feature: interactively exporting loaded model to binfile

[jon-chuang]

With #820, the loaded model may be different from the base model. It makes sense to be able to interactively export the currently loaded model to binfile.

Especially if one allows the option to linearly interpolate multiple LoRA files - i.e. LoRA mixology for a unique LLM personality.

[MillionthOdin16]

If you're familiar with the mixing of loras I think it would be helpful for a lot of people here if you could link some resources on it. I've heard you can do some pretty cool things, but I'm not very familiar.

[jon-chuang]

I'm not that familiar, it simply occured to me as a possibility.

Linearly interpolating between base model and LoRA is already a standard feature: https://huggingface.co/docs/diffusers/main/en/training/lora

Anw, linearly interpolating is tracked here: #905

This issue is about exporting loaded (and modified e.g. fine-tuned) model to binfile.

[github-actions]

This issue was closed because it has been inactive for 14 days since being marked as stale.