[v3.0.0] response is interrupted if switiching to an other chat session

[SuperUserNameMan]

Bug Report

If i select a previous chat session to read it while the LLM is answering into an other chat session, then, its answer is interrupted instead of continuing in the background.

Steps to Reproduce

1. Aks a question to the model. (edit : use CPU device)
2. Switch to an previous chat session.
3. Come back

Expected behavior

Keep generating in the background ?
or resume ?

Your Environment

• GPT4All version: 3.0.0
• Operating System: Linux
  - Generation speed : 8 tok/s using CPU