Wiki not specific enough about installing offline models

[DeutscheGabanna]

The wiki says that I merely need to add the model by name from Huggingace in http://localhost:42110/server/admin/database/chatmodel/. Looking through the server logs, Khoj does not download the model after adding it to the model list. When I try to chat, it understandably complains:

No file found in deepseek-ai/deepseek-llm-7b-chat that match *Q4_K_M.gguf

So, am I forced to use ollama for downloading the model? If not, why does the manual not tell me where to put the model or how it is downloaded?
P.S I used Docker (actually, podman, but it's the same thing basically) to set up Khoj.

[DeutscheGabanna]

This is the only snippet that I found vaguely referring to downloading models, but it uses gpt4all which seems like not my exact situation:

I haven't tested it but you maybe able to use any model in gguf format. You'll need to manually download a chat model in gguf format into the ~/.cache/gpt4all/ folder
#574 (reply in thread)