New Llama 2 based Hermes and Wizard models

[cjcarroll012]

Feature request

Can you please update the GPT4ALL chat JSON file to support the new Hermes and Wizard models built on LLAMA 2?

Motivation

Using GPT4ALL

Your contribution

Awareness. Really love gpt4all.

[clearsitedesigns]

This would be pretty sweet:)

[cosmic-snow]

For 7B and 13B Llama 2 models these just need a proper JSON entry in models.json. Bigger models need architecture support, though.

Any takers? All you need to do is side load one of these and make sure it works, then add an appropriate JSON entry.

[Sidx-sys]

I was adding these models in the models.json for testing. To clarify, @cjcarroll012, by Wizard Models, do you mean this one https://huggingface.co/WizardLM/WizardLM-13B-V1.2?

[activescott]

I was looking at this one and found the following models already integrated into the models json:

• NousResearch/Nous-Hermes-Llama2-13b: done
• WizardLM/WizardLM-13B-V1.2: done

Ones still missing are:

• NousResearch/Nous-Hermes-llama-2-7b in GGUF from TheBloke at https://huggingface.co/TheBloke/Nous-Hermes-Llama-2-7B-GGUF - PR at #1753