A Practical Debugging Tool for Training Deep Neural Networks
![](https://www.deeplearningdaily.com/wp-content/uploads/2021/09/a-practical-debugging-tool-for-training-deep-neural-networks_6137e3d333d21-375x210.jpeg)
pip install ‘git+https://github.com/f-dangel/cockpit.git’
Cockpit is a visual and statistical debugger specifically designed for deep learning. Training a deep neural network is often a pain! Successfully training such a network usually requires either years of intuition or expensive parameter searches involving lots of trial and error. Traditional debuggers provide only limited help: They can find syntactical errors but not training bugs such as ill-chosen learning rates. Cockpit offers a closer, more meaningful look into the training process with multiple well-chosen instruments.
![CockpitAnimation](https://github.com/f-dangel/cockpit/raw/main/docs/source/_static/showcase.gif)
Installation
To install Cockpit simply run
pip install 'git+https://github.com/f-dangel/cockpit.git'
Documentation
The documentation provides a full tutorial on how to get started using Cockpit as well as a detailed documentation of its API.
Experiments
To showcase the capabilities of Cockpit we performed several experiments illustrating the usefulness