Fix some typos

Author: alexwong

docs/custom_service.md

@@ -17,7 +17,7 @@ The custom service file should define a method that acts as an entry point for e
 The function can have any name, not necessarily handle, however this function should accept, the following parameters
 
 * **data** - The input data from the incoming request
-* **context** - Is the MMS [context](https://github.com/vrakesh/mxnet-model-server/blob/master/mms/context.py) information passed for use with the custom service if required. 
+* **context** - Is the MMS [context](https://github.com/awslabs/mxnet-model-server/blob/master/mms/context.py) information passed for use with the custom service if required. 
 
 
 The signature of a entry point function is:

examples/densenet_pytorch/README.md

@@ -1,4 +1,4 @@
-# PyToch serving  
+# PyTorch serving  
 This example shows how to serve PyTorch trained models for flower species recognition..
 The custom handler is implemented in `densenet_service.py`.
 For simplicity, we'll use a pre-trained model. For simplicity we will use docker container to run Model Server.