Does anyone have a good guide to running an agent using a local LLM (ollama or llama.cpp)? I can get the basic text responses running but the generating code etc just falls over and crashes
Commit
1944aa0
i faced this same kind of issue when i tried it some time back
Managed to get this working last night using ollama and opencode. Was cool letting an agent do stuff but the model I was using was a bit poor. Things are a bit more manual like you have to enable tooling in the config and bump up the context but seemed fast enough for what I needed.
commit 1944aa0