Revert "fix(router.py): return model alias w/ underlying deployment on router.get_model_list()"

This reverts commit 638896309c.

litellm/proxy/_new_secret_config.yaml

@@ -1,9 +1,16 @@
 model_list:
-  - model_name: "gpt-turbo"
+  - model_name: "anthropic/claude-3-5-sonnet-20240620"
     litellm_params:
-      model: azure/chatgpt-v-2
-      api_key: os.environ/AZURE_API_KEY
-      api_base: os.environ/AZURE_API_BASE
+      model: anthropic/claude-3-5-sonnet-20240620
+      # api_base: http://0.0.0.0:9000 
+  - model_name: gpt-3.5-turbo
+    litellm_params:
+      model: openai/*
 
-router_settings:
-  model_group_alias: {"gpt-4": "gpt-turbo"} 
+litellm_settings:
+  success_callback: ["s3"]
+  s3_callback_params:
+    s3_bucket_name: litellm-logs   # AWS Bucket Name for S3
+    s3_region_name: us-west-2              # AWS Region Name for S3
+    s3_aws_access_key_id: os.environ/AWS_ACCESS_KEY_ID  # us os.environ/<variable name> to pass environment variables. This is AWS Access Key ID for S3
+    s3_aws_secret_access_key: os.environ/AWS_SECRET_ACCESS_KEY  # AWS Secret Access Key for S3

litellm/proxy/health_check.py

@@ -3,7 +3,7 @@
 import asyncio
 import logging
 import random
-from typing import List, Optional
+from typing import Optional
 
 import litellm
 from litellm._logging import print_verbose
@@ -36,25 +36,6 @@ def _clean_endpoint_data(endpoint_data: dict, details: Optional[bool] = True):
     )
 
 
-def filter_deployments_by_id(
-    model_list: List,
-) -> List:
-    seen_ids = set()
-    filtered_deployments = []
-
-    for deployment in model_list:
-        _model_info = deployment.get("model_info") or {}
-        _id = _model_info.get("id") or None
-        if _id is None:
-            continue
-
-        if _id not in seen_ids:
-            seen_ids.add(_id)
-            filtered_deployments.append(deployment)
-
-    return filtered_deployments
-
-
 async def _perform_health_check(model_list: list, details: Optional[bool] = True):
     """
     Perform a health check for each model in the list.
@@ -124,9 +105,6 @@ async def perform_health_check(
             _new_model_list = [x for x in model_list if x["model_name"] == model]
         model_list = _new_model_list
 
-    model_list = filter_deployments_by_id(
-        model_list=model_list
-    )  # filter duplicate deployments (e.g. when model alias'es are used)
     healthy_endpoints, unhealthy_endpoints = await _perform_health_check(
         model_list, details
     )

litellm/router.py

@@ -4556,27 +4556,6 @@ class Router:
                     ids.append(id)
         return ids
 
-    def _get_all_deployments(
-        self, model_name: str, model_alias: Optional[str] = None
-    ) -> List[DeploymentTypedDict]:
-        """
-        Return all deployments of a model name
-
-        Used for accurate 'get_model_list'.
-        """
-
-        returned_models: List[DeploymentTypedDict] = []
-        for model in self.model_list:
-            if model["model_name"] == model_name:
-                if model_alias is not None:
-                    alias_model = copy.deepcopy(model)
-                    alias_model["model_name"] = model_name
-                    returned_models.append(alias_model)
-                else:
-                    returned_models.append(model)
-
-        return returned_models
-
     def get_model_names(self) -> List[str]:
         """
         Returns all possible model names for router.
@@ -4588,18 +4567,15 @@ class Router:
     def get_model_list(
         self, model_name: Optional[str] = None
     ) -> Optional[List[DeploymentTypedDict]]:
-        """
-        Includes router model_group_alias'es as well
-        """
         if hasattr(self, "model_list"):
             returned_models: List[DeploymentTypedDict] = []
 
             for model_alias, model_value in self.model_group_alias.items():
-                returned_models.extend(
-                    self._get_all_deployments(
-                        model_name=model_value, model_alias=model_alias
-                    )
+                model_alias_item = DeploymentTypedDict(
+                    model_name=model_alias,
+                    litellm_params=LiteLLMParamsTypedDict(model=model_value),
                 )
+                returned_models.append(model_alias_item)
 
             if model_name is None:
                 returned_models += self.model_list
@@ -4607,7 +4583,8 @@ class Router:
                 return returned_models
 
             for model in self.model_list:
-                returned_models.extend(self._get_all_deployments(model_name=model_name))
+                if model["model_name"] == model_name:
+                    returned_models.append(model)
 
             return returned_models
         return None

proxy_server_config.yaml

@@ -148,7 +148,6 @@ router_settings:
   redis_password: os.environ/REDIS_PASSWORD
   redis_port: os.environ/REDIS_PORT
   enable_pre_call_checks: true
-  model_group_alias: {"my-special-fake-model-alias-name": "fake-openai-endpoint-3"} 
 
 general_settings: 
   master_key: sk-1234 # [OPTIONAL] Use to enforce auth on proxy. See - https://docs.litellm.ai/docs/proxy/virtual_keys