[keriati1]

We actually run already in-house ollama server prototype for coding assistance with deepseek coder and it is pretty good. Now if we would get a model for this, that is on chatgpt 4 level, I would be super happy.

[eurekin]

Did you finetune a model?

[keriati1]

No, we went with RAG pipeline approach as we assume things change too fast.

[eurekin]

Thanks! Any details how you chunk and find the relevant code?

Or how you deal with context length? I.e. do you send anything other than the current file? How is the prompt constructed?