Description
Deep Learning Profiler (DLProf) is a tool for profiling deep learning models to help data scientists understand and improve performance of their models visually via Tensorboard or by analyzing text reports. It also helps understand resource usage when models are trained.
Key Features
The key features of DLProf v0.4.0/r19.09:
- CLI improvements: Report generation has changed, output options have been improved to make it easier for users to generate detailed reports.
- Tensorboard Improvements:
- Model summary tab now shows a kernel summary report that details GPU time summary for kernels in use.
- Iterations summary tab now shows operation names correlated with the kernels used in detail.
Known Issues
- This software is only accessible in the NGC TensorFlow container.