mirror of
https://github.com/meta-llama/llama-stack.git
synced 2025-12-06 10:37:22 +00:00
feat(qdrant): implement hybrid and keyword search support (#4006)
# What does this PR do? - Part of #3009 - Implement hybrid search using Qdrant's native query filtering - Add keyword search support - Update test suites to include qdrant for keyword and hybrid modes <!-- If resolving an issue, uncomment and update the line below --> <!-- Closes #[issue-number] --> ## Test Plan <!-- Describe the tests you ran to verify your changes with result summaries. *Provide clear instructions so the plan can be easily re-executed.* --> ``` pytest -sv tests/unit/providers/vector_io/ ....... ============================================================================================== slowest 10 durations =============================================================================================== 0.20s call tests/unit/providers/vector_io/test_vector_io_openai_vector_stores.py::test_max_concurrent_files_per_batch[qdrant] 0.20s call tests/unit/providers/vector_io/test_vector_io_openai_vector_stores.py::test_max_concurrent_files_per_batch[pgvector] 0.20s call tests/unit/providers/vector_io/test_vector_io_openai_vector_stores.py::test_max_concurrent_files_per_batch[sqlite_vec] 0.20s call tests/unit/providers/vector_io/test_vector_io_openai_vector_stores.py::test_max_concurrent_files_per_batch[faiss] 0.06s setup tests/unit/providers/vector_io/test_vector_io_openai_vector_stores.py::test_insert_chunks_with_missing_document_id[pgvector] 0.04s call tests/unit/providers/vector_io/test_sqlite_vec.py::test_query_chunks_hybrid_tie_breaking 0.04s call tests/unit/providers/vector_io/test_sqlite_vec.py::test_query_chunks_hybrid_weighted_reranker_parametrization 0.03s call tests/unit/providers/vector_io/test_sqlite_vec.py::test_query_chunks_hybrid_score_selection 0.03s call tests/unit/providers/vector_io/test_sqlite_vec.py::test_query_chunks_hybrid_edge_cases 0.03s setup tests/unit/providers/vector_io/test_faiss.py::test_faiss_query_vector_returns_infinity_when_query_and_embedding_are_identical ======================================================================================== 180 passed, 47 warnings in 2.78s ========================================================================================= ``` Signed-off-by: Varsha Prasad Narsing <varshaprasad96@gmail.com> Co-authored-by: Francisco Javier Arceo <arceofrancisco@gmail.com>
This commit is contained in:
Varsha
GitHub
parent
5873a316db
commit
743683ba26
3 changed files with 226 additions and 3 deletions
|
|
@ -129,7 +129,63 @@ class QdrantIndex(EmbeddingIndex):
|
|||
return QueryChunksResponse(chunks=chunks, scores=scores)
|
||||
|
||||
async def query_keyword(self, query_string: str, k: int, score_threshold: float) -> QueryChunksResponse:
|
||||
raise NotImplementedError("Keyword search is not supported in Qdrant")
|
||||
"""
|
||||
Performs keyword-based search using Qdrant's MatchText filter.
|
||||
|
||||
Uses Qdrant's query_filter with MatchText to search for chunks containing
|
||||
the specified text query string in the chunk content.
|
||||
|
||||
Args:
|
||||
query_string: The text query for keyword search
|
||||
k: Number of results to return
|
||||
score_threshold: Minimum similarity score threshold
|
||||
|
||||
Returns:
|
||||
QueryChunksResponse with chunks and scores matching the keyword query
|
||||
"""
|
||||
try:
|
||||
results = (
|
||||
await self.client.query_points(
|
||||
collection_name=self.collection_name,
|
||||
query_filter=models.Filter(
|
||||
must=[
|
||||
models.FieldCondition(
|
||||
key="chunk_content.content", match=models.MatchText(text=query_string)
|
||||
)
|
||||
]
|
||||
),
|
||||
limit=k,
|
||||
with_payload=True,
|
||||
with_vectors=False,
|
||||
score_threshold=score_threshold,
|
||||
)
|
||||
).points
|
||||
except Exception as e:
|
||||
log.error(f"Error querying keyword search in Qdrant collection {self.collection_name}: {e}")
|
||||
raise
|
||||
|
||||
chunks, scores = [], []
|
||||
for point in results:
|
||||
if not isinstance(point, models.ScoredPoint):
|
||||
raise RuntimeError(f"Expected ScoredPoint from Qdrant query, got {type(point).__name__}")
|
||||
if point.payload is None:
|
||||
raise RuntimeError("Qdrant query returned point with no payload")
|
||||
|
||||
try:
|
||||
chunk = Chunk(**point.payload["chunk_content"])
|
||||
except Exception:
|
||||
chunk_id = point.payload.get(CHUNK_ID_KEY, "unknown") if point.payload else "unknown"
|
||||
point_id = getattr(point, "id", "unknown")
|
||||
log.exception(
|
||||
f"Failed to parse chunk in collection {self.collection_name}: "
|
||||
f"chunk_id={chunk_id}, point_id={point_id}"
|
||||
)
|
||||
continue
|
||||
|
||||
chunks.append(chunk)
|
||||
scores.append(point.score)
|
||||
|
||||
return QueryChunksResponse(chunks=chunks, scores=scores)
|
||||
|
||||
async def query_hybrid(
|
||||
self,
|
||||
|
|
@ -140,7 +196,66 @@ class QdrantIndex(EmbeddingIndex):
|
|||
reranker_type: str,
|
||||
reranker_params: dict[str, Any] | None = None,
|
||||
) -> QueryChunksResponse:
|
||||
raise NotImplementedError("Hybrid search is not supported in Qdrant")
|
||||
"""
|
||||
Hybrid search combining vector similarity and keyword filtering in a single query.
|
||||
|
||||
Uses Qdrant's native capability to combine a vector query with a query_filter,
|
||||
allowing vector similarity search to be filtered by keyword matches in one call.
|
||||
|
||||
Args:
|
||||
embedding: The query embedding vector
|
||||
query_string: The text query for keyword filtering
|
||||
k: Number of results to return
|
||||
score_threshold: Minimum similarity score threshold
|
||||
reranker_type: Not used with this approach, but kept for API compatibility
|
||||
reranker_params: Not used with this approach, but kept for API compatibility
|
||||
|
||||
Returns:
|
||||
QueryChunksResponse with filtered vector search results
|
||||
"""
|
||||
try:
|
||||
results = (
|
||||
await self.client.query_points(
|
||||
collection_name=self.collection_name,
|
||||
query=embedding.tolist(),
|
||||
query_filter=models.Filter(
|
||||
must=[
|
||||
models.FieldCondition(
|
||||
key="chunk_content.content", match=models.MatchText(text=query_string)
|
||||
)
|
||||
]
|
||||
),
|
||||
limit=k,
|
||||
with_payload=True,
|
||||
score_threshold=score_threshold,
|
||||
)
|
||||
).points
|
||||
except Exception as e:
|
||||
log.error(f"Error querying hybrid search in Qdrant collection {self.collection_name}: {e}")
|
||||
raise
|
||||
|
||||
chunks, scores = [], []
|
||||
for point in results:
|
||||
if not isinstance(point, models.ScoredPoint):
|
||||
raise RuntimeError(f"Expected ScoredPoint from Qdrant query, got {type(point).__name__}")
|
||||
if point.payload is None:
|
||||
raise RuntimeError("Qdrant query returned point with no payload")
|
||||
|
||||
try:
|
||||
chunk = Chunk(**point.payload["chunk_content"])
|
||||
except Exception:
|
||||
chunk_id = point.payload.get(CHUNK_ID_KEY, "unknown") if point.payload else "unknown"
|
||||
point_id = getattr(point, "id", "unknown")
|
||||
log.exception(
|
||||
f"Failed to parse chunk in collection {self.collection_name}: "
|
||||
f"chunk_id={chunk_id}, point_id={point_id}"
|
||||
)
|
||||
continue
|
||||
|
||||
chunks.append(chunk)
|
||||
scores.append(point.score)
|
||||
|
||||
return QueryChunksResponse(chunks=chunks, scores=scores)
|
||||
|
||||
async def delete(self):
|
||||
await self.client.delete_collection(collection_name=self.collection_name)
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue