707313eb47
SAE+ConverterMasked now offer default values such as mode, erode, blur, values based on model settings.
iperov2019-02-03 21:44:30 +04:00
e226ab5385
SAE: increased speed of training by 10-18%, increased clipping border mask in full face mode results better transition of cheeks, default archi now 'df'
iperov2019-02-03 20:32:46 +04:00
41abda42d2
SampleProcessor now uses CUBIC interpolation (was LANCSOZ4) in order to reduce CPU usage
iperov2019-02-03 19:19:42 +04:00
66390335d6
upd readme: added SAE asian face preview
iperov2019-01-22 13:15:25 +04:00
2571f23bc4
Extractor added new feature --manual-output-debug-fix, Performs manual reextract input-dir frames which were deleted from [output_dir]_debug\ dir.
iperov2019-01-22 12:46:54 +04:00
b6c4171ea1
optimizations of nnlib and SampleGeneratorFace, refactorings
iperov2019-01-22 11:52:04 +04:00
2de45083a4
temporally forcing requirement of numpy version to 1.15.4
iperov2019-01-22 11:16:55 +04:00
3bbc008207
change jpeg quality to 50%
iperov2019-01-22 10:05:20 +04:00
8016ee32cc
change aligned_debug files from png to jpg
iperov2019-01-22 09:55:41 +04:00
fe66b6b2a1
SAE: added 'learn mask' option, default:Y, SAE: added clipping mask at borders to remove artifact lines.
iperov2019-01-19 16:15:49 +04:00
946688567d
enhanced SAE model. You should to restart training. new default style power = 2.0 fix DF default batch sizes. upd readme
iperov2019-01-17 21:41:40 +04:00
1f2b1481ef
now you can train models on multiple GPU's on same workspace without cloning any folders. Model files names will be prefixed with GPU index if GPU choosed explicitly on train/convert start. if you leave GPU idx choice default, then best GPU idx will be choosed and model file names will not contain index prefix. It gives you possibility to train same fake with various models or options on multiple GPUs.
iperov2019-01-14 10:48:23 +04:00
bf5282f7ec
if you have multi GPU in a system, you can choose what GPU to use on model start
iperov2019-01-10 19:22:54 +04:00
48d0123f0b
SAE: new options face_style_power, bg_style_power instead of style_power. Zero - means dont use style. SAE: new option 'lighter_encoder'. Now model settings can be overrided by pressing enter in 2 seconds while model loading (works on Windows). Removed all MultiGPU models, because keras multi_gpu in fact doesn't work.
iperov2019-01-10 18:28:12 +04:00
iperov