Enforce saving at end of training if saving option chosen

src/transformers/trainer_callback.py

@@ -544,6 +544,9 @@ def on_step_end(self, args: TrainingArguments, state: TrainerState, control: Tra
         # End training
         if state.global_step >= state.max_steps:
             control.should_training_stop = True
+            # Save the model at the end if we have a save strategy
+            if args.save_strategy != IntervalStrategy.NO:
+                control.should_save = True
 
         return control
 

src/transformers/training_args.py

@@ -333,6 +333,9 @@ class TrainingArguments:
                 - `"no"`: No save is done during training.
                 - `"epoch"`: Save is done at the end of each epoch.
                 - `"steps"`: Save is done every `save_steps`.
+
+                If `"epoch"` or `"steps"` is chosen, saving will also be performed at the
+                very end of training, always.
         save_steps (`int` or `float`, *optional*, defaults to 500):
             Number of updates steps before two checkpoint saves if `save_strategy="steps"`. Should be an integer or a
             float in range `[0,1)`. If smaller than 1, will be interpreted as ratio of total training steps.

tests/trainer/test_trainer.py

@@ -128,6 +128,7 @@
     if is_safetensors_available():
         import safetensors.torch
 
+
 # for version specific tests in TrainerIntegrationTest
 require_accelerate_version_min_0_28 = partial(require_accelerate, min_version="0.28")
 GRAD_ACCUM_KWARGS_VERSION_AVAILABLE = is_accelerate_available("0.28")
@@ -1968,6 +1969,56 @@ def test_safe_checkpoints(self):
                     tmpdir, 5, int(self.n_epochs * 64 / self.batch_size), False, safe_weights=save_safetensors
                 )
 
+    def test_load_best_model_with_save(self):
+        with tempfile.TemporaryDirectory() as tmpdir:
+            trainer = get_regression_trainer(
+                output_dir=tmpdir,
+                save_steps=5,
+                evaluation_strategy="steps",
+                eval_steps=5,
+                max_steps=9,
+            )
+            trainer.train()
+            # Check that we have the last known step:
+            assert os.path.exists(
+                os.path.join(tmpdir, f"checkpoint-{trainer.state.max_steps}")
+            ), f"Could not find checkpoint-{trainer.state.max_steps}"
+            # And then check the last step
+            assert os.path.exists(os.path.join(tmpdir, "checkpoint-9")), "Could not find checkpoint-9"
+
+        # Now test that using a limit works
+        # Should result in:
+        # - save at step 5 (but is deleted)
+        # - save at step 10 (loaded in at the end when `load_best_model=True`)
+        # - save at step 11
+        with tempfile.TemporaryDirectory() as tmpdir:
+            trainer = get_regression_trainer(
+                output_dir=tmpdir,
+                save_steps=5,
+                evaluation_strategy="steps",
+                eval_steps=5,
+                load_best_model_at_end=True,
+                save_total_limit=2,
+                max_steps=11,
+            )
+            trainer.train()
+            # Check that we have the last known step:
+            assert os.path.exists(os.path.join(tmpdir, "checkpoint-11")), "Could not find checkpoint-11"
+            # And then check the last multiple
+            assert os.path.exists(os.path.join(tmpdir, "checkpoint-10")), "Could not find checkpoint-10"
+            # Finally check that we don't have an old one
+            assert not os.path.exists(os.path.join(tmpdir, "checkpoint-5")), "Found checkpoint-5, limit not respected"
+
+            # Finally check that the right model was loaded in, checkpoint-10
+            # this goes by the last `eval` step check to do so, so it won't be
+            # the last model *saved*
+            model_state = trainer.model.state_dict()
+            final_model_weights = safetensors.torch.load_file(
+                os.path.join(tmpdir, "checkpoint-10", "model.safetensors")
+            )
+            for k, v in model_state.items():
+                assert torch.allclose(v, final_model_weights[k]), f"{k} is not the same"
+
     @require_torch_multi_accelerator
     def test_run_seq2seq_double_train_wrap_once(self):
         # test that we don't wrap the model more than once

tests/trainer/test_trainer_callback.py

@@ -153,7 +153,7 @@ def get_expected_events(self, trainer):
                     expected_events.append("on_log")
                 if trainer.args.eval_strategy == IntervalStrategy.STEPS and step % trainer.args.eval_steps == 0:
                     expected_events += evaluation_events.copy()
-                if step % trainer.args.save_steps == 0:
+                if step % trainer.args.save_steps == 0 or step == trainer.state.max_steps:
                     expected_events.append("on_save")
             expected_events.append("on_epoch_end")
             if trainer.args.eval_strategy == IntervalStrategy.EPOCH: