!13225 Get a string path when the summary path is a list

From: @ouwenchang
Reviewed-by: @yelihua,@lixiaohui33
Signed-off-by: @lixiaohui33

mindspore/train/callback/_summary_collector.py

@@ -20,6 +20,7 @@ import json
 from json.decoder import JSONDecodeError
 
 from importlib import import_module
+from collections.abc import Iterable
 
 import numpy as np
 
@@ -842,12 +843,21 @@ class SummaryCollector(Callback):
         dataset_file_set = (dataset_package.MindDataset, dataset_package.ManifestDataset)
         dataset_files_set = (dataset_package.TFRecordDataset, dataset_package.TextFileDataset)
 
+        dataset_path = ''
+
         if isinstance(output_dataset, dataset_file_set):
-            return output_dataset.dataset_file
+            dataset_path = output_dataset.dataset_file
         if isinstance(output_dataset, dataset_dir_set):
-            return output_dataset.dataset_dir
+            dataset_path = output_dataset.dataset_dir
         if isinstance(output_dataset, dataset_files_set):
-            return output_dataset.dataset_files[0]
+            dataset_path = output_dataset.dataset_files[0]
+
+        if dataset_path:
+            if isinstance(dataset_path, str):
+                return dataset_path
+            if isinstance(dataset_path, Iterable):
+                return list(dataset_path)[0]
+
         return self._get_dataset_path(output_dataset.children[0])
 
     @staticmethod

mindspore/train/summary/_summary_adapter.py

@@ -22,6 +22,7 @@ from PIL import Image
 from mindspore import log as logger
 from mindspore import context
 from mindspore.communication.management import get_rank
+from mindspore.communication.management import GlobalComm
 
 from ..._checkparam import Validator
 from ..anf_ir_pb2 import DataType, ModelProto
@@ -57,7 +58,7 @@ def get_event_file_name(prefix, suffix, time_second):
 
     device_num = context.get_auto_parallel_context('device_num')
     device_id = context.get_context('device_id')
-    if device_num > 1:
+    if device_num > 1 or GlobalComm.WORLD_COMM_GROUP == 'nccl_world_group':
         # Notice:
         # In GPU distribute training scene, get_context('device_id') will not work,
         # so we use get_rank instead of get_context.