The meaning of beta and gamma in the code

[madoka109]

parser.add_argument('--beta', default=1.0, type=float,
help='supervise loss weight')
parser.add_argument('--gamma', default=1.0, type=float,
help='paco loss')
I found these two parameters in the code, and they are used in losses.py, but I can't understand the using of them, are they mentioned in the paper?
If you can explain them to me, I will very appreciate on you, thank you!

[jiequancui]

Hi,

Thanks for your question.

Please ignore the beta and gamma hyper-parameters. They are set to 1.0 in our all experiments.

The only meaningful hyper-parameter of paco loss is alpha which is described in our paper.

[madoka109]

Thanks for your reply!
Besides, I have another question about c_y in above formula. What does c_y specifically mean?

[jiequancui]

Hi,

The c_y is a learnable sample for class y. The learnable samples are implemented with a linear fc weights.

[madoka109]

I understand, Thanks!