Fixed passing scheduler-specific kwargs via TrainingArguments lr_scheduler_kwargs (#27595)

* Fix passing scheduler-specific kwargs through TrainingArguments `lr_scheduler_kwargs`

* Added test for lr_scheduler_kwargs

src/transformers/trainer.py

@@ -1111,7 +1111,7 @@ class Trainer:
                 optimizer=self.optimizer if optimizer is None else optimizer,
                 num_warmup_steps=self.args.get_warmup_steps(num_training_steps),
                 num_training_steps=num_training_steps,
-                **self.args.lr_scheduler_kwargs,
+                scheduler_specific_kwargs=self.args.lr_scheduler_kwargs,
             )
             self._created_lr_scheduler = True
         return self.lr_scheduler

tests/trainer/test_trainer.py

@@ -39,6 +39,7 @@ from transformers import (
     IntervalStrategy,
     PretrainedConfig,
     TrainingArguments,
+    get_polynomial_decay_schedule_with_warmup,
     is_torch_available,
     logging,
 )
@@ -643,6 +644,33 @@ class TrainerIntegrationPrerunTest(TestCasePlus, TrainerIntegrationCommon):
         self.assertFalse(torch.allclose(trainer.model.b, b))
         self.assertEqual(trainer.optimizer.state_dict()["param_groups"][0]["lr"], 1.0)
 
+    def test_lr_scheduler_kwargs(self):
+        # test scheduler kwargs passed via TrainingArguments
+        train_dataset = RegressionDataset()
+        model = RegressionModel()
+        num_steps, num_warmup_steps = 10, 2
+        extra_kwargs = {"power": 5.0, "lr_end": 1e-5}  # Non-default arguments
+        args = TrainingArguments(
+            "./regression",
+            lr_scheduler_type="polynomial",
+            lr_scheduler_kwargs=extra_kwargs,
+            learning_rate=0.2,
+            warmup_steps=num_warmup_steps,
+        )
+        trainer = Trainer(model, args, train_dataset=train_dataset)
+        trainer.create_optimizer_and_scheduler(num_training_steps=num_steps)
+
+        # Checking that the scheduler was created
+        self.assertIsNotNone(trainer.lr_scheduler)
+
+        # Checking that the correct args were passed
+        sched1 = trainer.lr_scheduler
+        sched2 = get_polynomial_decay_schedule_with_warmup(
+            trainer.optimizer, num_warmup_steps=num_warmup_steps, num_training_steps=num_steps, **extra_kwargs
+        )
+        self.assertEqual(sched1.lr_lambdas[0].args, sched2.lr_lambdas[0].args)
+        self.assertEqual(sched1.lr_lambdas[0].keywords, sched2.lr_lambdas[0].keywords)
+
     def test_reduce_lr_on_plateau_args(self):
         # test passed arguments for a custom ReduceLROnPlateau scheduler
         train_dataset = RegressionDataset(length=64)