fix debug completed logging (#1897)

mmdet/utils/contextmanagers.py

@@ -14,8 +14,8 @@ DEBUG_COMPLETED_TIME = bool(os.environ.get('DEBUG_COMPLETED_TIME', False))
 
 
 @contextlib.asynccontextmanager
-async def completed(trace_name="",
-                    name="",
+async def completed(trace_name='',
+                    name='',
                     sleep_interval=0.05,
                     streams: List[torch.cuda.Stream] = None):
     """
@@ -42,7 +42,7 @@ async def completed(trace_name="",
         stream_before_context_switch.record_event(start)
 
         cpu_start = time.monotonic()
-    logger.debug("%s %s starting, streams: %s", trace_name, name, streams)
+    logger.debug('%s %s starting, streams: %s', trace_name, name, streams)
     grad_enabled_before = torch.is_grad_enabled()
     try:
         yield
@@ -60,37 +60,41 @@ async def completed(trace_name="",
 
         # observed change of torch.is_grad_enabled() during concurrent run of
         # async_test_bboxes code
-        assert grad_enabled_before == grad_enabled_after, \
-            "Unexpected is_grad_enabled() value change"
+        assert (grad_enabled_before == grad_enabled_after
+                ), 'Unexpected is_grad_enabled() value change'
 
         are_done = [e.query() for e in end_events]
-        logger.debug("%s %s completed: %s streams: %s", trace_name, name,
+        logger.debug('%s %s completed: %s streams: %s', trace_name, name,
                      are_done, streams)
         with torch.cuda.stream(stream_before_context_switch):
             while not all(are_done):
                 await asyncio.sleep(sleep_interval)
                 are_done = [e.query() for e in end_events]
-                logger.debug("%s %s completed: %s streams: %s", trace_name,
-                             name, are_done, streams)
+                logger.debug(
+                    '%s %s completed: %s streams: %s',
+                    trace_name,
+                    name,
+                    are_done,
+                    streams,
+                )
 
         current_stream = torch.cuda.current_stream()
         assert current_stream == stream_before_context_switch
 
         if DEBUG_COMPLETED_TIME:
             cpu_time = (cpu_end - cpu_start) * 1000
-            stream_times_ms = ""
+            stream_times_ms = ''
             for i, stream in enumerate(streams):
                 elapsed_time = start.elapsed_time(end_events[i])
-                stream_times_ms += " {stream} {elapsed_time:.2f} ms".format(
-                    stream, elapsed_time)
-            logger.info("{trace_name} {name} cpu_time {cpu_time:.2f} ms",
-                        trace_name, name, cpu_time, stream_times_ms)
+                stream_times_ms += ' {} {:.2f} ms'.format(stream, elapsed_time)
+            logger.info('%s %s %.2f ms %s', trace_name, name, cpu_time,
+                        stream_times_ms)
 
 
 @contextlib.asynccontextmanager
 async def concurrent(streamqueue: asyncio.Queue,
-                     trace_name="concurrent",
-                     name="stream"):
+                     trace_name='concurrent',
+                     name='stream'):
     """Run code concurrently in different streams.
 
     :param streamqueue: asyncio.Queue instance.
@@ -110,12 +114,12 @@ async def concurrent(streamqueue: asyncio.Queue,
 
         try:
             with torch.cuda.stream(stream):
-                logger.debug("%s %s is starting, stream: %s", trace_name, name,
+                logger.debug('%s %s is starting, stream: %s', trace_name, name,
                              stream)
                 yield
                 current = torch.cuda.current_stream()
                 assert current == stream
-                logger.debug("%s %s has finished, stream: %s", trace_name,
+                logger.debug('%s %s has finished, stream: %s', trace_name,
                              name, stream)
         finally:
             streamqueue.task_done()