Skip to content

ASR Benchmark Overview

The ASR Benchmark functionality simplifies the evaluation of Automatic Speech Recognition (ASR) models under various scenarios. It supports user-defined benchmarks and comes with out-of-the-box text normalization for Arabic and English.

Key Features

  1. Custom Benchmark Creation

    • Design benchmarks tailored to your specific ASR use cases.
    • Quickly import open-source datasets for immediate testing.

  2. Normalization & Scoring

    • Automatic text normalization for Arabic and English.
    • Streamlined scoring to assess model accuracy and performance.

Next Steps

  • LLM Fine-Tuning UI – Explore how to fine-tune LLMs before comparing ASR performance.
  • Benchmarks UI – Learn how to manage and monitor custom benchmark packages and runs.
  • Inference UI – Deploy your ASR models and evaluate them in real-world conditions.