mobileML 1.0

mobileML provides a platform for utilizing neural networks effectively and efficiently.

MAIN FEATURES

• LEARN about neural networks
  • Access an extensive Help Center containing detailed tutorials on using mobileML and the fundamental mathematics of neural networks.
  • Explore and download sample networks from MagMHJ.com to gain a practical understanding of neural networks within mobileML.
• DESIGN your own neural networks
  • Create customized, fully-connected neural networks with tailored dimensions to suit your needs.
  • Incorporate input nodes that can be updated with various data sources such as sensor data, stock prices, cryptocurrency prices, microphone data, and more.
  • Utilize HealthKit data to design networks geared towards enhancing health management effectively.
  • Quickly generate common neural network structures by applying presets.
  • Edit values of individual nodes, edges, and activation functions within each layer.
• TRAIN neural networks
  • Live train networks to produce extensive datasets.
  • Leverage HealthKit data for training to develop networks aimed at simplifying health management tasks.
  • Opt for automatic control of learning rate during network training as needed.
  • Retrain networks using previously saved datasets or import datasets from external sources.
  • Export datasets as text files for easy sharing with peers or use in other network configurations.
• RUN neural networks and automate tasks
  • Automatically execute network propagation whenever input nodes are updated.
  • Assign actions such as sending Bluetooth messages or launching websites to different output nodes.
  • Seamlessly chain different networks together by associating actions that trigger the execution of other interconnected networks.
• SHARE networks easily with AirDrop, messaging & social media
  • Pack and share networks (including images and associated actions) across various platforms like iMessage and AirDrop effortlessly.
• EXPORT networks as Python scripts
  • Easily convert networks into PyTorch (Python 3) scripts for running them on diverse platforms.
  • Create packaged PyTorch scripts along with companion datasets to streamline training processes on different platforms efficiently.
  • mobileML serves as an excellent prototyping tool for developing enterprise-grade machine learning solutions.