Skip to content

Latest commit

 

History

History
32 lines (22 loc) · 969 Bytes

README.md

File metadata and controls

32 lines (22 loc) · 969 Bytes

ml-bench 🔬

A tool for benchmarking inference time of onnx models using the OpenVINO runtime.

The benchmarks are run using criterion, running a forward pass on the model --samples times. Once the benchmark is complete, the results are printed to stdout.

Usage

./ml-bench <model> [--samples <num_samples>]

Building

macOS

When compiling from macOS to x86-64-unknown-linux-gnu, create a .cargo/config.toml file with the following content:

[target.x86_64-unknown-linux-gnu]
linker = "x86_64-unknown-linux-gnu-gcc"
ar = "x86_64-unknown-linux-gnu-ar"

[env]
TARGET_CC = "x86_64-unknown-linux-gnu-gcc"
TARGET_CXX = "x86_64-unknown-linux-gnu-g++"
CC = "/opt/homebrew/opt/x86_64-unknown-linux-gnu/bin/x86_64-linux-gnu-gcc"
CXX = "/opt/homebrew/opt/x86_64-unknown-linux-gnu/bin/x86_64-linux-gnu-g++"