diff --git a/README.md b/README.md
index 01d2670..30ee3d5 100644
--- a/README.md
+++ b/README.md
@@ -30,7 +30,9 @@ Figure 1. Validation Curve Achieved for differnet Regularizers Proposed
 - [x] Mutual Coherence Based
 - [x] Restricted Isometry (**Best Performing** )
 
-## Usage Wide-Resnet CIFAR
+### Usage Wide-Resnet CIFAR
+For CIFAR datasets,we choose Wide Resnet Arch. with a depth of 28 and Kernel width of 10,which
+gives the best results for comparable number parameters for any Wide-Resnet Model. 
 To train on Cifar-10 using 2 gpu:
 
 ```bash
@@ -46,10 +48,13 @@ CUDA_VISIBLE_DEVICES=6,7 python train_n.py --ngpu 2 --dataset cifar100
 After train phase, you can check saved model in the ```runs``` folder.
 
 ## Usage Wide-Resnet SVHN
+For SVHN datasets,we choose Wide Resnet Arch. with a depth of 16 and Kernel width of 8,which
+gives the best results for comparable number parameters for any Wide-Resnet Model. 
 ``` bash
 CUDA_VISIBEL_DEVICES=0 python train.py --dataset svhn --model wideresnet --learning_rate 0.01 --epochs 160
 ```
 
+## Result
 | **Network** | **CIFAR-10** | **CIFAR-100** | **SVHN** |
 | ----------- | ------------ | ------------- | -------- |
 | WideResNet  | 4.16       | 20.50          | 1.60     |