SCRFD人脸检测关键点模型 示例報錯 about modelscope HOT 6 CLOSED

thank you for reporting the issue. this is a known regession that is caused by our last release, a fix will be available this week.

from modelscope.

thank you for reporting the issue. this is a known regession that is caused by our last release, a fix will be available this week.

Thanks! I have updated modelscope, the new issue generate:

from modelscope.

thank you for reporting the issue. this is a known regession that is caused by our last release, a fix will be available this week.

Thanks! I have updated modelscope, the new issue generate:

try to delete the local model cache and run it again？

from modelscope.

please delete the entire local model cache folder:

and try again

from modelscope.

please delete the entire local model cache folder:

and try again

Yes, it works. After I assign path to 'damo/cv_resnet_facedetection_scrfd10gkps', it is normal. But it is so strange. When I run the following code, it also works.

from modelscope.pipelines import pipeline
from modelscope.utils.constant import Tasks

p = pipeline('face-detection', 'cache_dir/damo/cv_resnet_facedetection_scrfd10gkps')

result = p('face_detection.png',)

print(result)

from modelscope.

for this code

p = pipeline('face-detection', 'cache_dir/damo/cv_resnet_facedetection_scrfd10gkps')

the second parameter is model_id_or_model_dir, when model id (damo/cv_resnet_facedetection_scrfd10gkps) is supplied, it will fetch the correct version of the model from modelscope hub, when a local model_dir is provided, it will use the local dir path, in this case, user would be responsbile for enesuring that the correct version of modelscope library is used for the model dir.

from modelscope.