Readme: Installation instructions

[natrius]

"Ich bin grad zpatschert dafür..."

I downloaded GPT4ll recently and would like to try Zicklein with it, but unfortunately i'm a bit stumped on how to get it into GPT4all so i can try it.
It might be worth expanding the readme on that part?

[avocardio]

I have not heard of GPT4all to be honest, but i'll look into it, sounds interesting!

[natrius]

Apology, here is the link https://gpt4all.io/index.html to the site.

[phil2sat]

Same for me, relatively new to all the stuff.

To explain gpt4all uses llama ggml q4_0 models in a single bin file.
Like: wizardlm-13b-v1.1-superhot-8k.ggmlv3.q4_0.bin
So we need a way to get the three models into one, thats what my zpatchert Hirn (smached brain) tells me.

From what i understood it's,
the base model: decapoda-research/llama-7b-hf
the stanford mod: tloen/alpaca-lora-7b
and the German mod: avocardio/alpaca-lora-7b-german-base-52k

but since im no developer i can't write a python-converter.py which does this for us and also i guess my hardware isn't able to fight it.

If it would, i guess running the generator.py would the correct way to get it.

Helping hands needed