Fixes #490

`EarlyStopping` should check the metric of interest `on_validation_end` rather than `on_epoch_end`. In a normal scenario, this does not cause a problem, but in combination with `check_val_every_n_epoch>1` in the `Trainer` it results in a warning or in a `RuntimeError` depending on `strict`.
Lightning-AI · May 25, 2020 · 006a067 · 006a067
1 parent 8ca8336
commit 006a067
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/pytorch_lightning/callbacks/early_stopping.py b/pytorch_lightning/callbacks/early_stopping.py
@@ -109,7 +109,7 @@ def on_train_start(self, trainer, pl_module):
         self.stopped_epoch = 0
         self.best = torch_inf if self.monitor_op == torch.lt else -torch_inf
 
-    def on_epoch_end(self, trainer, pl_module):
+    def on_validation_end(self, trainer, pl_module):
         logs = trainer.callback_metrics
         stop_training = False
         if not self._validate_condition_metric(logs):