enforce vertex streaming to use sse for streaming

litellm/proxy/proxy_config.yaml

@@ -4,15 +4,6 @@ model_list:
       model: openai/gpt-4o
       api_key: os.environ/OPENAI_API_KEY
 
-
-router_settings:
-  provider_budget_config: 
-    openai: 
-      budget_limit: 0.000000000001 # float of $ value budget for time period
-      time_period: 1d # can be 1d, 2d, 30d 
-    azure:
-      budget_limit: 100
-      time_period: 1d
-
-litellm_settings:
-  callbacks: ["prometheus"]
+default_vertex_config:
+  vertex_project: "adroit-crow-413218"
+  vertex_location: "us-central1"

litellm/proxy/vertex_ai_endpoints/vertex_endpoints.py

@@ -194,14 +194,16 @@ async def vertex_proxy_route(
     verbose_proxy_logger.debug("updated url %s", updated_url)
 
     ## check for streaming
+    target = str(updated_url)
     is_streaming_request = False
     if "stream" in str(updated_url):
         is_streaming_request = True
+        target += "?alt=sse"
 
     ## CREATE PASS-THROUGH
     endpoint_func = create_pass_through_route(
         endpoint=endpoint,
-        target=str(updated_url),
+        target=target,
         custom_headers=headers,
     )  # dynamically construct pass-through endpoint based on incoming path
     received_value = await endpoint_func(