Support use of AzureOpenAI proxy by prepdocs (#1760)

* Support custom URL for prepdocs

* update upload

* pwsh

* pwsh

* Move logic to embeddings

* Keep service variable

* Make mypy happy

* Fix tests

* Test coverage amount

* Update E2E

Author: pamelafox

.github/workflows/python-test.yaml

@@ -58,7 +58,7 @@ jobs:
           run: black . --check --verbose
         - name: Run Python tests
           if: runner.os != 'Windows'
-          run: python3 -m pytest -s -vv --cov --cov-fail-under=87
+          run: python3 -m pytest -s -vv --cov --cov-fail-under=86
         - name: Run E2E tests with Playwright
           id: e2e
           if: runner.os != 'Windows'

app/backend/app.py

@@ -400,6 +400,7 @@ async def setup_clients():
         os.getenv("AZURE_OPENAI_CHATGPT_DEPLOYMENT") if OPENAI_HOST.startswith("azure") else None
     )
     AZURE_OPENAI_EMB_DEPLOYMENT = os.getenv("AZURE_OPENAI_EMB_DEPLOYMENT") if OPENAI_HOST.startswith("azure") else None
+    AZURE_OPENAI_CUSTOM_URL = os.getenv("AZURE_OPENAI_CUSTOM_URL")
     AZURE_VISION_ENDPOINT = os.getenv("AZURE_VISION_ENDPOINT", "")
     # Used only with non-Azure OpenAI deployments
     OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
@@ -499,6 +500,7 @@ async def setup_clients():
             openai_host=OPENAI_HOST,
             openai_model_name=OPENAI_EMB_MODEL,
             openai_service=AZURE_OPENAI_SERVICE,
+            openai_custom_url=AZURE_OPENAI_CUSTOM_URL,
             openai_deployment=AZURE_OPENAI_EMB_DEPLOYMENT,
             openai_dimensions=OPENAI_EMB_DIMENSIONS,
             openai_key=clean_key_if_exists(OPENAI_API_KEY),
@@ -527,12 +529,14 @@ async def setup_clients():
 
     if OPENAI_HOST.startswith("azure"):
         api_version = os.getenv("AZURE_OPENAI_API_VERSION") or "2024-03-01-preview"
-
         if OPENAI_HOST == "azure_custom":
-            endpoint = os.environ["AZURE_OPENAI_CUSTOM_URL"]
+            if not AZURE_OPENAI_CUSTOM_URL:
+                raise ValueError("AZURE_OPENAI_CUSTOM_URL must be set when OPENAI_HOST is azure_custom")
+            endpoint = AZURE_OPENAI_CUSTOM_URL
         else:
+            if not AZURE_OPENAI_SERVICE:
+                raise ValueError("AZURE_OPENAI_SERVICE must be set when OPENAI_HOST is azure")
             endpoint = f"https://{AZURE_OPENAI_SERVICE}.openai.azure.com"
-
         if api_key := os.getenv("AZURE_OPENAI_API_KEY"):
             openai_client = AsyncAzureOpenAI(api_version=api_version, azure_endpoint=endpoint, api_key=api_key)
         else:

app/backend/prepdocs.py

@@ -109,6 +109,7 @@ def setup_embeddings_service(
     openai_host: str,
     openai_model_name: str,
     openai_service: Union[str, None],
+    openai_custom_url: Union[str, None],
     openai_deployment: Union[str, None],
     openai_dimensions: int,
     openai_key: Union[str, None],
@@ -126,6 +127,7 @@ def setup_embeddings_service(
         )
         return AzureOpenAIEmbeddingService(
             open_ai_service=openai_service,
+            open_ai_custom_url=openai_custom_url,
             open_ai_deployment=openai_deployment,
             open_ai_model_name=openai_model_name,
             open_ai_dimensions=openai_dimensions,
@@ -309,10 +311,16 @@ async def main(strategy: Strategy, setup_index: bool = True):
     parser.add_argument(
         "--disablebatchvectors", action="store_true", help="Don't compute embeddings in batch for the sections"
     )
+
+    parser.add_argument(
+        "--openaicustomurl",
+        required=False,
+        help="Optional. Use this custom OpenAI URL instead of the default OpenAI URL",
+    )
     parser.add_argument(
         "--openaikey",
         required=False,
-        help="Optional. Use this Azure OpenAI account key instead of the current user identity to login (use az login to set current user for Azure). This is required only when using non-Azure endpoints.",
+        help="Optional. Use this OpenAI account key instead of the current Azure user identity to login.",
     )
     parser.add_argument("--openaiorg", required=False, help="This is required only when using non-Azure endpoints.")
     parser.add_argument(
@@ -419,6 +427,7 @@ async def main(strategy: Strategy, setup_index: bool = True):
         openai_host=args.openaihost,
         openai_model_name=args.openaimodelname,
         openai_service=args.openaiservice,
+        openai_custom_url=args.openaicustomurl,
         openai_deployment=args.openaideployment,
         openai_dimensions=args.openaidimensions,
         openai_key=clean_key_if_exists(args.openaikey),

app/backend/prepdocslib/embeddings.py

@@ -164,10 +164,17 @@ def __init__(
         open_ai_model_name: str,
         open_ai_dimensions: int,
         credential: Union[AsyncTokenCredential, AzureKeyCredential],
+        open_ai_custom_url: Union[str, None] = None,
         disable_batch: bool = False,
     ):
         super().__init__(open_ai_model_name, open_ai_dimensions, disable_batch)
         self.open_ai_service = open_ai_service
+        if open_ai_service:
+            self.open_ai_endpoint = f"https://{open_ai_service}.openai.azure.com"
+        elif open_ai_custom_url:
+            self.open_ai_endpoint = open_ai_custom_url
+        else:
+            raise ValueError("Either open_ai_service or open_ai_custom_url must be provided")
         self.open_ai_deployment = open_ai_deployment
         self.credential = credential
 
@@ -187,7 +194,7 @@ class AuthArgs(TypedDict, total=False):
             raise TypeError("Invalid credential type")
 
         return AsyncAzureOpenAI(
-            azure_endpoint=f"https://{self.open_ai_service}.openai.azure.com",
+            azure_endpoint=self.open_ai_endpoint,
             azure_deployment=self.open_ai_deployment,
             api_version="2023-05-15",
             **auth_args,

scripts/prepdocs.ps1

@@ -65,6 +65,12 @@ if ($env:USE_FEATURE_INT_VECTORIZATION) {
   $integratedVectorizationArg = "--useintvectorization $env:USE_FEATURE_INT_VECTORIZATION"
 }
 
+if ($env:AZURE_OPENAI_API_KEY) {
+  $openaiApiKey = $env:AZURE_OPENAI_API_KEY
+} else {
+  $openaiApiKey = $env:OPENAI_API_KEY
+}
+
 $cwd = (Get-Location)
 $dataArg = "`"$cwd/data/*`""
 
@@ -75,6 +81,7 @@ $argumentList = "./app/backend/prepdocs.py $dataArg --verbose " + `
 "$searchAnalyzerNameArg " + `
 "--openaihost `"$env:OPENAI_HOST`" --openaimodelname `"$env:AZURE_OPENAI_EMB_MODEL_NAME`" $openaiDimensionsArg " + `
 "--openaiservice `"$env:AZURE_OPENAI_SERVICE`" --openaideployment `"$env:AZURE_OPENAI_EMB_DEPLOYMENT`" " + `
+"--openaicustomurl `"$env:OPENAI_CUSTOM_URL`" " + `
 "--openaikey `"$env:OPENAI_API_KEY`" --openaiorg `"$env:OPENAI_ORGANIZATION`" " + `
 "--documentintelligenceservice $env:AZURE_DOCUMENTINTELLIGENCE_SERVICE " + `
 "$searchImagesArg $visionEndpointArg " + `

scripts/prepdocs.sh

@@ -63,6 +63,11 @@ if [ -n "$USE_FEATURE_INT_VECTORIZATION" ]; then
   integratedVectorizationArg="--useintvectorization $USE_FEATURE_INT_VECTORIZATION"
 fi
 
+if [ -n "$AZURE_OPENAI_API_KEY" ]; then
+  openAiApiKeyArg="$AZURE_OPENAI_API_KEY"
+elif [ -n "$OPENAI_API_KEY" ]; then
+  openAiApiKeyArg="$OPENAI_API_KEY"
+fi
 
 ./.venv/bin/python ./app/backend/prepdocs.py './data/*' --verbose \
 --subscriptionid $AZURE_SUBSCRIPTION_ID  \
@@ -71,7 +76,8 @@ fi
 $searchAnalyzerNameArg \
 --openaihost "$OPENAI_HOST" --openaimodelname "$AZURE_OPENAI_EMB_MODEL_NAME" $openAiDimensionsArg \
 --openaiservice "$AZURE_OPENAI_SERVICE" --openaideployment "$AZURE_OPENAI_EMB_DEPLOYMENT"  \
---openaikey "$OPENAI_API_KEY" --openaiorg "$OPENAI_ORGANIZATION" \
+--openaicustomurl "$AZURE_OPENAI_CUSTOM_URL" \
+--openaikey $openAiApiKeyArg --openaiorg "$OPENAI_ORGANIZATION" \
 --documentintelligenceservice "$AZURE_DOCUMENTINTELLIGENCE_SERVICE" \
 $searchImagesArg $visionEndpointArg \
 $adlsGen2StorageAccountArg $adlsGen2FilesystemArg $adlsGen2FilesystemPathArg \

tests/e2e.py

@@ -54,6 +54,7 @@ def run_server(port: int):
             "AZURE_SEARCH_SERVICE": "test-search-service",
             "AZURE_SPEECH_SERVICE_ID": "test-id",
             "AZURE_SPEECH_SERVICE_LOCATION": "eastus",
+            "AZURE_OPENAI_SERVICE": "test-openai-service",
             "AZURE_OPENAI_CHATGPT_MODEL": "gpt-35-turbo",
         },
         clear=True,

tests/test_app_config.py

@@ -13,6 +13,7 @@ def minimal_env(monkeypatch):
         monkeypatch.setenv("AZURE_STORAGE_CONTAINER", "test-storage-container")
         monkeypatch.setenv("AZURE_SEARCH_INDEX", "test-search-index")
         monkeypatch.setenv("AZURE_SEARCH_SERVICE", "test-search-service")
+        monkeypatch.setenv("AZURE_OPENAI_SERVICE", "test-openai-service")
         monkeypatch.setenv("AZURE_OPENAI_CHATGPT_MODEL", "gpt-35-turbo")
         yield