Resources

View closed (9)

Request for instructions to correctly convert a model to fp8 with appropriate scaling

#39 opened 5 days ago by

Pbgsgxyx

Upload 1000014456.jpg

#38 opened 12 days ago by

Wittyobligations7854

Create Png 2.0

#37 opened 12 days ago by

Wittyobligations7854

Png

#36 opened 12 days ago by

Wittyobligations7854

omg yet again , explanation needed.

#35 opened 17 days ago by

pikkaa

How to build text edit dataset？

#34 opened 19 days ago by

zqw

Request: Lightning LoRA optimized for Qwen Image Edit 2509 FP8 non-scaled model

#33 opened 19 days ago by

Mkeni

lora vs scaled model !?

#32 opened 21 days ago by

pikkaa

Can we get the 4-step bf16 checkpoint to train Lora directly?

#28 opened about 1 month ago by

samliu99

Images snapping back

#27 opened about 2 months ago by

lhucklen

Qwen Image Edit Lightning Lora blocks certain model capabilities, particularly related to style

👍 1

#26 opened about 2 months ago by

TerminalCalm

Questions about distillation algorithms

#25 opened about 2 months ago by

min123456

Can anyone tell me if this file is a LoRa or a model?

#24 opened 2 months ago by

Zahar007

Which LORA V2.0 should i use for this setup?

#23 opened 2 months ago by

ZeroCool22

Requesting Lighting Lora update for Qwen Edit Plus (2509)

➕ 31

#22 opened 2 months ago by

ruleez

Quantization issue with particles on loras, especially with 1.1 and 2.0 versions.

#20 opened 2 months ago by

anyMODE

asking about base model

#19 opened 2 months ago by

pawarispp

About training data?

#17 opened 3 months ago by

animemory

Qwen-Image-Edit-Inpainting

#16 opened 3 months ago by

cugtsw

workflow for Qwen-Image-Edit-Lightning-4steps

#15 opened 3 months ago by

Alan8989

Balck Images

#14 opened 3 months ago by

primalfeasible

Qwen-Image-Lightning-8steps-V1.0 Bf16 Request

#13 opened 3 months ago by

boricuapab

bf16?

#12 opened 3 months ago by

alexloops

4 step lora Blurry images

#11 opened 3 months ago by

GrzesMis

8steps-v1.1 very slow

#10 opened 3 months ago by

jgabbay

Is the v1.1 8-step LoRA actually v1.1 4-step?

👀 ❤️ 1

#9 opened 3 months ago by

xzuyn

which qwen_image models does this work with?

#8 opened 3 months ago by

nikkosad

Low variance across seeds in 4-step

#6 opened 3 months ago by

uayo

Is 8step model higher quality than the 4step one ?

#4 opened 4 months ago by

NielsGx

Request to Build a Qwen‑Image‑Lightning‑Style Model Based on Skywork‑UniPic

#3 opened 4 months ago by

PlayAI