promptgen majinai unsafe LLM Model

Introduction

The PROMPTGEN-MAJINAI-UNSAFE model is a fine-tuned version of the distilgpt2 model, specifically tailored for text generation tasks. It has been fine-tuned for 40 epochs on a dataset comprised of 825 prompts sourced from majinai.art. The model processes prompts with weights and emphasis removed and includes negative prompts.

Architecture

The model is based on the distilgpt2 architecture, which is a lighter variant of the GPT-2 model. It operates within the Hugging Face Transformers library and is implemented using PyTorch.

Training

The model underwent fine-tuning over 40 epochs using prompts collected from majinai.art. This process involved stripping weights and emphasis to maintain a specific focus on the input prompts. Negative prompts are also part of the training set, enhancing the model's versatility in generating text.

Guide: Running Locally

Clone the Repository:

Clone the stable diffusion web UI prompt generator repository:

git clone https://github.com/AUTOMATIC1111/stable-diffusion-webui-promptgen.git

Install Dependencies:
- Navigate to the cloned directory and install necessary dependencies:
```
cd stable-diffusion-webui-promptgen
pip install -r requirements.txt
```
Run the Model:
- Execute the main script to start generating text based on input prompts:
```
python run.py
```

Cloud GPUs

For improved performance, especially with larger datasets or more demanding applications, consider using cloud GPUs from providers such as AWS, Google Cloud, or Azure. These platforms offer access to high-performance computing resources that can significantly expedite training and inference processes.

License

This model is distributed under the MIT License, permitting open use and modification with proper attribution.

More Related APIs in Text Generation