config.toml.example

# ModelicaRag configuration example
#
# Copy this file to config.toml and edit the paths for your machine.
# config.toml is listed in .gitignore and will not be committed.

# ---------------------------------------------------------------------------
# Option A: GitHub Models (recommended — no local server required)
#
# Free for all GitHub users: 150 embedding requests/day.
# Requires a GitHub personal access token (no special scopes needed).
# Set GITHUB_TOKEN in your shell — do not put the token in this file.
#
#   export GITHUB_TOKEN=github_pat_...
#
# [embeddings]
# backend    = "github_models"
# model      = "text-embedding-3-small"   # or "text-embedding-3-large"
# batch_size = 32
#
# ---------------------------------------------------------------------------
# Option B: Ollama (local, no API key needed)
#
# Install Ollama (https://ollama.com), then pull the model:
#   ollama pull nomic-embed-text
#
# ---------------------------------------------------------------------------
# Option C: llama-server (local, GGUF model file required)
#
# Build llama.cpp and download a GGUF embedding model.
# Julia starts the server automatically when needed.
# ---------------------------------------------------------------------------

[embeddings]
backend    = "ollama"                  # "github_models", "ollama", or "llama"
url        = "http://localhost:11434"  # Ollama URL  (use "http://localhost:8080" for llama backend)
model      = "nomic-embed-text"        # model name; for github_models use "text-embedding-3-small"
batch_size = 32                        # number of chunks sent per embedding request

# ---------------------------------------------------------------------------
# llama-server settings
#
# Only used when backend = "llama".
# Julia starts the server automatically and stops it on exit.
# ---------------------------------------------------------------------------

[server]
llama_server = "~/llama.cpp/build/bin/llama-server"                # path to the llama-server binary
model_path   = "~/llama.cpp/models/Qwen3-Embedding-8B-Q8_0.gguf"  # path to GGUF model file

# ---------------------------------------------------------------------------
# Index database
#
# SQLite file that stores all parsed chunks and their embedding vectors.
# The directory is created automatically on first run.
# ---------------------------------------------------------------------------

[store]
path = "data/index.db"   # can be absolute or relative to the project root

# ---------------------------------------------------------------------------
# Modelica library to index
#
# root       — top-level directory; all .mo files underneath are indexed.
#              Point this at your Modelica Standard Library, e.g.:
#                /usr/share/openmodelica/libraries/Modelica 4.0.0
#              The indexer is incremental: only changed files are re-embedded
#              on subsequent runs.
#
# extensions — file extensions to include.
# ---------------------------------------------------------------------------

[codebase]
root       = "/path/to/Modelica/library"   # e.g. /usr/share/openmodelica/libraries/Modelica 4.0.0
extensions = [".mo"]