---
tags:
- stable-diffusion-xl
- stable-diffusion-xl-diffusers
- text-to-image
- diffusers
- lora
- template:sd-lora
widget:
- text: A photo of <s0><s1> a man wearing headphones and a blue shirt
  output:
    url: image-0.png
- text: A photo of <s0><s1> a bald man wearing glasses and a white t - shirt
  output:
    url: image-1.png
- text: A photo of <s0><s1> a man with glasses and a beard smiles
  output:
    url: image-2.png
- text: A photo of <s0><s1> a bald man with glasses and a colorful shirt
  output:
    url: image-3.png
- text: A photo of <s0><s1> a man with glasses and a hat wearing an orange cap
  output:
    url: image-4.png
- text: A photo of <s0><s1> a man wearing glasses and a yellow hat taking a selfie
  output:
    url: image-5.png
base_model: stabilityai/stable-diffusion-xl-base-1.0
instance_prompt: A photo of <s0><s1>
license: openrail++
inference:
  parameters:
    num_inference_steps: 30
    scheduler: EulerAncestralDiscreteScheduler
---

# SDXL LoRA DreamBooth - multimodalart/poli-100-multiplier-face

<Gallery />

## Model description

### These are multimodalart/poli-100-multiplier-face LoRA adaption weights for stabilityai/stable-diffusion-xl-base-1.0.

## Trigger words

To trigger image generation of trained concept(or concepts) replace each concept identifier in you prompt with the new inserted tokens:

to trigger concept `TOK` → use `<s0><s1>` in your prompt 



## Use it with the [🧨 diffusers library](https://github.com/huggingface/diffusers)

```py
from diffusers import AutoPipelineForText2Image
import torch
from huggingface_hub import hf_hub_download
from safetensors.torch import load_file
        
pipeline = AutoPipelineForText2Image.from_pretrained('stabilityai/stable-diffusion-xl-base-1.0', torch_dtype=torch.float16).to('cuda')
pipeline.load_lora_weights('multimodalart/poli-100-multiplier-face', weight_name='pytorch_lora_weights.safetensors')
embedding_path = hf_hub_download(repo_id='multimodalart/poli-100-multiplier-face', filename="embeddings.safetensors", repo_type="model")
state_dict = load_file(embedding_path)
pipeline.load_textual_inversion(state_dict["clip_l"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder, tokenizer=pipeline.tokenizer)
pipeline.load_textual_inversion(state_dict["clip_g"], token=["<s0>", "<s1>"], text_encoder=pipeline.text_encoder_2, tokenizer=pipeline.tokenizer_2)
        
image = pipeline('A photo of <s0><s1>').images[0]
```

For more details, including weighting, merging and fusing LoRAs, check the [documentation on loading LoRAs in diffusers](https://huggingface.co/docs/diffusers/main/en/using-diffusers/loading_adapters)

## Download model

### Use it with UIs such as AUTOMATIC1111, Comfy UI, SD.Next, Invoke

- Download the LoRA *.safetensors [here](/multimodalart/poli-100-multiplier-face/blob/main/pytorch_lora_weights.safetensors). Rename it and place it on your Lora folder.
- Download the text embeddings *.safetensors [here](/multimodalart/poli-100-multiplier-face/blob/main/embeddings.safetensors). Rename it and place it on it on your embeddings folder.

All [Files & versions](/multimodalart/poli-100-multiplier-face/tree/main).

## Details

The weights were trained using [🧨 diffusers Advanced Dreambooth Training Script](https://github.com/huggingface/diffusers/blob/main/examples/advanced_diffusion_training/train_dreambooth_lora_sdxl_advanced.py).

LoRA for the text encoder was enabled. False.

Pivotal tuning was enabled: True.

Special VAE used for training: madebyollin/sdxl-vae-fp16-fix.