支持中文吗？

[JingKun Hao]

[Sanchit Gandhi]

It does not, but you have two options for training a Chinese-compatible model:

1. Follow the distillation instructions in the training folder and train on the Mandarin split of the Common Voice dataset https://github.com/huggingface/distil-whisper/tree/main/training
2. Fine-tune the pre-trained checkpoint on Mandarin (instructions for this can also be found under the training folder)

=> 1 will get you the best results, but is a little bit more involved than 2

Disclaimer: the following translation is generated using Google Translate

它没有，但你有两种选择来训练中文兼容模型：

1. 按照训练文件夹中的蒸馏说明，对 Common Voice 数据集的普通话分割进行训练 https://github.com/huggingface/distil-whisper/tree/main/training
2. 微调普通话的预训练检查点（也可以在训练文件夹下找到相关说明）

=> 1 会给你最好的结果，但比 2 稍微复杂一些

免责声明：使用 Google 翻译翻译回复

[JingKun Hao]

thanks your recommendation

[shuaijiang]

you can refer to https://huggingface.co/BELLE-2/Belle-distilwhisper-large-v2-zh
support Chinese based on distil-whipser-larger-v2

[JingKun Hao]

thanks