[Bug]

Prerequisite

• I have searched Issues and Discussions but cannot get the expected help.
• The bug has not been fixed in the latest version(https://github.com/open-mmlab/mmengine).

Environment

---

UnicodeDecodeError Traceback (most recent call last)
Cell In[11], line 6
3 from mmengine.config import Config
4 from mmengine.registry import init_default_scope
----> 6 cfg = Config.fromfile('rtmdet_tiny_1xb12-40e_cat.py')
8 init_default_scope(cfg.get('default_scope', 'mmdet'))
10 dataset = DATASETS.build(cfg.train_dataloader.dataset)

File ~\anaconda3\lib\site-packages\mmengine\config\config.py:178, in Config.fromfile(filename, use_predefined_variables, import_custom_modules, use_environment_variables)
165 """Build a Config instance from config file.
166
167 Args:
(...)
175 Config: Config instance built from config file.
176 """
177 filename = str(filename) if isinstance(filename, Path) else filename
--> 178 cfg_dict, cfg_text, env_variables = Config._file2dict(
179 filename, use_predefined_variables, use_environment_variables)
180 if import_custom_modules and cfg_dict.get('custom_imports', None):
181 try:

File ~\anaconda3\lib\site-packages\mmengine\config\config.py:471, in Config._file2dict(filename, use_predefined_variables, use_environment_variables)
469 # Substitute predefined variables
470 if use_predefined_variables:
--> 471 Config._substitute_predefined_vars(filename,
472 temp_config_file.name)
473 else:
474 shutil.copyfile(filename, temp_config_file.name)

File ~\anaconda3\lib\site-packages\mmengine\config\config.py:301, in Config._substitute_predefined_vars(filename, temp_config_name)
295 support_templates = dict(
296 fileDirname=file_dirname,
297 fileBasename=file_basename,
298 fileBasenameNoExtension=file_basename_no_extension,
299 fileExtname=file_extname)
300 with open(filename, encoding='utf-8') as f:
--> 301 config_file = f.read()
302 for key, value in support_templates.items():
303 regexp = r'{{\s*' + str(key) + r'\s*}}'

File ~\anaconda3\lib\codecs.py:322, in BufferedIncrementalDecoder.decode(self, input, final)
319 def decode(self, input, final=False):
320 # decode input (taking the buffer into account)
321 data = self.buffer + input
--> 322 (result, consumed) = self._buffer_decode(data, self.errors, final)
323 # keep undecoded input until the next call
324 self.buffer = data[consumed:]

UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb7 in position 96: invalid start byte

Reproduces the problem - code sample

#-- coding: UTF-8--
from mmdet.registry import DATASETS, VISUALIZERS
from mmengine.config import Config
from mmengine.registry import init_default_scope

cfg = Config.fromfile('rtmdet_tiny_1xb12-40e_cat.py')

init_default_scope(cfg.get('default_scope', 'mmdet'))

dataset = DATASETS.build(cfg.train_dataloader.dataset)
visualizer = VISUALIZERS.build(cfg.visualizer)
visualizer.dataset_meta = dataset.metainfo

plt.figure(figsize=(16, 5))

只可视化前 8 张图片

for i in range(8):
item=dataset[i]

img = item['inputs'].permute(1, 2, 0).numpy()
data_sample = item['data_samples'].numpy()
gt_instances = data_sample.gt_instances
img_path = osp.basename(item['data_samples'].img_path)

gt_bboxes = gt_instances.get('bboxes', None)
gt_instances.bboxes = gt_bboxes.tensor
data_sample.gt_instances = gt_instances

visualizer.add_datasample(
osp.basename(img_path),
img,
data_sample,
draw_pred=False,
show=False)
drawed_image=visualizer.get_image()

plt.subplot(2, 4, i+1)
plt.imshow(drawed_image[..., [2, 1, 0]])
plt.title(f"{osp.basename(img_path)}")
plt.xticks([])
plt.yticks([])

plt.tight_layout()

Reproduces the problem - command or script

run

Reproduces the problem - error message

UnicodeDecodeError: 'utf-8' codec can't decode byte 0xb7 in position 96: invalid start byte

Additional information

1.如何解决？
2.为什么报错？

Start with mmengine/config/config.py, especially Config.fromfile and _substitute_predefined_vars, using the reported Config.fromfile call and rtmdet_tiny_1xb12-40e_cat.py to reproduce the UnicodeDecodeError. Done should mean the encoding failure is understood and the reported configuration-loading behavior is verified against a regression check.