Update autoregressive to support type inputs

implement clip-guided generation (and never use it...)

support latents into the diffusion decoder

Update API to have more expressive interface for controlling various generation knobs

Modifications to support "v1.5"

Some renaming