minimaxir/textgenrnn
A Python module for training character-level and word-level RNNs to generate text on any dataset.
Not currently ranked — collecting fresh signals.
star history
textgenrnn is a Keras/TensorFlow wrapper that simplifies training recurrent neural networks for text generation. It supports both character-level and word-level generation, configurable RNN sizes and layers, bidirectional processing, and GPU acceleration via CuDNN. The library implements attention-weighting and skip-embedding techniques to improve training speed and output quality.
Frequently asked
- What is minimaxir/textgenrnn?
- A Python module for training character-level and word-level RNNs to generate text on any dataset.
- Is textgenrnn open source?
- Yes — minimaxir/textgenrnn is an open-source project tracked on heatdrop.
- What language is textgenrnn written in?
- minimaxir/textgenrnn is primarily written in Python.
- How popular is textgenrnn?
- minimaxir/textgenrnn has 4.9k stars on GitHub.
- Where can I find textgenrnn?
- minimaxir/textgenrnn is on GitHub at https://github.com/minimaxir/textgenrnn.