text conditioning

[nikuson]

I would like to understand how to do something like Glide in audio diffusion.

[flavioschneider]

It's not yet possible to do without changing the model. Text conditioning is currently a work in progress.

[jbmaxwell]

I'm training on the text conditioning model, using a tweaked diff_textcond_test.yaml file, and I'm just wondering how I might make it stereo, rather than mono? I've been looking at other stereo (channels = 2) examples in exp, but I'm pretty certain I'll smash things to a buh-zillion pieces trying to just hack it... :)

Any tips appreciated.