Fix val_progress_bar total with num_sanity_val_steps (#3751)

* Fix val_progress_bar total with num_sanity_val_steps

* chlog

* Fix val_progress_bar total with num_sanity_val_steps

* move test

* replaced with sanity flag and suggestions

CHANGELOG.md

@@ -93,9 +93,11 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 - Fixed determinism in `DDPSpawnBackend` when using `seed_everything` in main process ([#3335](https://github.com/PyTorchLightning/pytorch-lightning/pull/3335))
 
-- Fixed `ModelCheckpoint` `period` to actually save every `period` epochs ([3630](https://github.com/PyTorchLightning/pytorch-lightning/pull/3630))
+- Fixed `ModelCheckpoint` `period` to actually save every `period` epochs ([#3630](https://github.com/PyTorchLightning/pytorch-lightning/pull/3630))
 
-- Fixed `ModelCheckpoint` with `save_top_k=-1` option not tracking the best models when a monitor metric is available ([3735](https://github.com/PyTorchLightning/pytorch-lightning/pull/3735))
+- Fixed `val_progress_bar` total with `num_sanity_val_steps` ([#3751](https://github.com/PyTorchLightning/pytorch-lightning/pull/3751))
+
+- Fixed `ModelCheckpoint` with `save_top_k=-1` option not tracking the best models when a monitor metric is available ([#3735](https://github.com/PyTorchLightning/pytorch-lightning/pull/3735))
 
 - Fixed counter-intuitive error being thrown in `Accuracy` metric for zero target tensor ([#3764](https://github.com/PyTorchLightning/pytorch-lightning/pull/3764))
 

pytorch_lightning/callbacks/progress.py

@@ -340,8 +340,9 @@ class ProgressBar(ProgressBarBase):
 
     def on_validation_start(self, trainer, pl_module):
         super().on_validation_start(trainer, pl_module)
-        self.val_progress_bar = self.init_validation_tqdm()
+        if not trainer.running_sanity_check:
-        self.val_progress_bar.total = convert_inf(self.total_val_batches)
+            self.val_progress_bar = self.init_validation_tqdm()
+            self.val_progress_bar.total = convert_inf(self.total_val_batches)
 
     def on_validation_batch_end(self, trainer, pl_module, batch, batch_idx, dataloader_idx):
         super().on_validation_batch_end(trainer, pl_module, batch, batch_idx, dataloader_idx)

tests/callbacks/test_progress_bar.py

@@ -193,3 +193,37 @@ def test_progress_bar_progress_refresh(tmpdir, refresh_rate):
 
     trainer.test(model)
     assert progress_bar.test_batches_seen == progress_bar.total_test_batches
+
+
+@pytest.mark.parametrize(['limit_val_batches', 'expected'], [
+    pytest.param(0, 0),
+    pytest.param(5, 7),
+])
+def test_num_sanity_val_steps_progress_bar(tmpdir, limit_val_batches, expected):
+    """
+    Test val_progress_bar total with 'num_sanity_val_steps' Trainer argument.
+    """
+    class CurrentProgressBar(ProgressBar):
+        def __init__(self):
+            super().__init__()
+            self.val_progress_bar_total = 0
+
+        def on_validation_epoch_end(self, trainer, pl_module):
+            self.val_progress_bar_total += trainer.progress_bar_callback.val_progress_bar.total
+
+    model = EvalModelTemplate()
+    progress_bar = CurrentProgressBar()
+
+    trainer = Trainer(
+        default_root_dir=tmpdir,
+        max_epochs=1,
+        num_sanity_val_steps=2,
+        limit_train_batches=0,
+        limit_val_batches=limit_val_batches,
+        callbacks=[progress_bar],
+        logger=False,
+        checkpoint_callback=False,
+        early_stop_callback=False,
+    )
+    trainer.fit(model)
+    assert trainer.progress_bar_callback.val_progress_bar_total == expected

tests/trainer/test_trainer.py

@@ -957,7 +957,6 @@ def test_num_sanity_val_steps(tmpdir, limit_val_batches):
         max_steps=1,
     )
     assert trainer.num_sanity_val_steps == num_sanity_val_steps
-    val_dataloaders = model.val_dataloader__multiple_mixed_length()
 
 
 @pytest.mark.parametrize(['limit_val_batches'], [
@@ -981,7 +980,6 @@ def test_num_sanity_val_steps_neg_one(tmpdir, limit_val_batches):
         max_steps=1,
     )
     assert trainer.num_sanity_val_steps == float('inf')
-    val_dataloaders = model.val_dataloader__multiple()
 
 
 @pytest.mark.parametrize("trainer_kwargs,expected", [