Palm1r
Contact to local ollama server lost after some hours of Workstation in sleep mode
Describe the bug QodeAssist could does not get answers from local Ollama server after workstation was woken up from sleep.
To Reproduce I had an active Qt Creator session with an active chat via QodeAssist configured with a local Ollama server. The GPU had the model still loaded, when the workstation was sent into sleep. The next day just after the workstation was woke up, the GPU still had the model in RAM, but shortly after is was removed. Then I started a new chat (I pressed "clear"). There were no answers send by ollama server, the GPU was not filled with the model. When I restarted QtCreator, everything worked again.
Expected behavior Should still work without restart of QtCreator. After restart, the context has to be created new.
Log Additional context Sorry, log was not activated.
