Add cognitive services endpoint to config parameters (#6)

Co-authored-by: Josh Bradley <[email protected]>

Author: dorbaker

backend/graphrag-wheel/graphrag-0.0.1-py3-none-any.whl

@@ -2,5 +2,5 @@ This graphrag wheel file was built from the following repo
 
 https://github.com/microsoft/graphrag
 
-on commit hash 6afec2ae421eabf468e1da8595e1e23d4bda5ecb
+on commit hash a389514b4723189803097dc0d602b50d139ee92c
 

backend/graphrag-wheel/note.txt

@@ -4,9 +4,9 @@
 # this yaml file serves as a configuration template for the graphrag indexing jobs
 # some values are hardcoded while others will be dynamically set
 input:
-  type: text
+  type: blob
+  file_type: text
   file_pattern: .*\.txt$
-  storage_type: blob
   storage_account_blob_url: $STORAGE_ACCOUNT_BLOB_URL
   container_name: PLACEHOLDER
   base_dir: .
@@ -35,6 +35,7 @@ llm:
   api_version: $GRAPHRAG_API_VERSION
   model: $GRAPHRAG_LLM_MODEL
   deployment_name: $GRAPHRAG_LLM_DEPLOYMENT_NAME
+  cognitive_services_endpoint: $GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT
   model_supports_json: True
   tokens_per_minute: 80000
   requests_per_minute: 480
@@ -56,6 +57,7 @@ embeddings:
     batch_size: 16
     model: $GRAPHRAG_EMBEDDING_MODEL
     deployment_name: $GRAPHRAG_EMBEDDING_DEPLOYMENT_NAME
+    cognitive_services_endpoint: $GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT
     tokens_per_minute: 350000
     concurrent_requests: 25
     requests_per_minute: 2100

backend/src/api/pipeline_settings.yaml

@@ -44,7 +44,8 @@
 )
 
 cognitive_services_endpoint = os.environ.get(
-    "COGNITIVE_SERVICES_ENDPOINT", "https://cognitiveservices.azure.com/.default"
+    "GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT",
+    "https://cognitiveservices.azure.com/.default",
 )
 token_provider = get_bearer_token_provider(
     DefaultAzureCredential(), cognitive_services_endpoint

backend/src/meta_agent/community/retrieve.py

@@ -19,7 +19,8 @@
 )
 
 cognitive_services_endpoint = os.environ.get(
-    "COGNITIVE_SERVICES_ENDPOINT", "https://cognitiveservices.azure.com/.default"
+    "GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT",
+    "https://cognitiveservices.azure.com/.default",
 )
 token_provider = get_bearer_token_provider(
     DefaultAzureCredential(), cognitive_services_endpoint

backend/src/meta_agent/global_search/retrieve.py

@@ -78,8 +78,9 @@ In the `deploy.parameters.json` file, provide values for the following required
 `GRAPHRAG_EMBEDDING_DEPLOYMENT_NAME` |                                    | Yes | Deployment name of the Azure OpenAI embedding model.
 `APIM_NAME`                          |                                    | No  | Hostname of the API. Must be a globally unique name. The API will be accessible at `https://<APIM_NAME>.azure-api.net`. If not provided a unique name will be generated.
 `RESOURCE_BASE_NAME`                 |                                    | No  | Suffix to apply to all azure resource names. If not provided a unique suffix will be generated.
-`AISEARCH_ENDPOINT_SUFFIX`           |                                    | No  | Suffix to apply to AI search endpoint. Will default to `search.windows.net` for Azure commercial cloud but should be defined for deployments in other Azure clouds.
-`REPORTERS`                          |                                    | No  | The type of logging to enable. If not provided, logging will be saved to a file in Azure Storage and to the console in AKS
+`AISEARCH_ENDPOINT_SUFFIX`           |                                    | No  | Suffix to apply to AI search endpoint. Will default to `search.windows.net` for Azure Commercial cloud but should be overriden for deployments in other Azure clouds.
+`REPORTERS`                          |                                    | No  | The type of logging to enable. If not provided, logging will be saved to a file in Azure Storage and to the console in AKS.
+`GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT` |                                  | No  | Endpoint for cognitive services identity authorization. Will default to `https://cognitiveservices.azure.com/.default` for Azure Commercial cloud but should be defined for deployments in other Azure clouds.
 
 ## 5. Deploy the solution accelerator
 ```

docs/DEPLOYMENT-GUIDE.md

@@ -11,6 +11,7 @@ AISEARCH_ENDPOINT_SUFFIX=""
 APIM_NAME=""
 RESOURCE_BASE_NAME=""
 REPORTERS=""
+GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT=""
 
 requiredParams=(
     CONTAINER_REGISTRY_SERVER
@@ -152,6 +153,10 @@ populateOptionalParams () {
         REPORTERS="blob,console"
         printf "\tsetting REPORTERS=blob,console\n"
     fi
+    if [ -z "$GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT" ]; then
+        GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT="https://cognitiveservices.azure.com/.default"
+        printf "\tsetting GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT=$GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT\n"
+    fi
     printf "Done.\n"
 }
 
@@ -353,6 +358,7 @@ deployHelmChart () {
         --set "graphragConfig.COSMOS_URI_ENDPOINT=$cosmosEndpoint" \
         --set "graphragConfig.GRAPHRAG_API_BASE=$GRAPHRAG_API_BASE" \
         --set "graphragConfig.GRAPHRAG_API_VERSION=$GRAPHRAG_API_VERSION" \
+	--set "graphragConfig.GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT=$GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT" \
         --set "graphragConfig.GRAPHRAG_LLM_MODEL=$GRAPHRAG_LLM_MODEL" \
         --set "graphragConfig.GRAPHRAG_LLM_DEPLOYMENT_NAME=$GRAPHRAG_LLM_DEPLOYMENT_NAME" \
         --set "graphragConfig.GRAPHRAG_EMBEDDING_MODEL=$GRAPHRAG_EMBEDDING_MODEL" \

infra/deploy.sh

@@ -31,6 +31,7 @@ graphragConfig:
   COSMOS_URI_ENDPOINT: ""
   GRAPHRAG_API_BASE: ""
   GRAPHRAG_API_VERSION: ""
+  GRAPHRAG_COGNITIVE_SERVICES_ENDPOINT: "https://cognitiveservices.azure.com/.default"
   GRAPHRAG_LLM_MODEL: ""
   GRAPHRAG_LLM_DEPLOYMENT_NAME: ""
   GRAPHRAG_EMBEDDING_MODEL: ""