Add health check monitoring for EXL2 errors (#206)

* Add health check monitoring for EXL2 errors

* Health: Format and change status code

A status code of 503 makes more sense to use.
---------

backends/exllamav2/model.py

@@ -32,6 +32,8 @@ from typing import List, Optional, Union
 
 from ruamel.yaml import YAML
 
+from common.health import HealthManager
+
 from backends.exllamav2.grammar import (
     ExLlamaV2Grammar,
     clear_grammar_func_cache,
@@ -1373,6 +1375,8 @@ class ExllamaV2Container:
             )
             asyncio.ensure_future(self.create_generator())
 
+            await HealthManager.add_unhealthy_event(ex)
+
             raise ex
         finally:
             # Log generation options to console