Error when running Devika with local models.

[LeonMatch]

Hello,
I am trying to run Devika with local models that I pulled from Ollama.
I am getting error:
ERROR: Inference failed. took too long. Model: OLLAMA, Model ID: codellama:7b-code.

Update:
After re-installing devika, I and got another error:
root: ERROR : llama runner process has terminated: exit status 0xc0000142
Then restarted virtual environment (created with anaconda), and restarted devika.
Now back to the previous error:
root: ERROR : Inference failed. took too long. Model: OLLAMA, Model ID: codegemma:2b

What can be wrong?