Schockingly close to plugins of early 2000’s. this is next level. Love it... keep experimenting and if poss - make it even more easier for no dev folk like hands on creative people to get involved without having to spend hours, days to install software. Plss
-
-
-
I am not the developer behind the code used here... Just to make that clear. I used https://github.com/NVlabs/stylegan which was written by Tero Karras from
@nvidia
Kraj razgovora
Novi razgovor -
-
-
Wooow insane ! How much time does it took to render the video ?
-
It´s quite fast. About 10 FPS on my setup
Kraj razgovora
Novi razgovor -
-
-
Really cool! How large was the data set?
-
55000 images
Kraj razgovora
Novi razgovor -
-
-
Impressive. If only I could afford to spend $6,399.00 for a single Nvidia Tesla v100, I could get similar results after just 41 days and 4 hours of training the model. Nvidia, how about demonstrating a version of StyleGAN that can train on a GTX 1080ti?
-
You absolutely can use GTX1080TI. Use 512x512 or less and you are good to go. Only for 1024x1024 you need more then 12GB GPU RAM. But you need time for the training. 7-12 days for 512x512
- Još 2 druga odgovora
Novi razgovor -
-
-
Hi! Do the images for training come from a ceratin dataset or the pics you used are personal? If that's the case, please, could you provide few hints on how you organized your dataset prior training? Kind regards.
-
Images are taken from the photobooth in the in the entrance hall of
@zkmkarlsruhe. To use your own dataset with stylegan take a look at https://github.com/NVlabs/stylegan/blob/master/dataset_tool.py … like this: python3 dataset_tool.py create_from_images tfrecord_dir image_dir - Još 1 odgovor
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.