I would like to use AI to generate articles to feed my link building software.
I know I may be able to use a Open AI key right away, but for cost consideration, do you have any guide on how to setup a free AI model on a VPS and let SEO Content machine communicate with it and generate articles and images?
I saw you posted a video a few weeks ago, but with AI progressing every week, what is the current best setup procedure for text generation and image generation that works with SEO Content Machine?
I am thinking about llama 3.2 as I heard it is quite powerful, but at this stage I am confused which version to use etc
I would start with groq ai, as its good and free right now, but with some call limitations.
If you want to go down the VPS route, know that you will need to spend the big bucks as large LLAMA models need 12GB+ of ram etc.
You also want a GPU accelerated VPS otherwise expect slow response times. Just CPU only you can end up with responses times going to 30-60s for one answer to your prompt.
Right now the big LLAMA models need a lot of hardware juice, but know that as time rolls around models are getting smaller and more efficient and eventually we might be even able to run them on a browser locally with GPT3.5 like answers.
For VPS Text → Run LMStudio, load model and click start.
Which model? Try the biggest ones first that will load on your VPS. Then go smaller if you don’t have enough ram or it takes to long for the AI to return a response.
Bigger = More parameter ie LLAMA 7B, 13B, 70B
For Images → Use comfyui, download stable diffusion model, set up an api
The setup for local AI images is much more complicated, I am currently playing around with them to see how to setup an API server for SCM to communicate with.
I will release a walk through with instructions to set it up with SCM.
