VIAME Training Pipeline Issues

VIAME & DIVE
1

Hello,

I have installed VIAME using the Windows zip file on GitHub then extracted everything. Annotating video footage has gone well, but when I have tried to train a model using annotated footage, a bunch of errors pop up through the process: some Python dependencies have been unable to be found, files can’t be found, etc. After the training has finished, I am unable to export the model, getting the error that yolo.weights is missing. I have tried to deploy the model as well, but quit after 5 days of loading. Do I need to install Python and its libraries?

Thank you!

Example Errors:

image
image1404×473 27.7 KB

(FileNotFoundError: [Errno 2] No such file or directory: ‘deep_training\fit\runs\viame-netharn-detector\nyqgnref\monitor\layers\module.detector.backbone.conv1\layer-module.detector.backbone.conv1-epoch_0_init.jpg’)

2

In your attached screenshot all of that text are just warnings, and can be ignored. You do not need to install Python, as it is included in the installation.

You do need to have installed adequate NVIDIA GPU drivers installed for training.

To see the status of training, I would need to see the full log underneath this (it’s best to copy and paste the full log into a text file and send it)

3

Thanks for getting back to me! I attached the log.

(Attachment 0011_and_0033_trained_netharn_cfrnn_log.txt is missing)