Spaces:

HuggingFaceH4

/

starchat-playground

like 795

Runtime error

same question.

In the model card page, click on 'Deploy' and then select 'Inference API.' This will provide you with an API_TOKEN and API_URL. To obtain the HF_TOKEN, go to your account settings and navigate to 'Access Tokens'.

@jsr90 you are right, the information can be found there, but as far as I understand you can't use the model via the Inference API for free:

The model HuggingFaceH4/starchat-alpha is too large to be loaded automatically (31GB > 10GB). For commercial use please use PRO spaces (https://huggingface.co/spaces) or Inference Endpoints (https://huggingface.co/inference-endpoints)

I don't want to run it commercially, I just want to build a demo. But it seems you actually have to run your own Inference Endpoint or upgrade your space to have a proper GPU.

I hope I'm missing an option 🙏