Introduction

Models

Available Rapidly models with file sizes and real-time factors.

The following table shows the currently available models along with CPU efficiency, file size, and latencies:

Model File NameFile SizeReal-time Factor*
speech-denoise-dereverb-96ms.v1.0.rapidly926 KB27x
speech-denoise-dereverb-32ms.v1.0.rapidly854 KB25x
speech-denoise-dereverb-micro-32ms.v1.0.rapidly241 KB115x
speech-denoise-dereverb-21ms.v1.0.rapidly851 KB12x
speech-denoise-dereverb-11ms.v1.0.rapidly615 KB12x
speech-denoise-96ms.v1.0.rapidly925 KB29x
speech-denoise-32ms.v1.0.rapidly854 KB27x
speech-denoise-micro-32ms.v1.0.rapidly241 KB125x
speech-denoise-21ms.v1.0.rapidly851 KB13x
speech-denoise-11ms.v1.0.rapidly615 KB13x

*The real-time factor is measured on a single core of an AMD RYZEN AI MAX+ 395.

Quick Start

Install the Rapidly SDK and run your first denoise in minutes.

Inference

Platforms supported by the Rapidly Engine and how it accelerates inference across them.