---
license: other
license_name: flux-1-dev-non-commercial-license
license_link: https://huggingface.co/black-forest-labs/FLUX.1-dev/blob/main/LICENSE.md
language:
  - en
pipeline_tag: text-to-image
tags:
  - Flux.1-dev
  - LoRA
  - Stable Diffusion
  - image-generation
---

# FLUX.1-dev-LoRA-collections
This repository contains popular LoRAs for FLUX.1-dev model trained by our users at [Shakker AI](https://www.shakker.ai/).


# Model Cards
- [FLUX-dev-lora-Black_Myth_Wukong_hyperrealism_v1.safetensors](https://huggingface.co/Shakker-Labs/FLUX.1-dev-LoRA-collections/blob/main/FLUX-dev-lora-Black_Myth_Wukong_hyperrealism_v1.safetensors)

1. A character LoRA for Black Myth: Wukong, by [xiongmaowf](https://www.shakker.ai/userpage/8f34045c526845c09d9601fafb3bfc4b/publish).
2. The recommended LoRA scale is 0.8, with `aiyouxiketang` as tagger word. 
3. Try testing prompts: `a man in armor with a beard and a beard`, `a man in a costume holding a ball`.

- [FLUX-dev-lora-asian_woman_v0.1.safetensors](https://huggingface.co/Shakker-Labs/FLUX.1-dev-LoRA-collections/blob/main/FLUX-dev-lora-asian_woman_v0.1.safetensors)

1. A style LoRA for asian woman, by [fok](https://www.shakker.ai/userpage/fc4c167ca7194520a0acaf3b36f7b19d/publish).
2. The recommended LoRA scale is 0.4-0.65, with `a young woman` as tagger word. 

- [FLUX-dev-lora-Cartoonillustration_Cartoon_v1.safetensors](https://huggingface.co/Shakker-Labs/FLUX.1-dev-LoRA-collections/blob/main/FLUX-dev-lora-Cartoonillustration_Cartoon_v1.safetensors)

1. A style LoRA for cartoon illustration, by [DIO成](https://www.shakker.ai/userpage/cdb49d6218324b5a85894fe147215295/publish).
2. The recommended LoRA scale is 1.4-2.0, with `oil painting` as tagger word. 


- [FLUX-dev-lora-dream_v1.safetensors](https://huggingface.co/Shakker-Labs/FLUX.1-dev-LoRA-collections/blob/main/FLUX-dev-lora-dream_v1.safetensors)
 
1. A style LoRA for art illustration. By [DIO成](https://www.shakker.ai/userpage/cdb49d6218324b5a85894fe147215295/publish).
2. The recommended LoRA scale is 0.8-1.5, no tagger word. 

# Inference
```python
import torch
from diffusers import FluxPipeline

pipe = FluxPipeline.from_pretrained("black-forest-labs/FLUX.1-dev", torch_dtype=torch.bfloat16)

# Black Myth: Wukong
pipe.load_lora_weights('Shakker-Labs/FLUX.1-dev-LoRA-collections', weight_name='FLUX-dev-lora-Black_Myth_Wukong_hyperrealism_v1.safetensors')
pipe.fuse_lora(lora_scale=0.8)
pipe.to("cuda")

prompt = "aiyouxiketang, a man in armor with a beard and a beard"

image = pipe(
    prompt, 
    num_inference_steps=24, 
    guidance_scale=5.0,
    generator=torch.Generator("cpu").manual_seed(0)
).images[0]
```

# Acknowledgements
All these models are trained by our copyrighted users. We release these model with their permission.