Reformat multi-line logs and docstrings (#1489)

mmdeploy/apis/visualize.py

@@ -30,7 +30,7 @@ def visualize_model(model_cfg: Union[str, mmcv.Config],
         >>> img = 'demo.jpg'
         >>> device = 'cpu'
         >>> visualize_model(model_cfg, deploy_cfg, model, \
-            img, device, show_result=True)
+        >>>    img, device, show_result=True)
 
     Args:
         model_cfg (str | mmcv.Config): Model config file or Config object.
@@ -59,8 +59,8 @@ def visualize_model(model_cfg: Union[str, mmcv.Config],
 
     if isinstance(model, (list, tuple)):
         assert len(model) > 0, 'Model should have at least one element.'
-        assert all([isinstance(m, str) for m in model]), 'All elements in the \
-            list should be str'
+        assert all([isinstance(m, str) for m in model]), \
+            'All elements in the list should be str'
 
         if backend == Backend.PYTORCH:
             model = task_processor.init_pytorch_model(model[0])

mmdeploy/backend/base/base_wrapper.py

@@ -54,9 +54,8 @@ class BaseWrapper(torch.nn.Module, metaclass=ABCMeta):
         """Set the output names."""
         self._output_names = value
 
-    def output_to_list(self, output_dict: Dict[str,
-                                               torch.Tensor]) -> \
-            List[torch.Tensor]:
+    def output_to_list(
+            self, output_dict: Dict[str, torch.Tensor]) -> List[torch.Tensor]:
         """Convert the output dict of forward() to a tensor list.
 
         Args:

mmdeploy/backend/onnxruntime/wrapper.py

@@ -44,11 +44,11 @@ class ORTWrapper(BaseWrapper):
         logger = get_root_logger()
         if osp.exists(ort_custom_op_path):
             session_options.register_custom_ops_library(ort_custom_op_path)
-            logger.info(f'Successfully loaded onnxruntime custom ops from \
-            {ort_custom_op_path}')
+            logger.info('Successfully loaded onnxruntime custom ops from '
+                        f'{ort_custom_op_path}')
         else:
-            logger.warning(f'The library of onnxruntime custom ops does \
-            not exist: {ort_custom_op_path}')
+            logger.warning('The library of onnxruntime custom ops does '
+                           f'not exist: {ort_custom_op_path}')
         device_id = parse_device_id(device)
         providers = ['CPUExecutionProvider'] \
             if device == 'cpu' else \

mmdeploy/backend/tensorrt/init_plugins.py

@@ -34,6 +34,6 @@ def load_tensorrt_plugin() -> bool:
         logger.info(f'Successfully loaded tensorrt plugins from {lib_path}')
         success = True
     else:
-        logger.warning(f'Could not load the library of tensorrt plugins. \
-            Because the file does not exist: {lib_path}')
+        logger.warning(f'Could not load the library of tensorrt plugins.'
+                       f'Because the file does not exist: {lib_path}')
     return success

mmdeploy/backend/tensorrt/onnx2tensorrt.py

@@ -68,8 +68,8 @@ def onnx2tensorrt(work_dir: str,
     if not isinstance(input_shapes, Dict):
         input_shapes = dict(zip(input_names, input_shapes))
 
-    assert device.startswith('cuda'), f'TensorRT requires cuda device, \
-        but given: {device}'
+    assert device.startswith('cuda'), 'TensorRT requires cuda device,' \
+        f'but given: {device}'
 
     device_id = parse_device_id(device)
     assert save_file.endswith(

mmdeploy/backend/tensorrt/wrapper.py

@@ -84,8 +84,8 @@ class TRTWrapper(BaseWrapper):
             self.engine = load(engine)
 
         if not isinstance(self.engine, trt.ICudaEngine):
-            raise TypeError(f'`engine` should be str or trt.ICudaEngine, \
-                but given: {type(self.engine)}')
+            raise TypeError('`engine` should be str or trt.ICudaEngine,'
+                            f'but given: {type(self.engine)}')
 
         self._register_state_dict_hook(TRTWrapper.__on_state_dict)
         self.context = self.engine.create_execution_context()

mmdeploy/codebase/mmcls/deploy/classification.py

@@ -300,8 +300,8 @@ class Classification(BaseTask):
         if 'topk' not in postprocess:
             topk = (1, )
             logger = get_root_logger()
-            logger.warning('no topk in postprocess config, using default \
-                 topk value.')
+            logger.warning('no topk in postprocess config, using default'
+                           'topk value.')
         else:
             topk = postprocess.topk
         postprocess.topk = max(topk)

mmdeploy/codebase/mmcls/deploy/classification_model.py

@@ -178,8 +178,8 @@ def get_classes_from_config(model_cfg: Union[str, mmcv.Config]):
 
     if class_names is None:
         logger = get_root_logger()
-        logger.warning(f'Use generated class names, because \
-            it failed to parse CLASSES from config: {data_cfg}')
+        logger.warning(f'Use generated class names, because '
+                       f'it failed to parse CLASSES from config: {data_cfg}')
         num_classes = model_cfg.model.head.num_classes
         class_names = [str(i) for i in range(num_classes)]
     return class_names

mmdeploy/codebase/mmdet/deploy/object_detection_model.py

@@ -583,8 +583,8 @@ class NCNNEnd2EndModel(End2EndModel):
                  device: str, class_names: Sequence[str],
                  model_cfg: Union[str, mmcv.Config],
                  deploy_cfg: Union[str, mmcv.Config], **kwargs):
-        assert backend == Backend.NCNN, f'only supported ncnn, but give \
-            {backend.value}'
+        assert backend == Backend.NCNN, 'only supported ncnn, but give' \
+            f'{backend.value}'
 
         super(NCNNEnd2EndModel,
               self).__init__(backend, backend_files, device, class_names,
@@ -669,8 +669,8 @@ class RKNNModel(End2EndModel):
                  device: str, class_names: Sequence[str],
                  model_cfg: Union[str, mmcv.Config],
                  deploy_cfg: Union[str, mmcv.Config], **kwargs):
-        assert backend == Backend.RKNN, f'only supported RKNN, but give \
-            {backend.value}'
+        assert backend == Backend.RKNN, 'only supported RKNN, but give' \
+            f'{backend.value}'
 
         super(RKNNModel, self).__init__(backend, backend_files, device,
                                         class_names, deploy_cfg, **kwargs)

mmdeploy/codebase/mmdet/models/backbones.py

@@ -41,8 +41,8 @@ def focus__forward__ncnn(ctx, self, x):
         x (Tensor): The calculated tensor with shape (N, 4*C, H//2, W//2).
     """
     batch_size, c, h, w = x.shape
-    assert h % 2 == 0 and w % 2 == 0, f'focus for yolox needs even feature\
-        height and width, got {(h, w)}.'
+    assert h % 2 == 0 and w % 2 == 0, 'focus for yolox needs even feature' \
+        f'height and width, got {(h, w)}.'
 
     x = x.reshape(batch_size, c * h, 1, w)
     _b, _c, _h, _w = x.shape

mmdeploy/codebase/mmdet/models/dense_heads/base_dense_head.py

@@ -356,8 +356,8 @@ def base_dense_head__get_bboxes__ncnn(ctx,
     """
     assert len(cls_scores) == len(bbox_preds)
     deploy_cfg = ctx.cfg
-    assert not is_dynamic_shape(deploy_cfg), 'base_dense_head for ncnn\
-        only supports static shape.'
+    assert not is_dynamic_shape(deploy_cfg), 'base_dense_head for ncnn' \
+        'only supports static shape.'
 
     if score_factors is None:
         # e.g. Retina, FreeAnchor, Foveabox, etc.
@@ -367,8 +367,8 @@ def base_dense_head__get_bboxes__ncnn(ctx,
         with_score_factors = True
         assert len(cls_scores) == len(score_factors)
     batch_size = cls_scores[0].shape[0]
-    assert batch_size == 1, f'ncnn deployment requires batch size 1, \
-        got {batch_size}.'
+    assert batch_size == 1, 'ncnn deployment requires batch size 1,' \
+        f'got {batch_size}.'
 
     num_levels = len(cls_scores)
     if with_score_factors:
@@ -384,13 +384,15 @@ def base_dense_head__get_bboxes__ncnn(ctx,
             vars = torch.tensor([normalizer, normalizer, 1, 1],
                                 dtype=torch.float32)
         else:
-            assert len(normalizer) == 4, f'normalizer of tblr must be 4,\
-                got {len(normalizer)}'
+            assert len(normalizer) == 4, 'normalizer of tblr must be 4,' \
+                f' got {len(normalizer)}'
 
-            assert (normalizer[0] == normalizer[1] and normalizer[2]
-                    == normalizer[3]), 'normalizer between top \
-                        and bottom, left and right must be the same value, or \
-                        we can not transform it to delta_xywh format.'
+            assert (
+                normalizer[0] == normalizer[1]
+                and normalizer[2] == normalizer[3]
+            ), 'normalizer between top and bottom,' \
+                'left and right must be the same value, or' \
+                'we can not transform it to delta_xywh format.'
 
             vars = torch.tensor([normalizer[0], normalizer[2], 1, 1],
                                 dtype=torch.float32)

mmdeploy/codebase/mmdet/models/dense_heads/yolo_head.py

@@ -80,7 +80,7 @@ def yolov3_head__get_bboxes(ctx,
         pred_map = pred_map.permute(0, 2, 3,
                                     1).reshape(batch_size, -1, self.num_attrib)
         # Inplace operation like
-        # ```pred_map[..., :2] = \torch.sigmoid(pred_map[..., :2])```
+        # ```pred_map[..., :2] = torch.sigmoid(pred_map[..., :2])```
         # would create constant tensor when exporting to onnx
         pred_map_conf = torch.sigmoid(pred_map[..., :2])
         pred_map_rest = pred_map[..., 2:]

mmdeploy/codebase/mmdet/models/dense_heads/yolox_head.py

@@ -188,8 +188,9 @@ def yolox_head__get_bboxes__ncnn(ctx,
     batch_mlvl_scores = torch.cat([dummy_cls_scores, cls_scores], dim=2)
     score_factor = torch.cat(flatten_objectness, dim=1).sigmoid()
     flatten_bbox_preds = torch.cat(flatten_bbox_preds, dim=1)
-    assert flatten_priors.shape[-1] == 4, f'yolox needs (B, N, 4) priors, got\
-        (B, N, {flatten_priors.shape[-1]})'
+    assert flatten_priors.shape[-1] == 4, \
+        'yolox needs (B, N, 4) priors, got ' \
+        f'(B, N, {flatten_priors.shape[-1]})'
     prior_box_x1 = (flatten_priors[:, :, 0:1] - flatten_priors[:, :, 2:3] / 2)\
         / img_width
     prior_box_y1 = (flatten_priors[:, :, 1:2] - flatten_priors[:, :, 3:4] / 2)\

mmdeploy/codebase/mmocr/deploy/text_recognition.py

@@ -53,8 +53,8 @@ def process_model_config(model_cfg: mmcv.Config,
                 test_pipeline[1].transforms[0].height = input_shape[1]
                 test_pipeline[1].transforms[0].max_width = input_shape[0]
             else:
-                raise ValueError(f'Transforms[0] should be ResizeOCR, but got\
-                        {test_pipeline[1].transforms[0].type}')
+                raise ValueError('Transforms[0] should be ResizeOCR, but got'
+                                 f'{test_pipeline[1].transforms[0].type}')
         else:
             test_pipeline[1].update(resize)
     model_cfg.data.test.pipeline = test_pipeline

mmdeploy/codebase/mmrotate/models/roi_heads/gv_ratio_roi_head.py

@@ -25,11 +25,11 @@ def gv_ratio_roi_head__simple_test_bboxes(ctx,
             Default: False.
 
     Returns:
-        tuple[list[Tensor], list[Tensor]]: The first list contains \
-            the boxes of the corresponding image in a batch, each \
-            tensor has the shape (num_boxes, 6) and last dimension \
-            6 represent (x, y, w, h, theta, score). Each Tensor \
-            in the second list is the labels with shape (num_boxes, ). \
+        tuple[list[Tensor], list[Tensor]]: The first list contains
+            the boxes of the corresponding image in a batch, each
+            tensor has the shape (num_boxes, 6) and last dimension
+            6 represent (x, y, w, h, theta, score). Each Tensor
+            in the second list is the labels with shape (num_boxes, ).
             The length of both lists should be equal to batch_size.
     """
 

mmdeploy/codebase/mmrotate/models/roi_heads/oriented_standard_roi_head.py

@@ -57,11 +57,11 @@ def oriented_standard_roi_head__simple_test_bboxes(ctx,
             Default: False.
 
     Returns:
-        tuple[list[Tensor], list[Tensor]]: The first list contains \
-            the boxes of the corresponding image in a batch, each \
-            tensor has the shape (num_boxes, 6) and last dimension \
-            6 represent (x, y, w, h, theta, score). Each Tensor \
-            in the second list is the labels with shape (num_boxes, ). \
+        tuple[list[Tensor], list[Tensor]]: The first list contains
+            the boxes of the corresponding image in a batch, each
+            tensor has the shape (num_boxes, 6) and last dimension
+            6 represent (x, y, w, h, theta, score). Each Tensor
+            in the second list is the labels with shape (num_boxes, ).
             The length of both lists should be equal to batch_size.
     """
 

mmdeploy/core/optimizers/function_marker.py

@@ -204,9 +204,9 @@ def mark(func_name: Optional[str] = None,
 
     Args:
         func_name (str): The name of the function where marks come from.
-        inputs (Sequence[str]): The input names of the marks. The final name \
+        inputs (Sequence[str]): The input names of the marks. The final name
             might have suffix if inputs is list or dictionary.
-        outputs (Sequence[str]): The output names of the marks. The final \
+        outputs (Sequence[str]): The output names of the marks. The final
             name might have suffix if outputs is list or dictionary.
 
     Returns:

mmdeploy/core/rewriters/symbolic_rewriter.py

@@ -19,8 +19,7 @@ class SymbolicRewriter:
     recover after exiting the RewriteContext.
 
     Examples:
-        >>> @SYMBOLIC_REWRITER.register_symbolic('squeeze', \
-        >>> is_pytorch=True)
+        >>> @SYMBOLIC_REWRITER.register_symbolic('squeeze', is_pytorch=True)
         >>> def squeeze_default(ctx, g, self, dim=None):
         >>>     if dim is None:
         >>>         dims = []
@@ -49,9 +48,9 @@ class SymbolicRewriter:
         Args:
             func_name (str): The function name/path to override the symbolic.
             backend (str): The rewriter will be activated on which backend.
-            is_pytorch (bool): Enable this flag if func_name is the name of \
+            is_pytorch (bool): Enable this flag if func_name is the name of
                 a pytorch builtin function.
-            arg_descriptors (Sequence[str]): The argument descriptors of the \
+            arg_descriptors (Sequence[str]): The argument descriptors of the
                 symbol.
             ir (IR): The rewriter will be activated on which IR.
             extra_checkers (Checker | List[Checker] | None): Other requirements

tests/test_ops/test_ops.py

@@ -584,8 +584,8 @@ def test_topk(backend,
         input = torch.rand(1, 8, 12, 17)
     else:
         input = input_list[0]
-    assert input.shape[0] == 1, (f'ncnn batch must be 1, \
-        but got {input.shape[0]}')
+    assert input.shape[0] == 1, ('ncnn batch must be 1, '
+                                 f'but got {input.shape[0]}')
 
     def topk_function(inputs):
         return torch.Tensor.topk(inputs, k, dim, largest, sorted)
@@ -635,11 +635,11 @@ def test_shape(backend,
     else:
         input = input_list[0]
         assert input.dim() == dim + 1, 'input.dim() must equal to dim + 1'
-        assert tuple(input.shape) == orig_shape, 'input.shape must the \
-            same as orig_shape'
+        assert tuple(input.shape) == orig_shape, ('input.shape must the '
+                                                  'same as orig_shape')
 
-    assert input.shape[0] == 1, (f'ncnn batch must be 1, \
-        but got {input.shape[0]}')
+    assert input.shape[0] == 1, ('ncnn batch must be 1, '
+                                 f'but got {input.shape[0]}')
 
     shape_node = make_node('Shape', input_names, output_names)
     assert len(input_names) == 1, 'length of input_names must be 1'
@@ -688,13 +688,13 @@ def test_constantofshape(backend,
         input = input_list[0]
         assert input.dim() == dim + 1, 'input.dim() must equal to dim + 1'
         assert tuple(input.shape) == (n, c, h,
-                                      w)[-dim - 1:], 'input.shape must the \
-            same as orig_shape'
+                                      w)[-dim - 1:], ('input.shape must the '
+                                                      'same as orig_shape')
 
-    assert input.shape[0] == 1, (f'ncnn input batch must be 1, \
-        got {input.shape[0]}')
-    assert input[0][0] == 1, (f'ncnn output mat batch must be 1, \
-        got {input[0][0]}')
+    assert input.shape[0] == 1, ('ncnn input batch must be 1, '
+                                 f'got {input.shape[0]}')
+    assert input[0][0] == 1, ('ncnn output mat batch must be 1, '
+                              f'got {input[0][0]}')
 
     constantofshape_node = make_node(
         'ConstantOfShape', input_names, output_names, value=float(val))
@@ -745,10 +745,10 @@ def test_gather(backend,
     else:
         data = input_list[0]
         indice = input_list[1]
-    assert data.shape[0] == 1, (f'ncnn batch must be 1, \
-        but got {data.shape[0]}')
-    assert indice.shape[0] == 1, (f'ncnn batch must be 1, \
-        but got {indice.shape[0]}')
+    assert data.shape[0] == 1, ('ncnn batch must be 1, '
+                                f'but got {data.shape[0]}')
+    assert indice.shape[0] == 1, ('ncnn batch must be 1, '
+                                  f'but got {indice.shape[0]}')
 
     gather_node = make_node('Gather', input_names, output_names, axis=axis + 1)
     gather_graph = make_graph([gather_node], 'gather_graph', [
@@ -766,8 +766,8 @@ def test_gather(backend,
     import importlib
 
     import onnxruntime
-    assert importlib.util.find_spec('onnxruntime') is not None, 'onnxruntime \
-         not installed.'
+    assert importlib.util.find_spec('onnxruntime') is not None, \
+        'onnxruntime not installed.'
 
     import numpy as np
     session = onnxruntime.InferenceSession(gather_model.SerializeToString())
@@ -795,11 +795,12 @@ def test_tensorslice(backend, dim, input_list=None, save_dir=None):
         input = torch.rand((8, 12, 17)[-dim:]).unsqueeze(0)
     else:
         input = input_list[0]
-        assert input.dim() == dim + 1, f'input.dim() must equal to \
-            dim + 1, expected: {dim + 1}, got: {input.dim()}'
+        assert input.dim() == dim + 1, ('input.dim() must equal to '
+                                        f'dim + 1, expected: {dim + 1}, '
+                                        f'got: {input.dim()}')
 
-    assert input.shape[0] == 1, (f'ncnn batch must be 1, \
-        but got {input.shape[0]}')
+    assert input.shape[0] == 1, ('ncnn batch must be 1, '
+                                 f'but got {input.shape[0]}')
 
     def tensorslice_function(inputs):
         if dim == 1:
@@ -835,10 +836,10 @@ def test_expand(backend,
     else:
         input = input_list[0]
         target = input_list[1]
-    assert input.shape[0] == 1, (f'ncnn batch must be 1, \
-        but not {input.shape[0]}')
-    assert target.shape[0] == 1, (f'ncnn batch must be 1, \
-        but not {target.shape[0]}')
+    assert input.shape[0] == 1, (f'ncnn batch must be 1, '
+                                 f'but not {input.shape[0]}')
+    assert target.shape[0] == 1, (f'ncnn batch must be 1, '
+                                  f'but not {target.shape[0]}')
 
     def expand_function(input, target):
         return input.expand_as(target)

tools/deploy.py

@@ -34,8 +34,8 @@ def parse_args():
         help='the dir to save logs and models')
     parser.add_argument(
         '--calib-dataset-cfg',
-        help='dataset config path used to calibrate in int8 mode. If not \
-            specified, it will use "val" dataset in model config instead.',
+        help=('dataset config path used to calibrate in int8 mode. If not '
+              'specified, it will use "val" dataset in model config instead.'),
         default=None)
     parser.add_argument(
         '--device', help='device used for conversion', default='cpu')
@@ -198,9 +198,9 @@ def main():
         assert len(model_params) == len(ir_files)
 
         from mmdeploy.apis.tensorrt import is_available as trt_is_available
-        assert trt_is_available(
-        ), 'TensorRT is not available,' \
-            + ' please install TensorRT and build TensorRT custom ops first.'
+        assert trt_is_available(), (
+            'TensorRT is not available,'
+            ' please install TensorRT and build TensorRT custom ops first.')
 
         from mmdeploy.apis.tensorrt import onnx2tensorrt
         PIPELINE_MANAGER.enable_multiprocess(True, [onnx2tensorrt])
@@ -229,9 +229,9 @@ def main():
         from mmdeploy.apis.ncnn import is_available as is_available_ncnn
 
         if not is_available_ncnn():
-            logger.error('ncnn support is not available, please make sure \
-                1) `mmdeploy_onnx2ncnn` existed in `PATH` \
-                2) python import ncnn success')
+            logger.error('ncnn support is not available, please make sure:\n'
+                         '1) `mmdeploy_onnx2ncnn` existed in `PATH`\n'
+                         '2) python import ncnn success')
             exit(1)
 
         import mmdeploy.apis.ncnn as ncnn_api
@@ -279,9 +279,10 @@ def main():
         from mmdeploy.apis.snpe import is_available as is_available
 
         if not is_available():
-            logger.error('snpe support is not available, please check \
-                1) `snpe-onnx-to-dlc` existed in `PATH` 2) snpe only support \
-                    ubuntu18.04')
+            logger.error('snpe support is not available, please check\n'
+                         '1) `snpe-onnx-to-dlc` existed in `PATH`\n'
+                         '2) snpe only support\n'
+                         'ubuntu18.04')
             exit(1)
 
         import mmdeploy.apis.snpe as snpe_api