fix(pattern_match_deployments.py): default to user input if unable to map based on wildcards (#6646)

litellm/router_utils/pattern_match_deployments.py

@@ -117,11 +117,26 @@ class PatternMatchRouter:
 
         E.g.:
 
+        Case 1:
         model_name: llmengine/* (can be any regex pattern or wildcard pattern)
         litellm_params:
             model: openai/*
 
         if model_name = "llmengine/foo" -> model = "openai/foo"
+
+        Case 2:
+        model_name: llmengine/fo::*::static::*
+        litellm_params:
+            model: openai/fo::*::static::*
+
+        if model_name = "llmengine/foo::bar::static::baz" -> model = "openai/foo::bar::static::baz"
+
+        Case 3:
+        model_name: *meta.llama3*
+        litellm_params:
+            model: bedrock/meta.llama3*
+
+        if model_name = "hello-world-meta.llama3-70b" -> model = "bedrock/meta.llama3-70b"
         """
 
         ## BASE CASE: if the deployment model name does not contain a wildcard, return the deployment model name
@@ -134,10 +149,9 @@ class PatternMatchRouter:
         dynamic_segments = matched_pattern.groups()
 
         if len(dynamic_segments) > wildcard_count:
-            raise ValueError(
+            return (
-                f"More wildcards in the deployment model name than the pattern. Wildcard count: {wildcard_count}, dynamic segments count: {len(dynamic_segments)}"
+                matched_pattern.string
-            )
+            )  # default to the user input, if unable to map based on wildcards.
-
         # Replace the corresponding wildcards in the litellm model pattern with extracted segments
         for segment in dynamic_segments:
             litellm_deployment_litellm_model = litellm_deployment_litellm_model.replace(

tests/local_testing/test_router_pattern_matching.py

@@ -182,6 +182,14 @@ async def test_route_with_no_matching_pattern():
     )
     assert result.choices[0].message.content == "Works"
 
+    ## WORKS
+    result = await router.acompletion(
+        model="meta.llama3-70b-instruct-v1:0",
+        messages=[{"role": "user", "content": "Hello, world!"}],
+        mock_response="Works",
+    )
+    assert result.choices[0].message.content == "Works"
+
     ## FAILS
     with pytest.raises(litellm.BadRequestError) as e:
         await router.acompletion(
@@ -198,6 +206,7 @@ async def test_route_with_no_matching_pattern():
             input="Hello, world!",
         )
 
+
 def test_router_pattern_match_e2e():
     """
     Tests the end to end flow of the router
@@ -228,4 +237,3 @@ def test_router_pattern_match_e2e():
             "model": "gpt-4o",
             "messages": [{"role": "user", "content": "Hello, how are you?"}],
         }
-

tests/router_unit_tests/test_router_helper_utils.py

@@ -960,6 +960,18 @@ def test_replace_model_in_jsonl(model_list):
             "openai/gpt-3.5-turbo",
             "openai/gpt-3.5-turbo",
         ),
+        (
+            "bedrock/meta.llama3-70b",
+            "*meta.llama3*",
+            "bedrock/meta.llama3-*",
+            "bedrock/meta.llama3-70b",
+        ),
+        (
+            "meta.llama3-70b",
+            "*meta.llama3*",
+            "bedrock/meta.llama3-*",
+            "meta.llama3-70b",
+        ),
     ],
 )
 def test_pattern_match_deployment_set_model_name(
@@ -1000,6 +1012,7 @@ def test_pattern_match_deployment_set_model_name(
     for model in updated_models:
         assert model["litellm_params"]["model"] == expected_model
 
+
 @pytest.mark.asyncio
 async def test_pass_through_moderation_endpoint_factory(model_list):
     router = Router(model_list=model_list)