Update readme

README.md

@@ -1,10 +1,14 @@
-# Alexnet-tf2
+# Alexnet
 
-Implemented "ImageNet Classification with Deep Convolutional Neural Networks (NIPS 2012)" a.k.a. Alexnet using tf2.
-(I have implemented their customized nesterov gradient descent optimizer, but do not implemented Distributed learning design.)
+Implemented "ImageNet Classification with Deep Convolutional Neural Networks (NIPS 2012)" a.k.a. Alexnet using tensorflow2.
 
-I have confirmed that Top-1 validation error using 1-CNN model is achieved 40.8% as described in Table 2 in the paper.
+# Note
 
-![top1](figure/top1.png)
+- I have implemented their customized nesterov gradient descent optimizer, but do not design distribution learning.
 
-![top5](figure/top5.png)
+- As shown below, the top-1 validation error using 1-CNN model achieved 39.98, similar to that described in Table 2 of the paper (40.7).
+
+    ![top1](figure/top1.png)
+
+- Top 5 validation error acheived 5.77 because I have used only 10 classes to train and validate the model.
+    ![top5](figure/top5.png)