update timer (#1179)

mmdeploy/codebase/mmocr/deploy/text_detection.py

@@ -36,9 +36,10 @@ def process_model_config(model_cfg: mmengine.Config,
             pipeline[i].meta_keys = tuple(j for j in pipeline[i].meta_keys
                                           if j != 'instances')
         # for static exporting
-        if input_shape is not None and transform.type == 'Resize':
-            pipeline[i].keep_ratio = False
-            pipeline[i].scale = tuple(input_shape)
+        if input_shape is not None:
+            if transform.type in ('Resize', 'ShortScaleAspectJitter'):
+                pipeline[i] = mmengine.ConfigDict(
+                    dict(type='Resize', scale=input_shape, keep_ratio=False))
 
     pipeline = [
         transform for transform in pipeline

mmdeploy/utils/timer.py

@@ -2,6 +2,7 @@
 import time
 import warnings
 from contextlib import contextmanager
+from logging import Logger
 from typing import Optional
 
 import numpy as np
@@ -43,6 +44,7 @@ class TimeCounter:
                 log_interval=log_interval,
                 warmup=warmup,
                 with_sync=with_sync,
+                batch_size=1,
                 enable=False)
 
             def fun(*args, **kwargs):
@@ -51,6 +53,7 @@ class TimeCounter:
                 log_interval = cls.names[name]['log_interval']
                 warmup = cls.names[name]['warmup']
                 with_sync = cls.names[name]['with_sync']
+                batch_size = cls.names[name]['batch_size']
                 enable = cls.names[name]['enable']
 
                 count += 1
@@ -66,7 +69,7 @@ class TimeCounter:
                 if enable:
                     if with_sync and torch.cuda.is_available():
                         torch.cuda.synchronize()
-                    elapsed = time.perf_counter() - start_time
+                    elapsed = (time.perf_counter() - start_time) / batch_size
 
                 if enable and count > warmup:
                     execute_time.append(elapsed)
@@ -93,7 +96,9 @@ class TimeCounter:
                  log_interval: int = 1,
                  with_sync: bool = False,
                  file: Optional[str] = None,
-                 logger=None):
+                 logger: Optional[Logger] = None,
+                 batch_size: int = 1,
+                 **kwargs):
         """Activate the time counter.
 
         Args:
@@ -105,6 +110,8 @@ class TimeCounter:
                 default False.
             file (str | None): The file to save output messages. The default
                 is `None`.
+            logger (Logger): The logger for the timer. Default to None.
+            batch_size (int): The batch size. Default to 1.
         """
         assert warmup >= 1
         if logger is None:
@@ -118,12 +125,14 @@ class TimeCounter:
             cls.names[func_name]['warmup'] = warmup
             cls.names[func_name]['log_interval'] = log_interval
             cls.names[func_name]['with_sync'] = with_sync
+            cls.names[func_name]['batch_size'] = batch_size
             cls.names[func_name]['enable'] = True
         else:
             for name in cls.names:
                 cls.names[name]['warmup'] = warmup
                 cls.names[name]['log_interval'] = log_interval
                 cls.names[name]['with_sync'] = with_sync
+                cls.names[name]['batch_size'] = batch_size
                 cls.names[name]['enable'] = True
         yield
         if func_name is not None:

tools/profiler.py

@@ -112,12 +112,12 @@ def main():
     is_pytorch = model_ext in ['.pth', '.pt']
     if is_pytorch:
         # load pytorch model
-        model = task_processor.init_pytorch_model(args.model[0])
+        model = task_processor.build_pytorch_model(args.model[0])
         model = TorchWrapper(model)
         backend = Backend.PYTORCH.value
     else:
         # load the model of the backend
-        model = task_processor.init_backend_model(args.model)
+        model = task_processor.build_backend_model(args.model)
         backend = get_backend(deploy_cfg).value
 
     model = model.eval().to(args.device)
@@ -140,11 +140,17 @@ def main():
         ]
     image_files = image_files[:total_nrof_image]
     with TimeCounter.activate(
-            warmup=args.warmup, log_interval=20, with_sync=with_sync):
+            warmup=args.warmup,
+            log_interval=20,
+            with_sync=with_sync,
+            batch_size=args.batch_size):
         for i in range(0, total_nrof_image, args.batch_size):
             batch_files = image_files[i:(i + args.batch_size)]
-            data, _ = task_processor.create_input(batch_files, input_shape)
-            task_processor.run_inference(model, data)
+            data, _ = task_processor.create_input(
+                batch_files,
+                input_shape,
+                data_preprocessor=getattr(model, 'data_preprocessor', None))
+            model.test_step(data)
 
     print('----- Settings:')
     settings = PrettyTable()

tools/test.py

@@ -143,7 +143,8 @@ def main():
                 warmup=args.warmup,
                 log_interval=args.log_interval,
                 with_sync=with_sync,
-                file=args.log2file):
+                file=args.log2file,
+                batch_size=test_dataloader['batch_size']):
             runner.test()
     else:
         runner.test()