Playing around with ZImage Turbo and put together this merge of models.
Requires VAE and Text Encoder.