b13b124
|
|
_base_ = './deeplabv3_r50-d8_769x769_80k_cityscapes.py'
model = dict(
pretrained='torchvision://resnet18',
backbone=dict(type='ResNet', depth=18),
decode_head=dict(
in_channels=512,
channels=128,
),
auxiliary_head=dict(in_channels=256, channels=64))
|
