Give more useful `metric_for_best_model` errors (#31450)

Give more useful metric_for_best_model errors

src/transformers/trainer.py

@@ -2749,7 +2749,13 @@ class Trainer:
             metric_to_check = self.args.metric_for_best_model
             if not metric_to_check.startswith("eval_"):
                 metric_to_check = f"eval_{metric_to_check}"
-            self.lr_scheduler.step(metrics[metric_to_check])
+            try:
+                self.lr_scheduler.step(metrics[metric_to_check])
+            except KeyError as exc:
+                raise KeyError(
+                    f"The `metric_for_best_model` training argument is set to '{metric_to_check}', which is not found in the evaluation metrics. "
+                    f"The available evaluation metrics are: {list(metrics.keys())}. Consider changing the `metric_for_best_model` via the TrainingArguments."
+                ) from exc
         return metrics
 
     def _maybe_log_save_evaluate(self, tr_loss, grad_norm, model, trial, epoch, ignore_keys_for_eval):
@@ -2873,7 +2879,13 @@ class Trainer:
             metric_to_check = self.args.metric_for_best_model
             if not metric_to_check.startswith("eval_"):
                 metric_to_check = f"eval_{metric_to_check}"
-            metric_value = metrics[metric_to_check]
+            try:
+                metric_value = metrics[metric_to_check]
+            except KeyError as exc:
+                raise KeyError(
+                    f"The `metric_for_best_model` training argument is set to '{metric_to_check}', which is not found in the evaluation metrics. "
+                    f"The available evaluation metrics are: {list(metrics.keys())}. Consider changing the `metric_for_best_model` via the TrainingArguments."
+                ) from exc
 
             operator = np.greater if self.args.greater_is_better else np.less
             if (