Replies: 2 comments
-
ok so i used --no-half-vae and it worked again! |
Beta Was this translation helpful? Give feedback.
0 replies
-
nvm the black image is back again |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
the extension used to work just fine but then it started to generate black image. i'm not sure if i'm doing anything wrong
I'm using --disable-nan-check --xformers --theme dark
i have a GTX1650 so 4GB Vram
these are the setting that used to work for me
Weights loaded in 19.9s (load weights from disk: 1.9s, apply weights to model: 9.0s, apply half(): 0.1s, load VAE: 3.5s, hijack: 0.2s, move model to device: 5.1s).
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [02:12<00:00, 6.61s/it]
[Tiled VAE]: input_size: torch.Size([1, 4, 64, 96]), tile_size: 48, padding: 11████████| 20/20 [01:42<00:00, 5.43s/it]
[Tiled VAE]: split to 1x2 = 2 tiles. Optimal tile size 48x48, original tile size 48x48
[Tiled VAE]: Fast mode enabled, estimating group norm parameters on 48 x 32 image
[Tiled VAE]: Executing Decoder Task Queue: 100%|█████████████████████████████████████| 246/246 [01:12<00:00, 3.41it/s]
[Tiled VAE]: Done in 89.029s, max VRAM alloc 2430.140 MB
Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [03:14<00:00, 9.73s/it]
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [02:00<00:00, 6.05s/it]
[Tiled VAE]: input_size: torch.Size([1, 4, 64, 96]), tile_size: 48, padding: 11████████| 20/20 [01:55<00:00, 6.13s/it]
[Tiled VAE]: split to 1x2 = 2 tiles. Optimal tile size 48x48, original tile size 48x48
[Tiled VAE]: Fast mode enabled, estimating group norm parameters on 48 x 32 image
[Tiled VAE]: Executing Decoder Task Queue: 100%|█████████████████████████████████████| 246/246 [00:03<00:00, 73.31it/s]
[Tiled VAE]: Done in 9.067s, max VRAM alloc 2430.140 MB
Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [02:07<00:00, 6.36s/it]
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [02:02<00:00, 6.11s/it]
Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [02:59<00:00, 8.99s/it]
100%|██████████████████████████████████████████████████████████████████████████████████| 20/20 [02:02<00:00, 6.13s/it]
[Tiled VAE]: input_size: torch.Size([1, 4, 64, 96]), tile_size: 48, padding: 11████████| 20/20 [01:56<00:00, 6.12s/it]
[Tiled VAE]: split to 1x2 = 2 tiles. Optimal tile size 48x48, original tile size 48x48
[Tiled VAE]: Fast mode enabled, estimating group norm parameters on 48 x 32 image
[Tiled VAE]: Executing Decoder Task Queue: 100%|█████████████████████████████████████| 246/246 [00:03<00:00, 70.64it/s]
[Tiled VAE]: Done in 9.012s, max VRAM alloc 2430.140 MB
Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [02:08<00:00, 6.43s/it]
Total progress: 100%|██████████████████████████████████████████████████████████████████| 20/20 [02:08<00:00, 6.12s/it]
Beta Was this translation helpful? Give feedback.
All reactions