# IBM/pytorch-seq2seq

An open source framework for seq2seq models in PyTorch.

Repository: https://github.com/IBM/pytorch-seq2seq
Canonical: https://ross.abutalabs.com/products/ibm-pytorch-seq2seq
Homepage: https://ibm.github.io/pytorch-seq2seq/public/index.html
Language: Python
License: Apache-2.0
License Family: permissive
Topics: pytorch, seq2seq, deeplearning
Archived: true
Last push: 2025-09-17T19:46:43+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 42, release rhythm 8, longevity 100
- inputs: {"age_days": 3353, "days_push": 350, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1515, forks 367 (observed 2026-08-28T04:04:56.524934+00:00)

## What it is
An open-source framework for building sequence-to-sequence (seq2seq) models in PyTorch, with modular and extensible components for models, training, inference, and checkpoints. It targets research and experimentation on encoder-decoder architectures such as machine translation and conversational models.

## Use cases
- train a seq2seq model for machine translation
- build an encoder-decoder model with attention in PyTorch
- experiment with sequence-to-sequence architectures
- train a conversational chatbot model
- use pre-trained word embeddings in a seq2seq model
- benchmark seq2seq models on WMT translation tasks

## When to choose
- you want a modular, extensible PyTorch framework specifically for seq2seq models
- you are researching or teaching encoder-decoder architectures
- you need a lightweight library rather than a full training platform

## When to avoid
- you need modern transformer implementations with active development
- you want a high-level framework like Hugging Face Transformers with pretrained models
- you need production-grade NLP pipelines rather than research code

## Facets
- artifact type: framework
- maturity: maintenance
- function: machine-learning, deep-learning
- domain: machine-learning
- platform: python
- tags: seq2seq, pytorch, encoder-decoder, attention, machine-translation, natural-language-processing

## Member repositories
- IBM/pytorch-seq2seq (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:56.524934+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 2026-08-30T04:32:08.125297+00:00, confidence not recorded.
  - readme: https://github.com/IBM/pytorch-seq2seq (fetched 2026-08-28T04:04:56.524934+00:00, sha cb2f3de90f5d)
  - homepage: https://ibm.github.io/pytorch-seq2seq/public/index.html (fetched 2026-08-29T11:35:40.609668+00:00, sha 09e90a97a614)
- Data as of 2026-08-30T08:39:29.467469+00:00.