# tensorflow/ranking

Learning to Rank in TensorFlow

Repository: https://github.com/tensorflow/ranking
Canonical: https://ross.abutalabs.com/products/ranking
Language: Python
License: Apache-2.0
License Family: permissive
Topics: ranking, machine-learning, deep-learning, information-retrieval, learning-to-rank, recommender-systems
Archived: true
Last push: 2024-03-18T20:31:57+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2830, "days_push": 898, "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 2773, forks 477 (observed 2026-08-28T04:07:19.097748+00:00)

## What it is
TensorFlow Ranking is a Python library for building Learning-to-Rank (LTR) models on top of TensorFlow. It provides pointwise, pairwise, and listwise loss functions, ranking metrics like MRR and NDCG, groupwise scoring functions, LambdaLoss, and unbiased learning-to-rank techniques.

## Use cases
- train a learning-to-rank model for search results
- optimize NDCG directly with LambdaLoss
- evaluate ranking models with MRR and NDCG metrics
- build unbiased ranking models from biased click feedback
- score document lists with groupwise scoring functions
- train a ranking model for a recommender system

## When to choose
- you need deep learning based ranking models in TensorFlow
- you want state-of-the-art LTR losses and metrics out of the box
- you are doing academic or industrial research on learning-to-rank

## When to avoid
- your project uses PyTorch or JAX instead of TensorFlow
- you need simple classical LTR algorithms like LambdaMART without deep learning
- you need a actively evolving library with frequent releases

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, search-engine
- domain: machine-learning
- platform: python
- tags: learning-to-rank, tensorflow, ranking-metrics, lambdaloss, ndcg, mrr, unbiased-ltr, search, information-retrieval, recommender-systems, linux, gpu

## Member repositories
- tensorflow/ranking (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:19.097748+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 2026-08-30T08:16:36.710909+00:00, confidence not recorded.
  - readme: https://github.com/tensorflow/ranking (fetched 2026-08-28T04:07:19.097748+00:00, sha a17b5031795f)
- Data as of 2026-08-30T08:39:29.467469+00:00.