ray_algorithm_config.patch

--- .env/lib/python3.12/site-packages/ray/rllib/algorithms/algorithm_config.py	2024-12-04 09:56:08.215374535 +0100
+++ .env/lib/python3.12/site-packages/ray/rllib/algorithms/algorithm_config_new.py	2024-12-09 12:38:21.850533422 +0100
@@ -4352,14 +4352,7 @@
         # If `evaluation_num_env_runners` > 0, warn if `evaluation_interval` is 0 or
         # None.
         if self.evaluation_num_env_runners > 0 and not self.evaluation_interval:
-            logger.warning(
-                f"You have specified {self.evaluation_num_env_runners} "
-                "evaluation workers, but your `evaluation_interval` is 0 or None! "
-                "Therefore, evaluation doesn't occur automatically with each"
-                " call to `Algorithm.train()`. Instead, you have to call "
-                "`Algorithm.evaluate()` manually in order to trigger an "
-                "evaluation run."
-            )
+            pass
         # If `evaluation_num_env_runners=0` and
         # `evaluation_parallel_to_training=True`, warn that you need
         # at least one remote eval worker for parallel training and