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_whisper.py

@@ -11,7 +11,10 @@ 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="WhisperDeployment", num_replicas=1)
@@ -42,15 +45,18 @@ class WhisperDeployment:
         print("Whisper model loaded successfully")
 
         # MLflow metrics
-        self._mlflow = InferenceLogger(
-            experiment_name="ray-serve-whisper",
-            run_name=f"whisper-{self.model_size}",
-            tags={"model.name": f"whisper-{self.model_size}", "model.framework": "faster-whisper", "device": self.device},
-            flush_every=5,
-        )
-        self._mlflow.initialize(
-            params={"model_size": self.model_size, "device": self.device, "compute_type": self.compute_type}
-        )
+        if InferenceLogger is not None:
+            self._mlflow = InferenceLogger(
+                experiment_name="ray-serve-whisper",
+                run_name=f"whisper-{self.model_size}",
+                tags={"model.name": f"whisper-{self.model_size}", "model.framework": "faster-whisper", "device": self.device},
+                flush_every=5,
+            )
+            self._mlflow.initialize(
+                params={"model_size": self.model_size, "device": self.device, "compute_type": self.compute_type}
+            )
+        else:
+            self._mlflow = None
 
     async def __call__(self, request: dict[str, Any]) -> dict[str, Any]:
         """
@@ -146,12 +152,13 @@ class WhisperDeployment:
             }
 
         # Log to MLflow
-        self._mlflow.log_request(
-            latency_s=time.time() - _start,
-            audio_duration_s=info.duration,
-            segments=len(segment_list),
-            realtime_factor=(time.time() - _start) / info.duration if info.duration > 0 else 0,
-        )
+        if self._mlflow:
+            self._mlflow.log_request(
+                latency_s=time.time() - _start,
+                audio_duration_s=info.duration,
+                segments=len(segment_list),
+                realtime_factor=(time.time() - _start) / info.duration if info.duration > 0 else 0,
+            )
 
         # Default JSON format (OpenAI-compatible)
         return {