diff --git a/speech/speech_synthesis/vqmivc/pytorch/README.md b/speech/speech_synthesis/vqmivc/pytorch/README.md
index 5c1b1d3adcf82bb8c1060619a3854daf4388b890..00dcdfb60fcd519b353e51190177ba369a83eb80 100644
--- a/speech/speech_synthesis/vqmivc/pytorch/README.md
+++ b/speech/speech_synthesis/vqmivc/pytorch/README.md
@@ -7,33 +7,34 @@ One-shot voice conversion (VC), which performs conversion across arbitrary speak
 
 ## Step 1: Preparing datasets
 
-```
-$ mkdir -p /home/data/vqmivc
-$ cd /home/data/vqmivc
-$ wget https://datashare.ed.ac.uk/bitstream/handle/10283/2651/VCTK-Corpus.zip 
-$ unzip VCTK-Corpus.zip
+```shell
+mkdir -p /home/data/vqmivc/
+cd /home/data/vqmivc/
+wget https://datashare.ed.ac.uk/bitstream/handle/10283/3443/VCTK-Corpus-0.92.zip
+unzip VCTK-Corpus-0.92.zip
 ```
 
 ## Step 2: Preprocess
 
-```
-$ cd vqmivc
-$ pip3 install -r requirements_bi.txt
-$ python3 preprocess.py
-$ ln -s /home/data/vqmivc/data data 
+```shell
+cd ${DEEPSPARKHUB_ROOT}/speech/speech_synthesis/vqmivc/pytorch/
+pip3 install -r requirements_bi.txt
+ln -s /home/data/vqmivc/data data
+python3 preprocess.py
 ```
 
 ## Step 3: Training
 
 * Training with mutual information minimization (MIM):
 
-```
-$ python3 train.py use_CSMI=True use_CPMI=True use_PSMI=True
+```shell
+python3 train.py use_CSMI=True use_CPMI=True use_PSMI=True
 ```
 
 * Training without MIM:
-```
-$ python3 train.py use_CSMI=False use_CPMI=False use_PSMI=False 
+
+```shell
+python3 train.py use_CSMI=False use_CPMI=False use_PSMI=False 
 ```
 
 ## Results on BI-V100
@@ -43,4 +44,4 @@ $ python3 train.py use_CSMI=False use_CPMI=False use_PSMI=False
 | BI      |0.635|1.062 |0.453 |401.693 |110.958|2.653E-4|0.052|0.001|219.895|0.021|4.315|
 
 ## Reference
-https://github.com/Wendison/VQMIVC
\ No newline at end of file
+https://github.com/Wendison/VQMIVC