feat: add llama 3.1 8b and 70b

[sozercan]

What this PR does / why we need it:

• add llama 3.1 8b and 70b
• deprecate llama3 and llama 2 models

Which issue(s) this PR fixes (optional, using fixes #<issue number>(, fixes #<issue_number>, ...) format, will close the issue(s) when the PR gets merged):
Fixes #

Special notes for your reviewer:

[reneleonhardt]

Thank you, it's a superb model!
I can't find new / updated images, does every user have to build llama-3.1-8b-instruct himself instead of the usual docker run instructions?
https://github.com/sozercan/aikit#cpu
https://github.com/sozercan/aikit/pkgs/container/llama3
https://github.com/sozercan?tab=packages

[sozercan]

@reneleonhardt they are available now: https://sozercan.github.io/aikit/docs/premade-models