version conflict in colab

[miladlink]

Hi
thanks to neuralmagic team for great repo
i was using this repo for Sparse Transfer Learning With YOLOv5 in google colab
but after installation ddint workout any commands including !parseml.yolov5.train ... or !sparseml.yolov5.export_onnx ... because of version conflict.
please help me to use this repo in google colab invironment

[rahul-tuli]

Hey @miladlink, Thank you for reaching out with this issue, could you provide some more information about the following,

1. The exact install command you used, for example pip install sparseml[torchvision] if you used that
2. The python version you used
3. The sparseml version installed, you can use pip show sparseml to do that
4. The exact command which failed, with the Traceback.

This will help in getting to the root cause.

[miladlink]

thanks for your reply i just had used pip install parseml your hint([torchvision]) make code works and i trained a pruned and quantized model using transfer learning with recepie but after training the saved .pt weight didnt export to onnx with sparseml.yolov5.export_onnx ... the error is about loading state dict from saved weight could you please tell me why? i just do anything to documentations thanks to neuralmagic team best regards

…

On Wed, Sep 14, 2022, 18:14 Rahul Tuli ***@***.***> wrote: Hey @miladlink <https://github.com/miladlink>, Thank you for reaching out with this issue, could you provide some more information about the following, 1. The exact install command you used, for example pip install sparseml[torchvision] if you used that 2. The python version you used 3. The sparseml version installed, you can use pip show sparseml to do that 4. The exact command which failed, with the Traceback. This information will be helpful in getting to the root cause. — Reply to this email directly, view it on GitHub <#1034 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ATPFP33MRDHKEQXNKNHDFFDV6HJCJANCNFSM6AAAAAAQML3KA4> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[rahul-tuli]

Hi @miladlink; Hopefully this issue is resolved now, if you still see a issue while exporting, kindly open a new issue with the relevant stack trace. Closing this one out for now.