.codegraph.toml.example•3.69 kB
# CodeGraph Configuration File
# Copy this to .codegraph.toml or ~/.codegraph/config.toml and customize
# ============================================================================
# Embedding Configuration
# ============================================================================
[embedding]
# Provider: "auto", "onnx", "ollama", "openai", or "lmstudio"
# "auto" will detect available models automatically
# "lmstudio" recommended for MLX + Flash Attention 2 (macOS)
provider = "lmstudio"
# Model path or identifier
# For ONNX: Absolute path to model directory (auto-detected from HuggingFace cache)
# For Ollama: Model name (e.g., "all-minilm:latest")
# For LM Studio: Model name (e.g., "jinaai/jina-embeddings-v3")
# For OpenAI: Model name (e.g., "text-embedding-3-small")
# Recommended: jinaai/jina-embeddings-v3 (1536-dim, optimized for code)
model = "jinaai/jina-embeddings-v3"
# LM Studio URL (default port 1234)
lmstudio_url = "http://localhost:1234"
# Ollama URL (only used if provider is "ollama")
ollama_url = "http://localhost:11434"
# OpenAI API key (only used if provider is "openai")
# Can also be set via OPENAI_API_KEY environment variable
# openai_api_key = "sk-..."
# Embedding dimension (1536 for jina-code-embeddings-1.5b, 384 for all-MiniLM)
dimension = 1536
# Batch size for embedding generation (GPU optimization)
batch_size = 64
# ============================================================================
# LLM Configuration (for insights generation)
# ============================================================================
[llm]
# Enable LLM insights (false = context-only mode for agents like Claude/GPT-4)
# Set to false for maximum speed if using an external agent
enabled = false
# LLM provider: "ollama" or "lmstudio"
# "lmstudio" recommended for MLX + Flash Attention 2 (macOS)
provider = "lmstudio"
# LLM model identifier
# For LM Studio: lmstudio-community/DeepSeek-Coder-V2-Lite-Instruct-GGUF/DeepSeek-Coder-V2-Lite-Instruct-Q4_K_M.gguf
# For Ollama: Model name (e.g., "qwen2.5-coder:14b", "codellama:13b")
# Recommended: DeepSeek Coder v2 Lite Instruct Q4_K_M (superior performance)
model = "lmstudio-community/DeepSeek-Coder-V2-Lite-Instruct-GGUF"
# LM Studio URL (default port 1234)
lmstudio_url = "http://localhost:1234"
# Ollama URL
ollama_url = "http://localhost:11434"
# Context window size (tokens)
# DeepSeek Coder v2 Lite: 32768 tokens
context_window = 32000
# Temperature for generation (0.0 = deterministic, 1.0 = creative)
temperature = 0.1
# Insights mode: "context-only", "balanced", or "deep"
# - context-only: Return context only (fastest, for agents)
# - balanced: Process top 10 files with LLM (good speed/quality)
# - deep: Process all reranked files (comprehensive)
insights_mode = "context-only"
# ============================================================================
# Performance Configuration
# ============================================================================
[performance]
# Number of worker threads (defaults to CPU count)
num_threads = 0 # 0 = auto-detect
# Cache size in MB
cache_size_mb = 512
# Enable GPU acceleration (requires CUDA/Metal support)
enable_gpu = false
# Maximum concurrent requests for embedding/LLM
max_concurrent_requests = 4
# ============================================================================
# Logging Configuration
# ============================================================================
[logging]
# Log level: "trace", "debug", "info", "warn", "error"
# Use "warn" during indexing for clean TUI output (recommended)
# Use "info" for development/debugging
level = "warn"
# Log format: "pretty", "json", "compact"
format = "pretty"