add device_id to env when profiling

RELEASE.md

@@ -20,7 +20,7 @@
 ## Bugfixes
 * Fix FileNotFound exception by adding robust check for summary watcher ([!281](https://gitee.com/mindspore/mindinsight/pulls/281)).
 * UI fix operator table sort jump problem ([!283](https://gitee.com/mindspore/mindinsight/pulls/283)).
-* Dataset serializer return schema json str when schema type is 'mindspore.dataset.engine.Schema' ([!2185](https://gitee.com/mindspore/mindspore/pulls/2185)).
+* Dataset serializer return schema json str when schema type is `mindspore.dataset.engine.Schema` ([!2185](https://gitee.com/mindspore/mindspore/pulls/2185)).
 
 
 ## Thanks to our Contributors

mindinsight/profiler/profiling.py

@@ -92,6 +92,7 @@ class Profiler:
         os.environ['PROFILING_MODE'] = 'true'
         os.environ['PROFILING_OPTIONS'] = 'training_trace:task_trace'
         os.environ['MINDDATA_PROFILING_DIR'] = self._output_path
+        os.environ['DEVICE_ID'] = self._dev_id
         # use context interface to open profiling, for the new mindspore version(after 2020.5.21)
         try:
             import mindspore.context as context