refactor to begin to productionize some of the pipelines

[free-soellingeraj]

• creates data bunch user interface
• pairs together the waveform and the sample_rate which are commonly used together
• moves some of the dictionaries into classes

[wooters]

Hey Aaron! I just took a look and it seems like a great start. Using torch.Tensor to store the audio in memory seems reasonable to me. I'm not a software engineer, so I can't make any knowledgable comments about how you should test your code. Anyway, I think you should move forward with what you have here and see how well it works for your use case. :-)