May be you can think of these based on your requirements:

1. Mate tools

2. BioKIT - For biomedical text

3. SEMAFOR - the parser requires 8GB of RAM

4. SENNA: A Fast Semantic Role Labeling (SRL) Tool

Also there is a comparison done on some of these SRL tools....maybe this too can be useful and help you to decide which one is best for you:

http://www.kenvanharen.com/2012/11/comparison-of-semantic-role-labelers.html

Thanks Renu.

I am using the praticnlptools, an old python package, in a research on critical discourse analysis. It is good, but not well documented.

https://pypi.python.org/pypi/practnlptools/1.0

https://pypi.python.org/pypi/practnlptools/1.0