A Clear step by step local Installation Method would be nice.
The Readme.md is a little cryptic and could be clearer about setting this up locally and getting it to run.
In the interests of making this technology more widely available to non-coders can StabilityAi please do a step by step install video with a Gradio or ComfyUi interface.
I have git cloned stable_audio_tools-0.0.16
I have downloaded model and .cfg
- Where does the model get put?
- Does it need to be re-named?
- How exactly does one go about getting the Hugging Face token?
I don't want to sound ungrateful, but given the brain-power that must exist within StabilityAi surely accessibility of your work to your market would be a priority.
I appreciate this all takes time and effort and I want to thank StabilityAi for their desire to keep Ai as Open as possible.
I added some more documentation for the Gradio interface to the stable-audio-tools README, easiest way to get it going in the repo root is to run python3 ./run_gradio.py --pretrained-name stabilityai/stable-audio-open-1.0. That will pull down the weights from Hugging Face and start the demo automatically.
You'll want to run huggingface-cli login on the command line first to add your API key for Hugging Face
Thanks for your reply Fauno15.
Saganaki22 has created a one click installer, which has made things a lot easier for me.
https://github.com/Saganaki22/StableAudioWebUI/releases/tag/latest