Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

InstructS2S-Eval

Benchmarks

Task NameDataset NameSOTA ResultTrend
Speech-to-speech conversationInstructS2S-Eval En (test)
GPT Score (S)4.02
4
Speech-to-speech conversationInstructS2S-Eval En→En
C-MOS4.17
3
Speech-to-speech conversationInstructS2S-Eval En→Zh
C-MOS3.5
2
Showing 3 of 3 rows