Jukka Seppänen PRO
Kijai
AI & ML interests
StableDiffusion, ComfyUI
Recent Activity
updated
a model
about 24 hours ago
Kijai/WanVideo_comfy_fp8_scaled
new activity
3 days ago
Kijai/LTXV2_comfy:22th January - Tiny VAE and live previews test workflow
new activity
4 days ago
Kijai/LTXV2_comfy:Garbled Output.
Organizations
22th January - Tiny VAE and live previews test workflow
👍
4
26
#41 opened 7 days ago
by
RuneXX
Garbled Output.
3
#45 opened 5 days ago
by
iceknightnine
FP4 - Special Node to Load It?
6
#43 opened 6 days ago
by
ScramboSplinergy
"Missing weight for layer gemma3_12b.transformer.model.layers.0.self_attn.q_proj"
7
#4 opened 19 days ago
by
MrRyukami
'VAE' object has no attribute 'latent_frequency_bins'
10
#10 opened 19 days ago
by
noone-ai
fp4 transfomer only?
🚀
1
4
#12 opened 18 days ago
by
shadovv76
VAE update (13th of jan)
👍
3
21
#23 opened 16 days ago
by
RuneXX
VAELoaderKJ Roadblock
9
#27 opened 16 days ago
by
Metryman55
Could you give more info about the wrong VAE? since i used the old VAE and nothing suspicious?
2
#22 opened 16 days ago
by
komixenon
FP16/32 VAE
1
#17 opened 17 days ago
by
GoreGoat
embeddings_connector: whats the difference between 2 files?
6
#19 opened 17 days ago
by
vvhitevvizard
How to use with Quantized Gemma 3 12b version?
5
#13 opened 18 days ago
by
Araxyllis
Works well
48
#1 opened 21 days ago
by
RuneXX
how to use the gguf model
27
#7 opened 20 days ago
by
ryuunami1
GGUF source model?
3
#8 opened 20 days ago
by
smhf72
DEAR KIJAI ,CAN YOU PUT THE Wan21-14B-SCAIL-preview_fp8_e4m3fn_scaled_KJ for e5m2 version please.Thanks
2
#41 opened about 1 month ago
by
bsarea
The uploaded lcm weights are not actually .safetensors files.
#3 opened about 2 months ago
by
Kijai
Speed optimization is possible with low impact on vram (Teacache, Efficient Time Embeddings, batching forward passes)
1
#111 opened about 2 months ago
by
Jd1911
Differences between same diffusion models from comfy-org repo
1
#38 opened about 2 months ago
by
AbsAnon