The paramters in optimizer

jiwoon-ahn / psa

Learning Pixel-level Semantic Affinity with Image-level Supervision for Weakly Supervised Semantic Segmentation, CVPR 2018

MIT License

385 stars 62 forks source link

The paramters in optimizer #28

Open lzyhha opened 3 years ago

lzyhha commented 3 years ago

Hello, I note that the order of paramters (params lr wd) in PolyOptimizer is different from official SGD(params lr momentum). So I think the value of wd will actually be assigned to momentum. Is it so?

class PolyOptimizer(torch.optim.SGD):

    def __init__(self, params, lr, weight_decay, max_step, momentum=0.9):
        super().__init__(params, lr, weight_decay)