Wanted to try compressing multiple prompts down into one.

First time training a LoRA and I missed the mark a little but I still find the result somewhat convenient and wanted to share it.

I find the most success at low cfg (I default to 3) and 0.7 lora strength. Please do share your own results!

Any feedback on this model or advice on training is much appreciated and I intend to update the model in future.