OpenAssistant

/

oasst-sft-4-pythia-12b-epoch-3.5

like 355

@DeeJaye Hey, have you found an answer to this question?

@samyar03 No, I haven't found an answer to it yet. Though, I'm thinking it doesn't work that way, but more or less like a model already trained with a large amount of data, it can be downloaded locally (if necessary), and used for my projects. But i'm open to hearing your feedback on it.

@DeeJaye Hm. Because I was hoping to use this LLM as a fallback for a chatbot that I'm developing, and I'd think having an API would make things easier, but it looks like I might have to try and create my own. Do you have any other thoughts on how to approach this?

@samyar03 It is possible to use the HuggingFace inference API to serve the model (payed).

How can I get OpenAssistant API to use for my project?

I'm toying around with it, created a demo repo, as starting point to learn, also to share:

https://github.com/morning-sunn/open-assistant-demo