diff --git a/index/main.py b/index/main.py index df4575a..5ae3343 100644 --- a/index/main.py +++ b/index/main.py @@ -74,7 +74,7 @@ class SparseVector(BaseModel): values: list[float] -CHUNK_SIZE = 256 +CHUNK_SIZE = 384 OVERLAP_SIZE = 128 SPARSE_MODEL_NAME = "Qdrant/bm25" FASTEMBED_CACHE_PATH = "/models/fastembed" diff --git a/search/main.py b/search/main.py index a55a466..7f9facb 100644 --- a/search/main.py +++ b/search/main.py @@ -161,10 +161,10 @@ async def lifespan(app: FastAPI): app = FastAPI(title="Search Service", version="0.1.0", lifespan=lifespan) -DENSE_PREFETCH_K = 80 -SPARSE_PREFETCH_K = 200 -RETRIEVE_K = 150 -RERANK_LIMIT = 10 +DENSE_PREFETCH_K = 50 +SPARSE_PREFETCH_K = 150 +RETRIEVE_K = 100 +RERANK_LIMIT = 15 async def embed_dense(client: httpx.AsyncClient, text: str) -> list[float]: