Skip to content

Configuration

Kritarth-Dandapat edited this page Apr 11, 2026 · 1 revision

Configuration

super-ollama config file

Resolved in order:

  1. $XDG_CONFIG_HOME/super-ollama/config.toml when XDG_CONFIG_HOME is set
  2. Otherwise ~/.super-ollama/config.toml

Fields (TOML):

Field Purpose
default_model Model name when --model is omitted (built-in default: gemma3:1b)

Logging

Behavior Condition
WARN and above on stderr Default (quiet inference logs)
Same levels as upstream envconfig.LogLevel() OLLAMA_DEBUG set to any non-empty value

Model files

Models live in the usual Ollama layout on disk. Use the ollama binary from this repo (or another install) for ollama pull <model>.

Environment variables

Inference and hardware behavior follow Ollama env vars (OLLAMA_*) where the shared engine reads them. See upstream documentation for GPU, context, and debug options.

Clone this wiki locally