Closed XuKun2424 closed 2 years ago
refer to this article
Inherit a base swin-transformer configuration file, and add this.
_base_ = ['../_base_/models/swin_transformer/tiny_224.py']
backbone=dict(
init_cfg = dict(
type='Pretrained',
checkpoint="https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925-66df6be6.pth",
prefix='backbone')
)
refer to this article
Inherit a base swin-transformer configuration file, and add this.
_base_ = ['../_base_/models/swin_transformer/tiny_224.py'] backbone=dict( init_cfg = dict( type='Pretrained', checkpoint="https://download.openmmlab.com/mmclassification/v0/swin-transformer/swin_tiny_224_b16x64_300e_imagenet_20210616_090925-66df6be6.pth", prefix='backbone') )
Haha, thank my brother
推荐使用英语模板 General question,以便你的问题帮助更多人。
首先确认以下内容
描述你遇到的问题
self = SwinTransformer(**extra_config) self._load_from_state_dict()
相关信息
pip list | grep "mmcv\|mmcls\|^torch"
命令的输出 [填写这里]mmcls
文件夹下的代码做了其他相关的修改,请在这里写明 [填写这里]