>your primary goal is getting off of OpenAI as soon as you possibly can I have calculated that LLAMA running model on AWS/CLOUD, even vast.ai with 3090 is much

>your primary goal is getting off of OpenAI as soon as you possibly can I have calculated that LLAMA running model on AWS/CLOUD, even vast.ai with 3090 is much more expensive then openai. Even with collocation 3090 as they charge absurd amount of money for power (hardware free) Only case that this is cheaper few times, is if I would host it in my house as I have fiber, and not that expensive power.