djlbench is a command line tool that allows you to benchmark the model on all different platforms for single-thread/multi-thread inference performance.

Currently djlbench support the models from the following framework:

  • PyTorch
  • TensorFlow
  • Apachmark MXNet
  • PaddlePaddle
  • ONNXRuntime
  • TensorFlow Lite
  • Neo DLR
  • XGBoost

Get it from the Snap Store

Search for another snap, or go back to the homepage.