https://github.com/daspartho/prompt-extend

extending stable diffusion prompts with suitable style cues using text generation
https://github.com/daspartho/prompt-extend

deep-learning gpt-2 huggingface-spaces huggingface-transformers machine-learning nlp prompt stable-diffusion text-generation

Last synced: 3 months ago
JSON representation

extending stable diffusion prompts with suitable style cues using text generation

• Host: GitHub
• URL: https://github.com/daspartho/prompt-extend
• Owner: daspartho
• License: apache-2.0
• Created: 2022-11-01T06:48:27.000Z (about 2 years ago)
• Default Branch: main
• Last Pushed: 2022-12-20T19:36:54.000Z (almost 2 years ago)
• Last Synced: 2024-04-16T17:34:16.254Z (7 months ago)
• Topics: deep-learning, gpt-2, huggingface-spaces, huggingface-transformers, machine-learning, nlp, prompt, stable-diffusion, text-generation
• Language: Jupyter Notebook
• Homepage: https://huggingface.co/spaces/daspartho/prompt-extend
• Size: 4.29 MB
• Stars: 173
• Watchers: 4
• Forks: 8
• Open Issues: 3
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

• awesome-generative-ai - daspartho/prompt-extend

README

        
# Prompt Extend

[![Generic badge](https://img.shields.io/badge/🤗-Open%20in%20Spaces-blue.svg)](https://huggingface.co/spaces/daspartho/prompt-extend)

Text generation model for generating suitable style cues given the main idea for a prompt.

Play with it on [HuggingFace Space](https://huggingface.co/spaces/daspartho/prompt-extend). 

[Tokenizer](https://huggingface.co/daspartho/prompt-tokenizer) and [Model](https://huggingface.co/daspartho/prompt-extend) on HuggingFace Hub.

# Usage

```python

from transformers import pipeline

text_pipe = pipeline('text-generation', model='daspartho/prompt-extend')

prompt = "peaceful village landscape"

extended_prompt = text_pipe(prompt, num_return_sequences=1)[0]["generated_text"]

print(extended_prompt)

```

```python

from transformers import AutoTokenizer, AutoModelForCausalLM

tokenizer = AutoTokenizer.from_pretrained("daspartho/prompt-extend")

model = AutoModelForCausalLM.from_pretrained("daspartho/prompt-extend")

```

# Code

- [dataset.ipynb](https://github.com/daspartho/prompt-extend/blob/main/dataset.ipynb): Creating a [subset dataset](https://huggingface.co/datasets/daspartho/stable-diffusion-prompts) of [diffusiondb](https://huggingface.co/datasets/poloclub/diffusiondb) consisting of just unique prompts.

- [tokenizer.ipynb](https://github.com/daspartho/prompt-extend/blob/main/tokenizer.ipynb): Training a new tokenizer on the [dataset](https://huggingface.co/datasets/daspartho/stable-diffusion-prompts) of stable diffusion prompts.

- [model.ipynb](https://github.com/daspartho/prompt-extend/blob/main/model.ipynb): Training a GPT-2 model on the [dataset](https://huggingface.co/datasets/daspartho/stable-diffusion-prompts) of stable diffusion prompts using the [trained tokenizer](https://huggingface.co/daspartho/prompt-tokenizer).

- [inference.ipynb](https://github.com/daspartho/prompt-extend/blob/main/inference.ipynb): Trying the prompts generated using the [trained model](https://huggingface.co/daspartho/prompt-extend) with stable diffusion.

- [app.ipynb](https://github.com/daspartho/prompt-extend/blob/main/app.ipynb): Gradio app for the model.

# Examples

![](examples/0.png)

![](examples/1.png)

![](examples/2.png)

![](examples/3.png)

![](examples/4.png)

# Contributing

If you want to contribute code, simply create a pull request. If you have an idea, create an issue and the developers will look into it!