A Python library for building custom machine learning models for tasks like image classification, object detection, and recommendations.
Turi Create is a Python library developed by Apple that simplifies the creation of custom machine learning models for tasks such as image classification, object detection, recommender systems, and activity classification. It enables developers to build and deploy ML models without requiring extensive machine learning expertise, and exports models to Core ML for use in Apple applications.
Developers and data scientists looking to integrate machine learning features like image recognition or recommendations into applications, particularly those targeting Apple platforms (iOS, macOS, watchOS, tvOS).
It offers an easy-to-use, task-oriented API with built-in visualizations and support for multiple data types, streamlining the ML workflow and reducing the complexity typically associated with custom model development.
Turi Create simplifies the development of custom machine learning models.
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Offers a high-level API for common ML tasks like image classification and recommender systems, allowing models to be built with just a few lines of code as shown in the example.
Includes streaming visualizations directly within the library, enabling developers to explore and understand datasets without external tools.
Handles diverse data types including images, audio, video, text, and sensor data, making it versatile for various applications beyond tabular data.
Exports trained models to Core ML format for easy deployment on iOS, macOS, watchOS, and tvOS apps, as emphasized in the iOS app example.
Requires x86_64 architecture and lacks native Windows support—only works via WSL—restricting accessibility for developers on non-compatible systems.
Abstracts away low-level algorithmic details, making it unsuitable for advanced ML research or projects requiring fine-grained control over model architectures.
GPU acceleration varies by operating system and model type, as shown in the GPU support table, leading to unpredictable performance gains across different tasks.