Maskrcnn-benchmark: Class advice impossible in Python3. Use the @implementer class decorator instead.

[kumi123]

❓ Questions and Help

I got an error like ‘’Class advice impossible in Python3. Use the @implementer class decorator instead.'' what should i do ?

[fmassa]

Are you sure this is related to maskrcnn-benchmark?

[Godricly]

I got same issue with apex. A fix can be found here.:smile:

[qihao-huang]

I got same issue with apex. A fix can be found here.

works well with your method. thx

[zhixuanli]

I got same issue with apex. A fix can be found here.

works well with your method. thx

I got same issue with apex. A fix can be found here.

@pietern @sotte @Celebio @martinruenz @Iwontbecreative

I have tried this way, but still remain this problem.
My environment is:

• pytorch 1.1.0
• CUDA 9.2 (I have tried 9.0 and 10.0 version too)

Can anyone help? Thanks a lot!!

[zhixuanli]

Finally solved by:

git clone https://github.com/NVIDIA/apex.git && cd apex && python setup.py install --cuda_ext --cpp_ext

[2mb]

I got same issue with apex. A fix can be found here.😄

I fixed my problem. thanks you.