README.md

metadata

title: Remrofsnart
emoji: 📚
colorFrom: yellow
colorTo: blue
sdk: gradio
sdk_version: 5.34.0
app_file: app.py
pinned: false
license: mit
short_description: previous token predictor

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference

Reverse GPT-2 Generator

This is a Gradio app that uses a reverse-trained GPT-2 model to generate text that would logically come before a given prompt.

Features

• Generate text that comes before your prompt
• Adjustable parameters:
  • Max tokens
  • Temperature
  • Top-k sampling
• User-friendly interface
• Real-time generation

Local Development

1. Clone this repository
2. Create a virtual environment:

   python -m venv venv
   source venv/bin/activate  # On Windows use: venv\Scripts\activate
   

3. Install dependencies:

   pip install -r requirements.txt
   

4. Run the app:

   python app.py
   

Usage

1. Enter a prompt (e.g., "And they lived happily ever after")
2. Adjust the generation parameters if desired
3. Click Submit to generate text

Model

The app uses a reverse-trained GPT-2 model hosted on HuggingFace. The model is automatically downloaded when the app starts.

License

MIT