On attempting to install the new update, there was a message saying could not install tensorflow 2.3.0 with conda, installing with pip.
Then there was a message saying could not install with pip, please install manually.
After restarting faceswap it failed to load properly, and I ended up deleting faceswap/all the conda files etc and reinstalled faceswap.
I then attempted to update again, which was successful in the fact I could then restart and open faceswap to use again, however it still says to install the new tensorflow 2.3.0 manually.
My question is, is it imperative that I install the new version of tensorflow? I don't have any idea on how to, Will I run into problems if I don't?
And if its a must, any tips on how to do so? Faceswap has an amazing gui so it's easy to use for someone like me who's not so savvy with the code/backend part of computing, knowing where to go to install something like that is beyond my knowledge!
My gpu is a gigabyte rtx 2060 super 8gb, and cpu is ryzen 3 2200 if that helps in any way!
Tensorflow 2.3.0
Read the FAQs and search the forum before posting a new topic.
Please mark any answers that fixed your problems so others can find the solutions.
Tensorflow 2.3.0
Assuming that the message said something like Tensorflow>2.2.0<=2.3.0 not available in Conda, installing with pip, then that is fine.
Top level, we like to use Conda to manage our packages, because it is good at keeping things isolated from the rest of your system. Unfortunately Conda still haven't released Tensorflow 2.2 for Conda, so we have to work around that (I won't go into the potential technical issues there, but let's just say it's not ideal, and has lead to some install issues for some people), but ultimately it means we pull tensorflow from another location, and it's required dependencies from within Conda...
Ultimately if the "error" was just during install, and everything is running fine now, then you should be ok. If you are getting some kind of error/warning when launching/running faceswap, then please paste it here, and also post the output from Help>Output System Information from within the GUI.
Re: Tensorflow 2.3.0
How are you installing Faceswap (we don't use TF2.3)?
Follow these steps and report back if you still have issues.
app.php/faqpage#f1r1
My word is final
Re: Tensorflow 2.3.0
I'm installing using the installer front the faceswap.dev website, setting up for nvidia gpu and letting it do its thing.
Reinstalling faceswap worked fine, and I don't seem to be having any more problems, although after reinstalling and then updating again it still came up with the same message, only this time update was successful and faceswap open and runs smoothly even with that message.
If its not an issue I won't worry about it. That being said, would there be a report on the update I can post just to show you what I mean? I'm not near my pc right now but can post later.
Re: Tensorflow 2.3.0
Assuming that the message said something like Tensorflow>2.2.0<=2.3.0 not available in Conda, installing with pip, then that is fine.
Top level, we like to use Conda to manage our packages, because it is good at keeping things isolated from the rest of your system. Unfortunately Conda still haven't released Tensorflow 2.2 for Conda, so we have to work around that (I won't go into the potential technical issues there, but let's just say it's not ideal, and has lead to some install issues for some people), but ultimately it means we pull tensorflow from another location, and it's required dependencies from within Conda...
Ultimately if the "error" was just during install, and everything is running fine now, then you should be ok. If you are getting some kind of error/warning when launching/running faceswap, then please paste it here, and also post the output from Help>Output System Information from within the GUI.
My word is final
Re: Tensorflow 2.3.0
That is exactly what the message said!
And everything is working fine now, was just checking I did everything right and I won't have problems from it in the future.
Thanks for clearing that up!
