spb/uqo-research-db

1# Embedding model configuration2embeddings:3  default_model: "all-MiniLM-L6-v2"45  models:6    all-MiniLM-L6-v2:7      provider: sentence-transformers8      dimension: 3849      max_tokens: 25610      description: "Fast, lightweight model for general semantic similarity"1112    all-mpnet-base-v2:13      provider: sentence-transformers14      dimension: 76815      max_tokens: 38416      description: "Higher quality embeddings, slower inference"1718    text-embedding-3-small:19      provider: openai20      dimension: 153621      max_tokens: 819122      description: "OpenAI small embedding model"2324    text-embedding-3-large:25      provider: openai26      dimension: 307227      max_tokens: 819128      description: "OpenAI large embedding model"2930  processing:31    batch_size: 25632    max_workers: 233    device: "mps"  # cpu, cuda, mps34    normalize: true35    show_progress: true3637  storage:38    base_dir: data/embeddings39    format: parquet40    compression: zstd41