Chroma Retriever¶

Use ChromaRetriever to evaluate ChromaDB collections.

Install¶

pip install "evret[chroma]"

Basic Usage¶

import chromadb
from evret.retrievers import ChromaRetriever

client = chromadb.PersistentClient(path="./chroma-data")

retriever = ChromaRetriever(
    collection_name="docs",
    client=client,
)

results = retriever.retrieve("ranking metrics for information retrieval", k=5)
for item in results:
    print(item.doc_id, item.score, item.metadata.get("document"))

By default, Chroma text query mode is used (query_texts).

Using a Query Encoder¶

def encode_query(query: str) -> list[float]:
    return embedding_model.embed_query(query)

retriever = ChromaRetriever(
    collection_name="docs",
    client=client,
    query_encoder=encode_query,
)

Notes¶

You can pass a Chroma collection directly as client
If query_encoder is set, Evret uses query_embeddings
Distance values are converted to scores with 1 / (1 + distance) by default