List your company

ResearchFormerly WizWand

OpenAudioBench

Benchmarks

TaskDatasetSOTA resultTrendResults

Speech Quality Evaluation

OpenAudioBench English subsets (test)

2.31WER

15

Text Quality Evaluation

OpenAudioBench English subsets (test)

7.43AlpacaEval

15

Efficiency Evaluation

OpenAudioBench English subsets (test)

2,153.43TPS

15

Spoken Question Answering

OpenAudioBench

74.82Accuracy

9

Speech Generation

OpenAudioBench (response)

4.32WER

7

Speech-to-Text Spoken Question Answering

OpenAudioBench S2T (test)

88.94AlpacaEval Score

7

Speech Generation

OpenAudioBench (subset)

0.57RTF

5

Spoken Question Answering

OpenAudioBench S2T

88.89AlpacaEval Score

4