diff --git a/README.md b/README.md
index 972e05c..813d158 100755
--- a/README.md
+++ b/README.md
@@ -14,3 +14,19 @@ Options to run:
 * -j Json file of the collection
 * -f Save image to the folder under the name of collection
 * -m Run in parallel, value is p for parallel running option, it will check the number of processor the machine has and will utilizes 75% of it.
+
+
+## How to train a binary classification model (AlexNet)
+- Copy trainning images to NEG and POS folder
+```bash
+cd AIDRImageFiltering/
+java -classpath /target/image-filtering-0.0.1-SNAPSHOT.jar hbku.qcri.sc.aidr.filtering.ImageFilteringModel -d toTrain/ -n 100 -b 60 -i 2 -s savedModel
+```
+
+Options to run:
+* -d dataset folder
+* -n total number of training images 
+* -b batch_size
+* -i interation
+* -l learing rate
+* -s save trained model to a folder 
diff --git a/src/main/java/hbku/qcri/sc/aidr/filtering/ImageFilteringModel.java b/src/main/java/hbku/qcri/sc/aidr/filtering/ImageFilteringModel.java
index cfc709a..2e01e2f 100755
--- a/src/main/java/hbku/qcri/sc/aidr/filtering/ImageFilteringModel.java
+++ b/src/main/java/hbku/qcri/sc/aidr/filtering/ImageFilteringModel.java
@@ -77,7 +77,7 @@ public void run(String[] args) throws Exception {
 		options.addOption("n", "num_examples", true, "Number of examples");
 		options.addOption("b", "batch_size", true, "batch_size");
 		options.addOption("i", "iterations", true, "interation");
-		options.addOption("l", "iterations", true, "learing rate");
+		options.addOption("l", "l_rate", true, "learing rate");
 		options.addOption("s", "saved_model", true, "saved_models");
 
 		CommandLine commandLine = parser.parse(options, args);