search

neuralmagic / deepsparse

Sparsity-aware deep learning inference runtime for CPUs
https://neuralmagic.com/deepsparse/
Other
2.98k stars 172 forks source link

How to use Benchmark feature on custom dataset #1256

Closed atultiwari closed 12 months ago

atultiwari commented 12 months ago

Hi, First of all thank you very much for this awesome project. I wanted to try it out for research purpose to compare its effectiveness with respect to GPU. I am trying to do this on custom dataset. This dataset has 10 classes. So there is folder named Dataset, in that there are two folders - Train & Test. In both of those there are 10 classes and each class has 200 images. Now I want to use benchmark on how much time does Yolo v8 take on CPU & on GPU, and how much time does Deepsparse takes. But I am unable to find any solution for that. Can you please guide me. thank you

Satrat commented 12 months ago

Hi @atultiwari, DeepSparse has built-in benchmarking you can invoke with deepsparse.benchmark_pipeline that will allow you to pass in custom input data. See the benchmark README for full instructions