Billy reads from ~/.localai/config.toml on startup. If it doesn’t exist, defaults are used.
url = "http://localhost:11434"
model = "qwen2.5-coder:14b"
history_file = "~/.localai/history.db"
| Key | Default | Description |
|---|
type | ollama | Backend to use. Currently only ollama. |
url | http://localhost:11434 | URL of the Ollama server |
| Key | Default | Description |
|---|
model | qwen2.5-coder:14b | Default model to use |
temperature | 0.7 | Sampling temperature (0.0 = deterministic, 1.0 = creative) |
| Key | Default | Description |
|---|
history_file | ~/.localai/history.db | SQLite database for history & memories |
All config values can be overridden:
BILLY_MODEL=llama3.2 billy
BILLY_BACKEND_URL=http://192.168.1.10:11434 billy