How to control the model parameters when jointly optimize 'alpha' and 'W' over the union of the training and validation sets using coordinate descent.?

quark0 / darts

Differentiable architecture search for convolutional and recurrent networks

https://arxiv.org/abs/1806.09055

Apache License 2.0

3.92k stars 843 forks source link

How to control the model parameters when jointly optimize 'alpha' and 'W' over the union of the training and validation sets using coordinate descent.? #140

Open larenzhang opened 4 years ago

larenzhang commented 4 years ago

The paper reported the model parameters is 3.1M if α and w are jointly optimized over the union of the training and validation sets using coordinate descent.

I want to know that the model parameters directly searched out is 3.1M or after modifying the number of conv channels？

Look forward to your reply.