would be really helpful if we could 1-click upload to huggingface (just set HF_TOKEN in the UI and upload).
The repo might look something like:
README.md
acidhouse-step=1000-epoch=20.safetensors
acidhouse-step=1000-epoch=41.safetensors
...
Where the README.md contains the header:
Which will make sure they get tagged as underfit, which then can be used to explore all the published LoRAs tagged as underfit (using huggingface api) and maybe other stableaudio3 apps could implement a way to
See example: https://huggingface.co/api/models?filter=fluxgym
Then any stableaudio client can easily implement a feature to discover and download underfit loras
would be really helpful if we could 1-click upload to huggingface (just set HF_TOKEN in the UI and upload).
The repo might look something like:
Where the
README.mdcontains the header:Which will make sure they get tagged as underfit, which then can be used to explore all the published LoRAs tagged as
underfit(using huggingface api) and maybe other stableaudio3 apps could implement a way toThen any stableaudio client can easily implement a feature to discover and download underfit loras