Update WRITEUP.md

Corrected a typographical error

WRITEUP.md

@@ -11,8 +11,6 @@ OpenVINO toolkit uses the model optimizer to convert the known layers in a model
 
 The process behind converting custom layers involves a preliminary check for unsupported layers using model and the device type, via the query_network method on the IECore class. If there are unsupported layers, a custom layer can then be added with the add_extension method using the required library and deployment device type. 
 
-Some of the potential reasons for handling custom layers are...
-
 ## Comparing Model Performance
 
 My method(s) to compare models before and after conversion to Intermediate Representations
@@ -36,4 +34,4 @@ The second use case can be very useful, especially in limiting the spread of inf
 
 No model is hundred percent accurate. The model deployed in this project has a mean average precision of 22 COCO mAP(^1). It could be seen that the model was not always identifying a particular person who was dressed in an all black attire and had his hair styled in a particular unique way. 
 
-The model would perform poorly in poor lighting conditions. The model should not be expected to detect people when the ambience is dark. 
+The model would perform poorly in poor lighting conditions. The model should not be expected to detect people when the ambience is dark.