checkpoints

[Parisa-Bagherzadeh]

How can I create the checkpoints for each dataset of speakers separately in different folders?

[MaxMax2016]

train each dataset of speakers separately；Train all dataset of speakers at once , they share a single checkpoint.

[Parisa-Bagherzadeh]

I have 5 speakers, speaker0, speaker1,... how i can train and save the checkpoints for each in separate folders?

[MaxMax2016]

when train speaker0, delete data of speaker1 speaker2~~~ or move to somewhere

[Parisa-Bagherzadeh]

/

train each dataset of speakers separately；Train all dataset of speakers at once , they share a single checkpoint.

A single checkpoint for all speakers?! wouldn't that give bad results after inference?

[MaxMax2016]

A single checkpoint for all speakers?! yes if they are all hight quality, every thing will be ok.