How to quickly train the model based on the original model for the audio data of the new speaker？ - Githubissues

mravanelli / SincNet

SincNet is a neural architecture for efficiently processing raw audio samples.

MIT License

1.13k stars 262 forks source link

How to quickly train the model based on the original model for the audio data of the new speaker？ #91

Open joewale opened 4 years ago

joewale commented 4 years ago

How to quickly train the model based on the original model for the audio data of the new speaker

briverse17 commented 4 years ago

What do you actually mean by "quickly" and "data of the new speaker"?