[Question] Why Training Adapter
Hi, I am fairly new to training a distilled model.
Could someone explain why do we need a training adapter to train a LoRA?
Really appreciated your answer.
I added some more information to the model card that should clarify it. Let me know if you have more questions @evilperson068
where do you place the adapter? what folder
I assume this is built into AIT, so you just train your model and call it a day. But do you then need to use this model in ComfyUI when using your trained LoRA? If so, at what strength?
where do you place the adapter? what folder
It will download automatically.
I assume this is built into AIT, so you just train your model and call it a day. But do you then need to use this model in ComfyUI when using your trained LoRA? If so, at what strength?
The model is only used during training. On inference in comfy, you can just do their default turbo inference with your trained Lora attached and it will function in the turbo model.
hi，i am a freshman, i should use zimage_turbo_training_adapter before i generate my image data or only in training my new lora model?@evilperson068,thanks