models/DMD2 Speed LoRA [SDXL, Pony, Illustrious] - dmd2_sdxl_4step_lora

DMD2 Speed LoRA [SDXL, Pony, Illustrious] - dmd2_sdxl_4step_lora

7/27/2025

11:19:04 AM

Related Keywords & Tags

6tz,distribution matching distillation,dmd2,dmd2_sdxl_4step_lora,fast image synthesis,hyper,lcm,lora,sdxl 1.0,speed,stable diffusion,tool,turbo

A dark hooded figure with glowing red eyes leans over a group of small white creatures called poros, set against a fisheye distorted circular purple vortex background with a wooden interior room.

Silhouette of a female character with flowing hair and neon highlights standing in a doorway with a colorful blue and pink splash background, showcasing a dynamic pose in high heels and an oversized shirt

Anime-style landscape depicting an ancient harbor with mountains and a reflective river under soft bloom lighting and a mint color theme.

A golden-leaved autumn forest with tall trees casting hard-edged shadows on the sandy ground, reflecting in water puddles within a wetland landscape.

Watercolor painting of a standing guinea pig with orange and white fur being brushed on the head with a green-tipped paintbrush.

A mystical black jackal with glowing red eyes and violet-blue flowing fur in a dynamic pose, splashing water painted in vibrant watercolor style with transparent layers and caustics.

Black and white close-up portrait of a woman with short hair, a bored expression, shown in silhouette and strong backlight, with abstract geometric lines and city buildings in the background, in a minimalist and symmetrical style.

Panoramic dark fantasy scene featuring a massive triangle shaped spaceship hovering over a fog-filled dystopian city with gothic pyramids and a crescent moon in the background, under soft cinematic light.

Close-up illustration of a young woman with blue hair and freckles wearing pink goggles in ukiyo-e watercolor style.

Close-up low angle view of a robot with glowing blue eyes and an LCD screen on its chest, surrounded by glowing butterflies against a dark background with glitch effects and multiple illuminated blue screens.

Anime-style elf girl with brown streaked hair and green eyes, sitting on the floor eating cookies with cheeks puffed, holding a cookie and surrounded by cookie jars and crumbs.

Recommended Prompts

A closeup of a SFW

Recommended Negative Prompts

NSFW, naked, nude, watermark, text, simple background

Recommended Parameters

samplers

LCM, lcm

steps

cfg

1.4 - 1.8

other models

Checkpoints\Z\ILXL\WIP\Perfection_ILXL_Realistic_3.2.safetensors (83a0ec21ca), Checkpoints\Checkpoints\SDXL\Photorealistic\juggernautXL_v9Rundiffusionphoto2.safetensors (c9e3e68f89)

Tips

Recommended LoRA weight: 0.7, but feel free to adjust.

Use the LoRA with any SDXL or similar checkpoint using the recommended settings.

You can merge this LoRA into a checkpoint to include its effects directly.

If results appear washed out or bland, try increasing the CFG scale.

Version Highlights

dmd2_sdxl_4step_lora

dmd2_sdxl_4step_lora_fp32

Creator Sponsors

This is the official DMD2 implementation LoRA from the Distribution Matching Distillation for Fast Image Synthesis paper.

Original model at Huggingface tianweiy/DMD2.

Full credit goes to Tianwei Yin.

This is the official DMD2 implementation LoRA from the Improved Distribution Matching Distillation for Fast Image Synthesis paper.

dmd2_sdxl_4step_lora

Original model at Huggingface tianweiy/DMD2.

Recommended Settings

Steps: 8

LoRA weight: 0.7

Sampler: LCM

Scheduler: Normal, Simple

CFG: 1.4 - 1.8

I use similar settings on both a low resolution base image, and the upscaled refinement pass (hires fix).

Usage

Load the LoRA like any other. I recommend a weight of 0.7, but feel free to play around.

You should be able to use it with any SDXL or similar checkpoint. Use the settings above.

You can also merge this into a checkpoint and save that, to include the result of the LoRA into the checkpoint directly. This means you no longer need to load this as a LoRA, and you can still use it with LCM and low steps for good result.

The LoRA itself doesn't add any speed improvement, it just allows you to generate at fewer steps, with better result.

If the results are washed out and bland, try increasing the CFG.

Licence

Creative Commons Attribution Non Commercial 4.0

I am not the creator of the model, just uploading the model here to make it easier to download and reference in creations. Full credit goes to Tianwei Yin.