migrate to aiohttp

litellm/llms/openai_like/chat/handler.py

@@ -7,6 +7,7 @@ For handling OpenAI-like chat completions, like IBM WatsonX, etc.
 import json
 from typing import Any, Callable, Optional, Union
 
+import aiohttp
 import httpx
 
 import litellm
@@ -160,7 +161,7 @@ class OpenAILikeChatHandler(OpenAILikeBase):
         model_response: ModelResponse,
         custom_llm_provider: str,
         print_verbose: Callable,
-        client: Optional[AsyncHTTPHandler],
+        client: Optional[aiohttp.ClientSession],
         encoding,
         api_key,
         logging_obj,
@@ -171,33 +172,32 @@ class OpenAILikeChatHandler(OpenAILikeBase):
         litellm_params=None,
         logger_fn=None,
         headers={},
-        timeout: Optional[Union[float, httpx.Timeout]] = None,
+        timeout: Optional[Union[float, aiohttp.ClientTimeout]] = None,
         json_mode: bool = False,
     ) -> ModelResponse:
         if timeout is None:
-            timeout = httpx.Timeout(timeout=600.0, connect=5.0)
+            timeout = aiohttp.ClientTimeout(total=600.0, connect=5.0)
 
         if client is None:
-            client = litellm.module_level_aclient
+            client = aiohttp.ClientSession(timeout=timeout)
 
         try:
-            response = await client.post(
-                api_base, headers=headers, data=json.dumps(data), timeout=timeout
-            )
-            response.raise_for_status()
-        except httpx.HTTPStatusError as e:
+            async with client.post(
+                api_base, headers=headers, data=json.dumps(data)
+            ) as response:
+                response.raise_for_status()
+                response_data = await response.json()
+        except aiohttp.ClientResponseError as e:
             raise OpenAILikeError(
-                status_code=e.response.status_code,
-                message=e.response.text,
+                status_code=e.status,
+                message=e.message,
             )
-        except httpx.TimeoutException:
-            raise OpenAILikeError(status_code=408, message="Timeout error occurred.")
         except Exception as e:
             raise OpenAILikeError(status_code=500, message=str(e))
 
         return OpenAILikeChatConfig._transform_response(
             model=model,
-            response=response,
+            response=response_data,
             model_response=model_response,
             stream=stream,
             logging_obj=logging_obj,
@@ -331,9 +331,9 @@ class OpenAILikeChatHandler(OpenAILikeBase):
                     litellm_params=litellm_params,
                     logger_fn=logger_fn,
                     headers=headers,
-                    timeout=timeout,
+                    timeout=timeout,  # type: ignore
                     base_model=base_model,
-                    client=client,
+                    client=client,  # type: ignore
                 )
         else:
             ## COMPLETION CALL

litellm/main.py

@@ -474,6 +474,7 @@ async def acompletion(
             or custom_llm_provider == "clarifai"
             or custom_llm_provider == "watsonx"
             or custom_llm_provider == "cloudflare"
+            or custom_llm_provider == "openai_like"
             or custom_llm_provider in litellm.openai_compatible_providers
             or custom_llm_provider in litellm._custom_providers
         ):  # currently implemented aiohttp calls for just azure, openai, hf, ollama, vertex ai soon all.
@@ -2822,6 +2823,18 @@ def completion(  # type: ignore # noqa: PLR0915
                 )
                 return response
             response = model_response
+        elif custom_llm_provider == "openai_like":
+            response = openai_like_chat.completion(
+                model=model,
+                messages=messages,
+                api_base=api_base,
+                model_response=model_response,
+                print_verbose=print_verbose,
+                optional_params=optional_params,
+                litellm_params=litellm_params,
+                logger_fn=logger_fn,
+                encoding=encoding,
+            )
         elif custom_llm_provider == "petals" or model in litellm.petals_models:
             api_base = api_base or litellm.api_base
 

litellm/proxy/proxy_config.yaml

@@ -1,33 +1,6 @@
 model_list:
-  - model_name: "openai/*"
+  - model_name: "openai_like/*"
     litellm_params:
-      model: "openai/*"
-      api_key: os.environ/OPENAI_API_KEY
-  - model_name: "azure/*"
-    litellm_params:
-      model: azure/chatgpt-v-2
-      api_base: https://openai-gpt-4-test-v-1.openai.azure.com/
-      client_id: os.environ/AZURE_CLIENT_ID
-      azure_username: os.environ/AZURE_USERNAME
-      azure_password: os.environ/AZURE_PASSWORD
-litellm_settings:
-  callbacks: ["datadog"]
-
-
-general_settings: 
-  alerting: ["pagerduty"]
-  alerting_args:
-    failure_threshold: 4  # Number of requests failing in a window
-    failure_threshold_window_seconds: 10  # Window in seconds
-
-    # Requests hanging threshold
-    hanging_threshold_seconds: 0.0000001  # Number of seconds of waiting for a response before a request is considered hanging
-    hanging_threshold_window_seconds: 10  # Window in seconds
-  key_management_system: "hashicorp_vault"
-
-# For /fine_tuning/jobs endpoints
-finetune_settings:
-  - custom_llm_provider: "vertex_ai"
-    vertex_project: "adroit-crow-413218"
-    vertex_location: "us-central1"
-    vertex_credentials: "/Users/ishaanjaffer/Downloads/adroit-crow-413218-a956eef1a2a8.json"
+      model: openai_like/*
+      api_base: https://exampleopenaiendpoint-production.up.railway.app/
+      api_key: "ishaan"