bug: `cortex run` flakiness

[freelerobot]

Problem Statement

1. cortex run mistral

Model loaded!
Inorder to exit, type `exit()`

2. Chat a bit and quit interactive shell
3. Run it again
4. sudo cortex run mistral

Starting server ...
20240919 15:54:32.723277 UTC 33913 INFO  Host: 127.0.0.1 Port: 3928
 - main.cc:32

Server started

Model loaded!
Inorder to exit, type `exit()`

5. Model loads & chats just fine

Expected

Expected: the same stdout experience as the first time I invoked cortex run
But I got: "Starting Server..." for all subsequent invocations

Questions

What could be causing this inconsistency in the run step?

OS: Mac

[vansangpfiev]

cortex run is the command chain which includes:

• cortex engines install
• cortex pull
• cortex start
• cortex models start
• cortex chat
  In this situation, the cortex models start doesn't work correctly. The CLI doesn't check if the model was loaded before starting it, and the engine doesn't ignore it when starting it again.
  The issue will be fixed on the CLI side first, and then on the engines side.