---
license: creativeml-openrail-m
tags:
- pytorch
- diffusers
- stable-diffusion
- text-to-image
- diffusion-models-class
- dreambooth-hackathon
- wildcard
datasets:
- TrpFrog/trpfrog-icons
widget:
- text: a photo of trpfrog
  output:
    url: output3.png
- text: a photo of trpfrog in spring
  output:
    url: spring.png
- text: a photo of trpfrog in summer
  output:
    url: summer.png
- text: a photo of trpfrog in autumn
  output:
    url: autumn.png
- text: a photo of trpfrog in winter
  output:
    url: winter.png
- text: a photo of trpfrog in fire
  output:
    url: output.png
- text: a photo of trpfrog growing shiitake mushrooms
  output:
    url: output2.png

---

# TrpFrog Diffusion
<Gallery />

This is a Stable Diffusion model fine-tuned on the trpfrog concept with DreamBooth. It can be used by modifying the `instance_prompt`: **a photo of trpfrog**

## Description
TrpFrog ([@Trpfrog](https://twitter.com/trpfrog?s=21&t=GcDGy74adYhOBYxX9HloOg)) is a famous Japanese Twitter comedian.

Below is a screenshot of the site he is creating.
![](trpfrog.png)

This is a fine-tuned model of Stable Diffusion with the powerful help of Dreambooth. 
The model is made specifically for TrpFrog and faithfully reproduces its appearance.



## Usage
First install the necessary packages.

```bash
pip install diffusers transformers accelerate scipy safetensors
```

The image can then be generated with the following script.

```python
import torch
from diffusers import StableDiffusionPipeline

keyword = "trpfrog"
prompt = f"a photo of {keyword}"

model_id = "Prgckwb/trpfrog-diffusion"

pipe = StableDiffusionPipeline.from_pretrained(
    model_id, torch_dtype=torch.float16
).to("cuda")

image = pipe(prompt).images[0]
image.save("trpfrog.jpg")
```