Changes to the `BatchMatrixMultiply` benchmarking code:

google / XNNPACK

High-efficiency floating-point neural network inference operators for mobile, server, and Web

Other

1.89k stars 376 forks source link

Closed copybara-service[bot] closed 3 weeks ago

copybara-service[bot] commented 3 weeks ago

Changes to the BatchMatrixMultiply benchmarking code: