test

# What does this PR do?


## Test Plan
# What does this PR do?


## Test Plan
# What does this PR do?


## Test Plan
Completes the refactoring started in previous commit by:

1. **Fix library client** (critical): Add logic to detect Pydantic model parameters
   and construct them properly from request bodies. The key fix is to NOT exclude
   any params when converting the body for Pydantic models - we need all fields
   to pass to the Pydantic constructor.

   Before: _convert_body excluded all params, leaving body empty for Pydantic construction
   After: Check for Pydantic params first, skip exclusion, construct model with full body

2. **Update remaining providers** to use new Pydantic-based signatures:
   - litellm_openai_mixin: Extract extra fields via __pydantic_extra__
   - databricks: Use TYPE_CHECKING import for params type
   - llama_openai_compat: Use TYPE_CHECKING import for params type
   - sentence_transformers: Update method signatures to use params

3. **Update unit tests** to use new Pydantic signature:
   - test_openai_mixin.py: Use OpenAIChatCompletionRequestParams

This fixes test failures where the library client was trying to construct
Pydantic models with empty dictionaries.
The previous fix had a bug: it called _convert_body() which only keeps fields
that match function parameter names. For Pydantic methods with signature:
  openai_chat_completion(params: OpenAIChatCompletionRequestParams)

The signature only has 'params', but the body has 'model', 'messages', etc.
So _convert_body() returned an empty dict.

Fix: Skip _convert_body() entirely for Pydantic params. Use the raw body
directly to construct the Pydantic model (after stripping NOT_GIVENs).

This properly fixes the ValidationError where required fields were missing.
The streaming code path (_call_streaming) had the same issue as non-streaming:
it called _convert_body() which returned empty dict for Pydantic params.

Applied the same fix as commit 7476c0ae:
- Detect Pydantic model parameters before body conversion
- Skip _convert_body() for Pydantic params
- Construct Pydantic model directly from raw body (after stripping NOT_GIVENs)

This fixes streaming endpoints like openai_chat_completion with stream=True.
The streaming code path (_call_streaming) had the same issue as non-streaming:
it called _convert_body() which returned empty dict for Pydantic params.

Applied the same fix as commit 7476c0ae:
- Detect Pydantic model parameters before body conversion
- Skip _convert_body() for Pydantic params
- Construct Pydantic model directly from raw body (after stripping NOT_GIVENs)

This fixes streaming endpoints like openai_chat_completion with stream=True.

tests/unit/providers/inference/test_remote_vllm.py

@@ -186,3 +186,47 @@ async def test_openai_chat_completion_is_async(vllm_inference_adapter):
 
         assert mock_create_client.call_count == 4  # no cheating
         assert total_time < (sleep_time * 2), f"Total time taken: {total_time}s exceeded expected max"
+
+
+async def test_extra_body_forwarding(vllm_inference_adapter):
+    """
+    Test that extra_body parameters (e.g., chat_template_kwargs) are correctly
+    forwarded to the underlying OpenAI client.
+    """
+    mock_model = Model(identifier="mock-model", provider_resource_id="mock-model", provider_id="vllm-inference")
+    vllm_inference_adapter.model_store.get_model.return_value = mock_model
+
+    with patch.object(VLLMInferenceAdapter, "client", new_callable=PropertyMock) as mock_client_property:
+        mock_client = MagicMock()
+        mock_client.chat.completions.create = AsyncMock(
+            return_value=OpenAIChatCompletion(
+                id="chatcmpl-abc123",
+                created=1,
+                model="mock-model",
+                choices=[
+                    OpenAIChoice(
+                        message=OpenAIAssistantMessageParam(
+                            content="test response",
+                        ),
+                        finish_reason="stop",
+                        index=0,
+                    )
+                ],
+            )
+        )
+        mock_client_property.return_value = mock_client
+
+        # Test with chat_template_kwargs for Granite thinking mode
+        await vllm_inference_adapter.openai_chat_completion(
+            "mock-model",
+            messages=[],
+            stream=False,
+            chat_template_kwargs={"thinking": True},
+        )
+
+        # Verify that the client was called with extra_body containing chat_template_kwargs
+        mock_client.chat.completions.create.assert_called_once()
+        call_kwargs = mock_client.chat.completions.create.call_args.kwargs
+        assert "extra_body" in call_kwargs
+        assert "chat_template_kwargs" in call_kwargs["extra_body"]
+        assert call_kwargs["extra_body"]["chat_template_kwargs"] == {"thinking": True}

tests/unit/providers/utils/inference/test_openai_mixin.py

@@ -12,7 +12,7 @@ from unittest.mock import AsyncMock, MagicMock, Mock, PropertyMock, patch
 import pytest
 from pydantic import BaseModel, Field
 
-from llama_stack.apis.inference import Model, OpenAIUserMessageParam
+from llama_stack.apis.inference import Model, OpenAIChatCompletionRequestParams, OpenAIUserMessageParam
 from llama_stack.apis.models import ModelType
 from llama_stack.core.request_headers import request_provider_data_context
 from llama_stack.providers.utils.inference.model_registry import RemoteInferenceProviderConfig
@@ -271,7 +271,8 @@ class TestOpenAIMixinImagePreprocessing:
             with patch("llama_stack.providers.utils.inference.openai_mixin.localize_image_content") as mock_localize:
                 mock_localize.return_value = (b"fake_image_data", "jpeg")
 
-                await mixin.openai_chat_completion(model="test-model", messages=[message])
+                params = OpenAIChatCompletionRequestParams(model="test-model", messages=[message])
+                await mixin.openai_chat_completion(params)
 
             mock_localize.assert_called_once_with("http://example.com/image.jpg")
 
@@ -303,7 +304,8 @@ class TestOpenAIMixinImagePreprocessing:
 
         with patch.object(type(mixin), "client", new_callable=PropertyMock, return_value=mock_client):
             with patch("llama_stack.providers.utils.inference.openai_mixin.localize_image_content") as mock_localize:
-                await mixin.openai_chat_completion(model="test-model", messages=[message])
+                params = OpenAIChatCompletionRequestParams(model="test-model", messages=[message])
+                await mixin.openai_chat_completion(params)
 
             mock_localize.assert_not_called()