Best place to learn about Bayesian Nonparametric modelling?

Videolectures is your friend here. I recommend you do not try jumping in straight into nonparametric bayesian models, however, and make sure you're comfortable with their parametric equivalents first. A good reference is Chris Bishop's book, Machine Learning and Pattern Recognition, as David Mackay's book on Information Theory, Inference, and Learning Algorithms, which is free online.

In practice there's a huge difference in many ways between nonparametric bayesian approaches centered around Gaussian processes and nonparametric bayesian approaches centered around Dirichlet, Indian Buffet, Beta, and Poisson processes, though of course there are deep connections.

If what interests you is this latter group, I recommend checking out Edwin Chen's amazing blog post on infinite mixture models, Kevin Knight's funny tutorial on Bayesian inference with tears, Resnik's introduction to Gibbs sampling, Michael Jordan's lectures on CRPs and DPs, and any video lecture you can find from one of the Machine Learning summer schools (there are many, and they're really different and the style in some might suit you more than others).

If you're interested in the Gaussian Processes side of things, however, I really like Rasmussen's Gaussian Processes for Machine Learning book (also free online), and you can just follow the links from that website to get lighter and heavier introductions.