[Progress Report] Hybrid methods Fusing Audio And Text

[KatJHuang]

Some existing implementations

• https://arxiv.org/pdf/1810.04635.pdf <- looks similar to what we proposed; producing results from end layers of text and spectrogram networks
• http://aclweb.org/anthology/P18-1207 <- they do it at lower levels