Introduction to Tensorleap

Tensorleap is a debugging, observability, and explainability platform for Deep Neural Networks that allows data scientists to develop advanced Deep Learning models dramatically faster and with far better results.

It uses deep algorithms to provide insights into the model’s health, identify strengths and weaknesses and allow data scientists to fully understand what in the data caused the model to fail, and the exact reasons for the failures.

The platform includes tools to conduct guided error analysis, unit testing, and dataset architecture, enabling data scientists to build reliable and effective models they can trust.

Tensorleap integrates into any Deep Learning environment and can demonstrate value in a short time and with low effort.

QuickStart

• Setup Tensorleap

• Integrate an off-the-shelf example

Reference

• Resources Management

• Project

• Dataset

• Secret Manager

• Network

• Evaluate a Model

• Versions

• Analysis