PyTorch nightly breaks mmcv API
See original GitHub issueJust upgraded pytorch, but all mmlab code can not runing anylonger:
torch.__version__
'1.13.0.dev20220921+cu116'
error:
anaconda3/lib/python3.9/site-packages/mmcv/onnx/symbolic.py", line 10, in <module>
from torch.onnx.symbolic_registry import register_op
ModuleNotFoundError: No module named 'torch.onnx.symbolic_registry'
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "/home//dev/codes/work/mmkd/export_onnx.py", line 22, in <module>
raise NotImplementedError("please update mmcv to version>=1.0.4")
NotImplementedError: please update mmcv to version>=1.0.4
I have 1.6.1 installed (lastest)
Installing collected packages: mmcv-full
Successfully installed mmcv-full-1.6.1
Successfully installed mmcv-full.
Please fix the wrong import error and wrong error messages. Pytorch seems changed their API
Issue Analytics
- State:
- Created a year ago
- Comments:5 (1 by maintainers)
Top Results From Across the Web
mmcv Documentation
MMCV is a foundational library for computer vision research and supports many research projects as below:.
Read more >torch.fx — PyTorch 1.13 documentation
An example of using these APIs to append a torch.relu() call can be found below. ... We can see what's happening during the...
Read more >pytorch cuda error: no kernel image is available for execution ...
I have a problem about CUDA and PyTorch. I am trying to run a simple code from SinGAN implementation but I got a...
Read more >Message "note: This error originates from a subprocess, and is ...
I had Python 3.10 and had the same error installing psutil from pip. I fixed the problem by installing an older version of...
Read more >mmcv-nightly - PyPI
mmcv -nightly 1.1.0.dev20200823. pip install mmcv-nightly. Copy PIP instructions. Latest version. Released: Aug 23, 2020. OpenMMLab Computer Vision ...
Read more >
Top Related Medium Post
No results found
Top Related StackOverflow Question
No results found
Troubleshoot Live Code
Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free
Top Related Reddit Thread
No results found
Top Related Hackernoon Post
No results found
Top Related Tweet
No results found
Top Related Dev.to Post
No results found
Top Related Hashnode Post
No results found
I’d also suggest moving away from using private methods in symbolic_helper. They are private (leading with
_
) so their signatures and even existence are not guaranteed.I would just make a copy and use them here.
closed by #2294