Open cphoover opened 7 years ago
Split the training data into chunks assign chunks to each cpu. Parallelize the process of training the data and mapping.
I agree it'd be great.
I have no plan to work on this at the moment, not enough free time.
Free free to implement and send a pull request or fork though.
Split the training data into chunks assign chunks to each cpu. Parallelize the process of training the data and mapping.