Merge branch 'fea-sherlock' into devin_fea_sherlock_persistent_rag

morpheus/llm/nodes/retriever_node.py

@@ -23,6 +23,18 @@
 
 
 class RetrieverNode(LLMNodeBase):
+    """
+    Node for retrieving data from a vector database based on embeddings.
+
+    Parameters
+    ----------
+    embedding : typing.Callable[[list[str]], typing.Coroutine[typing.Any, typing.Any, list[list[float]]]] | None
+        Callable function for generating vector embeddings. Default is None.
+    service : VectorDBResourceService
+        Vector database resource service for executing similarity searches.
+    similarity_search_kwargs : dict
+        Additional keyword arguments for the similarity search.
+    """
 
     def __init__(
         self,
@@ -38,12 +50,33 @@ def __init__(
         self._similarity_search_kwargs = similarity_search_kwargs
 
     def get_input_names(self) -> list[str]:
+        """
+        Get the input names for the RetrieverNode.
+
+        Returns
+        -------
+        list[str]
+            List of input names for the RetrieverNode.
+        """
         if (self._embedding is None):
             return ["embedding"]
 
         return ["query"]
 
     async def execute(self, context: LLMContext):
+        """
+        Execute the retrieval process based on the provided context.
+
+        Parameters
+        ----------
+        context : LLMContext
+            Context object containing necessary information for execution.
+
+        Returns
+        -------
+        LLMContext
+            Updated context object after the execution.
+        """
 
         if (self._embedding is not None):
             # Get the keys from the task

tests/llm/nodes/test_llm_retriever_node.py

@@ -0,0 +1,49 @@
+# SPDX-FileCopyrightText: Copyright (c) 2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+import typing
+from unittest import mock
+
+import pytest
+
+from _utils.llm import execute_node
+from morpheus.llm import LLMNodeBase
+from morpheus.llm.nodes.retriever_node import RetrieverNode
+
+
+@pytest.mark.parametrize("embedding", [None, mock.AsyncMock()])
+def test_constructor(embedding: typing.Callable | None):
+    mock_vdb_service = mock.MagicMock()
+    node = RetrieverNode(embedding=embedding, service=mock_vdb_service)
+    assert isinstance(node, LLMNodeBase)
+
+
+@pytest.mark.parametrize("embedding, expected", [(None, ["embedding"]), (mock.AsyncMock(), ["query"])])
+def test_get_input_names(embedding: typing.Callable | None, expected: list[str]):
+    mock_vdb_service = mock.MagicMock()
+    node = RetrieverNode(embedding=embedding, service=mock_vdb_service)
+    assert node.get_input_names() == expected
+
+
+@pytest.mark.parametrize("embedding", [None, mock.AsyncMock(return_value=[[1.2, 2.3, 3.4], [4.5, 5.6, 6.7]])])
+def test_execute(embedding: mock.AsyncMock | None):
+    mock_vdb_service = mock.MagicMock()
+    mock_vdb_service.similarity_search = mock.AsyncMock(return_value=[[1, 2, 3], [4, 5, 6]])
+
+    node = RetrieverNode(embedding=embedding, service=mock_vdb_service)
+
+    expected_output = [[1, 2, 3], [4, 5, 6]]
+
+    assert execute_node(node, query=["query"]) == expected_output

tests/llm/nodes/test_llm_retriever_node_pipe.py

@@ -0,0 +1,126 @@
+# SPDX-FileCopyrightText: Copyright (c) 2023, NVIDIA CORPORATION & AFFILIATES. All rights reserved.
+# SPDX-License-Identifier: Apache-2.0
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from unittest import mock
+
+import pytest
+
+import cudf
+
+from morpheus.config import Config
+from morpheus.llm import LLMEngine
+from morpheus.llm.nodes.extracter_node import ExtracterNode
+from morpheus.llm.nodes.retriever_node import RetrieverNode
+from morpheus.llm.task_handlers.simple_task_handler import SimpleTaskHandler
+from morpheus.messages import ControlMessage
+from morpheus.pipeline.linear_pipeline import LinearPipeline
+from morpheus.service.vdb.milvus_vector_db_service import MilvusVectorDBResourceService
+from morpheus.service.vdb.milvus_vector_db_service import MilvusVectorDBService
+from morpheus.stages.input.in_memory_source_stage import InMemorySourceStage
+from morpheus.stages.llm.llm_engine_stage import LLMEngineStage
+from morpheus.stages.output.in_memory_sink_stage import InMemorySinkStage
+from morpheus.stages.preprocess.deserialize_stage import DeserializeStage
+
+
+@pytest.fixture(scope="module", name="milvus_service")
+def milvus_service_fixture(milvus_server_uri: str):
+    service = MilvusVectorDBService(uri=milvus_server_uri)
+    yield service
+
+
+def _build_engine(vdb_service, **similarity_search_kwargs) -> LLMEngine:
+    mock_embedding = mock.AsyncMock(return_value=[[1.2, 2.3, 3.4], [4.5, 5.6, 6.7]])
+    engine = LLMEngine()
+    engine.add_node("extracter", node=ExtracterNode())
+    engine.add_node("retriever",
+                    inputs=["/extracter"],
+                    node=RetrieverNode(service=vdb_service, embedding=mock_embedding, **similarity_search_kwargs))
+    engine.add_task_handler(inputs=["/retriever"], handler=SimpleTaskHandler())
+
+    return engine
+
+
+@pytest.mark.use_python
+def test_pipeline(config: Config):
+    expected_output = [[1, 2, 3], [4, 5, 6]]
+
+    values = {'prompt': ["prompt1", "prompt2"]}
+    input_df = cudf.DataFrame(values)
+    expected_df = input_df.copy(deep=True)
+    expected_df["response"] = expected_output
+
+    task_payload = {"task_type": "llm_engine", "task_dict": {"input_keys": sorted(values.keys())}}
+
+    mock_vdb_service = mock.MagicMock()
+    mock_vdb_service.similarity_search = mock.AsyncMock(return_value=[[1, 2, 3], [4, 5, 6]])
+
+    pipe = LinearPipeline(config)
+    pipe.set_source(InMemorySourceStage(config, dataframes=[input_df]))
+    pipe.add_stage(
+        DeserializeStage(config, message_type=ControlMessage, task_type="llm_engine", task_payload=task_payload))
+    pipe.add_stage(LLMEngineStage(config, engine=_build_engine(vdb_service=mock_vdb_service)))
+    sink = pipe.add_stage(InMemorySinkStage(config))
+
+    pipe.run()
+
+    message = sink.get_messages()[0]
+    assert isinstance(message, ControlMessage)
+    actual_df = message.payload().df
+
+    # Using equals, as CompareDataFrameStage fails to compare.
+    assert actual_df.to_pandas().equals(expected_df.to_pandas())
+
+
+@pytest.mark.use_python
+@pytest.mark.milvus
+def test_pipeline_with_milvus(config: Config,
+                              milvus_service: MilvusVectorDBService,
+                              idx_part_collection_config: dict,
+                              milvus_data: list[dict]):
+
+    collection_name = "test_retriever_node_collection"
+    # Make sure to drop any existing collection from previous runs.
+    milvus_service.drop(collection_name)
+    # Create a collection.
+    milvus_service.create(collection_name, **idx_part_collection_config)
+    resource_service: MilvusVectorDBResourceService = milvus_service.load_resource(name=collection_name)
+    # Insert data into collection
+    resource_service.insert(milvus_data)
+
+    # Define a similarity_search filter.
+    expr = "age==26 or age==27"
+
+    values = {'prompt': ["prompt1", "prompt2"]}
+    input_df = cudf.DataFrame(values)
+    expected_df = input_df.copy(deep=True)
+    expected_df["response"] = [[{'0': 27, '1': 2}, {'0': 26, '1': 1}], [{'0': 27, '1': 2}, {'0': 26, '1': 1}]]
+
+    task_payload = {"task_type": "llm_engine", "task_dict": {"input_keys": sorted(values.keys())}}
+
+    pipe = LinearPipeline(config)
+    pipe.set_source(InMemorySourceStage(config, dataframes=[input_df]))
+    pipe.add_stage(
+        DeserializeStage(config, message_type=ControlMessage, task_type="llm_engine", task_payload=task_payload))
+    pipe.add_stage(LLMEngineStage(config, engine=_build_engine(vdb_service=resource_service, expr=expr)))
+    sink = pipe.add_stage(InMemorySinkStage(config))
+
+    pipe.run()
+
+    message = sink.get_messages()[0]
+    assert isinstance(message, ControlMessage)
+    actual_df = message.payload().df
+
+    # Using equals, as CompareDataFrameStage fails to compare.
+    assert actual_df.to_pandas().equals(expected_df.to_pandas())