Jukka Seppänen PRO

Kijai's activity

• All
• Models
• Datasets
• Spaces
• Papers
• Collections
• Community
• Posts
• Upvotes
• Likes
• Articles

New activity in Kijai/LTXV2_comfy 21 days ago

22th January - Tiny VAE and live previews test workflow

👍 4

26

#41 opened 25 days ago by

RuneXX

New activity in Kijai/LTXV2_comfy 22 days ago

Garbled Output.

3

#45 opened 22 days ago by

iceknightnine

New activity in Kijai/LTXV2_comfy 24 days ago

FP4 - Special Node to Load It?

6

#43 opened 24 days ago by

ScramboSplinergy

New activity in Comfy-Org/ltx-2 29 days ago

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

7

#4 opened about 1 month ago by

MrRyukami

New activity in Kijai/LTXV2_comfy about 1 month ago

'VAE' object has no attribute 'latent_frequency_bins'

11

#10 opened about 1 month ago by

noone-ai

fp4 transfomer only?

🚀 1

4

#12 opened about 1 month ago by

shadovv76

VAE update (13th of jan)

👍 3

21

#23 opened about 1 month ago by

RuneXX

VAELoaderKJ Roadblock

10

#27 opened about 1 month ago by

Metryman55

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

2

#22 opened about 1 month ago by

komixenon

FP16/32 VAE

1

#17 opened about 1 month ago by

GoreGoat

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

👍 1

6

#21 opened about 1 month ago by

landon2022

embeddings_connector: whats the difference between 2 files?

6

#19 opened about 1 month ago by

vvhitevvizard

How to use with Quantized Gemma 3 12b version?

5

#13 opened about 1 month ago by

Araxyllis

Works well

48

#1 opened about 1 month ago by

RuneXX

how to use the gguf model

27

#7 opened about 1 month ago by

ryuunami1

GGUF source model?

3

#8 opened about 1 month ago by

smhf72

New activity in Kijai/WanVideo_comfy_fp8_scaled about 2 months ago

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

2

#41 opened about 2 months ago by

bsarea

New activity in jingheya/Lotus-2 2 months ago

The uploaded lcm weights are not actually .safetensors files.

#3 opened 2 months ago by

Kijai

New activity in Kijai/WanVideo_comfy 2 months ago

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

1

#111 opened 3 months ago by

Jd1911

New activity in Kijai/WanVideo_comfy_fp8_scaled 3 months ago

Differences between same diffusion models from comfy-org repo

1

#38 opened 3 months ago by

AbsAnon