Fix learning rate gap on resume (#9468)

Signed-off-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: UltralyticsAssistant <web@ultralytics.com>
Co-authored-by: Glenn Jocher <glenn.jocher@ultralytics.com>
Co-authored-by: EunChan Kim <eunchan@hanyang.ac.kr>
Co-authored-by: Lakshantha Dissanayake <lakshanthad@yahoo.com>
Co-authored-by: RizwanMunawar <chr043416@gmail.com>
Co-authored-by: gs80140 <gs80140@users.noreply.github.com>

ultralytics/engine/trainer.py

@@ -331,6 +331,10 @@ def _do_train(self, world_size=1):
         while True:
             self.epoch = epoch
             self.run_callbacks("on_train_epoch_start")
+            with warnings.catch_warnings():
+                warnings.simplefilter("ignore")  # suppress 'Detected lr_scheduler.step() before optimizer.step()'
+                self.scheduler.step()
+
             self.model.train()
             if RANK != -1:
                 self.train_loader.sampler.set_epoch(epoch)
@@ -426,15 +430,12 @@ def _do_train(self, world_size=1):
             t = time.time()
             self.epoch_time = t - self.epoch_time_start
             self.epoch_time_start = t
-            with warnings.catch_warnings():
-                warnings.simplefilter("ignore")  # suppress 'Detected lr_scheduler.step() before optimizer.step()'
-                if self.args.time:
-                    mean_epoch_time = (t - self.train_time_start) / (epoch - self.start_epoch + 1)
-                    self.epochs = self.args.epochs = math.ceil(self.args.time * 3600 / mean_epoch_time)
-                    self._setup_scheduler()
-                    self.scheduler.last_epoch = self.epoch  # do not move
-                    self.stop |= epoch >= self.epochs  # stop if exceeded epochs
-                self.scheduler.step()
+            if self.args.time:
+                mean_epoch_time = (t - self.train_time_start) / (epoch - self.start_epoch + 1)
+                self.epochs = self.args.epochs = math.ceil(self.args.time * 3600 / mean_epoch_time)
+                self._setup_scheduler()
+                self.scheduler.last_epoch = self.epoch  # do not move
+                self.stop |= epoch >= self.epochs  # stop if exceeded epochs
             self.run_callbacks("on_fit_epoch_end")
             torch.cuda.empty_cache()  # clear GPU memory at end of epoch, may help reduce CUDA out of memory errors
 

ultralytics/utils/torch_utils.py

@@ -16,7 +16,7 @@
 import torch.nn.functional as F
 import torchvision
 
-from ultralytics.utils import DEFAULT_CFG_DICT, DEFAULT_CFG_KEYS, LOGGER, __version__
+from ultralytics.utils import DEFAULT_CFG_DICT, DEFAULT_CFG_KEYS, LOGGER, colorstr, __version__
 from ultralytics.utils.checks import PYTHON_VERSION, check_version
 
 try:
@@ -614,8 +614,9 @@ def __call__(self, epoch, fitness):
         self.possible_stop = delta >= (self.patience - 1)  # possible stop may occur next epoch
         stop = delta >= self.patience  # stop training if patience exceeded
         if stop:
+            prefix = colorstr("EarlyStopping: ")
             LOGGER.info(
-                f"Stopping training early as no improvement observed in last {self.patience} epochs. "
+                f"{prefix}Training stopped early as no improvement observed in last {self.patience} epochs. "
                 f"Best results observed at epoch {self.best_epoch}, best model saved as best.pt.\n"
                 f"To update EarlyStopping(patience={self.patience}) pass a new patience value, "
                 f"i.e. `patience=300` or use `patience=0` to disable EarlyStopping."