chore: turn OpenAIMixin into a pydantic.BaseModel (#3671)

# What does this PR do?

- implement get_api_key instead of relying on
LiteLLMOpenAIMixin.get_api_key
 - remove use of LiteLLMOpenAIMixin
 - add default initialize/shutdown methods to OpenAIMixin
 - remove __init__s to allow proper pydantic construction
- remove dead code from vllm adapter and associated / duplicate unit
tests
 - update vllm adapter to use openaimixin for model registration
 - remove ModelRegistryHelper from fireworks & together adapters
 - remove Inference from nvidia adapter
 - complete type hints on embedding_model_metadata
- allow extra fields on OpenAIMixin, for model_store, __provider_id__,
etc
 - new recordings for ollama
 - enhance the list models error handling
- update cerebras (remove cerebras-cloud-sdk) and anthropic (custom
model listing) inference adapters
 - parametrized test_inference_client_caching
- remove cerebras, databricks, fireworks, together from blanket mypy
exclude
 - removed unnecessary litellm deps

## Test Plan

ci

llama_stack/providers/remote/inference/llama_openai_compat/__init__.py

@@ -4,14 +4,12 @@
 # This source code is licensed under the terms described in the LICENSE file in
 # the root directory of this source tree.
 
-from llama_stack.apis.inference import InferenceProvider
-
 from .config import LlamaCompatConfig
 
 
-async def get_adapter_impl(config: LlamaCompatConfig, _deps) -> InferenceProvider:
+async def get_adapter_impl(config: LlamaCompatConfig, _deps):
     # import dynamically so the import is used only when it is needed
     from .llama import LlamaCompatInferenceAdapter
 
-    adapter = LlamaCompatInferenceAdapter(config)
+    adapter = LlamaCompatInferenceAdapter(config=config)
     return adapter

llama_stack/providers/remote/inference/llama_openai_compat/llama.py

@@ -8,38 +8,21 @@ from typing import Any
 from llama_stack.apis.inference.inference import OpenAICompletion
 from llama_stack.log import get_logger
 from llama_stack.providers.remote.inference.llama_openai_compat.config import LlamaCompatConfig
-from llama_stack.providers.utils.inference.litellm_openai_mixin import LiteLLMOpenAIMixin
 from llama_stack.providers.utils.inference.openai_mixin import OpenAIMixin
 
 logger = get_logger(name=__name__, category="inference::llama_openai_compat")
 
 
-class LlamaCompatInferenceAdapter(OpenAIMixin, LiteLLMOpenAIMixin):
+class LlamaCompatInferenceAdapter(OpenAIMixin):
+    config: LlamaCompatConfig
+
+    provider_data_api_key_field: str = "llama_api_key"
     """
     Llama API Inference Adapter for Llama Stack.
-
-    Note: The inheritance order is important here. OpenAIMixin must come before
-    LiteLLMOpenAIMixin to ensure that OpenAIMixin.check_model_availability()
-    is used instead of ModelRegistryHelper.check_model_availability().
-
-    - OpenAIMixin.check_model_availability() queries the Llama API to check if a model exists
-    - ModelRegistryHelper.check_model_availability() (inherited by LiteLLMOpenAIMixin) just returns False and shows a warning
     """
 
-    _config: LlamaCompatConfig
-
-    def __init__(self, config: LlamaCompatConfig):
-        LiteLLMOpenAIMixin.__init__(
-            self,
-            litellm_provider_name="meta_llama",
-            api_key_from_config=config.api_key,
-            provider_data_api_key_field="llama_api_key",
-            openai_compat_api_base=config.openai_compat_api_base,
-        )
-        self.config = config
-
-    # Delegate the client data handling get_api_key method to LiteLLMOpenAIMixin
-    get_api_key = LiteLLMOpenAIMixin.get_api_key
+    def get_api_key(self) -> str:
+        return self.config.api_key or ""
 
     def get_base_url(self) -> str:
         """
@@ -49,12 +32,6 @@ class LlamaCompatInferenceAdapter(OpenAIMixin, LiteLLMOpenAIMixin):
         """
         return self.config.openai_compat_api_base
 
-    async def initialize(self):
-        await super().initialize()
-
-    async def shutdown(self):
-        await super().shutdown()
-
     async def openai_completion(
         self,
         model: str,