Skip to content

P
PaddleSeg

PaddlePaddle / PaddleSeg

通知 289
Star 8
Fork 1
P
PaddleSeg
提交 d9ac9ca8 编写于 作者: P pengmian
浏览文件
操作

add name bn

上级 8efa5b79
隐藏空白更改
内联 并排
Showing with 4 addition and 4 deletion
pretrained_model/download_model.py
浏览文件 @ d9ac9ca8
...@@ -57,7 +57,7 @@ model_urls = { ...@@ -57,7 +57,7 @@ model_urls = {
"https://paddleseg.bj.bcebos.com/models/unet_coco_v3.tgz", "https://paddleseg.bj.bcebos.com/models/unet_coco_v3.tgz",
"icnet_bn_cityscapes": "icnet_bn_cityscapes":
"https://paddleseg.bj.bcebos.com/models/icnet_cityscapes.tar.gz", "https://paddleseg.bj.bcebos.com/models/icnet_cityscapes.tar.gz",
"pspnet50_cityscapes": "pspnet50_bn_cityscapes":
"https://paddleseg.bj.bcebos.com/models/pspnet50_cityscapes.tgz" "https://paddleseg.bj.bcebos.com/models/pspnet50_cityscapes.tgz"
} }
......
turtorial/finetune_pspnet.md
浏览文件 @ d9ac9ca8
...@@ -21,7 +21,7 @@ python dataset/download_pet.py ...@@ -21,7 +21,7 @@ python dataset/download_pet.py
接着下载对应的预训练模型 接着下载对应的预训练模型
```shell ```shell
python pretrained_model/download_model.py pspnet50_cityscapes python pretrained_model/download_model.py pspnet50_bn_cityscapes
``` ```
## 三. 准备配置 ## 三. 准备配置
...@@ -117,5 +117,5 @@ python pdseg/eval.py --use_gpu --cfg ./configs/test_pet.yaml ...@@ -117,5 +117,5 @@ python pdseg/eval.py --use_gpu --cfg ./configs/test_pet.yaml
|预训练模型名称|BackBone|Norm|数据集|配置| |预训练模型名称|BackBone|Norm|数据集|配置|
|-|-|-|-|-| |-|-|-|-|-|
|pspnet50_cityscapes|ResNet50|bn|Cityscapes|MODEL.MODEL_NAME: pspnet <br> MODEL.DEFAULT_NORM_TYPE: bn <br> MODEL.PSPNET.LAYERS: 50| |pspnet50_bn_cityscapes|ResNet50|bn|Cityscapes|MODEL.MODEL_NAME: pspnet <br> MODEL.DEFAULT_NORM_TYPE: bn <br> MODEL.PSPNET.LAYERS: 50|
|pspnet101_cityscapes|ResNet101|bn|Cityscapes|MODEL.MODEL_NAME: pspnet <br> MODEL.DEFAULT_NORM_TYPE: bn <br> MODEL.PSPNET.LAYERS: 101| |pspnet101_bn_cityscapes|ResNet101|bn|Cityscapes|MODEL.MODEL_NAME: pspnet <br> MODEL.DEFAULT_NORM_TYPE: bn <br> MODEL.PSPNET.LAYERS: 101|
Markdown is supported
0% 或 .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑！
想要评论请 注册 或