深度学习网络层配置

在这个网络参数中,我们的输入是
def __init__(self, num_layer, layer_num, layer_filter, output_channel, deploy=False, **kwargs):

 

net = UnetFlexible(num_layer=cfg.seg_num_stage, layer_num=cfg.seg_layer_num, layer_filter=cfg.seg_layer_filter,
                   output_channel=cfg.net_output_channel, deploy=deploy)
网络层数
seg_num_stage = 6
每层的子层数目
seg_layer_num = [2, 2, 2, 2, 2, 2]
特征通道数
seg_layer_filter = [32, 32, 64, 64, 128, 128]
输出通道数
net_output_channel = 1

所以

网络层数
self.num_layer=cfg.seg_num_stage=seg_num_stage = 6
每层的子层数目
self.layer_num=cfg.seg_layer_num=seg_layer_num = [2, 2, 2, 2, 2, 2]
每层的特征通道数

self.layer_filter=cfg.seg_layer_filter=seg_layer_filter = [32, 32, 64, 64, 128, 128]
输出通道数
self.output_channel=cfg.net_output_channel=net_output_channel = 1

所以

self.num_layer==cfg.seg_num_stage=seg_num_stage = 6
len(self.layer_num)=6
len(self.layer_filter)=6
sum(self.layer_num)=12

对于 sum(self.layer_num),它将是 seg_layer_num 列表中所有元素的总和。在这种情况下,seg_layer_num 的元素是 [2, 2, 2, 2, 2, 2],所以 sum(self.layer_num) 的结果是 2 + 2 + 2 + 2 + 2 + 2 = 12

断言语句的一般形式是

assert <条件表达式>, <错误信息表达式>

assert 是一个断言语句,用于在代码中进行条件检查。它用于确保某个条件为真,如果条件为假,则会引发一个 AssertionError 异常。

当条件表达式为假时,断言会触发异常,并将错误信息表达式作为异常信息显示。如果条件表达式为真,则断言通过,程序继续执行。

你可能感兴趣的:(python,python)