cannot import name 'DwsConvBlock' from 'pytorchcv.models.common' #1664
Labels
bug
Something isn't working
Comments
|
The problem is that pytorchcv is just updated to v0.0.69 on colab. You can workaround this by downgrade the version to 0.0.67 with: Most notebooks work after this workaround, apart from an exception: AttributeError: ‘SimpleCNN’ object has no attribute ‘recursive_adaptation’ in the section Nested Dynamic Modules. But I'm afraid this is another problem. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
🐛 Describe the bug
Using the colab notebook "02_models.ipynb", I encountered this error while executing the 2nd code block in the statement, "from avalanche.models import SimpleCNN".
🐜 To Reproduce
Run the "02_models.ipynd" in Google Colab
🐝 Expected behavior
No Errors and SampleCNN model is printed.
🐞 Screenshots
🦋 Additional context
The error happens in "/usr/local/lib/python3.10/dist-packages/avalanche/models/mobilenetv1.py". The original statements were:
try:
from pytorchcv.models.mobilenet import DwsConvBlock
except Exception:
from pytorchcv.models.common import DwsConvBlock
"DwsConvBlock" is coded in common/conv.py, so the correct code should be:
try:
from pytorchcv.models.mobilenet import DwsConvBlock
except Exception:
from pytorchcv.models.common.conv import DwsConvBlock
The text was updated successfully, but these errors were encountered: