adding option

mattgotteiner · mattgotteiner · commit 79cb03daf644 · 2025-03-09T23:10:35.000-07:00
diff --git a/app/backend/app.py b/app/backend/app.py
@@ -71,6 +71,7 @@
     CONFIG_OPENAI_CLIENT,
     CONFIG_SEARCH_CLIENT,
     CONFIG_SEMANTIC_RANKER_DEPLOYED,
+    CONFIG_QUERY_REWRITING_ENABLED,
     CONFIG_SPEECH_INPUT_ENABLED,
     CONFIG_SPEECH_OUTPUT_AZURE_ENABLED,
     CONFIG_SPEECH_OUTPUT_BROWSER_ENABLED,
@@ -291,6 +292,7 @@ def config():
         {
             "showGPT4VOptions": current_app.config[CONFIG_GPT4V_DEPLOYED],
             "showSemanticRankerOption": current_app.config[CONFIG_SEMANTIC_RANKER_DEPLOYED],
+            "showQueryRewritingOption": current_app.config[CONFIG_QUERY_REWRITING_ENABLED],
             "showVectorOption": current_app.config[CONFIG_VECTOR_SEARCH_ENABLED],
             "showUserUpload": current_app.config[CONFIG_USER_UPLOAD_ENABLED],
             "showLanguagePicker": current_app.config[CONFIG_LANGUAGE_PICKER_ENABLED],
@@ -453,6 +455,7 @@ async def setup_clients():
     AZURE_SEARCH_QUERY_LANGUAGE = os.getenv("AZURE_SEARCH_QUERY_LANGUAGE") or "en-us"
     AZURE_SEARCH_QUERY_SPELLER = os.getenv("AZURE_SEARCH_QUERY_SPELLER") or "lexicon"
     AZURE_SEARCH_SEMANTIC_RANKER = os.getenv("AZURE_SEARCH_SEMANTIC_RANKER", "free").lower()
+    AZURE_SEARCH_QUERY_REWRITING = os.getenv("AZURE_SEARCH_QUERY_REWRITING", "false").lower()
 
     AZURE_SPEECH_SERVICE_ID = os.getenv("AZURE_SPEECH_SERVICE_ID")
     AZURE_SPEECH_SERVICE_LOCATION = os.getenv("AZURE_SPEECH_SERVICE_LOCATION")
@@ -634,6 +637,7 @@ async def setup_clients():
 
     current_app.config[CONFIG_GPT4V_DEPLOYED] = bool(USE_GPT4V)
     current_app.config[CONFIG_SEMANTIC_RANKER_DEPLOYED] = AZURE_SEARCH_SEMANTIC_RANKER != "disabled"
+    current_app.config[CONFIG_QUERY_REWRITING_ENABLED] = AZURE_SEARCH_QUERY_REWRITING == "true" and AZURE_SEARCH_SEMANTIC_RANKER != "disabled"
     current_app.config[CONFIG_VECTOR_SEARCH_ENABLED] = os.getenv("USE_VECTORS", "").lower() != "false"
     current_app.config[CONFIG_USER_UPLOAD_ENABLED] = bool(USE_USER_UPLOAD)
     current_app.config[CONFIG_LANGUAGE_PICKER_ENABLED] = ENABLE_LANGUAGE_PICKER
diff --git a/app/backend/approaches/approach.py b/app/backend/approaches/approach.py
@@ -149,6 +149,7 @@ async def search(
         use_semantic_captions: bool,
         minimum_search_score: Optional[float],
         minimum_reranker_score: Optional[float],
+        use_query_rewriting: Optional[bool]
     ) -> List[Document]:
         search_text = query_text if use_text_search else ""
         search_vectors = vectors if use_vector_search else []
@@ -158,6 +159,7 @@ async def search(
                 filter=filter,
                 top=top,
                 query_caption="extractive|highlight-false" if use_semantic_captions else None,
+                query_rewrites="generative" if use_query_rewriting else None,
                 vector_queries=search_vectors,
                 query_type=QueryType.SEMANTIC,
                 query_language=self.query_language,
diff --git a/app/backend/approaches/chatreadretrieveread.py b/app/backend/approaches/chatreadretrieveread.py
@@ -89,6 +89,7 @@ async def run_until_final_call(
         use_vector_search = overrides.get("retrieval_mode") in ["vectors", "hybrid", None]
         use_semantic_ranker = True if overrides.get("semantic_ranker") else False
         use_semantic_captions = True if overrides.get("semantic_captions") else False
+        use_query_rewriting = True if overrides.get("query_rewriting") else False
         top = overrides.get("top", 3)
         minimum_search_score = overrides.get("minimum_search_score", 0.0)
         minimum_reranker_score = overrides.get("minimum_reranker_score", 0.0)
@@ -147,6 +148,7 @@ async def run_until_final_call(
             use_semantic_captions,
             minimum_search_score,
             minimum_reranker_score,
+            use_query_rewriting
         )
 
         # STEP 3: Generate a contextual and content specific answer using the search results and chat history
diff --git a/app/backend/approaches/chatreadretrievereadvision.py b/app/backend/approaches/chatreadretrievereadvision.py
@@ -81,6 +81,7 @@ async def run_until_final_call(
         use_text_search = overrides.get("retrieval_mode") in ["text", "hybrid", None]
         use_vector_search = overrides.get("retrieval_mode") in ["vectors", "hybrid", None]
         use_semantic_ranker = True if overrides.get("semantic_ranker") else False
+        use_query_rewriting = True if overrides.get("query_rewriting") else False
         use_semantic_captions = True if overrides.get("semantic_captions") else False
         top = overrides.get("top", 3)
         minimum_search_score = overrides.get("minimum_search_score", 0.0)
@@ -151,6 +152,7 @@ async def run_until_final_call(
             use_semantic_captions,
             minimum_search_score,
             minimum_reranker_score,
+            use_query_rewriting
         )
 
         # STEP 3: Generate a contextual and content specific answer using the search results and chat history
diff --git a/app/backend/approaches/retrievethenread.py b/app/backend/approaches/retrievethenread.py
@@ -67,6 +67,7 @@ async def run(
         use_text_search = overrides.get("retrieval_mode") in ["text", "hybrid", None]
         use_vector_search = overrides.get("retrieval_mode") in ["vectors", "hybrid", None]
         use_semantic_ranker = True if overrides.get("semantic_ranker") else False
+        use_query_rewriting = True if overrides.get("query_rewriting") else False
         use_semantic_captions = True if overrides.get("semantic_captions") else False
         top = overrides.get("top", 3)
         minimum_search_score = overrides.get("minimum_search_score", 0.0)
@@ -89,6 +90,7 @@ async def run(
             use_semantic_captions,
             minimum_search_score,
             minimum_reranker_score,
+            use_query_rewriting
         )
 
         # Process results
diff --git a/app/backend/approaches/retrievethenreadvision.py b/app/backend/approaches/retrievethenreadvision.py
@@ -76,6 +76,7 @@ async def run(
         use_text_search = overrides.get("retrieval_mode") in ["text", "hybrid", None]
         use_vector_search = overrides.get("retrieval_mode") in ["vectors", "hybrid", None]
         use_semantic_ranker = True if overrides.get("semantic_ranker") else False
+        use_query_rewriting = True if overrides.get("query_rewriting") else False
         use_semantic_captions = True if overrides.get("semantic_captions") else False
         top = overrides.get("top", 3)
         minimum_search_score = overrides.get("minimum_search_score", 0.0)
@@ -108,6 +109,7 @@ async def run(
             use_semantic_captions,
             minimum_search_score,
             minimum_reranker_score,
+            use_query_rewriting
         )
 
         # Process results
diff --git a/app/backend/config.py b/app/backend/config.py
@@ -10,6 +10,7 @@
 CONFIG_AUTH_CLIENT = "auth_client"
 CONFIG_GPT4V_DEPLOYED = "gpt4v_deployed"
 CONFIG_SEMANTIC_RANKER_DEPLOYED = "semantic_ranker_deployed"
+CONFIG_QUERY_REWRITING_ENABLED = "query_rewriting_enabled"
 CONFIG_VECTOR_SEARCH_ENABLED = "vector_search_enabled"
 CONFIG_SEARCH_CLIENT = "search_client"
 CONFIG_OPENAI_CLIENT = "openai_client"
diff --git a/app/backend/requirements.in b/app/backend/requirements.in
@@ -7,7 +7,7 @@ tenacity
 azure-ai-documentintelligence==1.0.0b4
 azure-cognitiveservices-speech
 azure-cosmos
-azure-search-documents==11.6.0b6
+azure-search-documents==11.6.0b9
 azure-storage-blob
 azure-storage-file-datalake
 uvicorn
diff --git a/app/backend/requirements.txt b/app/backend/requirements.txt
@@ -57,7 +57,7 @@ azure-monitor-opentelemetry==1.6.1
     # via -r requirements.in
 azure-monitor-opentelemetry-exporter==1.0.0b32
     # via azure-monitor-opentelemetry
-azure-search-documents==11.6.0b6
+azure-search-documents==11.6.0b9
     # via -r requirements.in
 azure-storage-blob==12.22.0
     # via
diff --git a/app/frontend/src/api/models.ts b/app/frontend/src/api/models.ts
@@ -20,6 +20,7 @@ export type ChatAppRequestOverrides = {
     retrieval_mode?: RetrievalMode;
     semantic_ranker?: boolean;
     semantic_captions?: boolean;
+    query_rewriting?: boolean;
     include_category?: string;
     exclude_category?: string;
     seed?: number;
@@ -84,6 +85,7 @@ export type ChatAppRequest = {
 export type Config = {
     showGPT4VOptions: boolean;
     showSemanticRankerOption: boolean;
+    showQueryRewritingOption: boolean;
     showVectorOption: boolean;
     showUserUpload: boolean;
     showLanguagePicker: boolean;
diff --git a/app/frontend/src/components/Settings/Settings.tsx b/app/frontend/src/components/Settings/Settings.tsx
@@ -19,13 +19,15 @@ export interface SettingsProps {
     minimumRerankerScore: number;
     useSemanticRanker: boolean;
     useSemanticCaptions: boolean;
+    useQueryRewriting: boolean;
     excludeCategory: string;
     includeCategory: string;
     retrievalMode: RetrievalMode;
     useGPT4V: boolean;
     gpt4vInput: GPT4VInput;
     vectorFieldList: VectorFieldOptions[];
     showSemanticRankerOption: boolean;
+    showQueryRewritingOption: boolean;
     showGPT4VOptions: boolean;
     showVectorOption: boolean;
     useOidSecurityFilter: boolean;
@@ -51,13 +53,15 @@ export const Settings = ({
     minimumRerankerScore,
     useSemanticRanker,
     useSemanticCaptions,
+    useQueryRewriting,
     excludeCategory,
     includeCategory,
     retrievalMode,
     useGPT4V,
     gpt4vInput,
     vectorFieldList,
     showSemanticRankerOption,
+    showQueryRewritingOption,
     showGPT4VOptions,
     showVectorOption,
     useOidSecurityFilter,
@@ -94,6 +98,7 @@ export const Settings = ({
     const excludeCategoryFieldId = useId("excludeCategoryField");
     const semanticRankerId = useId("semanticRanker");
     const semanticRankerFieldId = useId("semanticRankerField");
+    const queryRewritingFieldId = useId("queryRewritingField");
     const semanticCaptionsId = useId("semanticCaptions");
     const semanticCaptionsFieldId = useId("semanticCaptionsField");
     const useOidSecurityFilterId = useId("useOidSecurityFilter");
@@ -239,6 +244,21 @@ export const Settings = ({
                 </>
             )}
 
+            {showQueryRewritingOption && (
+                <>
+                    <Checkbox
+                        id={queryRewritingFieldId}
+                        className={styles.settingsSeparator}
+                        checked={useQueryRewriting}
+                        disabled={!useSemanticRanker}
+                        label={t("labels.useQueryRewriting")}
+                        onChange={(_ev, checked) => onChange("useQueryRewriting", !!checked)}
+                        aria-labelledby={queryRewritingFieldId}
+                        onRenderLabel={props => renderLabel(props, queryRewritingFieldId, queryRewritingFieldId, t("helpTexts.useQueryRewriting"))}
+                    />
+                </>
+            )}
+
             {useLogin && (
                 <>
                     <Checkbox
diff --git a/app/frontend/src/locales/en/translation.json b/app/frontend/src/locales/en/translation.json
@@ -90,6 +90,7 @@
         "excludeCategory": "Exclude category",
         "useSemanticRanker": "Use semantic ranker for retrieval",
         "useSemanticCaptions": "Use semantic captions",
+        "useQueryRewriting": "Use query rewriting for retrieval",
         "useSuggestFollowupQuestions": "Suggest follow-up questions",
         "useGPT4V": "Use GPT vision model",
         "gpt4VInput": {
@@ -139,6 +140,8 @@
             "Specifies a category to exclude from the search results. There are no categories used in the default data set.",
          "useSemanticReranker":
             "Enables the Azure AI Search semantic ranker, a model that re-ranks search results based on semantic similarity to the user's query.",
+        "useQueryRewriting":
+            "Enables Azure AI Search query rewriting, a process that modifies the user's query to improve search results. Requires semantic ranker to be enabled.",
         "useSemanticCaptions":
              "Sends semantic captions to the LLM instead of the full search result. A semantic caption is extracted from a search result during the process of semantic ranking.",
         "suggestFollowupQuestions": "Asks the LLM to suggest follow-up questions based on the user's query.",
diff --git a/app/frontend/src/pages/ask/Ask.tsx b/app/frontend/src/pages/ask/Ask.tsx
@@ -32,6 +32,7 @@ export function Component(): JSX.Element {
     const [retrieveCount, setRetrieveCount] = useState<number>(3);
     const [useSemanticRanker, setUseSemanticRanker] = useState<boolean>(true);
     const [useSemanticCaptions, setUseSemanticCaptions] = useState<boolean>(false);
+    const [useQueryRewriting, setUseQueryRewriting] = useState<boolean>(false);
     const [useGPT4V, setUseGPT4V] = useState<boolean>(false);
     const [gpt4vInput, setGPT4VInput] = useState<GPT4VInput>(GPT4VInput.TextAndImages);
     const [includeCategory, setIncludeCategory] = useState<string>("");
@@ -42,6 +43,7 @@ export function Component(): JSX.Element {
     const [useGroupsSecurityFilter, setUseGroupsSecurityFilter] = useState<boolean>(false);
     const [showGPT4VOptions, setShowGPT4VOptions] = useState<boolean>(false);
     const [showSemanticRankerOption, setShowSemanticRankerOption] = useState<boolean>(false);
+    const [showQueryRewritingOption, setShowQueryRewritingOption] = useState<boolean>(false);
     const [showVectorOption, setShowVectorOption] = useState<boolean>(false);
     const [showUserUpload, setShowUserUpload] = useState<boolean>(false);
     const [showLanguagePicker, setshowLanguagePicker] = useState<boolean>(false);
@@ -78,6 +80,8 @@ export function Component(): JSX.Element {
             setShowGPT4VOptions(config.showGPT4VOptions);
             setUseSemanticRanker(config.showSemanticRankerOption);
             setShowSemanticRankerOption(config.showSemanticRankerOption);
+            setUseQueryRewriting(config.showQueryRewritingOption);
+            setShowQueryRewritingOption(config.showQueryRewritingOption);
             setShowVectorOption(config.showVectorOption);
             if (!config.showVectorOption) {
                 setRetrievalMode(RetrievalMode.Text);
@@ -126,6 +130,7 @@ export function Component(): JSX.Element {
                         retrieval_mode: retrievalMode,
                         semantic_ranker: useSemanticRanker,
                         semantic_captions: useSemanticCaptions,
+                        query_rewriting: useQueryRewriting,
                         use_oid_security_filter: useOidSecurityFilter,
                         use_groups_security_filter: useGroupsSecurityFilter,
                         vector_fields: vectorFieldList,
@@ -180,6 +185,9 @@ export function Component(): JSX.Element {
             case "useSemanticCaptions":
                 setUseSemanticCaptions(value);
                 break;
+            case "useQueryRewriting":
+                setUseQueryRewriting(value);
+                break;
             case "excludeCategory":
                 setExcludeCategory(value);
                 break;
@@ -321,13 +329,15 @@ export function Component(): JSX.Element {
                     minimumRerankerScore={minimumRerankerScore}
                     useSemanticRanker={useSemanticRanker}
                     useSemanticCaptions={useSemanticCaptions}
+                    useQueryRewriting={useQueryRewriting}
                     excludeCategory={excludeCategory}
                     includeCategory={includeCategory}
                     retrievalMode={retrievalMode}
                     useGPT4V={useGPT4V}
                     gpt4vInput={gpt4vInput}
                     vectorFieldList={vectorFieldList}
                     showSemanticRankerOption={showSemanticRankerOption}
+                    showQueryRewritingOption={showQueryRewritingOption}
                     showGPT4VOptions={showGPT4VOptions}
                     showVectorOption={showVectorOption}
                     useOidSecurityFilter={useOidSecurityFilter}
diff --git a/app/frontend/src/pages/chat/Chat.tsx b/app/frontend/src/pages/chat/Chat.tsx
@@ -48,6 +48,7 @@ const Chat = () => {
     const [retrieveCount, setRetrieveCount] = useState<number>(3);
     const [retrievalMode, setRetrievalMode] = useState<RetrievalMode>(RetrievalMode.Hybrid);
     const [useSemanticRanker, setUseSemanticRanker] = useState<boolean>(true);
+    const [useQueryRewriting, setUseQueryRewriting] = useState<boolean>(false);
     const [shouldStream, setShouldStream] = useState<boolean>(true);
     const [useSemanticCaptions, setUseSemanticCaptions] = useState<boolean>(false);
     const [includeCategory, setIncludeCategory] = useState<string>("");
@@ -76,6 +77,7 @@ const Chat = () => {
 
     const [showGPT4VOptions, setShowGPT4VOptions] = useState<boolean>(false);
     const [showSemanticRankerOption, setShowSemanticRankerOption] = useState<boolean>(false);
+    const [showQueryRewritingOption, setShowQueryRewritingOption] = useState<boolean>(false);
     const [showVectorOption, setShowVectorOption] = useState<boolean>(false);
     const [showUserUpload, setShowUserUpload] = useState<boolean>(false);
     const [showLanguagePicker, setshowLanguagePicker] = useState<boolean>(false);
@@ -100,6 +102,8 @@ const Chat = () => {
             setShowGPT4VOptions(config.showGPT4VOptions);
             setUseSemanticRanker(config.showSemanticRankerOption);
             setShowSemanticRankerOption(config.showSemanticRankerOption);
+            setUseQueryRewriting(config.showQueryRewritingOption);
+            setShowQueryRewritingOption(config.showQueryRewritingOption);
             setShowVectorOption(config.showVectorOption);
             if (!config.showVectorOption) {
                 setRetrievalMode(RetrievalMode.Text);
@@ -197,6 +201,7 @@ const Chat = () => {
                         retrieval_mode: retrievalMode,
                         semantic_ranker: useSemanticRanker,
                         semantic_captions: useSemanticCaptions,
+                        query_rewriting: useQueryRewriting,
                         suggest_followup_questions: useSuggestFollowupQuestions,
                         use_oid_security_filter: useOidSecurityFilter,
                         use_groups_security_filter: useGroupsSecurityFilter,
@@ -285,6 +290,9 @@ const Chat = () => {
             case "useSemanticRanker":
                 setUseSemanticRanker(value);
                 break;
+            case "useQueryRewriting":
+                setUseQueryRewriting(value);
+                break;
             case "useSemanticCaptions":
                 setUseSemanticCaptions(value);
                 break;
@@ -500,13 +508,15 @@ const Chat = () => {
                         minimumRerankerScore={minimumRerankerScore}
                         useSemanticRanker={useSemanticRanker}
                         useSemanticCaptions={useSemanticCaptions}
+                        useQueryRewriting={useQueryRewriting}
                         excludeCategory={excludeCategory}
                         includeCategory={includeCategory}
                         retrievalMode={retrievalMode}
                         useGPT4V={useGPT4V}
                         gpt4vInput={gpt4vInput}
                         vectorFieldList={vectorFieldList}
                         showSemanticRankerOption={showSemanticRankerOption}
+                        showQueryRewritingOption={showQueryRewritingOption}
                         showGPT4VOptions={showGPT4VOptions}
                         showVectorOption={showVectorOption}
                         useOidSecurityFilter={useOidSecurityFilter}
