# alibaba/EasyRec

A framework for large scale recommendation algorithms.

Repository: https://github.com/alibaba/EasyRec
Canonical: https://ross.abutalabs.com/products/easyrec
Language: Python
License: Apache-2.0
License Family: permissive
Topics: recommendation-algorithms, recommender-system, dssm, esmm, mind, deepfm, dlrm, autoint, din, eges, pdn, online-learning, deepmatching, multi-task-learning, automl, capsule-network, transformers-models, wide-and-deep, ctr-prediction
Last push: 2026-04-15T02:02:32+00:00

## Health v2 (maintenance only)
Score: 65/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 77, release rhythm 30, longevity 100
- inputs: {"age_days": 2095, "days_push": 141, "days_rel": 258, "gap_med": 197, "n_releases_24m": 2}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2356, forks 384 (observed 2026-08-28T04:06:40.641607+00:00)

## What it is
EasyRec is a TensorFlow-based framework from Alibaba for building large-scale deep learning recommendation models covering matching, ranking, and multi-task learning. It provides configuration-driven model building, hyperparameter tuning, large-scale embedding, online learning, and deployment support across local and Alibaba Cloud platforms.

## Use cases
- train ctr prediction models like deepfm and wide-and-deep
- build two-tower matching models for candidate generation
- run multi-task learning for recommendation ranking
- do online learning for recommender systems at scale
- hyperparameter search for recommendation models
- deploy large-scale embedding recommendation models

## When to choose
- you need production-scale deep learning recommendation models with config-driven workflows
- you use Alibaba Cloud data platforms like MaxCompute, EMR, or PAI
- you want built-in HPO, knowledge distillation, and online learning for recsys

## When to avoid
- you prefer PyTorch - consider TorchEasyRec instead
- you need a general-purpose ML framework outside recommendation
- you want a lightweight library without cloud platform integration

## Facets
- artifact type: framework
- maturity: active
- function: machine-learning, deep-learning, llm-training, data-science
- domain: machine-learning, large-language-models, data-science, e-commerce
- platform: python, cloud
- tags: recommender-system, ctr-prediction, deepfm, dssm, multi-task-learning, online-learning, tensorflow, automl, candidate-generation, ranking, linux, docker

## Member repositories
- alibaba/EasyRec (main) score 65

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:40.641607+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 2026-08-30T02:36:17.946800+00:00, confidence not recorded.
  - readme: https://github.com/alibaba/EasyRec (fetched 2026-08-28T04:06:40.641607+00:00, sha 5129639ab1ba)
- Data as of 2026-08-30T08:39:29.467469+00:00.