142 2

Jukka Seppänen PRO

Kijai

AI & ML interests

StableDiffusion, ComfyUI

Recent Activity

new activity about 2 hours ago

Kijai/LTXV2_comfy:22th January - Tiny VAE and live previews test workflow

new activity about 22 hours ago

Kijai/LTXV2_comfy:Garbled Output.

new activity 3 days ago

Kijai/LTXV2_comfy:FP4 - Special Node to Load It?

View all activity

Organizations

New activity in Kijai/LTXV2_comfy about 2 hours ago

22th January - Tiny VAE and live previews test workflow

👍 4

#41 opened 4 days ago by

RuneXX

New activity in Kijai/LTXV2_comfy about 22 hours ago

Garbled Output.

#45 opened 1 day ago by

iceknightnine

New activity in Kijai/LTXV2_comfy 3 days ago

FP4 - Special Node to Load It?

#43 opened 3 days ago by

ScramboSplinergy

New activity in Comfy-Org/ltx-2 8 days ago

"Missing weight for layer gemma3_12b.transformer.model.layers.0.self_attn.q_proj"

#4 opened 16 days ago by

MrRyukami

New activity in Kijai/LTXV2_comfy 11 days ago

'VAE' object has no attribute 'latent_frequency_bins'

#10 opened 16 days ago by

noone-ai

New activity in Kijai/LTXV2_comfy 12 days ago

fp4 transfomer only?

🚀 1

#12 opened 15 days ago by

shadovv76

VAE update (13th of jan)

👍 3

#23 opened 13 days ago by

RuneXX

VAELoaderKJ Roadblock

#27 opened 12 days ago by

Metryman55

New activity in Kijai/LTXV2_comfy 13 days ago

Could you give more info about the wrong VAE? since i used the old VAE and nothing suspicious?

#22 opened 13 days ago by

komixenon

FP16/32 VAE

#17 opened 14 days ago by

GoreGoat

Hello. KJ, thanks for you contribution. I noticed a resized_dynamic lora in the repo. what is it for? Thanks

👍 1

#21 opened 14 days ago by

landon2022

New activity in Kijai/LTXV2_comfy 14 days ago

embeddings_connector: whats the difference between 2 files?

#19 opened 14 days ago by

vvhitevvizard

How to use with Quantized Gemma 3 12b version?

#13 opened 15 days ago by

Araxyllis

New activity in Kijai/LTXV2_comfy 17 days ago

Works well

#1 opened 18 days ago by

RuneXX

how to use the gguf model

#7 opened 17 days ago by

ryuunami1

GGUF source model?

#8 opened 17 days ago by

smhf72

New activity in Kijai/WanVideo_comfy_fp8_scaled about 1 month ago

DEAR KIJAI ,CAN YOU PUT THE Wan21-14B-SCAIL-preview_fp8_e4m3fn_scaled_KJ for e5m2 version please.Thanks

#41 opened about 1 month ago by

bsarea

New activity in jingheya/Lotus-2 about 2 months ago

The uploaded lcm weights are not actually .safetensors files.

#3 opened about 2 months ago by

Kijai

New activity in Kijai/WanVideo_comfy about 2 months ago

Speed optimization is possible with low impact on vram (Teacache, Efficient Time Embeddings, batching forward passes)

#111 opened about 2 months ago by

Jd1911

New activity in Kijai/WanVideo_comfy_fp8_scaled about 2 months ago

Differences between same diffusion models from comfy-org repo

#38 opened about 2 months ago by

AbsAnon

Jukka Seppänen PRO

AI & ML interests

Recent Activity

Organizations

Kijai's activity

22th January - Tiny VAE and live previews test workflow

Garbled Output.

FP4 - Special Node to Load It?

"Missing weight for layer gemma3_12b.transformer.model.layers.0.self_attn.q_proj"

'VAE' object has no attribute 'latent_frequency_bins'

fp4 transfomer only?

VAE update (13th of jan)

VAELoaderKJ Roadblock

Could you give more info about the wrong VAE? since i used the old VAE and nothing suspicious?

FP16/32 VAE

Hello. KJ, thanks for you contribution. I noticed a resized_dynamic lora in the repo. what is it for? Thanks

embeddings_connector: whats the difference between 2 files?

How to use with Quantized Gemma 3 12b version?

Works well

how to use the gguf model

GGUF source model?

DEAR KIJAI ,CAN YOU PUT THE Wan21-14B-SCAIL-preview_fp8_e4m3fn_scaled_KJ for e5m2 version please.Thanks

The uploaded lcm weights are not actually .safetensors files.

Speed optimization is possible with low impact on vram (Teacache, Efficient Time Embeddings, batching forward passes)

Differences between same diffusion models from comfy-org repo