fix(main.py): pass extra headers through for async calls

litellm/main.py

@@ -294,6 +294,7 @@ async def acompletion(
         "api_version": api_version,
         "api_key": api_key,
         "model_list": model_list,
+        "extra_headers": extra_headers,
         "acompletion": True,  # assuming this is a required parameter
     }
     if custom_llm_provider is None:

litellm/proxy/_super_secret_config.yaml

@@ -24,8 +24,8 @@ general_settings:
   database_connection_timeout: 60
   health_check_interval: 300
   ui_access_mode: all
-litellm_settings:
-  json_logs: true
+# litellm_settings:
+#   json_logs: true
 model_list:
 - litellm_params:
     api_base: http://0.0.0.0:8080
@@ -43,5 +43,14 @@ model_list:
     api_version: '2023-05-15'
     model: azure/chatgpt-v-2
   model_name: gpt-3.5-turbo
+- litellm_params:
+    model: anthropic.claude-3-sonnet-20240229-v1:0
+  model_name: bedrock-anthropic-claude-3
+- litellm_params:
+    api_base: https://openai-gpt-4-test-v-1.openai.azure.com/
+    api_key: os.environ/AZURE_API_KEY
+    api_version: '2023-05-15'
+    model: azure/chatgpt-v-2
+  model_name: gpt-3.5-turbo
 router_settings:
   enable_pre_call_checks: true

litellm/tests/test_bedrock_completion.py

@@ -13,7 +13,7 @@ import pytest
 import litellm
 from litellm import embedding, completion, completion_cost, Timeout, ModelResponse
 from litellm import RateLimitError
-from litellm.llms.custom_httpx.http_handler import HTTPHandler
+from litellm.llms.custom_httpx.http_handler import HTTPHandler, AsyncHTTPHandler
 from unittest.mock import patch, AsyncMock, Mock
 
 # litellm.num_retries = 3
@@ -519,20 +519,21 @@ def test_bedrock_ptu():
         mock_client_post.assert_called_once()
 
 
-def test_bedrock_extra_headers():
+@pytest.mark.asyncio
+async def test_bedrock_extra_headers():
     """
     Check if a url with 'modelId' passed in, is created correctly
 
     Reference: https://github.com/BerriAI/litellm/issues/3805
     """
-    client = HTTPHandler()
+    client = AsyncHTTPHandler()
 
-    with patch.object(client, "post", new=Mock()) as mock_client_post:
+    with patch.object(client, "post", new=AsyncMock()) as mock_client_post:
         litellm.set_verbose = True
         from openai.types.chat import ChatCompletion
 
         try:
-            response = litellm.completion(
+            response = await litellm.acompletion(
                 model="anthropic.claude-3-sonnet-20240229-v1:0",
                 messages=[{"role": "user", "content": "What's AWS?"}],
                 client=client,