# jcjohnson/torch-rnn

Efficient, reusable RNNs and LSTMs for torch

Repository: https://github.com/jcjohnson/torch-rnn
Canonical: https://ross.abutalabs.com/products/torch-rnn
Language: Lua
License: MIT
License Family: permissive
Last push: 2022-06-21T21:10:11+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 3853, "days_push": 1534, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2559, forks 508 (observed 2026-08-28T04:07:00.904290+00:00)

## What it is
torch-rnn provides high-performance, reusable RNN and LSTM modules for the Torch7 deep learning framework, used for character-level language modeling similar to char-rnn. It is significantly faster and more memory-efficient than char-rnn while having minimal dependencies.

## Use cases
- train a character-level language model
- run an RNN or LSTM in torch7
- generate text with a char-rnn style model
- benchmark LSTM implementations for speed and memory
- integrate reusable RNN modules into an existing torch project

## When to choose
- you are already working in the Torch7 ecosystem and need efficient RNN/LSTM modules
- you want a faster, lower-memory alternative to char-rnn
- you need GPU acceleration via CUDA or OpenCL for recurrent models

## When to avoid
- you are starting a new project - modern frameworks like PyTorch or TensorFlow are actively maintained instead
- you need transformer-based or large language model tooling
- you require long-term support, as Torch7 is a legacy framework

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training, benchmarking
- domain: deep-learning, machine-learning
- platform: lua
- tags: rnn, lstm, torch7, char-rnn, language-modeling, cuda, opencl, natural-language-processing, linux, macos, gpu, docker

## Member repositories
- jcjohnson/torch-rnn (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:00.904290+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:23:32.196152+00:00, confidence not recorded.
  - readme: https://github.com/jcjohnson/torch-rnn (fetched 2026-08-28T04:07:00.904290+00:00, sha 755732a1421f)
- Data as of 2026-08-30T08:39:29.467469+00:00.