larryvrh/MiniMax-H3-Turbo-Lora

#45 opened 1 day ago by

Dineshup99999

Update README.md

#44 opened about 1 month ago by

Whitericebb

MiniMax-H3 Turbo custom node still required?

👍 1

5

#43 opened about 1 month ago by

Specialk12

Ref2video Lora like v4 600 steps ema

➕ 2

#42 opened about 1 month ago by

StefanFalkok

Training method ？

🤝 1

#41 opened about 2 months ago by

gaga-lee97

Update README.md

#40 opened about 2 months ago by

gabrielprogm

4 steps

9

#39 opened about 2 months ago by

marin73tomas

只有fl2va才可以使用吗？ref2v有吗？

👍 1

#38 opened about 2 months ago by

AIGC-T

Is this still in active development or abundant

#37 opened about 2 months ago by

roypranto39

colorshift and oversharpen

5

#36 opened about 2 months ago by

Rkss

Amy

#33 opened about 2 months ago by

RodKykis

stable-diffusion.cpp experience

#32 opened about 2 months ago by

urtuuuu

MiniMax H3 Audio Fix and 4 step Turbo LoRA for Low VRAM GPUs

#31 opened 2 months ago by

LabMike3D

Unstable turbo lora

➕ 2

9

#30 opened 2 months ago by

dieginsa

Fix `RuntimeError` shape mismatch when using `ref2va` with audio inputs

👀 1

#29 opened 2 months ago by

cmhacks

Can't follow camera move in editing video while using refv2a

#28 opened 2 months ago by

Linger19

This LoRA is much better at rendering faces. Would it be possible to create a pruned version?

#27 opened 2 months ago by

moon1143

New 600-step V4 fixed my closeup face issues, it's perfect

🚀 2

#26 opened 2 months ago by

kipdg

minimax_h3_turbo_v4 comfyui compatability

#25 opened 2 months ago by

xdfin99

support for w4a8 convrot?

#24 opened 2 months ago by

kingp0dd

Issues regarding the first version of the LoRA: "minimax_h3_turbo_4step.safetensors"

🤗 2

#22 opened 2 months ago by

CXK996

Maintained LoRA acceleration effectiveness while ensuring high-quality testing; this has been completed

🤗 2

#21 opened 2 months ago by

ANe5s

Severe visual degradation, ghosting and audio corruption with MiniMax-H3 Turbo LoRA on pruned INT8 ConvRot, especially at 4 steps

#20 opened 2 months ago by

kangklx

850 LORA It seems to have produced a bizarre result.

#18 opened 2 months ago by

junguang

Results are better if we use native nodes than the turbo nodes

16

#17 opened 2 months ago by

Meryruizk

MiniMax H3 video VAE int8 convrot and 8-step Turbo LoRA

🤗🔥 5

10

#16 opened 2 months ago by

LabMike3D

ckpt850 not working?

8

#15 opened 2 months ago by

RenegadeR

Example video mashup created with 4step_ckpt500 variant!

❤️ 6

#13 opened 2 months ago by

reverentelusarca

Turbo node vs Lora node, beta vs normal (Turbo 500)

❤️👍 3

7

#12 opened 2 months ago by

Dagdg

【 MiniMaxH3TurboLoRA 】Do you need a Pruned LoRA ~

👍 1

15

#11 opened 2 months ago by

sunnyboxs

Does it support ref2v?

👍🔥 6

#10 opened 2 months ago by

Pevernow

[ERROR] ERROR lora diffusion_model.blocks.0.adaln_proj.linear.weight shape '[96768, 8]' is invalid for input of size 260112384

5

#9 opened 2 months ago by

sumirecccp

ComfyUI dual-clock Euler sampler for the Turbo LoRA (fixes 4-step audio noise)

🤗❤️ 7

#8 opened 2 months ago by

condayeeees

我来做点什么吧！ComfyUI教程都来看，别瞎几把搞了。

🤗 1

#7 opened 2 months ago by deleted

For ComfyUI users

🚀👍 14

23

#6 opened 2 months ago by

Meryruizk

HELP!! PLEASE!! Update: Problem solved! It's working perfectly now!

#5 opened 2 months ago by

lucas-ai26

About the training approach

#4 opened 2 months ago by

Alissonerdx

Convert to ComfyUI format ~

9

#3 opened 2 months ago by

sunnyboxs

Initial Tests Look Great!