[DI-206] feat: make Document Index search parameters optional (#1053)

* [DI-206] feat: make Document Index search parameters optional

This commit aligns optional parameters in the IL with the Document Index
API.

* [DI-206] docs: move DocumentIndexRetriever constructor args

- move init docstring to constructor
- move changes to correct place in changelog

CHANGELOG.md

@@ -1,13 +1,17 @@
 # Changelog
 ## Unreleased
 ### Features
-...
+- `min_score` and `max_results` are now optional parameters in `DocumentIndexClient.SearchQuery`.
+- `k` is now an optional parameter in `DocumentIndexRetriever`.
+
 ### Fixes
 ...
 ### Deprecations 
 ...
 ### Breaking Changes
- - The default model for `Llama3InstructModel` is now `llama-3.1-8b-instruct` instead of `llama-3-8b-instruct`. We also removed the llama3.0 models from the recommended models of the `Llama3InstructModel`.
+- The default model for `Llama3InstructModel` is now `llama-3.1-8b-instruct` instead of `llama-3-8b-instruct`. We also removed the llama3.0 models from the recommended models of the `Llama3InstructModel`.
+- The default value of `threshold` in the `DocumentIndexRetriever` has changed from `0.5` to `0.0`. This accommodates fusion scoring for searches over hybrid indexes.
+
 
 ## 6.0.0
 

src/documentation/document_index.ipynb

@@ -283,7 +283,6 @@
     "    namespace=NAMESPACE,\n",
     "    collection=COLLECTION,\n",
     "    k=5,\n",
-    "    threshold=0.5,\n",
     ")\n",
     "\n",
     "document_index_retriever.get_relevant_documents_with_scores(\n",
@@ -402,7 +401,6 @@
     "    namespace=NAMESPACE,\n",
     "    collection=COLLECTION,\n",
     "    k=5,\n",
-    "    threshold=0.2,\n",
     ")"
    ]
   },

src/intelligence_layer/connectors/document_index/document_index.py

@@ -255,8 +255,8 @@ class SearchQuery(BaseModel):
     """
 
     query: str
-    max_results: int = Field(..., ge=0)
-    min_score: float = Field(..., ge=0.0, le=1.0)
+    max_results: int = Field(ge=0, default=1)
+    min_score: float = Field(ge=0.0, le=1.0, default=0.0)
     filters: Optional[list[Filters]] = None
 
 

src/intelligence_layer/connectors/retrievers/document_index_retriever.py

@@ -20,16 +20,7 @@
 class DocumentIndexRetriever(BaseRetriever[DocumentPath]):
     """Search through documents within collections in the `DocumentIndexClient`.
 
-    We initialize this Retriever with a collection & namespace names, and we can find the documents in the collection
-    most semanticly similar to our query.
-
-    Args:
-        document_index: Client offering functionality for search.
-        index_name: The name of the index to be used.
-        namespace: The namespace within the `DocumentIndexClient` where all collections are stored.
-        collection: The collection within the namespace that holds the desired documents.
-        k: The (top) number of documents to be returned by search.
-        threshold: The mimumum value of cosine similarity between the query vector and the document vector.
+    This retriever lets you search for relevant documents in the given Document Index collection.
 
     Example:
     >>> import os
@@ -45,9 +36,21 @@ def __init__(
         index_name: str,
         namespace: str,
         collection: str,
-        k: int,
-        threshold: float = 0.5,
+        k: int = 1,
+        threshold: float = 0.0,
     ) -> None:
+        """Initialize the DocumentIndexRetriever.
+
+        Args:
+            document_index: The Document Index client.
+            index_name: The name of the Document Index index to use.
+            namespace: The Document Index namespace.
+            collection: The Document Index collection to use. This is the search context for the retriever.
+            k: The number of most-relevant documents to return when searching. Defaults to 1.
+            threshold: The minimum score for search results. For semantic indexes, this is the cosine
+                similarity between the query and the document chunk. For hybrid indexes, this corresponds
+                to fusion rank. Defaults to 0.0.
+        """
         self._document_index = document_index
         self._index_name = index_name
         self._collection_path = CollectionPath(