Skip to content

armgit / dspy

Public Imported from GitHub
0

DSPy: The framework for programming—not prompting—language models

Python

feature: add PgVectorRM for postgresql database retrieval

Brian Muhiaebe00686b7cf
diff --git a/dspy/retrieve/pgvector_rm.py b/dspy/retrieve/pgvector_rm.py
new file mode 100644
index 000000000..20b805dd7
--- /dev/null
+++ b/dspy/retrieve/pgvector_rm.py
@@ -0,0 +1,112 @@
+import dspy
+import openai
+import psycopg2
+from psycopg2 import sql
+from typing import List, Union, Optional
+from pgvector.psycopg2 import register_vector
+
+
+class PgVectorRM(dspy.Retrieve):
+ """
+ Implements a retriever that (as the name suggests) uses pgvector to retrieve passages,
+ using a raw SQL query and a postgresql connection managed by psycopg2.
+
+ It needs to register the pgvector extension with the psycopg2 connection
+
+ Returns a list of dspy.Example objects
+
+ Args:
+ db_url (str): A PostgreSQL database URL in psycopg2's DSN format
+ pg_table_name (Optional[str]): name of the table containing passages
+ openai_client (openai.OpenAI): OpenAI client to use for computing query embeddings
+ k (Optional[int]): Default number of top passages to retrieve. Defaults to 20
+ embedding_field (str = "embedding"): Field containing passage embeddings. Defaults to "embedding"
+ fields (List[str] = ['text']): Fields to retrieve from the table. Defaults to "text"
+
+ Examples:
+ Below is a code snippet that shows how to use PgVector as the default retriever
+
+ ```python
+ import dspy
+ import openai
+ import psycopg2
+
+ openai.api_key = os.environ.get("OPENAI_API_KEY", None)
+ openai_client = openai.OpenAI()
+
+ llm = dspy.OpenAI(model="gpt-3.5-turbo")
+
+ DATABASE_URL should be in the format postgresql://user:password@host/database
+ db_url=os.getenv("DATABASE_URL")
+
+ retriever_model = PgVectorRM(conn, openai_client=openai_client, "paragraphs", fields=["text", "document_id"], k=20)
+ dspy.settings.configure(lm=llm, rm=retriever_model)
+ ```
+
+ Below is a code snippet that shows how to use PgVector in the forward() function of a module
+ ```python
+ self.retrieve = PgVectorRM(db_url, openai_client=openai_client, "paragraphs", fields=["text", "document_id"], k=20)
+ ```
+ """
+ def __init__(
+ self,
+ db_url: str,
+ pg_table_name: Optional[str],
+ openai_client: openai.OpenAI,
+ k: Optional[int]=20,
+ embedding_field: str = "embedding",
+ fields: List[str] = ['text', 'document_id']
+ ):
+ """
+ k = 20 is the number of paragraphs to retrieve
+ """
+ self.openai_client = openai_client
+
+ self.conn = psycopg2.connect(db_url)
+ register_vector(self.conn)
+ self.pg_table_name = pg_table_name
+ self.fields = fields
+ self.embedding_field = embedding_field
+
+ super().__init__(k=k)
+
+ def forward(self, query: str, k: Optional[int]=20):
+ """Search with PgVector for self.k top passages for query
+
+ Args:
+ query (str): The query to search for
+ k (Optional[int]): The number of top passages to retrieve. Defaults to self.k
+ Returns:
+ dspy.Prediction: an object containing the retrieved passages.
+ """
+ # Embed query
+ query_embedding = self.openai_client.embeddings.create(
+ model="text-embedding-ada-002",
+ input=query,
+ encoding_format="float"
+ ).data[0].embedding
+
+ related_paragraphs = []
+
+ sql_query = sql.SQL(
+ "select {fields} from {table} order by {embedding_field} <-> {query_embedding}::vector limit {limit}").format(
+ fields=sql.SQL(',').join([
+ sql.Identifier(f)
+ for f in self.fields
+ ]),
+ embedding_field=sql.Identifier(self.embedding_field),
+ table=sql.Identifier(self.pg_table_name),
+ query_embedding=sql.Identifier('query_embedding'),
+ limit=sql.Identifier('limit')
+ )
+
+ with self.conn as conn:
+ with conn.cursor() as cur:
+ cur.execute(
+ sql_query,
+ (query_embedding, self.k))
+ rows = cur.fetchall()
+ for row in rows:
+ related_paragraphs.append(dspy.Example(long_text=row[0], document_id=row[1]))
+ # Return Prediction
+ return related_paragraphs
\ No newline at end of file
diff --git a/pyproject.toml b/pyproject.toml
index 91d847b3c..86c4fcae7 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -99,6 +99,8 @@ pydantic = {version = "*", optional = true}
autodoc_pydantic = {version = "*", optional = true}
sphinx-reredirects = {version = "^0.1.2", optional = true}
sphinx-automodapi = {version = "0.16.0", optional = true}
+pgvector = {version = "*", optional = true}
+psycopg2 = {version = "*", optional = true}
[tool.poetry.extras]
chromadb = ["chromadb"]
@@ -106,6 +108,7 @@ qdrant = ["qdrant-client", "fastembed"]
marqo = ["marqo"]
pinecone = ["pinecone-client"]
weaviate = ["weaviate-client"]
+pgvector = ["pgvector", "psycopg2"]
docs = [
"sphinx",
"furo",