tensorflow/recommenders

TensorFlow Recommenders is a library for building recommender system models using TensorFlow.

What it solves

It simplifies the process of building recommender system models, covering the entire workflow from data preparation and model formulation to training, evaluation, and deployment.

How it works

Built on Keras and TensorFlow, the library provides high-level abstractions for creating recommendation models. It allows developers to build factorization models by defining user and item representations (embeddings) and using specialized tasks (like retrieval tasks) to optimize the model's performance based on evaluation metrics.

Who it’s for

Developers and machine learning engineers who want to build recommendation systems using the TensorFlow ecosystem with a gentle learning curve and the flexibility to build complex models.

Highlights

  • Comprehensive workflow support: handles data preparation, training, and deployment.
  • Keras-based: leverages Keras for a gentle learning curve.
  • Flexible model building: supports the complex model architectures required for recommendation systems.

Related

  • Project
    tensorflow/privacy

    TensorFlow Privacy is a library that enables developers to train machine learning models with differential privacy, ensuring individual data points remain protected while maintaining model performance.

  • Project
    tdrussell/diffusion-pipe

    diffusion‑pipe is an open‑source training framework that uses DeepSpeed pipeline‑parallelism to train huge diffusion models (image, text‑to‑image, text‑to‑video, etc.) across multiple GPUs. It provides caching of latents/text embeddings, TensorBoard logging, checkpoint/resume, and a unified code path for both image and video models. Install via Conda, set up PyTorch/Flash‑Attention, edit an example TOML config, and launch with `deepspeed`. The repo supports a long list of recent diffusion families (SDXL, Flux, Qwen‑Image, MiniMax H3, etc.) and includes low‑VRAM tricks for training 14‑B‑parameter models on a single GPU.

  • Project
    pykeen/pykeen

    PyKEEN is a Python library for training and evaluating knowledge‑graph embedding models, offering dozens of models, many benchmark datasets, and a high‑level pipeline API.

  • Project
    SciML/SciMLBook

    An evolving open-source textbook based on MIT course material that teaches the integration of high-performance parallel computing with scientific machine learning.