Is this similar to fastllm?

fastllm targets the GPU, while colibri uses CPU inference only

	▲	aliljet 3 hours ago \| parent [-]
		I'd be curious about an.option that would allow glm use with a low end GPU like a 2080 ti...