Embedding Doesn't Work Without Manually Pulling Embedding Model

[anaseinea]

Describe the bug
Embedding Doesn't Work Without Manually Pulling Embedding Model

To Reproduce
start fresh, there will be no embedding model, open the embedding options and click "Embed workspace documents", check ollama logs and you will see a bunch of 404

Expected behavior
To pull the default embedding model or at least warn the user to do so.

API Provider
Ollama

Chat or Auto Complete?
Chat

Desktop (please complete the following information):

• OS: Windows

[rjmacarthy]

It's a good point, but we cannot force a user to add an embedding model or pull models for them when we don't know what provider they are using. Perhaps a warning if it doesn't work would be better?

[anaseinea]

Yes a warning is good.
There are other things not sure if I should open separate issues for:
1- The embedding model setting in the general settings is useless?
2- I couldn't see a button for toggling the embeddings (on/off button per message)
3- After clicking the embed button, chatting with the model didn't take into consideration the embeddings only when I add @general in the message and that would take a long time (possible re embedding everything?)