Javihaus/ndt

Neural Dimensionality Tracker (NDT) is a production-ready Python library for high-frequency monitoring of neural network representational dimensionality during Neural Networks training

/ 100

Experimental

This tool helps machine learning researchers and practitioners understand how deep neural networks learn by monitoring the complexity of their internal data representations during training. It takes your PyTorch neural network and training data, then outputs visualizations and data that show how the network's internal structure changes over time, including detecting sudden 'jumps' in complexity. This is for anyone who develops, researches, or fine-tunes deep learning models and wants to gain deeper insights into their learning process.

Use this if you need to understand the learning dynamics and internal representational changes of your deep neural networks at a very high frequency during training.

Not ideal if you are looking for a tool to simply monitor standard training metrics like loss and accuracy, or if you are not working with deep neural networks.

deep-learning-research neural-network-training model-interpretability machine-learning-engineering representation-learning

No Package No Dependents

Maintenance 6 / 25

Adoption 5 / 25

Maturity 13 / 25

Community 0 / 25

How are scores calculated?

Stars

Forks

—

Language

Python

License

MIT

Higher-rated alternatives

emlearn/emlearn

Machine Learning inference engine for Microcontrollers and Embedded devices

analogdevicesinc/ai8x-training

Model Training for ADI's MAX78000 and MAX78002 Edge AI Devices

DT42/BerryNet

Deep learning gateway on Raspberry Pi and other edge devices

eclypse-org/eclypse

An Edge-Cloud python platform for simulated (and emulated) runtime environments

SummerGift/EmbeddedSystem

:books: 计算机体系架构、嵌入式系统基础与主流编程语言相关内容总结

Explore ML Frameworks

All categories Trending ML Framework directory Insights