vsuthichai/paraphraser
A bidirectional LSTM encoder-decoder with attention that generates paraphrases of input sentences.

Not currently ranked — collecting fresh signals.
star history
This project provides a sentence-level paraphrase generation model built with TensorFlow. It uses a bidirectional LSTM encoder and LSTM decoder with attention mechanism to transform input sentences into paraphrases. The model was trained on aggregated public datasets including para-nmt-5m, Quora question pairs, SNLI, and Semeval, and evaluates performance using BLEU score.
Frequently asked
- What is vsuthichai/paraphraser?
- A bidirectional LSTM encoder-decoder with attention that generates paraphrases of input sentences.
- Is paraphraser open source?
- Yes — vsuthichai/paraphraser is open source, released under the MIT license.
- What language is paraphraser written in?
- vsuthichai/paraphraser is primarily written in Python.
- How popular is paraphraser?
- vsuthichai/paraphraser has 407 stars on GitHub.
- Where can I find paraphraser?
- vsuthichai/paraphraser is on GitHub at https://github.com/vsuthichai/paraphraser.