AI Atlasby Verbito
Explore › Run & train AI

Machine-learning toolkits

Build machine-learning models from scratch (for developers). Every tool here is free and open source.

Free alternative to DataRobot, Vertex AutoML, H2O Driverless AI, Dataiku

267 tools

  1. Build and train machine-learning models (for developers)
    ★ 201kstars
    tensorflow by tensorflow
    For developersCode libraryHealthy
  2. Build and train neural networks (for developers)
    ★ 105kstars
    pytorch by pytorch
    For developersCode libraryHealthy
  3. Build classic machine-learning models in Python
    ★ 68kstars
    scikit-learn by scikit-learn
    For developersCode libraryHealthy
  4. Build deep-learning models with simple Python code
    ★ 64.4kstars
    keras by keras-team
    For developersCode libraryHealthy
  5. Scale AI training and data jobs across many computers
    ★ 44.2kstars
    ray by ray-project
    For developersCode libraryHealthy
  6. Use hundreds of pretrained image models in PyTorch
    ★ 37.3kstars
    pytorch-image-models by huggingface
    For developersCode libraryHealthy
  7. Match images with text descriptions
    ★ 34.6kstars
    CLIP by openai
    For developersCode libraryFair
  8. Use Caffe, an early deep-learning framework
    ★ 34.6kstars
    caffe by BVLC
    For developersCode libraryUse with care
  9. See inside AI model files visually
    ★ 33.7kstars
    netron by lutzroeder
    No codingDesktop appHealthy
  10. Train translation and language models
    ★ 32.2kstars
    fairseq by facebookresearch
    For developersCode libraryUse with careArchived
  11. Build accurate prediction models from tables of data
    ★ 28.9kstars
    xgboost by dmlc
    For developersCode libraryHealthy
  12. Train deep learning models with less code
    ★ 28.3kstars
    fastai by fastai
    For developersCode libraryHealthy
  13. Explain why a machine-learning model made its prediction
    ★ 25.8kstars
    shap by shap
    For developersCode libraryHealthy
  14. Use Vision Transformers for image recognition
    ★ 25.6kstars
    vit-pytorch by lucidrains
    For developersCode libraryHealthy
  15. Build and train deep learning models with Baidu's framework
    ★ 24.1kstars
    Paddle by PaddlePaddle
    For developersCode libraryHealthy
  16. Build AI that learns from networks and graphs
    ★ 24.1kstars
    pytorch_geometric by pyg-team
    For developersCode libraryHealthy
  17. Move AI models between frameworks
    ★ 21.6kstars
    onnx by onnx
    For developersCode libraryHealthy
  18. Run and train AI models in JavaScript
    ★ 19.2kstars
    tfjs by tensorflow
    For developersCode libraryHealthy
  19. Build fast prediction models from tabular data
    ★ 18.9kstars
    LightGBM by lightgbm-org
    For developersCode libraryHealthy
  20. Use Microsoft's former deep learning toolkit
    ★ 17.6kstars
    CNTK by microsoft
    For developersCode libraryUse with careArchived
  21. Explore Google's deep learning models library
    ★ 17.5kstars
    tensor2tensor by tensorflow
    For developersCode libraryUse with careArchived
  22. Build deep learning models in Rust
    ★ 16.2kstars
    burn by tracel-ai
    For developersCode libraryHealthy
  23. Run machine learning efficiently on everyday hardware (for developers)
    ★ 15.6kstars
    ggml by ggml-org
    For developersCode libraryHealthy
  24. Find the best settings for your machine-learning models
    ★ 15kstars
    optuna by optuna
    For developersCode libraryHealthy
  25. Build neural networks in JavaScript
    ★ 14.9kstars
    brain.js by BrainJS
    For developersCode libraryFair
  26. Create more training images by transforming the ones you have
    ★ 14.7kstars
    imgaug by aleju
    For developersCode libraryFair
  27. Automate machine-learning model tuning
    ★ 14.4kstars
    nni by microsoft
    For developersCode libraryUse with careArchived
  28. Build deep learning models on graphs
    ★ 14.3kstars
    dgl by dmlc
    For developersCode libraryFair
  29. Build deep learning models in Java
    ★ 14.3kstars
    deeplearning4j by deeplearning4j
    For developersCode libraryHealthy
  30. Convert machine-learning code between frameworks
    ★ 14.2kstars
    ivy by unifyai
    For developersCode libraryHealthy
  31. Make AI models run fast on any hardware (for developers)
    ★ 13.9kstars
    tvm by apache
    For developersCode libraryHealthy
  32. Split text into tokens for AI models
    ★ 12.2kstars
    sentencepiece by google
    For developersCode libraryHealthy
  33. Build NLP research models
    ★ 11.9kstars
    allennlp by allenai
    For developersCode libraryUse with careArchived
  34. Split text into tokens fast for AI models
    ★ 11.2kstars
    tokenizers by huggingface
    For developersCode libraryHealthy
  35. Get accurate ML models in three lines of code
    ★ 10.8kstars
    autogluon by autogluon · Free alternative to Vertex AutoML, DataRobot
    For developersCode libraryHealthy
  36. Write fast GPU math for AI (for developers)
    ★ 10.7kstars
    cutlass by NVIDIA
    For developersCode libraryHealthy
  37. Let AI find the best machine-learning pipeline
    ★ 10.1kstars
    tpot by EpistasisLab
    For developersCode libraryFair
  38. Do data science on data you can't see directly
    ★ 10.1kstars
    PySyft by OpenMined
    For developersCode libraryHealthy
  39. Build neural networks with DeepMind's TensorFlow library
    ★ 10kstars
    sonnet by google-deepmind
    For developersCode libraryHealthy
  40. Build machine-learning models with very little code
    ★ 9.9kstars
    pycaret by pycaret · Free alternative to DataRobot
    For developersCode libraryFair

Showing 40 of 267. Search or open the full list for more.