包含2890名动画角色的VITS语音合成模型接下来,你们有角色模型追加计划吗?难道,为止2890名就完成呵?
VITS speech synthesis model with 2890 animated characters next, do you have an additional plan for the character model? Is it possible that the 2890 places have been completed so far?
Please answer the question.
この質問への回答↓
估计也不会加了
It is estimated that it will not be added.
想用别的数据覆盖到所有音,初步运行100多次,音色还没有完全确定然后换成其他的会好一些吗
If you want to cover all the tones with other data, it will be better to run it for more than 100 times initially, but the timbre has not been completely determined and then replaced with other ones.
在没玩vits不是有个优鸭行步态的网页,那个语音更多
Without playing vits, there is not a web page with excellent duck gait, which has more voice.
此前把数据切了十几分,发现就是有一块不对各种报错,扔了就好了,想继续尝试下一步
Before that, I cut the data for more than ten points, and found that there was an incorrect piece of error. Just throw it away. I want to continue to try the next step.
↓
開発基幹メンバー(モデレーターより上)から直々のお達しが来た