diff --git a/Mxnet-Scala/TrainQuantization/README.md b/Mxnet-Scala/TrainQuantization/README.md
index fc4aa19..5d1312f 100644
--- a/Mxnet-Scala/TrainQuantization/README.md
+++ b/Mxnet-Scala/TrainQuantization/README.md
@@ -1,5 +1,5 @@
 # MXNET-Scala TrainQuantization
-Simply implementation of Quantization Aware Training[1] with MXNet-scala module. 
+Simpfly implementation of Quantization Aware Training[1] with MXNet-scala module. 
 
 
 ## Setup
@@ -10,33 +10,33 @@ Tested on Ubuntu 14.04
 * sbt 0.13 http://www.scala-sbt.org/
 * Mxnet v1.4 https://github.com/dmlc/mxnet
 
-### steps
+### Build steps
 
 1, compile MXNet with CUDA, then compile the scala-pkg，doc： https://github.com/dmlc/mxnet/tree/master/scala-package
 
-2, in the Mxnet-Scala/TrainQuantization folder 
+2, under the Mxnet-Scala/TrainQuantization folder:
 ```bash
  mkdir lib;
+ ln -s $MXNET_HOME/scala-package/assembly/linux-x86_64-gpu/target/mxnet-full_2.11-linux-x86_64-gpu-1.5.0-SNAPSHOT.jar lib
 ```
-3, copy the compiled mxnet-full_2.11-linux-x86_64-gpu-1.5.0-SNAPSHOT.jar into lib folder;
 
-4, run `sbt` and then compile the project
+3, run `sbt` and then compile the project
 
 ## Train vgg on Cifar10
-Using the script `train_vgg16_cifar10.sh` under the scripts folder to train vgg net from scratch on Cifar10:
+Using the script `train_vgg16_cifar10.sh` under the scripts folder to train vgg from scratch on Cifar10:
 
 ```bash
 FINETUNE_MODEL_EPOCH=-1
 FINETUNE_MODEL_PREFIX=$ROOT/models/
 ```
-Or you can finetune with the provided pretrained model:
+Or you can finetune with the provided pretrain model:
 
 ```bash
 FINETUNE_MODEL_EPOCH=46
 FINETUNE_MODEL_PREFIX=$ROOT/models/cifar10_vgg16_acc_0.8772035
 ```
 
-I did not use any data augmentation and tune the hyper-parameters during training, the best accuracy I got was 0.877, worse than the best accracy 0.93 reported on Cifar10.
+I did not use any data augmentation and carefully tune the hyper-parameters during training, the best accuracy I got was 0.877, worse than the best accracy 0.93 reported on Cifar10.
 
 ## Train vgg with fake quantization on Cifar10
 Using the script `train_quantize_vgg16_cifar10.sh` under the scripts folder to train vgg with fake quantization on Cifar10,
@@ -47,7 +47,7 @@ FINETUNE_MODEL_EPOCH=46
 FINETUNE_MODEL_PREFIX=$ROOT/models/cifar10_vgg16_acc_0.8772035
 ```
 
-If everything goes right, you can get almost the same accuray with pretrained model after serveral epoch.
+If everything goes right, you should get almost the same accuray with pretrained model after serveral epoch.
 
 ## Test vgg with simulated quantization on Cifar10
 Using the script `test_quantize_vgg16_cifar10.sh` under the scripts folder to test pretrained fake quantization vgg with simulated quantization on Cifar10, you must provide the pretrained model:
@@ -58,7 +58,7 @@ FINETUNE_MODEL_PREFIX=$ROOT/models/cifar10_quantize_vgg16_acc_0.877504
 ```
 
 ## Warning
-Currently there is memory leak problems some where in the code, but I can't figure out the reason. You will see the memory usage keep increasing when you run the tranining script. So remenber to stop the traning script when memory usage is too high, and you can resume the training process with saved model previously.
+Currently there is memory leak some where in the code, but I can't figure out the reason. You will see the memory usage keep increasing when you run the tranining script. So remenber to stop the traning script when memory usage is too high, and you can resume the training process with saved model previously.
 
 ## Reference
 [1] Quantizing deep convolutional networks for efficient inference: A whitepaper. https://arxiv.org/pdf/1806.08342.pdf