b12746fb8a
readme: update face examples
iperov2020-12-20 22:28:48 +04:00
977d8a2d77
The load time of training data has been reduced significantly.
iperov2020-12-20 19:17:24 +04:00
554217d026
added "Windows 10 users IMPORTANT notice" on training start
iperov2020-12-20 16:55:25 +04:00
35cb35eb9f
upd windows magnet link
iperov2020-12-20 10:07:26 +04:00
db83a21244
Eyes priority is replaced with Eyes and mouth priority, Helps to fix eye problems during training like "alien eyes" and wrong eyes direction. Also makes the detail of the teeth higher.
iperov2020-12-20 09:45:22 +04:00
dd037d2dea
SAEHD: lr_dropout now can be used with AdaBelief
iperov2020-12-18 12:42:10 +04:00
704b5dc072
SAEHD: added new option Use AdaBelief optimizer? Experimental AdaBelief optimizer. It requires more VRAM, but the accuracy of the model is higher, and lr_dropout is not needed.
iperov2020-12-11 15:48:27 +04:00
b9c9e7cffd
fix depth_to_space for tf2.4.0. Removing compute_output_shape in leras, because it uses CPU device, which does not support all ops.
iperov2020-12-11 11:28:33 +04:00
3e7ee22ae3
Merger: fix load time of xseg if it has no model files
Colombo2020-07-29 00:41:53 +04:00
6134e57762
options_show_override{} for model summary in pretrain mode
Colombo2020-07-27 16:37:13 +04:00
1435dd3dd1
SAEHD: now random_warp is disabled for pretraining mode by default
Colombo2020-07-26 16:11:08 +04:00
ce95af9068
SAEHD: write_preview_history now works faster The frequency at which the preview is saved now depends on the resolution. For example 64x64 – every 10 iters. 448x448 – every 70 iters.
Colombo2020-07-18 11:03:13 +04:00
7da28b283d
Extractor: added ‘Image size’ option. The higher image size, the worse face-enhancer works. Use higher than 512 value only if the source image is sharp enough and the face does not need to be enhanced.
Colombo2020-06-24 11:45:23 +04:00
d78ec338c6
fix Subprocessor now writes an error if it fails to process the data
Colombo2020-06-24 11:44:58 +04:00
iperov