Deep Learning Profiler 19.07 Release Notes

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 v0.3.9/r19.07:
  • Ability to aggregate data per iteration: User can specify the iteration range to aggregate timing metrics for all reports by specifying start and stop iterations.
  • Tensor Core Report: DLProf can create a CSV report listing all unique Tensor Core kernels that were executed in the model, along with node and timing metric information.
  • Support for Tensorboard 1.14: Visualization component is now based on Tensorboard 1.14.

Known Issues

  • This is early version software. It is only accessible in the NGC TensorFlow container.

Resolved Issues

  • Bug fixes in timing data in Tensorboard.
  • Fix for models not showing Tensor Core usage.