fix: make mlflow_logger import optional with no-op fallback

The strixhalo LLM worker uses py_executable pointing to the Docker
image venv which doesn't have the updated ray-serve-apps package.
Wrap all InferenceLogger imports in try/except and guard usage with
None checks so apps degrade gracefully without MLflow logging.

ray_serve/serve_llm.py

@@ -10,7 +10,10 @@ from typing import Any
 
 from ray import serve
 
-from ray_serve.mlflow_logger import InferenceLogger
+try:
+    from ray_serve.mlflow_logger import InferenceLogger
+except ImportError:
+    InferenceLogger = None
 
 
 @serve.deployment(name="LLMDeployment", num_replicas=1)
@@ -40,19 +43,22 @@ class LLMDeployment:
         print(f"Model {self.model_id} async engine created")
 
         # MLflow metrics
-        self._mlflow = InferenceLogger(
-            experiment_name="ray-serve-llm",
-            run_name=f"llm-{self.model_id.split('/')[-1]}",
-            tags={"model.name": self.model_id, "model.framework": "vllm", "gpu": "strixhalo"},
-            flush_every=5,
-        )
-        self._mlflow.initialize(
-            params={
-                "model_id": self.model_id,
-                "max_model_len": str(self.max_model_len),
-                "gpu_memory_utilization": str(self.gpu_memory_utilization),
-            }
-        )
+        if InferenceLogger is not None:
+            self._mlflow = InferenceLogger(
+                experiment_name="ray-serve-llm",
+                run_name=f"llm-{self.model_id.split('/')[-1]}",
+                tags={"model.name": self.model_id, "model.framework": "vllm", "gpu": "strixhalo"},
+                flush_every=5,
+            )
+            self._mlflow.initialize(
+                params={
+                    "model_id": self.model_id,
+                    "max_model_len": str(self.max_model_len),
+                    "gpu_memory_utilization": str(self.gpu_memory_utilization),
+                }
+            )
+        else:
+            self._mlflow = None
 
     async def __call__(self, request: dict[str, Any]) -> dict[str, Any]:
         """
@@ -96,15 +102,16 @@ class LLMDeployment:
         completion_tokens = len(generated_text.split())
 
         # Log to MLflow
-        self._mlflow.log_request(
-            latency_s=latency,
-            prompt_tokens=prompt_tokens,
-            completion_tokens=completion_tokens,
-            total_tokens=prompt_tokens + completion_tokens,
-            tokens_per_second=completion_tokens / latency if latency > 0 else 0,
-            temperature=temperature,
-            max_tokens_requested=max_tokens,
-        )
+        if self._mlflow:
+            self._mlflow.log_request(
+                latency_s=latency,
+                prompt_tokens=prompt_tokens,
+                completion_tokens=completion_tokens,
+                total_tokens=prompt_tokens + completion_tokens,
+                tokens_per_second=completion_tokens / latency if latency > 0 else 0,
+                temperature=temperature,
+                max_tokens_requested=max_tokens,
+            )
 
         # Return OpenAI-compatible response
         return {