Implementing only CPU Usage

[AnimeshDeb]

Hello, I am using the node js binding for a llama 2 chat model I found on the hugging face website, specifically from TheBloke. I have implemented a chat bot using the llama-2-7b-chat.Q5_K_S.gguf model from https://huggingface.co/TheBloke/Llama-2-7B-Chat-GGUF . I want to ensure that only CPU is used when the AI model is running. Any advice on how to achieve this as I am relatively new to using AI models and optimizing their performance.

[giladgd]

@AnimeshDeb Using the version 3 beta you can sepcify gpu: false when you call getLlama({ ... }) to disable the GPU support, but I do not see any advantage to doing that if your goal is to optimize for inference performance