hunkim/word-rnn-tensorflow

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1294 stars · 484 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

A TensorFlow implementation of multi-layer recurrent neural networks (LSTM/RNN) for training word-level language models, adapted from char-rnn-tensorflow. It ships train and sample scripts that support greedy, weighted, and beam-search decoding for generating text from any training corpus.

Use cases

• train a word-level language model on my own text corpus
• generate shakespeare-like text with an LSTM
• learn how RNN language models are implemented in TensorFlow
• sample text from a trained LSTM using beam search
• compare word-level versus character-level language models
• build a multi-layer LSTM text generator in Python

When to choose

• you want a small, readable reference implementation of word-level LSTM language modeling
• you are working with legacy TensorFlow 1.x and need a working char-rnn-style trainer
• you want to experiment with beam search decoding for sequence generation

When to avoid

• you need a maintained codebase - it requires TensorFlow 1.1 and has seen no releases since 2019
• you want state-of-the-art language models such as transformers
• you need production-grade text generation - output quality is research-demo level

Facets

library · maturity abandoned

machine-learning deep-learning nlp machine-learning deep-learning python tensorflow lstm rnn language-model text-generation beam-search char-rnn recurrent-neural-networks shakespeare educational natural-language-processing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="hunkim/word-rnn-tensorflow")