-
512*512 is generally enough for video, but portrait needs 1024*1024 or 2048*2048.
-
Preview is not shown for certain nodes, until "restarted".
I tested with Ultimate SD Upscale and ImpactPack's FaceDetailer nodes.
While the preview is always shown for the KSampler (Efficient) nod…
-
Thanks for your contribution!
There is something I don't understand in the transform part:
it seems that the target of transforms.Resize(input_size1), CenterCrop(input_size2) and Upscale is the orig…
-
With audio as local input we could explore using wavenets for audio processing, can they be used to simulate reverbs, non-linear time-varying signal processing (chorus/flanger), etc?
Furthermore, I t…
-
**Is your feature request related to a problem? Please describe.**
I've been experimenting with the new shaders but I'm not really happy with the restore shaders. The normal shader introduces a massi…
-
I have done over 300+ trainings now (renting gpus) of different training parameters from both kohya and ai toolkit.
I get good to decent results now on both.
I believe if you implement these ko…
-
**Glide Version**:Glide 4.6.1
**Integration libraries**:Glide 4.6.1
**Issue details / Repro steps / Use case background**:
**Glide load line / `GlideModule` (if any) / list A…
-
I use this setting below to train flux lora:
```
accelerate launch --gpu_ids 0,1 --main_process_port 29502 --mixed_precision bf16 --num_cpu_threads_per_process=2 \
flux_train_network.py --pr…
-
I think you should add images like https://johnlewis.scene7.com/is/image/JohnLewis/004113259?scl=1&fmt=png-alpha to possibly upscaled images, because they contain word scl in url which I think means s…
ghost updated
3 years ago
-
Objective:
Create a Chainner script that will automatically combine 4 images into a 4x4 grid
I previously split an image up into 4 equal pieces, (1.jpg, 2.jpg, 3.jpg and 4.jpg) where 1.jpg is the …