Open jgrinaveckis opened 4 years ago
Hello,
For quite some time I can't model LogNormal AFT model, as no matter what learning rate, optimizer I select, gradient always explodes... Is there could be anything to this method specifically here?
Thanks for answer in advance!
Hello,
For quite some time I can't model LogNormal AFT model, as no matter what learning rate, optimizer I select, gradient always explodes... Is there could be anything to this method specifically here?
Thanks for answer in advance!