Fix: handle logging of scalars in Weights & Biases summary (#29612)

fix: handle logging of scalars in wandb summary

fixes:  #29430

src/transformers/integrations/integration_utils.py

@@ -802,13 +802,25 @@ class WandbCallback(TrainerCallback):
                 self._wandb.run.log_artifact(artifact)
 
     def on_log(self, args, state, control, model=None, logs=None, **kwargs):
+        single_value_scalars = [
+            "train_runtime",
+            "train_samples_per_second",
+            "train_steps_per_second",
+            "train_loss",
+            "total_flos",
+        ]
+
         if self._wandb is None:
             return
         if not self._initialized:
             self.setup(args, state, model)
         if state.is_world_process_zero:
-            logs = rewrite_logs(logs)
-            self._wandb.log({**logs, "train/global_step": state.global_step})
+            for k, v in logs.items():
+                if k in single_value_scalars:
+                    self._wandb.run.summary[k] = v
+            non_scalar_logs = {k: v for k, v in logs.items() if k not in single_value_scalars}
+            non_scalar_logs = rewrite_logs(non_scalar_logs)
+            self._wandb.log({**non_scalar_logs, "train/global_step": state.global_step})
 
     def on_save(self, args, state, control, **kwargs):
         if self._log_model == "checkpoint" and self._initialized and state.is_world_process_zero: