[examples] better model example (#10427)

* refactors

* typo

examples/seq2seq/run_seq2seq.py

@@ -572,7 +572,6 @@ def main():
         compute_metrics=compute_metrics if training_args.predict_with_generate else None,
     )
 
-    all_metrics = {}
     # Training
     if training_args.do_train:
         if last_checkpoint is not None:
@@ -589,13 +588,10 @@ def main():
             data_args.max_train_samples if data_args.max_train_samples is not None else len(train_dataset)
         )
         metrics["train_samples"] = min(max_train_samples, len(train_dataset))
-        if trainer.is_world_process_zero():
-            trainer.log_metrics("train", metrics)
-            trainer.save_metrics("train", metrics)
-            all_metrics.update(metrics)
 
-            # Need to save the state, since Trainer.save_model saves only the tokenizer with the model
-            trainer.state.save_to_json(os.path.join(training_args.output_dir, "trainer_state.json"))
+        trainer.log_metrics("train", metrics)
+        trainer.save_metrics("train", metrics)
+        trainer.save_state()
 
     # Evaluation
     results = {}
@@ -608,10 +604,8 @@ def main():
         max_val_samples = data_args.max_val_samples if data_args.max_val_samples is not None else len(eval_dataset)
         metrics["eval_samples"] = min(max_val_samples, len(eval_dataset))
 
-        if trainer.is_world_process_zero():
-            trainer.log_metrics("eval", metrics)
-            trainer.save_metrics("eval", metrics)
-            all_metrics.update(metrics)
+        trainer.log_metrics("eval", metrics)
+        trainer.save_metrics("eval", metrics)
 
     if training_args.do_predict:
         logger.info("*** Test ***")
@@ -626,11 +620,10 @@ def main():
         max_test_samples = data_args.max_test_samples if data_args.max_test_samples is not None else len(test_dataset)
         metrics["test_samples"] = min(max_test_samples, len(test_dataset))
 
-        if trainer.is_world_process_zero():
-            trainer.log_metrics("test", metrics)
-            trainer.save_metrics("test", metrics)
-            all_metrics.update(metrics)
+        trainer.log_metrics("test", metrics)
+        trainer.save_metrics("test", metrics)
 
+        if trainer.is_world_process_zero():
             if training_args.predict_with_generate:
                 test_preds = tokenizer.batch_decode(
                     test_results.predictions, skip_special_tokens=True, clean_up_tokenization_spaces=True
@@ -640,9 +633,6 @@ def main():
                 with open(output_test_preds_file, "w") as writer:
                     writer.write("\n".join(test_preds))
 
-    if trainer.is_world_process_zero():
-        trainer.save_metrics("all", metrics)
-
     return results
 
 

src/transformers/trainer.py

@@ -231,7 +231,7 @@ class Trainer:
 
     """
 
-    from .trainer_pt_utils import _get_learning_rate, log_metrics, metrics_format, save_metrics
+    from .trainer_pt_utils import _get_learning_rate, log_metrics, metrics_format, save_metrics, save_state
 
     def __init__(
         self,

src/transformers/trainer_pt_utils.py

@@ -599,12 +599,16 @@ def log_metrics(self, split, metrics):
     """
     Log metrics in a specially formatted way
 
+    Under distributed environment this is done only for a process with rank 0.
+
     Args:
         split (:obj:`str`):
             Mode/split name: one of ``train``, ``eval``, ``test``
         metrics (:obj:`Dict[str, float]`):
             The metrics returned from train/evaluate/predictmetrics: metrics dict
     """
+    if not self.is_world_process_zero():
+        return
 
     logger.info(f"***** {split} metrics *****")
     metrics_formatted = self.metrics_format(metrics)
@@ -614,16 +618,48 @@ def log_metrics(self, split, metrics):
         logger.info(f"  {key: <{k_width}} = {metrics_formatted[key]:>{v_width}}")
 
 
-def save_metrics(self, split, metrics):
+def save_metrics(self, split, metrics, combined=True):
     """
     Save metrics into a json file for that split, e.g. ``train_results.json``.
 
+    Under distributed environment this is done only for a process with rank 0.
+
     Args:
         split (:obj:`str`):
             Mode/split name: one of ``train``, ``eval``, ``test``, ``all``
         metrics (:obj:`Dict[str, float]`):
             The metrics returned from train/evaluate/predict
+        combined (:obj:`bool`, `optional`, defaults to :obj:`True`):
+            Creates combined metrics by updating ``all_results.json`` with metrics of this call
     """
+    if not self.is_world_process_zero():
+        return
+
     path = os.path.join(self.args.output_dir, f"{split}_results.json")
     with open(path, "w") as f:
         json.dump(metrics, f, indent=4, sort_keys=True)
+
+    if combined:
+        path = os.path.join(self.args.output_dir, "all_results.json")
+        if os.path.exists(path):
+            with open(path, "r") as f:
+                all_metrics = json.load(f)
+        else:
+            all_metrics = {}
+
+        all_metrics.update(metrics)
+        with open(path, "w") as f:
+            json.dump(all_metrics, f, indent=4, sort_keys=True)
+
+
+def save_state(self):
+    """
+    Saves the Trainer state, since Trainer.save_model saves only the tokenizer with the model
+
+    Under distributed environment this is done only for a process with rank 0.
+    """
+    if not self.is_world_process_zero():
+        return
+
+    path = os.path.join(self.args.output_dir, "trainer_state.json")
+    self.state.save_to_json(path)