Sd35m Sfwbooru Lokr

Developed by bghira

Text-to-image generation model based on Stable Diffusion 3.5 architecture, supporting high-quality image generation and image-to-image conversion

Image Generation Open Source License:Other #High-detail image generation #Multi-scene adaptation #Cyberpunk style

Downloads 317

Release Time : 2/10/2025

Model Overview

This model is a medium-scale version based on the Stable Diffusion 3.5 architecture, focusing on text-to-image generation and image-to-image conversion tasks, supporting image generation for various creative scenarios.

Model Features

High-quality image generation

Capable of generating high-resolution, detail-rich images

Diverse style support

Supports multiple artistic styles from surreal to photorealistic

LoRA support

Supports LoRA (Low-Rank Adaptation) fine-tuning technology

Broad scenario coverage

Can handle a wide range of themes from sci-fi, fantasy to real-world scenarios

Model Capabilities

Text-to-image generation

Image-to-image generation

Style transfer

Creative content generation

Scene construction

Use Cases

Creative arts

Concept art creation

Creating concept art for games, films and other media

Generating high-quality, style-consistent creative images

Illustration generation

Automatically generating illustrations based on text descriptions

Quickly producing visual content that meets requirements

Entertainment content

Comic creation

Generating comic scenes or characters

Such as the four-panel comic generation in the example

Game assets

Generating environment, character assets for game development

Rapid prototyping and concept validation

Commercial applications

Advertisement design

Generating advertisement concept images

Quick visualization of creative concepts

Product visualization

Creating concept images for products

Such as the textbook cover generation in the example

license: other base_model: "stabilityai/stable-diffusion-3.5-medium" tags:

sd3
sd3-diffusers
text-to-image
image-to-image
diffusers
simpletuner
not-for-all-audiences
lora
template:sd-lora
standard pipeline_tag: text-to-image inference: true widget:
text: 'unconditional (blank prompt)' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_0_0.png
text: 'Alien planet, strange rock formations, glowing plants, bizarre creatures, surreal atmosphere' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_1_0.png
text: 'Alien marketplace, bizarre creatures, exotic goods, vibrant colors, otherworldly atmosphere' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_2_0.png
text: 'Child holding a balloon, happy expression, colorful balloons, sunny day, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_3_0.png
text: 'a 4-panel comic strip showing an orange cat saying the words ''HELP'' and ''LASAGNA''' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_4_0.png
text: 'a hand is holding a comic book with a cover that reads ''The Adventures of Superhero''' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_5_0.png
text: 'Underground cave filled with crystals, glowing lights, reflective surfaces, fantasy environment, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_6_0.png
text: 'Bustling cyberpunk bazaar, vendors, neon signs, advanced tech, crowded, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_7_0.png
text: 'Cyberpunk hacker in a dark room, neon glow, multiple screens, intense focus, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_8_0.png
text: 'a cybernetic anne of green gables with neural implant and bio mech augmentations' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_9_0.png
text: 'Post-apocalyptic cityscape, ruined buildings, overgrown vegetation, dark and gritty, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_10_0.png
text: 'Magical castle in a lush forest, glowing windows, fantasy architecture, high resolution, detailed textures' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_11_0.png
text: 'Ruins of an ancient temple in an enchanted forest, glowing runes, mystical creatures, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_12_0.png
text: 'Mystical forest, glowing plants, fairies, magical creatures, fantasy art, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_13_0.png
text: 'Magical garden with glowing flowers, fairies, serene atmosphere, detailed plants, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_14_0.png
text: 'Whimsical garden filled with fairies, magical plants, sparkling lights, serene atmosphere, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_15_0.png
text: 'Majestic dragon soaring through the sky, detailed scales, dynamic pose, fantasy art, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_16_0.png
text: 'Fantasy world, floating islands in the sky, waterfalls, lush vegetation, detailed landscape, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_17_0.png
text: 'Futuristic city skyline at night, neon lights, cyberpunk style, high contrast, sharp focus' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_18_0.png
text: 'Space battle scene, starships fighting, laser beams, explosions, cosmic background' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_19_0.png
text: 'Abandoned fairground at night, eerie rides, ghostly figures, fog, dark atmosphere, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_20_0.png
text: 'Spooky haunted mansion on a hill, dark and eerie, glowing windows, ghostly atmosphere, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_21_0.png
text: 'a hardcover physics textbook that is called PHYSICS FOR DUMMIES' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_22_0.png
text: 'Epic medieval battle, knights in armor, dynamic action, detailed landscape, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_23_0.png
text: 'Bustling medieval market with merchants, knights, and jesters, vibrant colors, detailed' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_24_0.png
text: 'Cozy medieval tavern, warm firelight, adventurers drinking, detailed interior, rustic atmosphere' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_25_0.png
text: 'Futuristic city skyline at night, neon lights, cyberpunk style, high contrast, sharp focus' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_26_0.png
text: 'Forest with neon-lit trees, glowing plants, bioluminescence, surreal atmosphere, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_27_0.png
text: 'Bright neon sign in a busy city street, ''Open 24 Hours'', bold typography, glowing lights' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_28_0.png
text: 'Vibrant neon sign, ''Bar'', bold typography, dark background, glowing lights, detailed design' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_29_0.png
text: 'Pirate ship on the high seas, stormy weather, detailed sails, dramatic waves, photorealistic' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_30_0.png
text: 'Pirate discovering a treasure chest, detailed gold coins, tropical island, dramatic lighting' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_31_0.png
text: 'a photograph of a woman experiencing a psychedelic trip. trippy, 8k, uhd, fractal' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_32_0.png
text: 'Cozy cafe on a rainy day, people sipping coffee, warm lights, reflections on wet pavement, photorealistic' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_33_0.png
text: '1980s arcade, neon lights, vintage game machines, kids playing, vibrant colors, nostalgic atmosphere' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_34_0.png
text: '1980s game room with vintage arcade machines, neon lights, vibrant colors, nostalgic feel' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_35_0.png
text: 'Robot blacksmith forging metal, sparks flying, detailed workshop, futuristic and medieval blend' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_36_0.png
text: 'Sleek robot performing a dance, futuristic theater, holographic effects, detailed, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_37_0.png
text: 'High-tech factory where robots are assembled, detailed machinery, futuristic setting, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_38_0.png
text: 'Garden tended by robots, mechanical plants, colorful flowers, futuristic setting, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_39_0.png
text: 'Cute robotic pet, futuristic home, sleek design, detailed features, friendly and animated' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_40_0.png
text: 'cctv trail camera night time security picture of a wendigo in the woods' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_41_0.png
text: 'Astronaut exploring an alien planet, detailed landscape, futuristic suit, cosmic background' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_42_0.png
text: 'Futuristic space station orbiting a distant exoplanet, sleek design, detailed structures, cosmic backdrop' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_43_0.png
text: 'a person holding a sign that reads ''SOON''' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_44_0.png
text: 'Steampunk airship in the sky, intricate design, Victorian aesthetics, dynamic scene, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_45_0.png
text: 'Steampunk inventor in a workshop, intricate gadgets, Victorian attire, mechanical arm, goggles' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_46_0.png
text: 'Stormy ocean with towering waves, dramatic skies, detailed water, intense atmosphere, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_47_0.png
text: 'Dramatic stormy sea, lighthouse in the distance, lightning striking, dark clouds, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_48_0.png
text: 'Graffiti artist creating a mural, vibrant colors, urban setting, dynamic action, high resolution' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_49_0.png
text: 'Urban alleyway filled with vibrant graffiti art, tags and murals, realistic textures' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_50_0.png
text: 'Urban street sign, ''Main Street'', bold typography, realistic textures, weathered look' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_51_0.png
text: 'Classic car show with vintage vehicles, vibrant colors, nostalgic atmosphere, high detail' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_52_0.png
text: 'Retro diner sign, ''Joe''s Diner'', classic 1950s design, neon lights, weathered look' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_53_0.png
text: 'Vintage store sign with elaborate typography, ''Antique Shop'', hand-painted, weathered look' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_54_0.png
text: 'A photo-realistic image of a cat' parameters: negative_prompt: 'blurry, cropped, ugly' output: url: ./assets/image_55_0.png

sd35m-sfwbooru-lokr

This is a standard PEFT LoRA derived from stabilityai/stable-diffusion-3.5-medium.

The main validation prompt used during training was:

A photo-realistic image of a cat

Validation settings

CFG: 3.2
CFG Rescale: 0.0
Steps: 30
Sampler: FlowMatchEulerDiscreteScheduler
Seed: 42
Resolution: 1024x1024
Skip-layer guidance: skip_guidance_layers=[7, 8, 9],

Note: The validation settings are not necessarily the same as the training settings.

You can find some example images in the following gallery:

The text encoder was not trained. You may reuse the base model text encoder for inference.

Training settings

Training epochs: 4
Training steps: 264250
Learning rate: 1.0
- Learning rate schedule: cosine
- Warmup steps: 500000
Max grad value: 0.0
Effective batch size: 6
- Micro-batch size: 6
- Gradient accumulation steps: 1
- Number of GPUs: 1
Gradient checkpointing: True
Prediction type: flow-matching (extra parameters=['shift=3'])
Optimizer: prodigyd0=1e-8,eps=1e-8
Trainable parameter precision: Pure BF16
Base model precision: no_change
Caption dropout probability: 10.0%
LoRA Rank: 16
LoRA Alpha: None
LoRA Dropout: 0.1
LoRA initialisation style: default

Datasets

sfwbooru-crop

Repeats: 0
Total number of images: 363920
Total number of aspect buckets: 1
Resolution: 1.048576 megapixels
Cropped: True
Crop style: random
Crop aspect: square
Used for regularisation data: No

Inference

import torch
from diffusers import DiffusionPipeline

model_id = 'stabilityai/stable-diffusion-3.5-medium'
adapter_id = 'bghira/sd35m-sfwbooru-lokr'
pipeline = DiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.bfloat16) # loading directly in bf16
pipeline.load_lora_weights(adapter_id)

prompt = "A photo-realistic image of a cat"
negative_prompt = 'blurry, cropped, ugly'

## Optional: quantise the model to save on vram.
## Note: The model was not quantised during training, so it is not necessary to quantise it during inference time.
#from optimum.quanto import quantize, freeze, qint8
#quantize(pipeline.transformer, weights=qint8)
#freeze(pipeline.transformer)
    
pipeline.to('cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu') # the pipeline is already in its target precision level
model_output = pipeline(
    prompt=prompt,
    negative_prompt=negative_prompt,
    num_inference_steps=30,
    generator=torch.Generator(device='cuda' if torch.cuda.is_available() else 'mps' if torch.backends.mps.is_available() else 'cpu').manual_seed(42),
    width=1024,
    height=1024,
    guidance_scale=3.2,
    skip_guidance_layers=[7, 8, 9],
).images[0]

model_output.save("output.png", format="PNG")

Exponential Moving Average (EMA)

SimpleTuner generates a safetensors variant of the EMA weights and a pt file.

The safetensors file is intended to be used for inference, and the pt file is for continuing finetuning.

The EMA model may provide a more well-rounded result, but typically will feel undertrained compared to the full model as it is a running decayed average of the model weights.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご