> This sounds like what they call "Bamba-9B" is actually an 18B model quantised ...

		cubefox 7 months ago \| parent \| context \| favorite \| on: Bamba: An open-source LLM that crosses a transform... > This sounds like what they call "Bamba-9B" is actually an 18B model quantised to 8 bits. No it doesn't? The fact that it is 18 GB with 16 bit per parameter before quantization means that it is a 9B parameter model.

Ah thanks, I see where I got confused now.