FunAudioLLM / CosyVoice

Multi-lingual large voice generation model, providing inference, training and deployment full-stack ability.
https://funaudiollm.github.io/
Apache License 2.0
6.54k stars 703 forks source link

WER计算 #657

Open liuxin99 opened 1 week ago

liuxin99 commented 1 week ago

请问在aishell3-test上的CER的计算设置是什么呢? 我用CosyVoice-300M的inference_cross_lingual 推理计算了aishell3-test的CER,但是算的结果是8.1%,和论文的3.8%还有一定距离?是不是我的设置有问题?

aluminumbox commented 1 week ago

这个 @ZhihaoDU 可以麻烦回复一下