fix retrieval

aimakerspace/vectordatabase.py

@@ -1,5 +1,5 @@
 from qdrant_client import QdrantClient
-from qdrant_client.http import models
+from qdrant_client.models import VectorParams, Distance
 from typing import List, Dict, Optional
 import os
 
@@ -34,22 +34,6 @@ class VectorDatabase:
             f"Inserted {len(texts)} documents into collection '{self.collection_name}'."
         )
 
-    def create_collection(self):
-        # Define the vector size and distance metric
-        vector_size = 128  # Size of the vectors you plan to store
-        distance_metric = (
-            models.Distance.COSINE
-        )  # Distance metric (e.g., COSINE, EUCLID, DOT)
-
-        # Create the collection
-        self.client.create_collection(
-            collection_name=self.collection_name,
-            vectors_config=models.VectorParams(
-                size=vector_size,
-                distance=distance_metric,
-            ),
-        )
-
     def _delete_collection(self):
         """
         Delete a collection from the Qdrant database.
@@ -60,31 +44,25 @@ class VectorDatabase:
         # Check if the collection exists
         collections = self.client.get_collections()
         collection_names = [collection.name for collection in collections.collections]
+        print(f"Existing collections: {collection_names}")
 
         if self.collection_name in collection_names:
             # Delete the collection
-            self.client.delete_collection(collection_name)
-            print(f"Collection '{collection_name}' deleted.")
+            self.client.delete_collection(self.collection_name)
+            print(f"Collection '{self.collection_name}' deleted.")
         else:
-            print(f"Collection '{collection_name}' does not exist.")
+            print(f"Collection '{self.collection_name}' does not exist.")
 
     def search_similar(self, query_text: str):
         search_result = self.client.query(
             collection_name=self.collection_name,
             query_text=query_text,
-            limit=1,
+            limit=4,
         )
 
-        document = search_result[0].document
-
-        return document
-
-    def delete_collection(self):
-        """
-        Delete the Qdrant collection.
-        """
-        self.client.delete_collection(self.collection_name)
-        print(f"Deleted collection: {self.collection_name}")
+        documents = [item.document for item in search_result]
+        result = " ".join(document for document in documents if document)
+        return result
 
     def list_collections(self):
         """

app.py

@@ -31,8 +31,6 @@ class RetrievalAugmentedQAPipeline:
     def __init__(self, llm: ChatOpenAI(), vector_db_retriever: VectorDatabase) -> None:
         self.llm = llm
         self.vector_db_retriever = vector_db_retriever
-        self.vector_db_retriever.delete_collection()
-        self.vector_db_retriever.create_collection()
 
     async def arun_pipeline(self, user_query: str):
         context_data = self.vector_db_retriever.search_similar(user_query)
@@ -115,6 +113,7 @@ async def on_chat_start():
 
     # Create a dict vector store
     vector_db = VectorDatabase()
+    vector_db._delete_collection()
     vector_db.upsert_documents(texts)
     chat_openai = ChatOpenAI()
 

uv.lock

@@ -7,7 +7,6 @@ version = "0.1.0"
 source = { virtual = "." }
 dependencies = [
     { name = "chainlit" },
-    { name = "maturin" },
     { name = "numpy" },
     { name = "openai" },
     { name = "pydantic" },
@@ -19,7 +18,6 @@ dependencies = [
 [package.metadata]
 requires-dist = [
     { name = "chainlit", specifier = ">=2.0.4" },
-    { name = "maturin", specifier = ">=1.8.1" },
     { name = "numpy", specifier = ">=2.2.2" },
     { name = "openai", specifier = ">=1.59.9" },
     { name = "pydantic", specifier = "==2.10.1" },
@@ -535,26 +533,6 @@ wheels = [
     { url = "https://files.pythonhosted.org/packages/8e/25/5b300f0400078d9783fbe44d30fedd849a130fc3aff01f18278c12342b6f/marshmallow-3.25.1-py3-none-any.whl", hash = "sha256:ec5d00d873ce473b7f2ffcb7104286a376c354cab0c2fa12f5573dab03e87210", size = 49624 },
 ]
 
-[[package]]
-name = "maturin"
-version = "1.8.1"
-source = { registry = "https://pypi.org/simple" }
-sdist = { url = "https://files.pythonhosted.org/packages/9a/08/ccb0f917722a35ab0d758be9bb5edaf645c3a3d6170061f10d396ecd273f/maturin-1.8.1.tar.gz", hash = "sha256:49cd964aabf59f8b0a6969f9860d2cdf194ac331529caae14c884f5659568857", size = 197397 }
-wheels = [
-    { url = "https://files.pythonhosted.org/packages/4c/00/f34077315f34db8ad2ccf6bfe11b864ca27baab3a1320634da8e3cf89a48/maturin-1.8.1-py3-none-linux_armv6l.whl", hash = "sha256:7e590a23d9076b8a994f2e67bc63dc9a2d1c9a41b1e7b45ac354ba8275254e89", size = 7568415 },
-    { url = "https://files.pythonhosted.org/packages/5c/07/9219976135ce0cb32d2fa6ea5c6d0ad709013d9a17967312e149b98153a6/maturin-1.8.1-py3-none-macosx_10_12_x86_64.macosx_11_0_arm64.macosx_10_12_universal2.whl", hash = "sha256:8d8251a95682c83ea60988c804b620c181911cd824aa107b4a49ac5333c92968", size = 14527816 },
-    { url = "https://files.pythonhosted.org/packages/e6/04/fa009a00903acdd1785d58322193140bfe358595347c39f315112dabdf9e/maturin-1.8.1-py3-none-macosx_10_12_x86_64.whl", hash = "sha256:b9fc1a4354cac5e32c190410208039812ea88c4a36bd2b6499268ec49ef5de00", size = 7580446 },
-    { url = "https://files.pythonhosted.org/packages/9b/d4/414b2aab9bbfe88182b734d3aa1b4fef7d7701e50f6be48500378b8c8721/maturin-1.8.1-py3-none-manylinux_2_12_i686.manylinux2010_i686.musllinux_1_1_i686.whl", hash = "sha256:621e171c6b39f95f1d0df69a118416034fbd59c0f89dcaea8c2ea62019deecba", size = 7650535 },
-    { url = "https://files.pythonhosted.org/packages/f0/64/879418a8a0196013ec1fb19eada0781c04a30e8d6d9227e80f91275a4f5b/maturin-1.8.1-py3-none-manylinux_2_12_x86_64.manylinux2010_x86_64.musllinux_1_1_x86_64.whl", hash = "sha256:98f638739a5132962347871b85c91f525c9246ef4d99796ae98a2031e3df029f", size = 8006702 },
-    { url = "https://files.pythonhosted.org/packages/39/c2/605829324f8371294f70303aca130682df75318958efed246873d3d604ab/maturin-1.8.1-py3-none-manylinux_2_17_aarch64.manylinux2014_aarch64.musllinux_1_1_aarch64.whl", hash = "sha256:f9f5c47521924b6e515cbc652a042fe5f17f8747445be9d931048e5d8ddb50a4", size = 7368164 },
-    { url = "https://files.pythonhosted.org/packages/be/6c/30e136d397bb146b94b628c0ef7f17708281611b97849e2cf37847025ac7/maturin-1.8.1-py3-none-manylinux_2_17_armv7l.manylinux2014_armv7l.musllinux_1_1_armv7l.whl", hash = "sha256:0f4407c7353c31bfbb8cdeb82bc2170e474cbfb97b5ba27568f440c9d6c1fdd4", size = 7450889 },
-    { url = "https://files.pythonhosted.org/packages/1b/50/e1f5023512696d4e56096f702e2f68d6d9a30afe0a4eec82b0e27b8eb4e4/maturin-1.8.1-py3-none-manylinux_2_17_ppc64le.manylinux2014_ppc64le.musllinux_1_1_ppc64le.whl", hash = "sha256:ec49cd70cad3c389946c6e2bc0bd50772a7fcb463040dd800720345897eec9bf", size = 9585819 },
-    { url = "https://files.pythonhosted.org/packages/b7/80/b24b5248d89d2e5982553900237a337ea098ca9297b8369ca2aa95549e0f/maturin-1.8.1-py3-none-manylinux_2_17_s390x.manylinux2014_s390x.whl", hash = "sha256:c08767d794de8f8a11c5c8b1b47a4ff9fb6ae2d2d97679e27030f2f509c8c2a0", size = 10920801 },
-    { url = "https://files.pythonhosted.org/packages/6e/f4/8ede7a662fabf93456b44390a5ad22630e25fb5ddaecf787251071b2e143/maturin-1.8.1-py3-none-win32.whl", hash = "sha256:d678407713f3e10df33c5b3d7a343ec0551eb7f14d8ad9ba6febeb96f4e4c75c", size = 6873556 },
-    { url = "https://files.pythonhosted.org/packages/9c/22/757f093ed0e319e9648155b8c9d716765442bea5bc98ebc58ad4ad5b0524/maturin-1.8.1-py3-none-win_amd64.whl", hash = "sha256:a526f90fe0e5cb59ffb81f4ff547ddc42e823bbdeae4a31012c0893ca6dcaf46", size = 7823153 },
-    { url = "https://files.pythonhosted.org/packages/a4/f5/051413e04f6da25069db5e76759ecdb8cd2a8ab4a94045b5a3bf548c66fa/maturin-1.8.1-py3-none-win_arm64.whl", hash = "sha256:e95f077fd2ddd2f048182880eed458c308571a534be3eb2add4d3dac55bf57f4", size = 6552131 },
-]
-
 [[package]]
 name = "mmh3"
 version = "4.1.0"