Faceswap Forum

I have not used DLight for a long time (I'm not a fan of unbalanced models to be honest, I think they underperform). As I said before, I think it will run with mixed precision enabled (and potentially a low batch size) And what model would you recommend with quality/performance balance? Just after ...

I assume you are talking about on an AMD GPU.... If so, then the answer is "I don't know". I'm fairly certain you could get this running on an Nvidia GPU with Mixed Precision enabled. This probably wouldn't work on AMD though, however I haven't tested. The best thing you could do to make ...

torzdf wrote: ↑Sat Sep 02, 2023 10:48 am

I can't comment on the speed as I can't benchmark it. You are probably best off in opening a new topic in Training Discussion

Good afternoon all, I have a MAC with i7 and RX 590, But a question arises, is 30-50 iterations at basic settings and Dlight the norm?

So you selected "APPLE-SILICON" using the install script and it just worked? If so, I will update the instructions to reflect this for other x64/AMD users Yes, that's correct. But I don't understand if 30-50 iterations per 2 minutes is the norm when using Dlight with default functions? Bu...

GPU is used for training/extract/convert etc. If you run with loglevel VERBOSE then you should get an output from Tensorflow saying which device it is using Good afternoon, I managed to get the GPU to work as well. I installed miniconda3 for x86-64 platform, but the config to install faceswap I use...

Unfortunately I do not own a Mac with an Intel card to be able to tell you this. The original instructions are based on what should work based on the research I did. I'm not sure how many Faceswap users have Intel Macs with AMD cards, so you may be better off asking in a more general purpose Tensor...

Good afternoon, I have a mac with intel processor and amd rx 590 ( metal 2.0). The question is about the performance of the GPU. Ventura 5.1 system As I understood from the instructions, first I need to install faceswap with CPU configuration via github script for macos, and only then tenserflow-met...