[Fix] Fix configs of wrn models and ofd. (#361)

* 1.revise the configs of wrn22, wrn24, and wrn40. 2.revise the data_preprocessor of ofd_backbone_resnet50_resnet18_8xb16_cifar10

* 1.Add README for vanilla-wrm.

* 1.Revise readme of wrn

Co-authored-by: zhangzhongyu <[email protected]>

configs/distill/mmcls/ofd/ofd_backbone_resnet50_resnet18_8xb16_cifar10.py

@@ -10,10 +10,10 @@
     data_preprocessor=dict(
         type='ImgDataPreprocessor',
         # RGB format normalization parameters
-        mean=[123.675, 116.28, 103.53],
-        std=[58.395, 57.12, 57.375],
+        mean=[125.307, 122.961, 113.8575],
+        std=[51.5865, 50.847, 51.255],
         # convert image from BGR to RGB
-        bgr_to_rgb=True),
+        bgr_to_rgb=False),
     architecture=dict(
         cfg_path=  # noqa: E251
         'mmrazor::vanilla/mmcls/wide-resnet/wrn16-w2_b16x8_cifar10.py',

configs/vanilla/mmcls/wide-resnet/README.md

@@ -0,0 +1,34 @@
+# Wide-ResNet
+
+> [Wide Residual Networks](https://arxiv.org/abs/1605.07146)
+
+<!-- [ALGORITHM] -->
+
+## Abstract
+
+Deep residual networks were shown to be able to scale up to thousands of layers and still have improving performance. However, each fraction of a percent of improved accuracy costs nearly doubling the number of layers, and so training very deep residual networks has a problem of diminishing feature reuse, which makes these networks very slow to train. To tackle these problems, in this paper we conduct a detailed experimental study on the architecture of ResNet blocks, based on which we propose a novel architecture where we decrease depth and increase width of residual networks. We call the resulting network structures wide residual networks (WRNs) and show that these are far superior over their commonly used thin and very deep counterparts. For example, we demonstrate that even a simple 16-layer-deep wide residual network outperforms in accuracy and efficiency all previous deep residual networks, including thousand-layer-deep networks, achieving new state-of-the-art results on CIFAR, SVHN, COCO, and significant improvements on ImageNet.
+
+<div align=center>
+<img src="https://user-images.githubusercontent.com/26739999/156701329-2c7ec7bc-23da-401b-86bf-dea8567ccee8.png" width="90%"/>
+</div>
+
+## Results and models
+
+### Cifar10
+
+| Model  | Top-1 (%) |                Config                 |                                                                                                                                     Download                                                                                                                                      |
+| :----: | :-------: | :-----------------------------------: | :-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: |
+| WRN-16 |   93.04   | [config](./wrn16-w2_b16x8_cifar10.py) | [model](https://openmmlab-share.oss-cn-hangzhou.aliyuncs.com/mmrazor/v1/wide_resnet/wrn16_2_b16x8_cifar10_20220831_204709-446b466e.pth) \| [log](https://openmmlab-share.oss-cn-hangzhou.aliyuncs.com/mmrazor/v1/wide_resnet/wrn16_2_b16x8_cifar10_20220831_204709-446b466e.json) |
+| WRN-22 |           | [config](./wrn22-w4_b16x8_cifar10.py) |                                                                                                                             [model](<>) \| [log](<>)                                                                                                                              |
+| WRN-28 |   95.41   | [config](./wrn28-w4_b16x8_cifar10.py) | [model](https://openmmlab-share.oss-cn-hangzhou.aliyuncs.com/mmrazor/v1/wide_resnet/wrn28_4_b16x8_cifar10_20220831_173536-d6f8725c.pth) \| [log](https://openmmlab-share.oss-cn-hangzhou.aliyuncs.com/mmrazor/v1/wide_resnet/wrn28_4_b16x8_cifar10_20220831_173536-d6f8725c.json) |
+| WRN-40 |           | [config](./wrn40-w2_b16x8_cifar10.py) |                                                                                                                             [model](<>) \| [log](<>)                                                                                                                              |
+
+## Citation
+
+```bibtex
+@INPROCEEDINGS{Zagoruyko2016WRN,
+    author = {Sergey Zagoruyko and Nikos Komodakis},
+    title = {Wide Residual Networks},
+    booktitle = {BMVC},
+    year = {2016}}
+```

configs/vanilla/mmcls/wide-resnet/wrn22-w4_b16x8_cifar10.py

@@ -1,3 +1,3 @@
-_base_ = ['wrn16_2_b16x8_cifar10.py']
+_base_ = ['wrn16_w2_b16x8_cifar10.py']
 model = dict(
     backbone=dict(depth=22, widen_factor=4), head=dict(in_channels=256, ))

configs/vanilla/mmcls/wide-resnet/wrn28-w4_b16x8_cifar10.py

@@ -1,3 +1,3 @@
-_base_ = ['wrn16_2_b16x8_cifar10.py']
+_base_ = ['wrn16_w2_b16x8_cifar10.py']
 model = dict(
     backbone=dict(depth=28, widen_factor=4), head=dict(in_channels=256, ))

configs/vanilla/mmcls/wide-resnet/wrn40-w2_b16x8_cifar10.py

@@ -1,3 +1,3 @@
-_base_ = ['wrn16_2_b16x8_cifar10.py']
+_base_ = ['wrn16_w2_b16x8_cifar10.py']
 model = dict(
     backbone=dict(depth=40, widen_factor=2), head=dict(in_channels=128, ))