standard-tests[patch]: require model_name in response_metadata if returns_usage_metadata (#30497)

We are implementing a token-counting callback handler in
`langchain-core` that is intended to work with all chat models
supporting usage metadata. The callback will aggregate usage metadata by
model. This requires responses to include the model name in its
metadata.

To support this, if a model `returns_usage_metadata`, we check that it
includes a string model name in its `response_metadata` in the
`"model_name"` key.

More context: https://github.com/langchain-ai/langchain/pull/30487

libs/partners/mistralai/tests/integration_tests/test_chat_models.py

@@ -20,7 +20,7 @@ def test_stream() -> None:
     """Test streaming tokens from ChatMistralAI."""
     llm = ChatMistralAI()
 
-    for token in llm.stream("I'm Pickle Rick"):
+    for token in llm.stream("Hello"):
         assert isinstance(token.content, str)
 
 
@@ -30,16 +30,19 @@ async def test_astream() -> None:
 
     full: Optional[BaseMessageChunk] = None
     chunks_with_token_counts = 0
-    async for token in llm.astream("I'm Pickle Rick"):
+    chunks_with_response_metadata = 0
+    async for token in llm.astream("Hello"):
         assert isinstance(token, AIMessageChunk)
         assert isinstance(token.content, str)
         full = token if full is None else full + token
         if token.usage_metadata is not None:
             chunks_with_token_counts += 1
-    if chunks_with_token_counts != 1:
+        if token.response_metadata:
+            chunks_with_response_metadata += 1
+    if chunks_with_token_counts != 1 or chunks_with_response_metadata != 1:
         raise AssertionError(
-            "Expected exactly one chunk with token counts. "
-            "AIMessageChunk aggregation adds counts. Check that "
+            "Expected exactly one chunk with token counts or response_metadata. "
+            "AIMessageChunk aggregation adds / appends counts and metadata. Check that "
             "this is behaving properly."
         )
     assert isinstance(full, AIMessageChunk)
@@ -50,6 +53,8 @@ async def test_astream() -> None:
         full.usage_metadata["input_tokens"] + full.usage_metadata["output_tokens"]
         == full.usage_metadata["total_tokens"]
     )
+    assert isinstance(full.response_metadata["model_name"], str)
+    assert full.response_metadata["model_name"]
 
 
 async def test_abatch() -> None: