Open qingfusheng opened 3 months ago
Hello. I want to use the method on Chinese dataset, the pre-training model is not doing well. It prefer to generate "Fake" if the input is Chinese and "True" if input is English.
So If I want to retrain it with Chinese, What should I change? If the Whisper tiny.en is still work? Thank you so much,
Hello. I want to use the method on Chinese dataset, the pre-training model is not doing well. It prefer to generate "Fake" if the input is Chinese and "True" if input is English.