Question about the provided checkpoints - Chinese Language Compatibility

[yunrongma]

Were the pre-trained checkpoints specifically optimized or trained on Chinese audio/speech data? Will the performance be affected if I test it on Chinese language audio? Any insights or recommendations would be greatly appreciated.