backend/scripts/test_embedder.py

"""Smoke test for Modal-backed MiniCPM-Embedding."""

import sys
from pathlib import Path

sys.path.insert(0, str(Path(__file__).parent.parent))

from models.embedder import MiniCPMEmbedder


def main():
    print("Calling Modal Embedder...")
    embedder = MiniCPMEmbedder()
    dim = embedder.get_embedding_dim()
    assert dim == 2304, f"Expected 2304, got {dim}"

    doc_vec = embedder.embed_documents(["Revenue increased 12% year-over-year."])
    query_vec = embedder.embed_query("What was the revenue growth?")
    print(f"Document vector length: {len(doc_vec[0])}")
    print(f"Query vector length: {len(query_vec)}")
    assert len(doc_vec[0]) == 2304
    assert len(query_vec) == 2304
    print("Modal embedder test passed.")


if __name__ == "__main__":
    main()