVar
configparam.get(root, type(root), delete_key=True)
File "/usr/local/lib/python3.5/dist-packages/theano/configparser.py", line 335, in get
self.set(cls, val_str)
File "/usr/local/lib/python3.5/dist-packages/theano/configparser.py", line 346, in set
self.val = self.filter(val)
File "/usr/local/lib/python3.5/dist-packages/theano/configdefaults.py", line 116, in filter
'You are tring to use the old GPU back-end. '
ValueError: You are tring to use the old GPU back-end. It was removed from Theano. Use device=cuda* now. See https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28gpuarray%29 for more information.
since https://github.com/deepfakes/faceswap/commit/232d9313afabc075e28c1a42047b22e0568200e6 theano is required and will not work with the newest version.
May be specific to Ubuntu Linux but comes from the keras backend being imported in conv_utils.py.
Either we need a supported version added to requirements or find a way to remove the import of keras/backend